Information

How do detergents interfere with protein assays?


This has been getting me stuck. I've tried to understand what a detergent would do in an assay, but I can't figure out whether it would affect the protein or the reagent (say, in a Bradford assay).


Depending on the detergent, its concentration, and the exact assay being performed, it can affect both the protein and the assay reagent(s). Some detergents will bind the (usually colorimetric) reagent, or otherwise chemically react with it, giving high background to your assay and sometimes completely masking the specific signal of the assay itself. Many modern protein assays are tolerant of a fairly low level of various detergents, perhaps some but not others, and this underscores the necessity of running a buffer/diluent-only control to verify the background of your measurements. Other assays, like Bio-Rad's DC Protein Assay or the Pierce Detergent Compatible Bradford Assay Kit from ThermoFisher, are engineered to tolerate a certain (often fairly high) level of detergent without increasing the background of the assay.


6 Common Lab Uses of Detergents

Detergents are all around us in the lab – and that’s a good thing! Thanks to their chemical structure, detergents can solubilize and interact with many types of molecules, making them vital to research. To show you why detergents are such a good thing for scientists, we’ll go through six examples in molecular biology where detergents are essential!


Protein Quantitation Assays — Frequently Asked Questions

In any protein assay, the best protein to use as a standard is a purified preparation of the protein being assayed. In the absence of such an absolute reference protein, another protein must be selected as a relative standard. The best relative standard to use is the one that gives a color yield similar to that of the protein being assayed. Selecting such a protein standard is generally done empirically. Alternatively, if only relative protein values are desired, any purified protein may be selected as a standard. If a direct comparison of two different protein assays is being performed, the same standard should be used for both procedures. Bio-Rad offers two standards: bovine γ-globulin (standard I) and BSA (standard II). With the Quick Start Bradford and Bio-Rad protein assays, the dye color development is significantly greater with albumin compared to most other proteins, including γ-globulin (Figure 1). Therefore, the BSA standard would be appropriate if the sample contains primarily albumin, or if the protein being assayed gives a similar response to the dye. For a color response that is typical of many proteins, the γ-globulin standard is appropriate. The DC and RC DC protein assays show little difference in color development between BSA and γ-globulin. It is recommended, however, that the same standard be used if comparisons are to be made between different assays.

Fig. 1. Typical standard curves for the Bio-Rad protein assay: bovine serum albumin () and γ-globulin ( ).


BCA Protein Assays

    Also known as the Smith Assay, is a highly sensitive colorimetric assay that is compatible with detergent solubilized protein solutions. Available in two sizes with either BSA Protein Assay Standards or a unique Non-Animal Protein Assay Standard.
  • Micro BCA Protein Assay: A highly sensitive colorimetric assay that is compatible with detergent solubilized protein solutions. This modification of the BCA Protein Assay is suitable for dilute protein samples (0.5-20µg/ml).
  • Reducing Agent Compatible BCA Protein Assay: Based on our BCA Protein Assay and supplied with the Reducing Agent Compatibility Agent (RACA) that modifies reducing agents to limit their effect on the reduction of the assay&rsquos copper ions, preventing inhibition of the assay. The use of RACA allows for samples containing up to 5mM DTT, 10mM TCEP or 35mM ß-mercaptoethanol.

Materials and methods

The brominated detergents used in the present study, octaethylene-glycol-5,6-dibromododecylether (5,6-Br2C12E8, MW = 696 g.mol -1 ) and dibromo-dodecyl-β-D-maltoside (5,6-Br2DDM, MW = 668 g.mol -1 ) were synthesized as previously described [9,16], including in 14 C-labelled form. Non-brominated C12E8 (MW = 538 g.mol -1 ) was obtained from Nikko and non-brominated DDM (MW = 510 g.mol -1 ) was from Anatrace Inc.. In contrast to non-brominated compound, dry 5,6-Br2C12E8 had an oily and yellowish appearance (presumably due to the bromine atoms themselves, since 1 H NMR spectra confirmed the absence of detectable impurities), but this did not prevent complete solubilization at concentrations as high as 100 mg/ml (like the non-brominated C12E8). Detergent stock solutions were subsequently prepared in deionised water at concentrations of 100 mM (for instance, 55 mg/ml and 70 mg/mL for the unbrominated and brominated 5,6-Br2C12E8, respectively), together with more dilute solutions (e.g. 5.5 mM for C12E8, 10 mM for DDM), and more concentrated ones if desired (e.g. 100 mg/mL C12E8, i.e. 186 mM, or 200 mg/ml DDM i.e. 390 mM). These nominal values (±10%), based on weight measurements, were in agreement with independent estimates based on measurement of the polyethyleneglycol contents of the detergent via its reaction with ammonium cobalto-thiocyanate and extraction into ethylene dichloride [17–19].

The biological membranes used here were sarcoplasmic reticulum (SR) vesicles extracted from rabbit fast twitch muscle (and containing about 0.5 g lipids/g membrane proteins), as in [8] and [20]. SR vesicles were prepared during the period from 2006 to 2009 and stored at -80°C until being used, under conditions that do not alter their biological properties, i.e. in the presence of 0.3 M sucrose. Three different SR preps were used for the present experiments, which were performed over several years, with no loss of ATPase activity observed on this period. Membrane preparation from the rabbit was carried out in strict accordance with the recommendations and after agreement from the Ethic committee of the “Commissariat à l’Énergie Atomique et aux Énergies Alternatives” (CEA agreement #E 91 272 106 see S1 NC3Rs ARRIVE Guidelines Checklist). All surgery was performed after killing rabbit by bleeding after a blow to the neck with a metal bar, as quickly as possible to minimize suffering (for a detailed procedure of subsequent steps, see [21]). The concentration of SR membranes in each experiment is expressed in terms of their protein contents, i.e. in μg of protein/mL. In all experiments the buffer contained 100 mM KCl, 1 mM MgCl2 and 50 mM Tes-Tris at pH 7.5 and 20°C (designated as “pH 7.5 buffer”), together with 0.05 mM free Ca 2+ (0.1 mM total Ca 2+ and 0.05 mM EGTA) to optimize preservation of SERCA1a, the main protein of the SR membranes.

For each detergent, its critical micellar concentration (cmc) at pH 7.5 was estimated using 40 μM methyl orange, as previously described [22,23]. Light scattering (at 290 nm) and Trp fluorescence (λex and λem at 290 nm and 340 nm, respectively) of the membranes, in the absence or presence of detergent, were measured as previously described (see e.g. [15]), and sometimes simultaneously, using a Spex Fluorolog equipped with two independent monochromators in the “T” configuration. In some cases also, light scattering and Trp fluorescence changes were measured during slow continuous dilution inside the cuvette of the contents of a detergent-containing mechanically-driven syringe. In those cases, detergent delivery from the syringe (containing a 5.5 mM C12E8 or a 10 mM DDM solution) into the 2 ml spectrophotometer cuvette was performed at 400 μl/hour, resulting within half an hour in a final addition of 0.5 mM or 0.9 mM detergent together with an up to 10% dilution of the membranes. Note that this detergent delivery procedure not only makes it possible to collect a large number of data points, but also minimizes the artefactual transient solubilization of membranes which might occur when a droplet of concentrated detergent is added from an ordinary pipet [24].

Stopped-flow experiments were performed using a Biologic SFM 3 equipment (see e.g. [25]), but here with mixing in different volume-to-volume ratios of the contents of the two syringes. The nominal dead-time of the machine is about 3 ms. The excitation wavelength was 290 nm, and fluorescence emitted at 340 nm was detected using a combination of filters (MTO J324 + A340).

Rapid filtration measurements were performed using a Biologic equipment, as in [8], and Whatman GF/F glass fiber filters. Such filters have pore diameters larger than the typical diameter (0.06–0.3 μm) of SR vesicles, and therefore retain the SR membranes thanks to adsorption of these membranes onto the pore walls. The total volume of “wetting fluid” in such filters is

100 μL, but membranes are loaded onto the filter using a funnel of diameter smaller than the one of the filter, and they probably adsorb mainly onto the walls of the central pores of the filter, say, within

50 μL fluid. The diameter of the syringe delivering the perfusion fluid is intermediate. Perfusion rates were 2–4.5 ml/s (faster rates for shorter periods).

Remember that in the presence of a low, non-solubilizing (i.e. only “perturbing” [4,26]) concentration of detergent, the total concentration of this detergent is equal to the sum of its free concentration (in the water phase) and its bound concentration (inserted in the membranes, but here expressed per ml of water phase), the latter, at a given free detergent concentration, being dependent on the amount of membranes [27]). The detergent binding isotherms used here, for planning and roughly estimating free and bound concentrations of the brominated detergents under the various situations explored in Figs 2 and 5, were deduced (see e.g. [27,28]) from the membrane concentration-dependent shifts in the detergent-dependent light scattering curves of Figs 1 and 4. They were similar for both versions of each detergents, and consistent with the binding isotherms already published for the non-brominated versions of either C12E8, [4,8] or DDM [29].

(A) Detergent cmc (arrow) for C12E8 and 5,6-Br2C12E8, as determined from the spectral changes of 40 μM methyl orange (expressed as ΔA415 nm− ΔA500 nm, in absorbance units) in the presence of increasing concentrations of these detergents. (B) Perturbation by C12E8 and 5,6-Br2C12E8 of the 90° light scattering (at 290 nm) by SR vesicles (at 4 μg protein/ml). In Panels A and B, closed symbols correspond to 5,6-Br2C12E8, open symbols correspond to non-brominated C12E8. (C and D) Perturbation by C12E8 (C) and 5,6-Br2C12E8 (D) of light scattering by SR vesicles, as in Panel B, but here recorded upon continuous delivery of concentrated detergent from a small syringe into the spectrophotometer cuvette, and in the presence of different concentrations of membranes (20, 50 or 100 μg/ml of protein, short dash, long dash, and continuous lines, respectively). Recorded signals were not corrected for the resulting small dilution of membranes (10% at 0.5 mM detergent). The few data points for “negative” detergent concentrations correspond to data recorded before actuation of the syringe. (E and F) Perturbation by C12E8 (E) and 5,6-Br2C12E8 (F) of the intrinsic fluorescence signal for SR vesicles under the above conditions (continuous delivery of detergent and in the presence of different concentrations of membranes, again at 20, 50 or 100 μg/ml of protein). In Panels D and F, after addition of Br2C12E8 up to 0.44 mM to the 100 μg/ml SR suspension, increasing amounts of non-brominated C12E8 were finally added to the

2 ml suspension, up to about 6 mM (2, 2, 4, 8, 16, and finally 32 μl of a very concentrated solution of C12E8−100 mg/ml, i.e. 186 mM). Blank values (buffer only, in the absence of membranes) were subtracted from the “fluorescence” signal, but detergent-induced dilution and photolysis were not corrected for.

Kinetics of Trp fluorescence quenching or recovery, observed in stopped-flow experiments upon binding (A) or dissociation (B, C, D) of 5,6-Br2C12E8 to or from sarcoplasmic reticulum membranes. (A) Binding experiments. For each shot in these experiments, 36 μl of a suspension of SR membranes at 440 μg protein/ml was mixed with 164 μl (i.e. at a ratio of

1:4.5 vol:vol) of 5,6-Br2C12E8 at zero (top trace, “control”), 73 μM (intermediate trace) or 110 μM (bottom trace) total concentrations, resulting after mixing in final concentrations of 80 μg protein/ml and zero,

90 μM of final total detergent, respectively (i.e. 0,

57 μM of free detergent and 0,

0.61 mol bound detergent/mol membrane lipid, as estimated on the basis of the well-established binding characteristics for the non-brominated C12E8 [4,8] and the fairly similar properties for 5,6-Br2C12E8, see Methods. (B, C, D) Dissociation experiments. In all cases, a suspension of SR membranes at 440 μg protein/ml was first preincubated for a few minutes with 5,6-Br2C12E8 at a total concentration of 200 μM, out of which the free detergent concentration is only 46 μM, a non-solubilizing concentration (see Fig 1), and bound 5,6-Br2C12E8 is

0.52 mol bound detergent/mol membrane lipid. For Panel B experiment, this suspension was then mixed with buffer alone in the stopped-flow machine (lower trace), at the same vol:vol ratio (36 μl + 164 μl, i.e.

1:4.5) as in the binding experiments illustrated in Panel A (hence the similar Y-axis scales). A control was also included (upper trace in Panel B), in which non-brominated C12E8 at the same concentration was used, instead of 5,6-Br2C12E8. For Panel C experiment, a larger dilution factor was used, (20 μl + 180 μl, i.e. 1:9), so that the final concentration of membranes was smaller than for Panel B, hence the smaller Trp fluorescence signal (see Y-axis scales). For Panel D experiment, an even larger dilution factor was used (10 μl + 180 μl, i.e. 1:18). Final free and bound detergent concentrations after reaching equilibrium in the various situations are

9 μM and 1.5 μM (i.e., 0.26, 0.17 or 0.10 mol bound detergent/mol membrane lipid, respectively), estimated as above-mentioned. Note that similar dissociation experiments, repeated now starting from membranes preincubated with a lower total concentration of 5,6-Br2C12E8 (100 μM instead of 200 μM), also led to qualitatively similar recordings (but of course less quenched initial and final fluorescence levels). Traces were usually recorded with 1–2 ms electronic filtering, and correspond to the average of 5–8 shots in all cases, the “blank” optical signal (about 0.4 volts) measured in the total absence of membranes has been subtracted from the recorded signal.


Methods

FC12 H , FC12 D , DDM H , DDM D , OG H , OG D , LMNG, DMNG, CHAPS and CHAPSO were from Anatrace. C4C10 and C4C12 were from CALIXAR. BSA (Fraction V, MW = 65.5 kDa) and lysozyme were from Euromedex. Buffers and solutes, ovalbumin and dextran (80 kDa MW standard) were from Sigma-Aldrich.

MALDI-TOF MS

Mass spectrometry was performed using a Voyager-DE Pro MALDI-TOF Mass Spectrometer (AB Sciex, Framingham, MA) equipped with a nitrogen UV laser (λ = 337 nm, 3 ns pulse). The instrument was operated in the positive reflectron mode (mass accuracy: 0.008%) with an accelerating potential of 5 kV. All volumes were weighted on a precision scale to maximize accuracy and to compensate for pipetting errors. Detergents (measured and standard) are mixed typically in 50 μL, then 1 μL of the mixture is added of 9 μL of its optimized desorption-helper matrix solution: 10 g DHBA/L water with DDM, OG and CHAPS 10 g CHCA/L 50:50 acetonitrile:water for FC12 and C4C10/12 10 g THAP/L 30:70 acetonitrile:water for D/L-MNG 10 g 9AA/L 50:50 acetone:methanol for cholate and deoxycholate. Except for FC12 and cholate/deoxycholate, 1 μL of 10 g NaI/L acetone was added to the detergent/matrix mixture to produce MNa + cations. However, in the case of Lauryl- and Decyl-MNGs, the 2′,4′,6′-trihydroxyacetophenone (THAP) was found more suitable because of the need of lower energy transferred during the ionization process leading to a more stable signal. The 9-aminoacridine (9AA) was found to efficiently desorb the bile derivatives cholate and deoxycholate 48 .

One microliter of the final solution was laid on the MALDI target and air-dried before analysis. For each trial, mass spectra were obtained by accumulation of 3 series of 300 laser shots, each acquired in 3 distinct areas of the dried mixture. Samples containing high concentrations of imidazole were diluted 10 times before addition of the matrix for favoring crystallization of the mixture. For the analysis of samples containing D/L-MNG, 1 μL of tetrahydrofuran was added onto the dried spot to homogenize it, and air-dried again before data acquisition. Standard curves were fitted with SigmaPlot V12.5.

The optimum acceleration voltage was set to 5,000 V for all detergents tested. This voltage, lower than those used in classical MALDI-TOF MS, gives less energy in the ionization process, leading to stable ionization conditions. As an example, the calibration curve for DDM clearly became non-linear as the voltage increased from 5,000 to 25,000 kV (Supplementary Figure 2).

Detergents desorbed quite differently (Fig. 1b), with a high abundance for deuterated/protonated (H/D) FC12, OG and CHAPS/CHAPSO (2,000 to 20,000 counts) and rather low for (H/D) DDM or L/D-MNG (200 to 2,000 counts). Each triplicate series, displayed highly variable abundances (Fig. 1c). However, the relative laser desorption efficacies of the measured and standard detergents remained remarkably similar (Fig. 1d), resulting in measured to standard ratio close to the theoretical value. In most cases, the linear fit led to experimental slopes of 0.1, which indicate that the two detergents desorb similarly, whereas we reproducibly obtained 0.143 for LMNG, suggesting that it desorbs less efficiently than DMNG, but still proportionally.

Automation

For further high-throughput analysis, we automated the MALDI-TOF MS acquisition process, for treating up to 100 samples in a row, by using the automatic tool embedded in the Voyager 5.1 (Sciex) software. We tested the set up with DDM and FC12. The acceptance and adjustment threshold of the software were based on an interval of signal intensities such that the signals of the molecular ions 533/558 (DDM + Na + ) or 352/390 (FC12 + H + ) fall below a signal-to-noise ratio of 10. We got these settings by choosing an automatic set up of (i) the laser beam intensity value, (ii) the laser beam displacement over the spot, and (iii) the accumulation of nine series of 100 laser shots. In these conditions, the acceptance criteria were found within an interval of 1,000–30,000 counts for a m/z of 533 and 10,000–40,000 counts for a m/z of 352.

Repeatability and reproducibility

We checked the repeatability and reproducibility of the method for FC12 and DDM by measuring 3 independent experiments respectively on the same day and over three distinct days. The results satisfactorily showed an intra-day average coefficient of variation (CV) of 0.5–12.8% and 0.5–3.8%, and an inter-day CV of 9–18% and 4.0–6.0%, for FC12 and DDM respectively (Table 1).

Accessible Hydrophobic Surface (AHS) of the membrane domain calculation

Membrane size limits were determined by using the Orientation of Proteins in Membranes online server 49 , http://opm.phar.umich.edu/server.php 56 . PDB entries were submitted to the server which placed the proteins in the same orientation and allowed to depict them in a same scale and orientation (Supplementary Figure 5). The software Naccess, V2.1.1 50 , with a probe radius of 1.5 Å (C-C bond length, corresponding to the alkyl chain of detergents) was used to determine total AHS, of which membrane-AHS was extracted using a home-made software, measuring the accessible surface area only contained within the membranous limits determined by the OPM server above. All the PDB entries available for the type of membrane protein tested on this study were subjected to this analysis. Given all the different conformations sampled by the PDB entries, the error on the membrane-AHS calculation was less than 10%.

Cylinder-shape modeling of detergent belt surrounding membrane proteins

The volume of the detergent belt, Vbelt can be approximated to the sum of the volume Vdet occupied by each detergent molecule,

Vdet is calculated for a given detergent by using the program VOIDOO 51 (http://xray.bmc.uu.se/usf/voidoo.html) (Supplementary Table 4). The number of bound detergent molecules, N, is determined experimentally by MALDI-TOF MS.

Vbelt can also be approximated to a hollow cylinder surrounding the membrane region of the protein, the inner volume being occupied by the protein (scheme in Fig. 5c). It can be determined by equation (2) below:

where Rt (Å) is the radius of the whole cylinder including protein and detergent, Rp (Å) being the radius of the protein cylinder and H (Å) the thickness of the membrane bilayer. Rp, is obtained by averaging 5–10 distances throughout the membrane domain of a given protein at the inner and outer membrane boundaries, parallel to the membrane plane, using the coordinate files of a given protein for which the 3D-structure is known or that of close homologs and a software for visualizing the 3D structures such as SwissPDBviewer 52 (v4.1) or PyMOL (The PyMOL Molecular Graphics System, Version 1.8 Schrödinger, LLC.). H is determined by submitting the same coordinates to the OPM server 49 (http://opm.phar.umich.edu/server.php) that determines the membrane boundaries according to electrostatic potentials.

Rt can be deduced from equation (2) as follows:

Finally, the radius of the detergent belt, Rb, can be deduced with the equation (4) below:

The detergent occupied volume is then displayed as a cylinder of radius Rt and height H around the membrane protein using Pymol.


Membrane protein stability can be compromised by detergent interactions with the extramembranous soluble domains

Correspondence to: Christie G. Brouillette, 1025 18th Street South, 234 CBSE, Birmingham, AL 35294-4400. E-mail: [email protected] or John F. Hunt, Department of Biological Sciences, Columbia University, New York, New York. E-mail: [email protected] Search for more papers by this author

Department of Chemistry, University of Alabama at Birmingham, Birmingham, Alabama

Center for Structural Biology, University of Alabama at Birmingham, Birmingham, Alabama

Correspondence to: Christie G. Brouillette, 1025 18th Street South, 234 CBSE, Birmingham, AL 35294-4400. E-mail: [email protected] or John F. Hunt, Department of Biological Sciences, Columbia University, New York, New York. E-mail: [email protected] Search for more papers by this author

Department of Chemistry, University of Alabama at Birmingham, Birmingham, Alabama

Center for Structural Biology, University of Alabama at Birmingham, Birmingham, Alabama

Department of Biological Sciences, Columbia University, New York, New York

Center for Structural Biology, University of Alabama at Birmingham, Birmingham, Alabama

Center for Structural Biology, University of Alabama at Birmingham, Birmingham, Alabama

Department of Cell Biology and Biochemistry, Texas Tech University Health Sciences Center, Lubbock, Texas

Department of Biochemistry and Biophysics, The University of North Carolina at Chapel Hill, Chapel Hill, North Carolina

Cystic Fibrosis Treatment and Research Center, The University of North Carolina at Chapel Hill, Chapel Hill, North Carolina

Department of Medicine, University of Alabama at Birmingham, Birmingham, Alabama

Birmingham Veterans Affairs Medical Center, Research Service, Birmingham, Alabama

Center for Structural Biology, University of Alabama at Birmingham, Birmingham, Alabama

Department of Optometry, University of Alabama at Birmingham, Birmingham, Alabama

Department of Biochemistry and Biophysics, The University of North Carolina at Chapel Hill, Chapel Hill, North Carolina

Cystic Fibrosis Treatment and Research Center, The University of North Carolina at Chapel Hill, Chapel Hill, North Carolina

Department of Cell Biology and Biochemistry, Texas Tech University Health Sciences Center, Lubbock, Texas

Center for Membrane Protein Research, Texas Tech University Health Sciences Center, Lubbock, TX

Department of Biological Sciences, Columbia University, New York, New York

Correspondence to: Christie G. Brouillette, 1025 18th Street South, 234 CBSE, Birmingham, AL 35294-4400. E-mail: [email protected] or John F. Hunt, Department of Biological Sciences, Columbia University, New York, New York. E-mail: [email protected] Search for more papers by this author

Department of Chemistry, University of Alabama at Birmingham, Birmingham, Alabama

Center for Structural Biology, University of Alabama at Birmingham, Birmingham, Alabama

Correspondence to: Christie G. Brouillette, 1025 18th Street South, 234 CBSE, Birmingham, AL 35294-4400. E-mail: [email protected] or John F. Hunt, Department of Biological Sciences, Columbia University, New York, New York. E-mail: [email protected] Search for more papers by this author

Abstract

Detergent interaction with extramembranous soluble domains (ESDs) is not commonly considered an important determinant of integral membrane protein (IMP) behavior during purification and crystallization, even though ESDs contribute to the stability of many IMPs. Here we demonstrate that some generally nondenaturing detergents critically destabilize a model ESD, the first nucleotide-binding domain (NBD1) from the human cystic fibrosis transmembrane conductance regulator (CFTR), a model IMP. Notably, the detergents show equivalent trends in their influence on the stability of isolated NBD1 and full-length CFTR. We used differential scanning calorimetry (DSC) and circular dichroism (CD) spectroscopy to monitor changes in NBD1 stability and secondary structure, respectively, during titration with a series of detergents. Their effective harshness in these assays mirrors that widely accepted for their interaction with IMPs, i.e., anionic > zwitterionic > nonionic. It is noteworthy that including lipids or nonionic detergents is shown to mitigate detergent harshness, as will limiting contact time. We infer three thermodynamic mechanisms from the observed thermal destabilization by monomer or micelle: (i) binding to the unfolded state with no change in the native structure (all detergent classes) (ii) native state binding that alters thermodynamic properties and perhaps conformation (nonionic detergents) and (iii) detergent binding that directly leads to denaturation of the native state (anionic and zwitterionic). These results demonstrate that the accepted model for the harshness of detergents applies to their interaction with an ESD. It is concluded that destabilization of extramembranous soluble domains by specific detergents will influence the stability of some IMPs during purification.

Additional Supporting Information may be found in the online version of this article.

Filename Description
pro2460-sup-0001-suppinfo.doc2.1 MB Supplementary Information

Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.


What are efficient methods to detect protein stability?

Many thermal shift analysis procedures have utilized spectroscopy to assess protein stability. Typical applications have included observing the helix-coil transition in proteins upon thermal denaturation, as well as for monitoring the effects that buffers, such as sodium acetate or sodium chloride, have on overall protein thermal stability. Prior to the introduction of the PROTEOSTAT® Thermal Shift Stability Assay Kit, protein aggregation-based thermal denaturation assays routinely employed devices to raise the temperature of the protein sample in a stepwise fashion. Environmentally-sensitive dyes, such as ANS and SYPRO® Orange, were applied to the detection of protein unfolding in thermal shift assays by interacting with exposed hydrophobic regions generated by partial or full unfolding of proteins. In conjunction with monitoring of protein aggregation through light scattering or turbidity measurements, a sigmoidal curve would be recovered. Examination of the sigmoidal curve inflection point provided the Tm (melting point), or the temperature at which 50% of measured protein is unfolded. Early thermal shift assays were cumbersome, typically required relatively high concentrations of protein, and provided a very limited assay dynamic range. Additionally, required assay equipment was often expensive and unavailable in standard laboratories. Proteins containing easily accessible hydrophobic regions and detergent exposure often caused unwanted interactions with the aforementioned dyes, which produced high fluorescent background.

The PROTEOSTAT® Thermal Shift Stability Assay provides an improved thermal shift approach for assessment of protein stability through directly monitoring protein aggregation, rather than protein unfolding and by minimizing problems encountered with the cited environment-sensitive dyes, such as high background fluorescence from their interaction with detergents, membrane proteins, or hydrophobic compounds. The PROTEOSTAT® TS dye is a proprietary 488 nm excitable molecular rotor probe that, under standard aqueous conditions, is minimally fluorescent. Upon binding to the surface of aggregated proteins, the dye emits a strong red signal at

600 nm, thus providing a homogeneous assay for the analysis of protein stability. The temperature at which the bulk of the protein becomes aggregated can readily be identified by the thermal shift assay. The aggregation temperature is an indicator of protein stability and can be used to optimize conditions that enhance protein stability as well as to identify ligands or drugs that bind and confer structural stability to a protein of interest. Conditions that increase the aggregation temperature increase the stability of the protein. The assay facilitates understanding of the underlying mechanisms impacting protein stability, and because it is not dependent upon measuring exposed hydrophobic regions arising from protein unfolding, it is more tolerant of detergents, micelle formation and certain ligands and proteins possessing hydrophobic characteristics. Enzo’s PROTEOSTAT® Protein Refolding and Aggregation Sensing Kit can additionally be used to identify optimal protein refolding conditions. Conditions that increase the aggregation temperature, increase the stability of the protein.

Figure 1. Typical results of the PROTEOSTAT® Thermal Shift Stability Assay are shown for goat anti-mouse IgG (11.2 mg/ml at pH 7.4). Using a RT-PCR instrument programmed to ramp the temperature from 30° to 99°C at a 3°/minute rate, while reading the fluorescence continuously. The blue line represents the raw fluorescence data, and the red line shows the corresponding first derivative trace, highlighting the slope of the fluorescence intensity curve. The first derivative plot provides the aggregation temperature of the protein (Tagg: The point of maximal slope).

Figure 2. Typical data for the determination of the aggregation temperature of IgG (goat anti-mouse, 5.6mg/ml). Panel A is the raw data from the thermocycler. Panel B is a plot of dF/dT° or slope of the data shown in panel A. The dashed line indicates the aggregation temperature.

Figure 3. Ligand stabilization of protein structure. Carbonic anhydrase I (1 μM) was incubated with 0 to 62.5 mM of TFMSA (trifluoromethane sulfonamide) in 25 mM MES, 50 mM NaCl, pH 6.1. The PROTEOSTAT® Thermal Shift Stability Assay demonstrates that ligand binding increases protein thermal stability by an amount proportional to the concentration of the ligand.

Figure 4. Stability of a protein in different buffers. β-Lactoglobulin was diluted into 50 mM buffer at different pHs in the presence of 150 mM NaCl, and the aggregation temperature was determined using PROTEOSTAT® dye.


Enzo’s catalog of widely cited and thoroughly validated products includes our PROTEOSTAT® assays for contamination monitoring, stability, and process optimization. This kit has been designed for monitoring protein stability under systematic thermal stress conditions. Please check out our Bioprocess Optimization platform for more information or contact our Technical Support Team for further assistance.


Abstract

Detergent interaction with extramembranous soluble domains (ESDs) is not commonly considered an important determinant of integral membrane protein (IMP) behavior during purification and crystallization, even though ESDs contribute to the stability of many IMPs. Here we demonstrate that some generally nondenaturing detergents critically destabilize a model ESD, the first nucleotide-binding domain (NBD1) from the human cystic fibrosis transmembrane conductance regulator (CFTR), a model IMP. Notably, the detergents show equivalent trends in their influence on the stability of isolated NBD1 and full-length CFTR. We used differential scanning calorimetry (DSC) and circular dichroism (CD) spectroscopy to monitor changes in NBD1 stability and secondary structure, respectively, during titration with a series of detergents. Their effective harshness in these assays mirrors that widely accepted for their interaction with IMPs, i.e., anionic > zwitterionic > nonionic. It is noteworthy that including lipids or nonionic detergents is shown to mitigate detergent harshness, as will limiting contact time. We infer three thermodynamic mechanisms from the observed thermal destabilization by monomer or micelle: (i) binding to the unfolded state with no change in the native structure (all detergent classes) (ii) native state binding that alters thermodynamic properties and perhaps conformation (nonionic detergents) and (iii) detergent binding that directly leads to denaturation of the native state (anionic and zwitterionic). These results demonstrate that the accepted model for the harshness of detergents applies to their interaction with an ESD. It is concluded that destabilization of extramembranous soluble domains by specific detergents will influence the stability of some IMPs during purification.

Abbreviations


Specialty assays

Finally, for specialists who may be examining small peptides at low concentration or proteins associated with lipid bilayers that can interfere with standard assay reagents, there are several kits and protocols that allow reliable and reproducible measurement. One example is the CBQCA assay, designed for highly sensitive measurement of proteins in lipid solutions, which uses potassium cyanide (danger!) to stimulate reaction with amine groups, resulting in fluorescent excitation.

In addition to CBQCA, there is a seemingly limitless variety of specialty applications for protein quantification. The table below provides a summary of differences and compatibilities between the more standard assays. Although BCA is perhaps currently the most favored, there are rationales for choosing other assays based on factors such as buffers, time frames, detection limits, and wavelengths particular to spectrophotometer and microplate reader setups. What works best for the end user depends on considerations that are unique to every laboratory&rsquos field of interest. This table uses kits provided by Thermo Fisher as a guide, although several are also offered by Sigma-Aldrich and Bio-Rad, and the more adventurous and thrifty among you can always make your own.


Watch the video: Ποια απορρυπαντικα χρησιμοποιώ! (January 2022).