Membrane (and Soluble) Protein Stability and Binding Measurements

Subscriber access provided by University of South Dakota

Article

Membrane (and Soluble) Protein Stability and Binding Measurements in the Lipid Cubic Phase Using Label-free Differential Scanning Fluorimetry Coilín Boland, Samir Olatunji, Jonathan Bailey, Nicole Howe, Dietmar Weichert, Susan Kathleen Fetics, Xiaoxiao Yu, Javier Merino-Gracia, Clement Delsaut, and Martin Caffrey Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.8b03176 • Publication Date (Web): 05 Sep 2018 Downloaded from http://pubs.acs.org on September 6, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Page |1 Membrane (and Soluble) Protein Stability and Binding Measurements in the Lipid Cubic Phase Using Label-free Differential Scanning Fluorimetry Coilín Boland*, Samir Olatunji*, Jonathan Bailey*, Nicole Howe*, Dietmar Weichert, Susan Kathleen Fetics, Xiaoxiao Yu, Javier Merino-Gracia, Clement Delsaut and Martin Caffrey * These authors contributed equally to the work. School of Biochemistry and Immunology, Trinity College Dublin, Ireland Corresponding Author: Martin Caffrey ([email protected]) Abstract Label-free differential scanning fluorimetry (DSF) is a relatively new method for evaluating the stability of proteins. It can be used as a screening tool for downstream applications such as crystallization. The method is attractive in that it requires miniscule quantities of proteins, it can be performed using intrinsic tryptophan and tyrosine fluorescence, and, with the right equipment, it is easy to perform. To date, the method has been used with proteins in liquid solutions and dispersions. It was of interest to determine if DSF could be used with membrane proteins in the lipid cubic phase (LCP), which increasingly is being used for crystallization in support of structure-function studies. The cubic phase is viscous. Furthermore, in coexistence with excess aqueous solution, as happens during crystallization trials, it can become turbid and scatter light. The concern was that these features may render the mesophase unsuitable for DSF analysis. However, using lysozyme and four integral membrane proteins we demonstrate that the method works with all tested proteins in solution and in the LCP. Of note is the observation that some of the test membrane proteins are more stable while others are less so in the mesophase. The method also works in ligand binding measurements. Thus, DSF should prove useful as an analytical tool for identifying host and additive lipids, detergents, precipitants and chemical probes that support the generation of quality crystals by the cubic phase method. Microscale thermophoresis was used to supplement the DSF study and was also shown to work with proteins in the mesophase. Measurements with lysozyme highlight the utility of the cubic mesophase as a model system in which to perform confinement studies.

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 18

Page |2 Membrane proteins function in the cell as enzymes, receptors, transporters as well as playing structural roles. With more than 50 % of drugs on the market targeting membrane proteins, many are pharmacologically important.1,2 The 3D structure of a protein and the complexes it forms, is invaluable in terms of informing how it works and for structure-based drug discovery.3-5 Having such a structure blueprint in atomic detail is particularly advantageous and this is what macromolecular X-ray crystallography (MX) offers. However, MX requires crystals and one of the greatest challenges along the route to a structure is the generation of quality crystals. The challenge is particularly great with membrane proteins that naturally reside in a lipid bilayer.6 Removing the protein from its native membrane, usually with the aid of a surfactant or detergent, brings with it the risk of irreversibly damaging or altering the protein’s structure.7,8 Crystallization is a stochastic process that necessarily involves extensive screening. Screening takes place directly at the crystallization step but also at the various stages that lead to crystallization. Having available fast and efficient methods for high-throughput screening enormously facilitates the overall process of structure determination.9-12 Differential scanning fluorimetry (DSF) is a relatively new method with promise as a screening tool with which to find conditions where the protein is optimally stable and therefore, more likely to crystallise.12-20 The prospect was that DSF might facilitate the structure determination process at the protein characterization and stabilization stages. There are several methods available for crystallizing membrane proteins.21 The most successful by far, the in surfo method, generates crystals directly from a solution of the protein in a surfactant micelle. The alternative bilayer methods use proteins reconstituted in a membrane mimic for crystal growth. These include the bicelle, vesicle and lipid cubic phase (LCP) methods.21,22 The latter, also referred to as the in meso method, and the focus of this study, was introduced in 1996.23 Currently, the Protein Data Bank contains over 500 membrane protein records attributed to the method. Like any crystallization technique, the in meso method requires extensive screening. The goal of the current study was to determine if label-free DSF, which has been shown to work with membrane proteins in solution,24 could be used as a screening tool in support of in meso crystallization. Obviously, some of the stability screening ahead of crystallogenesis trials can be done with the protein in solution as a detergent micelle. However, some screening must be carried out with the protein in the sticky and viscous LCP itself. This presents two challenges. First, the mesophase must be transferred into a narrow bore glass or quartz capillary for analysis. Second, DSF involves heating the sample usually from 20 °C to 95 °C. While the mesophase prepared with the commonly used monoacylglycerol (see SI-Footnote 1), monoolein (9.9 MAG), remains in the cubic phase over much of this temperature range, the mesophase looses water at higher temperatures (Figure S-1).25 As a result, what starts out as a transparent cubic mesophase that is well suited to optical interrogation becomes cloudy upon heating due to shed water and the attendant scattering of light. The fear was that this dramatic change in light scattering properties of the sample during the course of a thermal ramp would interfere with the DSF measurement and render the data uninterpretable. Despite the challenges, there were grounds to be optimistic. To facilitate in meso crystallization, tools had been developed for conveniently making and handling the viscous and sticky mesophase.26 The sense was therefore that the DSF sample capillary, despite being of narrow bore, could be filled uniformly for analysis. It was less clear that the light scattering issue could be dealt with. However, considering how DSF works it became apparent that it may well cope with scattering provided the


Page 3 of 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Page |3 turbidity that developed was uniformly distributed throughout the sample and that it was not excessive to the point of triggering bulk phase separation. A more complete description of these points is provided in the Supporting Information (SI-Introduction). Spectroscopic and indeed fluorescence measurements on proteins in the cubic phase have been performed successfully in the past.14,27-33 These have been used to monitor protein stability and function in a membrane mimetic.14 However, they require large amounts of mesophase and valuable protein and do not lend themselves to high-throughput screening. Here we show that label-free DSF measurements can be made with tryptophan containing membrane proteins reconstituted in the lipid cubic mesophase. Four membrane proteins, three α-helical and one β-barrel protein, were included in the study. The method can be used to screen for conditions that optimize protein stability and ligand binding in solution and in the viscous and sticky cubic phase. Complementary measurements show that the related method of microscale thermophoresis (MST) can likewise be used with soluble and membrane proteins in the LCP. Lysozyme, proved invaluable as a soluble protein for exploratory work in this study. In so doing, the utility of the cubic phase as a model system for investigating the behaviour of soluble proteins in particular under conditions of confinement became apparent.

Methods Mesophase Preparation and Protein Reconstitution Hen egg white lysozyme was used as the reference soluble protein. For most applications a solution of the protein at 20 mg/mL in Milli-Q water was employed. Membrane proteins used included the alginate transporter (AlgE),34 the undecaprenyl-pyrophosphate phosphatase (BacA),35 the apolipoprotein Nacyltransferase (Lnt)36 and the lipoprotein signal peptidase II (LspA).37 The production and purification of these proteins have been detailed in the cited literature. Additional details are available in the Supporting Information (SI-Reagents, SI-Methods). Lysozyme was incorporated into the LCP following a well-established protocol.38 Briefly, this involved homogenizing two volumes of protein solution at 50 mg/mL in water with three volumes of 9.9 MAG.39 A similar protocol was used to reconstitute membrane proteins into the LCP. The final concentrations of membrane proteins in the mesophase was as follows: Lnt at 1 mg/mL, BacA at 0.8 mg/mL, AlgE at 0.5 mg/mL and LspA at 0.4 mg/mL. Mixed lipid systems were formulated, as described previously.26 Complexes with ligands were prepared by incubating the protein and ligand in a detergent micelle solution for 1 h at 4 °C37 ahead of use directly in DSF measurements or for use in mesophase preparation for subsequent DSF analysis. Differential Scanning Fluorimetry and Microscale Thermophoresis Liquid samples were loaded into DSF capillaries by capillary action. Mesophase samples were transferred into capillaries by positive displacement from one of the syringes of the coupled syringemixing device used to make the mesophase. For this purpose, homogenous mesophase was transferred to one of the syringes in the mixer, the empty syringe and coupler were disconnected from the loaded syringe which was then fitted with a 10 mm long 26-gauge Hamilton needle held in place by a knurled retaining nut. The needle was inserted into one end of the capillary and mesophase was transferred



Page 4 of 18

Page |4 from the syringe by pressing on the plunger while the capillary was held firmly against the retaining nut to prevent the mesophase from leaking. Approximately 8.5 µL mesophase was transferred into the capillary, the needle was withdrawn and the loaded capillary was positioned on the deck of the DSF instrument for analysis. DSF measurements were performed on a Prometheus NT.48 (NanoTemper Technologies) following manufacturer’s instructions at 1°C/min from 20°C to either 85°C, 90°C or 95°C. Cooling thermograms were recorded at 1°C/min immediately the high temperature limit was reached. Fluorescence measurements were made at an excitation wavelength of 280 nm. Emitted intensities at 330 nm (F330) and 350 nm (F350) were used to calculate the ratio F350/F330 - the DSF signal. Light scattering (LS) measurements were made in parallel. Water loss from samples during DSF measurements was quantified and shown to be insubstantial (Supporting Information and Table S-1). Samples used for MST were prepared as described above for DSF measurements in quartz capillaries (Monolith NT.LabelFree Zero Background MST Premium Coated Capillary) using a Monolith NT.LabelFree (201410-LF-N003) device (NanoTemper Technologies). Additional details are available in the Supporting Information (SI-Methods). Results DSF of Lysozyme in Aqueous Solution The objective of this study was to determine if useful DSF measurements could be made with membrane proteins reconstituted in the LCP as a screening tool primarily for subsequent crystallization trials. Before launching into these potentially more complicated measurements with membrane proteins, preliminary studies were carried out with a reference water-soluble protein. The protein of choice was lysozyme (Figure S-2a) which has been shown separately to be compatible with and to crystallize from the LCP.38 The thermal properties of lysozyme in solution have been studied extensively.40-43 It has been shown to undergo a reversible unfolding or melting transition (Tm) in the 70 to 80°C range.45-48 With six tryptophans and three tyrosines in a total of 147 residues, lysozyme in solution is known to have a Tm at the expected temperature that can be monitored by DSF44 (Figure S-2a). This was confirmed in the current study where a reversible transition with a Tm of ~76.4°C was recorded for the protein in aqueous solution (Figure S-3). The transition seen in the DSF thermogram is sharp and cooperative taking place over a 10°C interval. This recapitulates exactly what has been reported for lysozyme by differential scanning calorimetry (DSC) under similar sample and data collection conditions (Figure S-3h).43 The F330 and F350 thermograms for lysozyme are relatively featureless with fluorescence values dropping during the course of the ramp. A small change in slope is observed in both in the 70 to 80°C interval (Figure S-3b, d). These differ in detail reflecting the denaturation event and give rise to a distinct transition when plotted as the ratio F350/F330, the DSF signal. The light scattering (LS) thermogram is flat and featureless in the 20 – 95°C range in both heating and cooling directions (Figure S-3g). This suggests that the protein does not aggregate to a significant extent upon heating and cooling between 20 and 95°C, as observed previously.42


Page 5 of 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Page |5 DSF of Lysozyme in the LCP To determine if it were possible to make DSF measurements with a protein incorporated into the LCP, lysozyme was used for initial exploratory purposes. The study began by preparing the mesophase with 9.9 MAG at or close to full hydration, which, at 20 °C, occurs at a water content of ~40 %(w/w) (Figure S1 and Figure 1f). At this level of hydration, the mesophase exists as a pure cubic phase (space group, Pn3m). The optically clear sample was transferred to a capillary and was subjected to DSF measurements. The fact that the sample was optically clear meant that it consisted of a single phase with the protein incorporated into its aqueous channels. As observed with lysozyme in solution, the F330 and F350 thermograms for the mesophase samples were relatively featureless with a small change in slope occurring in the 70-80°C interval (Figure 1). This translated to a transition in the DSF thermogram with a Tm of 75.6 ± 0.1°C and a transition width of about 10°C, very similar to what was observed by DSF (and DSC43) for lysozyme in solution (Figure 1f). Furthermore, the transition was shown to be reversible with no evidence of hysteresis (Figure 1e). Additional measurements made with lysozyme in the LCP at hydration levels above and below 40 %(w/w) are described in the Supporting Information (SI-Results).



Page 6 of 18

Page |6

Figure 1. DSF and light scattering heating and cooling thermograms recorded with lysozyme in the LCP prepared with 9.9 MAG. a, DSF fluorescence ratio (F350/F330). b, F330 trace. c, DSF fluorescence ratio (F350/F330) of the heating thermogram to 85 °C followed by a cooling ramp (e). d, F350 trace. e, DSF fluorescence ratio (F350/F330) of the cooling thermogram to 20°C. f, Derivative plot of the data in (a) identifies the inflection point in the curve as a peak at 75.6°C. g, Light scattering profile. Thermograms were recorded using triplicate samples (blue, black, grey traces). h, A simplified version of the 9.9 MAG/water phase diagram in Figure S-1 with the 40 %(w/w) hydration isopleth identified as a dashed line.


Page 7 of 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Page |7 This result demonstrates definitively that DSF works for proteins dispersed in the LCP. Furthermore, the method is robust to dramatic changes in LS that were expected to and that did occur during the heating thermogram. This was apparent in the LS profile where scattering rose dramatically beginning at about 43°C and then levelled off at 70°C (Figure 1g). A slight change in slope was observed at about 86°C before the end of the thermogram at 95°C. Furthermore, after the DSF measurement the sample was opaque to the eye. These observations are expanded on in the Supporting Information (SI-Discussion). Control DSF measurements were made using 9.9 MAG mesophase at 40 %(w/w) water without added protein. As expected, the DSF thermogram is essentially flat and featureless in the 20°C to 95°C range (Figure S-4a). However, the LS profile (Figure S-4d) shows features that mimic those recorded for the lysozyme-laden mesophase and that reflect the protein’s response to temperature along the 40 %(w/w) hydration isopleth (Figure S-1). Thus, changes in scattering at about 30°C and 90°C correspond to the initiation of water shedding and to the cubic-to-HII phase transition, respectively. A final control measurement was made with lysozyme in the cubic phase using MST to determine if the protein was free to move in the mesophase. Mobility is expected and indeed had been shown previously in separate fluorescence recovery after photobleaching (FRAP) experiments.45 MST involves tracking the movement of the protein in the sample in response to a small temperature jump.46 The results (Figure S5) show clearly that lysozyme is free to move in the LCP and that its rate of movement is considerably less than that of the protein in bulk solution. These data provide confidence that the material being used for DSF is behaving as expected and that MST measurements can be performed with soluble proteins in the LCP. DSF Measurements with Membrane Proteins Lipoprotein N-Acyltransferase, Lnt. In meso Lnt is a bacterial membrane enzyme involved in lipoprotein posttranslational processing.47,48 Structures of the protein from Pseudomonas aeruginosa (LntPae) and from Escherichia coli (LntEco) have been solved using crystal grown by the in meso method in the case of both organisms (Figure S-2c) and by the in surfo method for the E. coli ortholog.36,49,50 LntEco and LntPae contain 16 and 21 tryptophans and 17 and 20 tyrosines, respectively. Both were used in the current study for in surfo and in meso DSF measurements. Based on the results obtained with lysozyme dispersed in meso, it came as no surprize that the DSF method also worked with Lnt reconstituted into the LCP. Initial measurements were made with LntEco in mesophase prepared with 9.9 MAG. The DSF thermogram showed a rising inflection with a Tm at 50.5°C (Figure 2a). The F330 and F350 thermograms consisted of monotonically decreasing fluorescence values with changes in slope at 30°C, 50 °C and 85°C (Figure 2b, c). The LS thermogram had a familiar shape with scattering dropping initially in the 20 to 30°C range, then rising from 30 to about 60°C beyond which it levelled off with an obvious change in slope at 85°C (Figure 2d). The transition observed in the DSF thermogram is presumably the denaturation transition undergone by Lnt in the mesophase and it took place over a 15°C range centered at ~50.5°C (Figure 2a). Changes in the F330, F350 and LS thermograms at 30°C and 85°C likely correspond to the water shedding and the cubic-Pn3m-to-HII phase transitions, respectively, as observed above with lysozyme in the mesophase. These agree with the T-C phase diagram for the 9.9 MAG/water system (Figure S-1).



Page 8 of 18

Page |8

Figure 2. DSF and light scattering heating thermograms recorded with LntEco in the LCP prepared with 9.9 MAG. a, DSF (F350/F330) thermogram. b, F330 thermogram. c, F350 thermogram. d, Light scattering thermogram. Thermograms were recorded using duplicate samples (blue, black traces). This result confirms the observation made previously with lysozyme that the LCP is a perfectly suitable system with which to perform DSF measurements with a view to establishing thermal stability and for screening purposes as applied to membrane proteins. Additional DSF measurements made with LntEco and with LntPae in the LCP subjected to a variety of treatments of relevance to crystallization are reported in the Supplemental Information (SI-Results 4). Lipoprotein N-Acyltransferase, Lnt. In Surfo For reference purposes, DSF measurements were performed on LntEco and LntPae in solution as protein-detergent micelles. Two detergents commonly used with membrane proteins were tested that included DDM and LMNG. The effect of additive lipids, PE, PG and EcoPL, was also examined. In the case of LntEco, two inactive mutant constructs, Cys387Ala and Cys387Ser, were included in the study.51 All thermograms were well behaved and showed clearly defined transitions in DSF plots (Figure S-6, Table S-2). The LS thermograms showed sharp rises in scattering that coincided, for the most part, with transitions observed in the corresponding DSF thermograms. These observations held regardless of the ortholog, construct, detergent or additive lipid used. Interestingly, they are in distinct contrast to what was observed with lysozyme where LS did not change in either heating or cooling thermograms. It would appear therefore that membrane proteins in detergent solution undergo aggregation upon thermal denaturation and this accounts for the increase in LS at or close to the Tm observed in the DSF thermograms.


Page 9 of 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Page |9 Lipoprotein Signal Peptidase II, LspA LspA catalyzes the second step in lipoprotein posttranslational processing in bacteria and is a potential target for antibiotic discovery (Figure S-2b).52 It is inhibited by the natural antibiotic globomycin.53 An Xray structure of LspA from P. aeruginosa in complex with globomycin was determined using crystals grown by the in meso method.37 In the current study, we set out to determine if DSF could be used to detect globomycin binding by LspA in solution and in the LCP. The results are shown in Figure 3. In the absence of globomycin, the DSF thermogram for LspA in detergent solution rises monotonically and is relatively featureless (Figure 3a). As soon as globomycin is added, an inflection appears in the DSF thermogram. The inflection point (Tm) rises with increasing globomycin concentration and eventually levels off. Quantitative binding analysis of the data provided a half minimal effective concentration (EC50) of 17 µM at a protein concentration of 12 µM (Figure 3c, Table S-3).

Figure 3. Effect of globomycin on the DSF heating thermograms of LspPae in detergent solution and in the LCP. a, DSF (F350/F330) thermograms in solution. b, DSF (F350/F330) thermograms in meso. c, Dose-response analysis for LspPae in solution. D, Dose-response analysis for LspPae in meso. Globomycin concentrations used and Tm values are included in Tables S-3 and S-4. To determine if globomycin binding to LspA could be measured with the protein reconstituted into the cubic phase DSF measurements were performed in 9.9 MAG, the host lipid system that provided in meso crystals and a structure of the LspA-globomycin complex (Figure S-2b).37 In the absence of globomycin, the DSF thermogram for the apo enzyme showed a barely perceptible inflection at ~39.6°C (Figure 3b). The inflection became more pronounced and the temperature at which it occurred rose with increasing globomycin concentration. The rise levelled off at ~48°C at high micromolar levels of globomycin.



Page 10 of 18

P a g e | 10 Binding analysis provided an EC50 of 45 µM at a protein concentration of 20 µM some two and a half times larger than that observed for LspA in detergent solution (Figure 3d).37 These EC50 values for the protein in solution in a detergent micelle or reconstituted into the bilayer of the cubic mesophase can be contrasted with the in vivo minimum inhibitory concentration (MIC) needed to stop visible E. coli growth of 12.5 µg/mL (20 µM). Interestingly, for P. aeruginosa the reported MIC is >100 µg/mL (>160 µM) which likely reflects this bacteria’s relatively impermeable outer membrane.54 These data show that DSF can be used to quantify binding with integral membrane proteins both in surfo and in meso when the binding event induces a measurable change in the presumed unfolding transition temperature of the protein. Ligand binding during mesophase formation can also been monitored as outlined in the Supporting Information (SI-Results). As was done with lysozyme in solution and in the cubic phase, control MST measurements were performed with LspA both in surfo and in meso. The data (Figure S-7) show that the protein is free to move in detergent micelles in solution and reconstituted into the lipid bilayer of the LCP. Clearly, MST is a viable method for use with membrane proteins in this viscous and sticky bicontinuous medium. Undecaprenyl-pyrophosphate Phosphatase, BacA BacA is a pyrophosphorylase involved in peptidogylcan synthesis.55 In E. coli, the enzyme has 273 residues, three of which are tryptophans and six are tyrosines (Figure S-2d). The structure of BacA has been determined using crystals grown by the in meso method.35 In DDM detergent solution, the protein has a DSF thermogram with a prominent transition at 79.3°C which is matched by a transition in the LS thermogram at the same temperature (Figure S-8a, b). Thus, aggregation as reflected in the rising LS signal correlates with the changing fluorescence that accompanies unfolding as observed in the DSF profile. The DSF thermogram also shows a change in slope at ~60°C. However, no corresponding change is seen in the LS ramp at this temperature. When the DSF measurements were repeated with BacA reconstituted into the LCP formed by 9.9 MAG, the lipid used for crystallogenesis and a crystal structure, the Tm observed was 81°C (Table S-5). The DSF thermogram had a change in slope at about 60 °C (data not shown), as was noted for the in surfo sample. However, neither the low nor the high temperature transition were visible in the LS thermogram. In this case, any aggregation that may have taken place upon unfolding was presumably masked by the LS change associated with the heating-induced shedding of water by the mesophase along this isopleth. Doping BacA with different phospholipids and other additives, as part of a screening exercise, resulted in shifts in the main Tm (Table S-5). These data suggest that DSF and LS may prove useful as screening tools to identify additive lipids and heavy atom reagents for use in de novo phasing. These issues are discussed further in the Supporting Information (SI-Results). Alginate Transporter, AlgE The alginate transporter, AlgE, in P. aeruginosa is a 490 residue β-barrel protein that resides in the outer membrane (Figure S-2e).34,56 It has 15 tryptophans and 15 tyrosines and its structure was solved using crystals grown by the in meso method. AlgE was included to provide contrast with the other, predominantly α-helical proteins, in this study. β-Barrel proteins are known for their stability. This was borne out in the DSF measurements performed with AlgE solubilized in the detergent LDAO where a Tm of 88.3°C was recorded (Figure 4a). As with Lnt, the transition was accompanied by an abrupt change in LS at the Tm suggesting that the protein in surfo aggregates upon thermal unfolding (Figure 4b).


Page 11 of 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


P a g e | 11 The DSF instrument in use in the current study has a ramp temperature range from 20°C to 95°C. We reasoned that if reconstituting the protein into the bilayer of the LCP were to stabilize the protein, a transition may no longer be visible given the native stability of the protein. Indeed, when this experiment was performed with AlgE reconstituted into the LCP no transition in the DSF thermogram from 20 to 95°C was observed (Figure 4c). Thus, either the protein had been stabilized to a temperature above 95°C or it had been destabilized to below 20°C. The latter is unlikely since a structure of AlgE was obtained using crystals grown in the LCP at 20°C,57 suggesting that the mesophase bilayer has a thermally stabilizing effect on the protein. We investigated this idea further based on the following reasoning. Sodium dodecyl sulfate (SDS) is a potent denaturing surfactant that can induce protein unfolding and destabilization. The thought was that SDS would destabilize AlgE and shift its denaturation transition to lower temperatures. Accordingly, the protein was incubated with SDS at 20°C for 30 min before running a DSF measurement (Figure 4d). The resulting thermograms show clearly that SDS had a profound destabilizing effect on AlgE reducing its Tm by 24°C to 64.1°C. Interestingly, the LS thermogram was flat suggesting that the protein did not aggregate significantly upon thermal unfolding in SDS and that the detergent likely contributes to keeping the denatured protein in solution above the Tm (Figure 4e). This is in contrast to the behaviour of the protein in LDAO where the protein aggregated upon denaturation (Figure 4b).

Figure 4. DSF and light scattering heating thermograms recorded with AlgE in LDAO detergent solution and in the LCP. Data are shown for AlgE that had (d, e, f) or had not been pre-treated with 1 %(w/v) SDS detergent (a, b, c). a, DSF (F350/F330) thermogram for AlgE in LDAO. b, Light scattering thermogram for AlgE in LDAO. c, DSF (F350/F330) thermogram for AlgE reconstituted into the LCP. d, DSF (F350/F330) thermogram for AlgE in SDS. e, Light scattering thermogram for AlgE in SDS. f, DSF (F350/F330) thermogram for AlgE in SDS reconstituted into the LCP. The black and blue traces correspond to duplicate samples.



Page 12 of 18

P a g e | 12 When the SDS-treated sample that had been incubated at 20°C was reconstituted into the 9.9 MAG mesophase and the DSF measurement was repeated, no transition was apparent in the 20-95°C range (Figure 4f). This suggested that the SDS in the sample, that presumably had destabilized the protein lowering its Tm to 64°C, had become diluted out into the bilayer of the mesophase and, as a result, had been drawn away from AlgE to restore its thermal stability. The possible beginnings of a transition are apparent in the DSF thermogram in the 92-95°C range. However, this could not be tracked to completion because the DSF instrument in use was not designed to scan beyond 95°C. If indeed it was the start of a transition, it means that the process of reconstitution into the lipid cubic phase can rescue proteins destabilized in harsh detergent environments. A similar type of rescue has been observed with the α-helical kinase, DgkA, unfolded and stripped of all lipid and detergent, upon reconstitution and refolding in the cubic mesophase.58

Discussion DSF of Membrane Proteins in the Lipid Cubic Phase The primary goal of this study was to determine if DSF could be used as a method for determining the stability of membrane proteins in the LCP as a screening tool for in meso crystallization. Four integral membrane proteins were tested and all provided results showing that DSF does work. Indeed, a transition can be seen in all cases despite the dramatic change in light scattering associated with the temperature-induced change in hydration level and mesophase transition. Thus, scattering that affects the entire emission spectrum effectively gets cancelled out, as expected, when the ratio DSF signal is calculated and plotted as a function of temperature. The DSF instrument also reports on LS in the sample and this, along with protein-free blanks, proved invaluable in interpreting the DSF profiles. In addition to the concern about light scattering, the possibility existed that during a ramp in temperature and the attendant water-shedding, the excess water would phase separate in bulk. If this were to happen, that part of the sample giving rise to the DSF signal may now take the form of bulk water devoid of protein-laden mesophase. This would render the DSF measurement useless. Fortunately, for all of the systems examined in this study that included 22 combinations of 4 membrane proteins and 19 MAG phase types and treatments, no such bulk phase separation and loss of DSF signal was observed. This means that the mesophase sheds water uniformly to form small well-dispersed particles that scatter light. Transition reversibility was investigated in heating followed by cooling ramps, and again, no issue with bulk phase separation was noted. These observations attest to the robust nature of the cubic mesophase system for DSF investigations. The DSF method was tested with four integral membrane proteins that included α-helical and β-barrel proteins. Within the α-helical group, BacA is predominantly membrane integral while Lnt and LspA both have intra- and extra-membrane domains as well as a mix of α-helices and β-sheet secondary structures (Figure S-2). In the case of LspA, it was examined with and without a bound inhibiting antibiotic, globomycin.37 The method worked with all four membrane proteins and in all cases, transitions were observed that made sense. The transition in the β-barrel protein, AlgE, approached the upper temperature limit of the DSF instrument and so was not extensively investigated. For the others, the effects of host lipids and other additives on Tm were quantified with ease. With LspA, it was possible to


Page 13 of 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


P a g e | 13 measure globomycin binding. These data show that the DSF method has generality when it comes to membrane protein and application types. DSF of Membrane Proteins In Surfo In parallel with the DSF study performed in the cubic mesophase, measurements were made with proteins solubilized in detergent micelles. The same four proteins were used and all behaved without issue in DSF measurements. Well-defined transitions were clearly visible that were shown to be sensitive to detergent type and to additives such as lipid, antibiotic, divalent cations and a heavy atom derivatizing reagent. Indeed, the shift in Tm was used to quantify binding between LspA and globomycin, as was done for the enzyme in meso. Membrane Protein Stability In Meso vs. In Surfo The expectation at the outset of this study was that membrane proteins would be more thermally stable reconstituted into the bilayer of the cubic phase than solubilized in a detergent micelle. The hypothesis was that a lipid bilayer, as opposed to a surfactant micelle in true solution, is more native membranelike.14 It is expected therefore to provide a more stabilizing environment that would show up as a higher Tm in a DSF thermogram. However, the results of this study involving four different membrane proteins provided mixed results. With AlgE, it seems likely that the mesophase stabilized the protein. With BacA, a similar stabilizing effect was recorded but its magnitude was small. In the case of Lnt and the LspAglobomycin complex, the micellarized form of the protein was the more stable. Whether a Tm value, as a measure of stability, determined on the basis of DSF translates from one system of reconstitution to another must remain in the realm of speculation. This is expanded on separately in the Supporting Information (SI-Discussion). The Cubic Mesophase as a Model System for Confinement Studies The fact that the DSF thermograms for lysozyme in solution and in the cubic phase under full hydration conditions are superimposable suggests that the protein is experiencing very similar chemical and physical environments in the two sample types, which represent extreme states of confinement. The similarity in the two environments is perhaps not surprising given that lysozyme produces identical crystals when grown in solution and in the cubic phase.38 This highlights the attractive features of the cubic phase created from MAGs as a model system in which to perform confinement studies.59,60 Undoubtedly, the minimal, if any, impact the mesophase has on a dissolved macromolecule like lysozyme reflects the relative size of the aqueous compartments the mesophase creates and the uncharged yet polar and water-like hydroxy group coating on the surface of the mesophase membrane.61 The cubic phase therefore has many features that make it an attractive model for investigating the properties of macromolecules in confinement. Intrinsically disordered proteins like the human amyloid α-synuclein, implicated in Parkinson’s disease, should prove particularly interesting in this regard given that their state of folding and aggregation propensity may change depending on the chemical properties and the crowded nature of their immediate environment.62-64 DSF and Protein Thermal Stability It is worth considering the individualistic nature of a DSF thermogram and its utility for comparative purposes. Thus, the argument can be made that a particular DSF thermogram simply reflects the number and location of tryptophans (see SI-Footnote 3) in the target protein, its structure and flexibility



Page 14 of 18

P a g e | 14 and the disparate environments to which the different tryptophans are exposed to during the course of a DSF measurement. For this reason, comparing the thermal stability of one protein type with that of another based on a DSF-defined Tm may not be useful. However, what can be useful - with caution - is the behaviour of a single protein type to different treatments to include changes in pH and buffer, salt and additive (precipitant, detergent, lipid, sulfhydryl reagent, chaotrope, etc.) type and concentration.14,15 In this way, it may prove useful to select conditions that shift a DSF transition to higher temperatures as those more suited for crystallization in a screening campaign. The validity of the approach would, of course, be strengthened if it were shown that the DSF transition under investigation corresponded to a change in the stability, solubility, structure, flexibility and/or activity of the protein. Agreement between any one or more of these indicators and the DSF transition, as observed for lysozyme in this study, would mean that DSF could be used with confidence as an alternative screening method that is simple, fast, efficient and informative. The role played by the host medium in thermal stability must also be taken into consideration, as commented on in the Supporting Information (SIDiscussion) Unfolding. Computational Insights Much progress has been made in understanding the structural, chemical and thermodynamic bases of protein folding and conformational stability.65 With tools such as DSF now available where unfolding transitions can be monitored simply, quickly and with miniscule amounts of label-free protein perhaps it is time to pay more attention to the unfolding process, and to aggregation that often follows, with a view to understanding and controlling both. Aggregates can form from native-like as well as from fully unfolded forms of a protein. Thus, any degree of unfolding may favour the aggregation process depending on chemical and environmental conditions. In some situations, the goal is to prevent or to slow these processes. In others, it is to accelerate and to favour both. Mechanistic understandings of these processes would have practical impact in areas ranging from protein production in a biochemistry laboratory, to biologics production in the pharmaceutical industry, misfolding in disease and clinical settings all the way to foods and feeds where shelf-life and organoleptic properties depend on the state of folding.66-68 Experimental studies aimed at establishing the principles of unfolding and denaturation would be enormously complemented by computational modelling which has had spectacular recent successes in understanding protein folding and predicting protein 3D structure.69 Applying these approaches to membrane proteins is not going to be easy. But with tools like DSF at hand, fundamental data can be generated efficiently for use in delivering on this grand challenge.

Acknowledgements We thank past and present members of the Membrane Structural and Functional Biology group for their assorted contributions and Drs. McDonnell and Hatty, NanoTemper Technologies Gmbh, for technical assistance and access to equipment. The work was funded by Science Foundation Ireland (12/IA/1255 and 16/IA/4435), the Cystic Fibrosis Foundation (CAFFRE16XX0), the German Research Foundation (WE 6084/1-1) and a Marie Skłodowska-Curie Actions Individual Fellowship (704848). Supporting Information Introduction, Footnotes, Additional reagents and methods, Results, Discussion, Supporting figures and tables


Page 15 of 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


P a g e | 15 References (1) Santos, R.; Ursu, O.; Gaulton, A.; Bento, A. P.; Donadi, R. S.; Bologa, C. G.; Karlsson, A.; Al-Lazikani, B.; Hersey, A.; Oprea, T. I.; Overington, J. P. Nat. Rev. Drug Discov. 2016, 16, 19-34. (2) Overington, J. P.; Al-Lazikani, B.; Hopkins, A. L. Nat. Rev. Drug Discov. 2006, 5, 993-996. (3) Holdgate, G. A.; Meek, T. D.; Grimley, R. L. Nat. Rev. Drug Discov. 2017, 17, 115-132. (4) Cooke, R. M.; Brown, A. J. H.; Marshall, F. H.; Mason, J. S. Drug Discov. Today 2015, 20, 1355-1364. (5) Manglik, A.; Lin, H.; Aryal, D. K.; McCorvy, J. D.; Dengler, D.; Corder, G.; Levit, A.; Kling, R. C.; Bernat, V.; Hübner, H.; Huang, X.-P.; Sassano, M. F.; Giguère, P. M.; Löber, S.; Da, D.; Scherrer, G.; Kobilka, B. K.; Gmeiner, P.; Roth, B. L.; Shoichet, B. K. Nature 2016, 537, 185-190. (6) Carpenter, E. P.; Beis, K.; Cameron, A. D.; Iwata, S. Curr. Opin. Struct. Biol. 2008, 18, 581-586. (7) Garavito, R. M.; Ferguson-Miller, S. J. Biol. Chem. 2001, 276, 32403-32406. (8) Privé, G. G. Methods 2007, 41, 388-397. (9) Sonoda, Y.; Cameron, A.; Newstead, S.; Omote, H.; Moriyama, Y.; Kasahara, M.; Iwata, S.; Drew, D. FEBS Lett. 2010, 584, 2539-2547. (10) Hardy, D.; Desuzinges Mandon, E.; Rothnie, A.; Jawhari, A. Methods 2018, doi: 10.1016/j ymeth.2018.02.017. (11) Senisterra, G.; Chau, I.; Vedadi, M. Assay Drug Dev. Technol. 2011, 10, 128-136. (12) Vedadi, M.; Arrowsmith, C. H.; Allali-Hassani, A.; Senisterra, G.; Wasney, G. A. J. Struct. Biol. 2010, 172, 107-119. (13) Niesen, F. H.; Berglund, H.; Vedadi, M. Nat. Protoc. 2007, 2, 2212-2221. (14) Liu, W.; Hanson, M. A.; Stevens, R. C.; Cherezov, V. Biophys. J. 2010, 98, 1539-1548. (15) Alexandrov, A. I.; Mileni, M.; Chien, E. Y. T.; Hanson, M. A.; Stevens, R. C. Structure 2008, 16, 351359. (16) Seabrook, S. A.; Newman, J. ACS Comb. Sci. 2013, 15, 387-392. (17) Ericsson, U. B.; Hallberg, B. M.; DeTitta, G. T.; Dekker, N.; Nordlund, P. Anal. Biochem. 2006, 357, 289-298. (18) Vedadi, M.; Niesen, F. H.; Allali-Hassani, A.; Fedorov, O. Y.; Finerty, P. J.; Wasney, G. A.; Yeung, R.; Arrowsmith, C.; Ball, L. J.; Berglund, H.; Hui, R.; Marsden, B. D.; Nordlund, P.; Sundstrom, M.; Weigelt, J.; Edwards, A. M. Proc. Natl. Acad. Sci. USA 2006, 103, 15835-1580. (19) Zucker, F. H.; Stewart, C.; Rosa, J. d.; Kim, J.; Zhang, L.; Xiao, L.; Ross, J.; Napuli, A. J.; Mueller, N.; Castaneda, L. J.; Nakazawa Hewitt, S. R.; Arakaki, T. L.; Larson, E. T.; Subramanian, E.; Verlinde, C. L. M. J.; Fan, E.; Buckner, F. S.; Van Voorhis, W. C.; Merritt, E. A.; Hol, W. G. J. J. Struct. Biol. 2010, 171, 64-73. (20) Kohlstaedt, M.; von der Hocht, I.; Hilbers, F.; Thielmann, Y.; Michel, H. Acta Crystallogr. D Biol. Crystallogr. 2015, 71, 1112-1122. (21) Moraes, I.; Evans, G.; Sanchez-Weatherby, J.; Newstead, S.; Stewart, P. D. S. Biochim. Biophys. Acta Biomembranes 2014, 1838, 78-87. (22) Caffrey, M. J. Struct. Biol. 2003, 142, 108-132. (23) Landau, E. M.; Rosenbusch, J. P. Proc. Natl. Acad. Sci. USA 1996, 93, 14532-14535. (24) Veith, K.; Martinez Molledo, M.; Almeida Hernandez, Y.; Josts, I.; Nitsche, J.; Löw, C.; Tidow, H. ChemBioChem 2017, 18, 1735-1742. (25) Qiu, H.; Caffrey, M. Biomaterials 2000, 21, 223-234. (26) Caffrey, M.; Cherezov, V. Nat. Protoc. 2009, 4, 706-713. (27) Hochkoeppler, A.; Landau Ehud, M.; Venturoli, G.; Zannoni, D.; Feick, R.; Luisi Pier, L. Biotechnol. Bioeng. 2004, 46, 93-98. (28) Portmann, M.; Landau, E. M.; Luisi, P. L. J. Phys. Chem. 1991, 95, 8437-8440. (29) Cherezov, V.; Yamashita, E.; Liu, W.; Zhalnina, M.; Cramer, W. A.; Caffrey, M. J. Mol. Biol. 2006, 364, 716-734.



Page 16 of 18

P a g e | 16 (30) Cherezov, V.; Clogston, J.; Papiz, M. Z.; Caffrey, M. J. Mol. Biol. 2006, 357, 1605-1618. (31) Fenalti, G.; Abola, E. E.; Wang, C.; Wu, B.; Cherezov, V. In Methods in Enzymology, Shukla, A. K., Ed.; Academic Press, 2015, 417-437. (32) Caffrey, M. Curr. Opin. Struct. Biol. 2000, 10, 486-497. (33) Liu, W.; Caffrey, M. J. Struct. Biol. 2005, 150, 23-40. (34) Tan, J.; Rouse, S. L.; Li, D.; Pye, V. E.; Vogeley, L.; Brinth, A. R.; El Arnaout, T.; Whitney, J. C.; Howell, P. L.; Sansom, M. S. P.; Caffrey, M. Acta Crystallogr. D Biol. Crystallogr. 2014, 70, 2054-2068. (35) El Ghachi, M.; Howe, N.; Huang, C.-Y.; Olieric, V.; Warshamanage, R.; Touzé, T.; Weichert, D.; Stansfeld, P. J.; Wang, M.; Kerff, F.; Caffrey, M. Nat. Commun. 2018, 9, 1078. (36) Wiktor, M.; Weichert, D.; Howe, N.; Huang, C.-Y.; Olieric, V.; Boland, C.; Bailey, J.; Vogeley, L.; Stansfeld, P. J.; Buddelmeijer, N.; Wang, M.; Caffrey, M. Nat. Commun. 2017, 8, 15952. (37) Vogeley, L.; El Arnaout, T.; Bailey, J.; Stansfeld, P. J.; Boland, C.; Caffrey, M. Science 2016, 351, 876880. (38) Aherne, M.; Lyons, J. A.; Caffrey, M. J. Appl. Crystallogr. 2012, 45, 1330-1333. (39) Cheng, A.; Hummel, B.; Qiu, H.; Caffrey, M. Chem. Phys. Lipids 1998, 95, 11-21. (40) Ueda, T.; Masumoto, K.; Ishibashi, R.; So, T.; Imoto, T. Protein Eng. Des. Sel. 2000, 13, 193-196. (41) Bye Jordan, W.; Falconer, R. J. Protein Sci. 2013, 22, 1563-1570. (42) Salvetti, G.; Tombari, E.; Mikheeva, L.; Johari, G. P. J. Phys. Chem. B 2002, 106, 6081-6087. (43) Blumlein, A.; McManus, J. J. Biochim. Biophys. Acta Proteins Proteom. 2013, 1834, 2064-2070. (44) Hofmann, L.; Gulati, S.; Sears, A.; Stewart, P. L.; Palczewski, K. Anal. Biochem. 2016, 499, 63-65. (45) Cherezov, V.; Liu, J.; Griffith, M.; Hanson, M. A.; Stevens, R. C. Cryst. Growth Des. 2008, 8, 43074315. (46) Jerabek-Willemsen, M.; André, T.; Wanner, R.; Roth, H. M.; Duhr, S.; Baaske, P.; Breitsprecher, D. J. Mol. Struct. 2014, 1077, 101-113. (47) Gupta, S. D.; Wu, H. C. FEMS Microbiol. Lett. 1991, 78, 37-41. (48) Hillmann, F.; Argentini, M.; Buddelmeijer, N. J. Biol. Chem. 2011, 286, 27936-27946. (49) Noland, C. L.; Kattke, M. D.; Diao, J.; Gloor, S. L.; Pantua, H.; Reichelt, M.; Katakam, A. K.; Yan, D.; Kang, J.; Zilberleyb, I.; Xu, M.; Kapadia, S. B.; Murray, J. M. Proc. Natl. Acad. Sci. USA 2017, 114, E6044E6053. (50) Lu, G.; Xu, Y.; Zhang, K.; Xiong, Y.; Li, H.; Cui, L.; Wang, X.; Lou, J.; Zhai, Y.; Sun, F.; Zhang, X. C. Nat. Commun. 2017, 8, 15948. (51) Buddelmeijer, N.; Young, R. Biochemistry 2010, 49, 341-346. (52) Buddelmeijer, N. FEMS Microbiol. Rev. 2015, 39, 246-261. (53) Dev, I. K.; Harvey, R. J.; Ray, P. H. J. Biol. Chem. 1985, 260, 5891-5894. (54) Kiho, T.; Nakayama, M.; Yasuda, K.; Miyakoshi, S.; Inukai, M.; Kogen, H. Bioorg. Med. Chem. 2004, 12, 337-361. (55) Manat, G.; El Ghachi, M.; Auger, R.; Baouche, K.; Olatunji, S.; Kerff, F.; Touzé, T.; Mengin-Lecreulx, D.; Bouhss, A. PLoS One 2015, 10, e0142870. (56) Rehm, B. H.; Boheim, G.; Tommassen, J.; Winkler, U. K. J. Bacteriol. 1994, 176, 5639-5647. (57) Ma, P.; Weichert, D.; Aleksandrov, L. A.; Jensen, T. J.; Riordan, J. R.; Liu, X.; Kobilka, B. K.; Caffrey, M. Nat. Protoc. 2017, 12, 1745-1762. (58) Li, D.; Caffrey, M. Sci. Rep. 2014, 4, 5806. (59) Zhou, H.-X.; Rivas, G.; Minton, A. P. Annu. Rev. Biophys. 2008, 37, 375-397. (60) Zhou, H.-X.; Dill, K. A. Biochemistry 2001, 40, 11289-11293. (61) Kulkarni, C. V.; Wachter, W.; Iglesias-Salto, G.; Engelskirchen, S.; Ahualli, S. Phys. Chem. Chem. Phys. 2011, 13, 3004-3021. (62) McNulty, B. C.; Young, G. B.; Pielak, G. J. J. Mol. Biol. 2006, 355, 893-897. (63) Munishkina, L. A.; Cooper, E. M.; Uversky, V. N.; Fink, A. L. J. Mol. Recognit. 2004, 17, 456-464.


Page 17 of 18 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


P a g e | 17 (64) Yamin, G.; Munishkina, L. A.; Karymov, M. A.; Lyubchenko, Y. L.; Uversky, V. N.; Fink, A. L. Biochemistry 2005, 44, 9096-9107. (65) Roman, A. E.; González Flecha, F. L. Biomolecules 2014, 4. (66) Lukacs, G. L.; Verkman, A. S. Trends Mol. Med. 2012, 18, 81-91. (67) Valastyan, J. S.; Lindquist, S. Dis. Models Mech. 2014, 7, 9-14. (68) Carpenter, J. F.; Manning, M. C.; Randolph, T. W. Curr. Protoc. Protein Sci. 2002, 27, 4.6.1-4.6.6. (69) Huang, P.-S.; Boyken, S. E.; Baker, D. Nature 2016, 537, 320-327.



Page 18 of 18

P a g e | 18

For Table of Contents Only


Membrane (and Soluble) Protein Stability and Binding Measurements

Recommend Documents