Determining protein stability in cell lysates by pulse proteolysis and Western blotting

Authors

  • Moon-Soo Kim,

    1. Department of Medicinal Chemistry and Molecular Pharmacology and Bindley Bioscience Center, Purdue University, 575 Stadium Mall Drive, West Lafayette, Indiana 47907
    Current affiliation:
    1. UC Davis Genome Center, Department of Pharmacology, 451 E. Health Sciences Drive, University of California, Davis, California 95616
    Search for more papers by this author
  • Jiao Song,

    1. Department of Medicinal Chemistry and Molecular Pharmacology and Bindley Bioscience Center, Purdue University, 575 Stadium Mall Drive, West Lafayette, Indiana 47907
    Search for more papers by this author
  • Chiwook Park

    Corresponding author
    1. Department of Medicinal Chemistry and Molecular Pharmacology and Bindley Bioscience Center, Purdue University, 575 Stadium Mall Drive, West Lafayette, Indiana 47907
    • 575 Stadium Mall Drive, Purdue University, West Lafayette, IN 47907
    Search for more papers by this author

Abstract

Proteins require proper conformational energetics to fold and to function correctly. Despite the importance of having information on conformational energetics, the investigation of thermodynamic stability has been limited to proteins, which can be easily expressed and purified. Many biologically important proteins are not suitable for conventional biophysical investigation because of the difficulty of expression and purification. As an effort to overcome this limitation, we have developed a method to determine the thermodynamic stability of low abundant proteins in cell lysates. Previously, it was demonstrated that protein stability can be determined quantitatively by measuring the fraction of folded proteins with a pulse of proteolysis (Pulse proteolysis). Here, we show that thermodynamic stability of low abundant proteins can be determined reliably in cell lysates by combining pulse proteolysis with quantitative Western blotting (Pulse and Western). To demonstrate the reliability of this method, we determined the thermodynamic stability of recombinant human H-ras added to lysates of E. coli and human Jurkat T cells. Comparison with the thermodynamic stability determined with pure H-ras revealed that Pulse and Western is a reliable way to monitor protein stability in cell lysates and the stability of H-ras is not affected by other proteins present in cell lysates. This method allows the investigation of conformational energetics of proteins in cell lysates without cloning, purification, or labeling.

Introduction

Protein homeostasis in cells through coordinated production and turnover is essential to maintaining cell viability.1 Optimal conformational energetics is required for the proper folding, function, and degradation of cellular proteins. Mutations that modify the energetic properties of a protein can be equally detrimental to its function as mutations which impair enzymatic activity or binding interactions. This phenomenon has been experimentally verified using a systematic analysis of random mutations in model proteins.2 In fact, most mutations, which inactivate proteins, occur at residues distant from active sites or binding regions. These mutations likely destabilize proteins to the extent that the proper structure or dynamics that are necessary for their function can not be achieved. Genetic screens of temperature-sensitive mutations also reveal the critical importance of the conformational energetics of proteins. The ability to rapidly, quantitatively, and accurately assess the conformational energetics of proteins would provide a powerful tool not only to biophysicists, but also to geneticists and cell biologists.

Current approaches to study conformational energetics of proteins necessitate cumbersome and time-consuming protein purification procedures. Cloning and heterologous expression are also commonly required to produce sufficient quantities of protein for analysis. These overexpression and purification procedures often involve trial and error optimization strategies. Therefore, new methods to study conformational energetics of proteins in situ without purification would greatly enhance our ability to obtain valuable energetic information on an array of proteins that would be challenging to study with current methodology.

Pulse proteolysis offers a powerful approach for studying the thermodynamic stability of unpurified proteins.3 By exploiting the difference in proteolytic susceptibility between folded and unfolded proteins, this method allows the facile determination of the fraction of folded proteins under a given condition. A mixture of folded and unfolded proteins is digested with an excess amount of a protease for only one minute so that unfolded proteins are completely digested while tightly folded proteins remain intact. After the reaction is quenched, the fraction of folded proteins is determined by SDS PAGE. This method has been successfully applied to determine thermodynamic stabilities and unfolding kinetics of proteins.3, 4 Because of the separation of proteins by SDS PAGE following proteolysis, the method is relatively tolerant to the presence of impurities and can be performed in cell lysates without protein purification, although reliable identification in SDS PAGE gels required that the protein of interest be highly abundant.

To overcome this limitation, we report here a method to determine thermodynamic stability of low abundant proteins in cell lysates by quantifying intact proteins after pulse proteolysis with Western blotting, dubbed “Pulse and Western.” We demonstrate the validity of the method with c-H-ras (H-ras) as a model protein. H-ras is a member of ras family, which are small GTPases, with critical functions in signal transduction pathways regulating cell growth and differentiation.5 Small GTPases exist in a complex with either GDP or GTP. When complexed with GTP, small GTPases are in their active conformation and interact with their downstream effectors. Hydrolysis of GTP results in an inactive GDP-bound complex. Mutations in ras proteins are frequently observed in many forms of human cancers.6 In mammalian cells, c-H-ras is anchored to membranes through post-translational modifications at a specific motif in its C-terminus.7 Soluble constructs of H-ras with the truncated C-terminus are frequent research subjects in structural biology and biophysics.8, 9 Thermodynamic stability of soluble H-ras has been rigorously investigated, and the role of ligand binding in its stability is well understood.10 As the ras proteins are key signaling molecules, antibodies for H-ras are readily available. These features make H-ras an ideal model protein for the implementation of Pulse and Western.

To avoid the complication of the effect of membrane anchoring on the thermodynamic stability of endogenous H-ras, we determined thermodynamic stabilities by Pulse and Western of soluble recombinant H-ras added to cell lysates and evaluated the results by comparing with the stabilities determined with the purified protein. The transition midpoint of equilibrium unfolding, Cm, was determined for H-ras added to an E. coli lysate to determine the thermodynamic stability of the protein in the presence of relatively inert protein background. The accuracy of Pulse and Western was assessed by comparing the result with the Cm values determined with purified H-ras. To assess the effect of interactions with potential binding partners on stability determination in mammalian cell lysates, we also determined the Cm value of soluble H-ras added to a Jurkat cell lysate. Jurkat cells express easily detectable levels of K-ras and N-ras, but undetectable levels of H-ras.11 This strain provides the opportunity to study the thermodynamic stability of H-ras in the context of a mammalian cell lysate without interference of the endogenous H-ras anchored to the membrane.

Results

Determination of Cm of purified H-ras by pulse proteolysis

The recombinant protein used for this study is a truncated version of human c-H-ras (residue 1–171), which does not have the C-terminal sequence necessary for membrane anchoring.12 The H-ras protein was expressed in E. coli and purified to homogeneity. The thermodynamic stability of purified H-ras was determined by pulse proteolysis. The thermodynamic stability of H-ras is dependent on the concentration of GDP and Mg2+ because H-ras exists as a ternary complex with these ligands.10 Therefore, 30 μM GDP and 5.0 mM Mg2+ were used as a standard condition for all of our assays. After equilibrium was established, each sample was digested with 0.2 mg/mL thermolysin for one minute. The amount of remaining intact protein in each reaction was determined by quantifying band intensities of H-ras on a SDS-PAGE gel stained with a fluorescent dye [Fig. 1(A)]. The intensities of intact H-ras protein decreased rapidly in the range of 3.9–4.3M urea, indicating a cooperative transition. The plot of the fraction of folded H-ras (Ffold) versus urea concentration was fit to a two-state equilibrium unfolding model [Fig. 1(B)].13 The Cm value of H-ras was determined to be 3.99 ± 0.02M. To confirm that only unfolded H-ras was digested during pulse proteolysis, H-ras equilibrated in 3.7M urea was incubated with 0.2 mg/mL thermolysin for 10 min [Fig. 1(B), inset]. Folded H-ras was clearly resistant to proteolysis under this condition, indicating that pulse proteolysis is a reliable way to determine Ffold of H-ras.

Figure 1.

Determination of Cm of purified H-ras by pulse proteolysis. (A) SDS-PAGE gel of purified H-ras after 1-min pulse proteolysis. The location of H-ras is marked by an arrow. Thermolysin forms bands with a molecular weight of ∼33 kDa. (B) Ffold of H-ras determined by pulse proteolysis in varying concentrations of urea. Cm (± SE) was determined to be 3.99 ± 0.02M by fitting to a two-state unfolding model. Inset: Proteolysis kinetics of H-ras in 3.7M urea was monitored for 10 min. The band intensity of the intact protein was plotted against incubation time. The standard error in Ffold determination by this method is typically less than 5%.

To confirm that H-ras reached conformational equilibrium following overnight incubation in urea, we determined its relaxation kinetic constant by pulse proteolysis at Cm (4.0M urea). Pulse proteolysis is also an effective way to determine relaxation kinetic constants when unfolding occurs on a slow time scale.4 Unfolding of H-ras was initiated by mixing the folded protein with urea to a final concentration of 4.0M. At each designated time point, pulse proteolysis was performed with a small fraction of the sample to digest unfolded protein. The amount of remaining folded protein at each time point was determined by measuring the band intensity of intact H-ras on a SDS PAGE gel. The relaxation kinetic constant was determined by fitting the intensity change over time to a first order rate equation (see Fig. 2). The half life of the relaxation in 4.0M urea was determined to be 3.9 ± 0.6 h, which indicates that overnight incubation (typically 18–20 h) is long enough for the reaction to reach equilibrium even at Cm where the relaxation is slowest.

Figure 2.

Relaxation kinetics of H-ras at its Cm in urea. The remaining folded H-ras was determined by pulse proteolysis at each time point after unfolding of the protein was initiated in 4.0M urea. The relaxation kinetic constant (± SE) was determined to be 0.18 ± 0.03 h−1 by fitting to a first-order kinetic equation. The standard error in Ffold determination by this method is typically less than 5%.

We also monitored the unfolding of H-ras by circular dichroism (CD) under identical conditions to verify the reliability of the pulse proteolysis result with H-ras. H-ras was incubated in different concentrations of urea (0–6M), and the ellipticity of each sample was determined at 222 nm. By fitting the ellipticity change to a two-state unfolding model, Cm and the m-value were determined to be 4.16 ± 0.05M and 1.74 ± 0.16 kcal/(mol · M) (Table I). The Cm value determined by pulse proteolysis (3.99 ± 0.02M) is in good agreement with the value determined by CD. Considering that the typical error in Cm determination by pulse proteolysis is ∼0.1M, the difference between the Cm values determined by the two methods seems to be within the experimental error. The global stability (ΔGunf°) of H-ras was determined by multiplying Cm from the pulse proteolysis experiment by the m-value (1.74 ± 0.16 kcal/(mol · M)) obtained from the equilibrium unfolding experiment by CD (Table I)*.

Table I. Cm Values of H-ras Determined by Pulse Proteolysis and CD Under Different Conditions
H-rasMethodCm (M)ΔGunf° (kcal/mol)
  1. The Cm values (±SE) were determined by fitting band intensities in different urea concentrations to a two-state equilibrium unfolding model. The concentrations of GDP and Mg2+ used in this study were 30 μM and 5 mM, respectively. The ΔGunf° values were calculated by multiplying the Cm values by the m-value determined by CD (1.74 kcal/(mol · M)).

H-ras onlyPulse proteolysis3.99 ± 0.026.94
H-ras onlyCD4.16 ± 0.057.24
In an E. coli cell lysatePulse proteolysis4.05 ± 0.037.05
In a Jurkat cell lysatePulse proteolysis4.00 ± 0.046.96

Determination of the stability of H-ras in an E. coli lysate by Pulse and Western

To determine the stability of H-ras at low concentration in the presence of background proteins, we chose an E. coli lysate as an inert control system. H-ras is a human protein and, therefore, is unlikely to have any specific interaction with proteins in an E. coli lysate.

Pulse proteolysis was performed on an E. coli lysate with 31 μg/mL H-ras and different concentrations of urea (0–6M) after overnight incubation. The total protein concentration in the cell lysate was ∼2 mg/mL. Therefore, H-ras was less than 2% (by weight) of the total protein in this assay. To prevent GDP from being hydrolyzed by enzymes in the lysate, we have used GDP-β-S, a non-hydrolyzable analogue of GDP for assays with cell lysates. We confirmed that GDP and GDP-β-S do not make any detectable deference in the thermodynamic stability of purified H-ras (Data not shown). Figure 3(A) shows the proteins remaining in the E. coli lysate after pulse proteolysis. H-ras typically runs as a band with a molecular weight of ∼19 kDa [Fig. 1(A)]. However, the band of H-ras cannot be identified because of the many background proteins in the cell lysate.

Figure 3.

Determination of Cm of purified H-ras added to an E. coli cell lysate by Pulse and Western. (A) SDS-PAGE gel stained with SyproRed, a fluorescent dye, to show the total proteins in the E. coli lysate containing 31 μg/mL of purified H-ras after pulse proteolysis in varying concentrations of urea. (B) Western blotting to detect intact H-ras in the E. coli lysate after pulse proteolysis. (C) Ffold of H-ras determined by quantifying band intensities of H-ras on the Western blot. Cm (± SE) was determined to be 4.05 ± 0.03M by fitting the plot to a two-state unfolding model. The standard error in Ffold determination by this method is typically less than 5%.

To quantify the remaining H-ras after pulse proteolysis in the lysate samples, we have performed Western blotting with an anti-H-ras primary antibody and a fluorescence-labeled secondary antibody. Fluorescence-based Western blotting allows quantitative and sensitive detection of H-ras. The linear dependence of the fluorescence intensity on the amount of H-ras was confirmed in advance by using known amounts of H-ras in the range of remaining protein concentrations after pulse proteolysis (Data not shown). After H-ras was transferred to the membrane, the protein was probed with a monoclonal antibody specific for H-ras. The Western blot visualizes the remaining H-ras specifically even in the presence of many background proteins in the cell lysate [Fig. 3(B)].

The band intensities determined by Western blotting were fit to two-state equilibrium unfolding model as used for purified H-ras [Fig. 3(C)]. The Cm value of H-ras was determined to be 4.05 ± 0.03M, which is in good agreement with Cm values determined by CD and pulse proteolysis with purified H-ras (Table I). This result shows that Pulse and Western is a reliable way to determine the stability of a protein in the presence of other background proteins.

Determination of stability of H-ras in a mammalian cell lysate by Pulse and Western

Unlike E. coli lysates, mammalian cell lysates have proteins which interact with H-ras. To determine if cellular proteins interfere with the stability determination of H-ras in mammalian cell lysates, we have monitored equilibrium unfolding of recombinant H-ras added to a Jurkat cell lysate by Pulse and Western. Because endogenous H-ras is anchored to cytosolic membranes through post-translational modifications, it is inappropriate to compare the thermodynamic stability of endogenous H-ras with that of recombinant H-ras. It is known that Jurkat cells do not express H-ras to a detectable level by Western blotting.11 Therefore, the Jurkat cell lysate offers a chance to assess the interference from proteins interacting with H-ras without the complication from endogenous H-ras.

Pulse proteolysis was performed on a Jurkat cell lysate with 31 μg/mL recombinant H-ras in varying concentrations of urea. The remaining H-ras was quantified with Western blotting [Fig. 4(A)]. Unlike H-ras in the E. coli lysate, the recombinant H-ras incubated in the Jurkat cell lysate looked somewhat diffusive on the Western blot. The smear of the band is not due to any proteins in the cell lysate cross-reacting with the antibody. The Jurkat cell lysate alone did not show any reactive band to the antibody used in this Western blotting [“−” lane in Fig. 4(A)] as reported previously.11 Also, H-ras incubated under the same condition without the cell lysate did not show any smear [“+” lane in Fig. 4(A)]. It is likely that a small population of H-ras was modified during incubation in the cell lysate. Nontheless, the integration of band intensities showed that the fraction of the protein modified was only ∼10% of the total protein. Even with this smearing and possible interactions with other cellular proteins, the stability of H-ras was not affected. The Cm value was determined to be 4.00 ± 0.04M, which is again in good agreement with the values determined with pure H-ras [Fig. 4(B) and Table I]. This result shows that transient interactions with other proteins in cell lysates do not affect significantly the thermodynamic stability of H-ras.

Figure 4.

Determination of Cm of purified H-ras added to a Jurkat cell lysate by Pulse and Western. (A) Western blotting to detect intact H-ras in a Jurkat cell lysate after pulse proteolysis. A Jurkat cell lysate without added H-ras (−) and H-ras alone (+) were run in the same concentration as controls. (B) Ffold of H-ras determined by quantifying band intensities of H-ras on the Western blot. Cm (± SE) was determined to be 4.00 ± 0.04M by fitting the plot to a two-state unfolding model. The standard error in Ffold determination by this method is typically less than 5%.

Discussion

Determining protein stability by Pulse and Western

The combination of pulse proteolysis and Western blotting provides an opportunity to determine protein stability in the presence of cellular background proteins. In this work, we validated this approach by determining the stability of recombinant H-ras in E. coli lysates and Jurkat cell lysates under conditions where H-ras was a minor protein component and therefore not identifiable on SDS PAGE gels [Fig. 3(A)]. By using an antibody specific for H-ras the amount of H-ras remaining after pulse proteolysis was determined quantitatively. The sensitivity and specificity of Pulse and Western allows for stability determination of a minor protein within a mixture of various proteins. This method also does not require engineering or labeling of proteins, and thus allows direct characterization of endogenous proteins.

Recently, several approaches have been developed to monitor protein stability in cell lysates, in cells, or even in vivo. SUPREX (stability of unpurified proteins from rates of H/D exchange) utilizes mass spectroscopy to monitor the change in protein masses from the exchange of amide protons with solvent deuterium upon unfolding.14 This method allows stability determination of proteins in living E. coli cells.15 However, because the method can monitor only proteins forming distinguishable peaks in a mass spectrum, the protein of interest needs to be over-expressed or endogenously highly abundant. Another notable method is specific tagging of proteins with a fluorescent moiety in cells. A fluorescein analogue containing two arsenoxide (FlAsH) can label proteins with Cys-Cys-X-X-Cys-Cys motif in cells.16 When a protein is engineered to have this motif and is labeled by using FlAsH, unfolding of the protein can be monitored in cells from the change in fluorescence intensity.17 This approach, however, requires cloning and engineering of the target protein. Though Pulse and Western does not allow the measurement of protein stability in vivo or in intact cells, in principle this method can be employed without cloning, labeling, or over-expression.

The successful implementation of Pulse and Western to monitor protein stability in cell lysates leads to exciting opportunities, which are not possible with conventional methods. Though over-expression and purification of proteins have become routine laboratory procedure, many proteins defy these approaches. Many human proteins with interesting biological function are still challenging to express and purify. The difficulty in expression and purification of membrane proteins is also well known. Pulse and Western offers a chance to study these challenging proteins without overexpression and protein purification. Conformational energetics of endogenous proteins can be investigated in cell lysates with an appropriate antibody. Pulse proteolysis is also useful to investigate folding/unfolding kinetics of proteins.4 We employed this method here to determine the relaxation kinetic constant of purified H-ras at its Cm (see Figure 2). Pulse and Western would allow the determination of folding and unfolding rates of endogenous proteins in cell lysates without purification.

The detection sensitivity of Pulse and Western is critical in investigating conformational energetics of endogenous proteins with low abundance. The minimum requirement of protein for Pulse and Western is actually determined by the sensitivity of Western blotting, which is dependent on several factors including the quality of the primary antibody and detection methodology. In theory, if a protein can be detected quantitatively by any Western blotting methodology, Pulse and Western can be used to study conformational energetics of the protein. In this study H-ras was quantified by fluorescence of the Cy5-tagged secondary antibody. This method is known to detect subnanogram quantities of proteins in a quantitative manner.18 By optimizing the Western blotting procedure and employing more sensitive detection methodology, Pulse and Western would readily be applicable to low abundant endogenous proteins.

The basic requirement of pulse proteolysis is that the folded protein is resistant to proteolysis.1, 19 However, if a protein contains proteolytically resistant domains, Pulse and Western still can be useful to investigate the conformational energetics of those domains. Because the equilibrium is established before proteolysis occurs, the determined stability of a domain would be the stability in the context of the full length protein. Additionally, the availability of antibodies to specific protein domains would allow the monitoring of individual domain unfolding.

Pulse and Western is only applicable when a specific antibody is available for a protein. However, tens of thousands of antibodies are available commercially and the generation of an antibody specific for a protein is possible when the sequence is known. Proteins are also frequently tagged with known epitope sequences recognizable by antibodies for facile detection or affinity purification.20 This epitope-tagging can also be applied to Pulse and Western, which then would not require a specific antibody for each protein. However, it should be confirmed that the epitope tags are resistant to pulse proteolysis and that epitope-tagging of a protein does not affect the energetic properties of the protein.

Protein stability in situ

Purification of the protein is always a prerequisite for conducting a rigorous biophysical analysis of a protein. Typical biophysical analyses provide ensemble-averaged measurements and impurities can contribute to these measurements. The investigation of a small amount of H-ras added to a cell lysate is an example of studying a protein in the presence of excess impurities. In the studies described here, H-ras was less than 2% of the total protein in the samples and cannot be studied by conventional biophysical methods. Pulse and Western offers a way to overcome this limitation of ensemble-averaged measurements by monitoring only one protein in the mixture.

Another issue of studying a protein in cell lysates is the possible interactions with other cellular proteins which may affect the thermodynamic stability of the protein of interest. A cell lystate from a mammalian cell line such as the human Jurkat cell line is likely to contain many potential proteins that interact with H-ras, including guanine nucleotide exchange factors (GEFs), GTPase activating proteins (GAPs), and downstream effector proteins. However, the presence of these interacting proteins apparently did not affect thermodynamic stability of H-ras under these experimental conditions.

How does the presence of interacting proteins in the cell lysate not interfere with the stability determination of H-ras? The degree of stabilization from binding is a function of the concentration of the ligand21:

equation image(1)

where ΔΔGunf° is the change in global stability, [L] is the ligand concentration, and Kd is the dissociation equilibrium constant of the complex. To obtain a significant increase in stability, [L] should be much greater than Kd. If [L] = Kd, the stabilization would be only 0.4 kcal/mol, which is within the usual error of protein stability determination. Considering the interactions between H-ras and its binding proteins are intrinsically transient, it is unlikely that [L] is much greater than Kd even in vivo. If [L] ≫ Kdin vivo, H-ras should exist as a stable complex with the binding partner. It has never been suggested that H-ras is a part of a stable protein complex in vivo. Though structures of complexes of H-ras and other proteins have been determined by X-ray crystallography,22, 23 the stable complexes are likely to form only under crystallization conditions with quite high protein concentrations. Moreover, when a lysate is prepared, the cellular components are diluted significantly. Typical concentrations of total effectors in cells are <1 μM and concentrations of GEFs and GAPs are <1 nM.24 When cell lysates are prepared, the concentration of these binding proteins are likely to decrease by several orders and, therefore, the chance to have [L] much greater than Kd is further decreased. Another factor to consider is that we have determined the stability of GDP-bound H-ras, which is the inactive form of the protein. When H-ras is bound to GDP, this protein does not interact with its downstream effector proteins or GAPs. Therefore, GDP-bound H-ras is relatively inert within a cellular environment.

This apparent inertness of the background proteins in H-ras stability determination should not be extrapolated to infer the stability of H-ras in vivo. As discussed above, the environment of cell lysates is clearly distinct from the environment in cells.25 Most of all, the cell is highly crowded with macromolecules whose concentration can be as high as 400 mg/mL.26 This “macromolecular crowding” may have significant consequences for the conformational energetics of proteins in cells.27 Proteins may also interact with many other cellular constituents including proteins, metabolites, nucleic acids and membranes. These interactions may affect the thermodynamic and kinetic stability of proteins. Therefore, it is necessary to discern protein stability determined in situ from protein stability in vivo.

Methods

Protein expression and purification

The C-terminal truncated version of human c-H-ras–(1-171)12 was expressed in E. coli BL21(DE3)pLysS under the control of T7 promoter. Cells were harvested after 3-h IPTG induction and resuspended in 20 mM Tris-HCl (pH 8.0) buffer containing 10 mM EDTA. After lysis by sonication, the inclusion bodies were collected by centrifugation. H-ras was then refolded from inclusion bodies as reported previously with some modifications.10, 28 H-ras in the inclusion body was solubilized in 20 mM Tris-HCl (pH 8.0) buffer containing 10 mM EDTA, 30 mM DTT, and 8M urea. After centrifugation, the solubilized inclusion body solution was added slowly to the refolding buffer (20 mM Tris-HCl, pH 8.0, 50 mM NaCl, 5 mM MgCl2, 1 mM DTT, 30 μM GDP, 1 mM PMSF, and 5% glycerol) and incubated overnight. After filtration, the refolding reaction was loaded onto DEAE Sepharose fast flow column (GE Healthcare, Piscataway, NJ) pre-equilibrated with 20 mM Tris-HCl (pH 8.0) buffer containing 5 mM MgCl2 and 1 mM DTT. Then H-ras was eluted with a linear gradient of NaCl (0–500 mM). Peak fractions containing H-ras were pooled and GDP was added to 30 μM. Subsequently, the pooled sample was loaded onto Superdex™ 75 gel filtration column (GE Healthcare, Piscataway, NJ) pre-equilibrated with 50 mM Tris-HCl (pH 8.0) buffer containing 100 mM NaCl, 5 mM MgCl2, and 1 mM DTT. Using AKTA FPLC system (GE Healthcare, Piscataway, NJ), H-ras was eluted from the column as a single peak. The protein purity was verified by SDS-PAGE. Protein concentration was obtained from the absorbance at 280 nm using an extinction coefficient of 2.4 × 104M/cm.29

Preparation of cell lysates

An E. coli lysate was prepared as reported previously.30 Briefly, E. coli K12 was grown overnight in Luria Bertani (LB) medium and harvested by centrifugation. The cell pellet was resuspended in 20 mM Tris-HCl, pH 8.0 and 50 mM NaCl and then lysed by sonication. The cell lysate was dialyzed initially in 20 mM Tris-HCl, pH 8.0 and 500 mM NaCl and then again in 20 mM Tris-HCl, pH 8.0 and 50 mM NaCl. Membrane fractions were removed by ultracentrifugation.

Jurkat T cell line (E6.1) was grown at 37°C in an atmosphere containing 5% CO2 in Roseland Park Memorial Institute (RPMI) 1640 media supplemented with 7.5% heat inactivated fetal bovine serum,12.5 μM Hepes (pH 7.4), 12 μM sodium bicarbonate, 50 μM 2-β-mercaptoethanol, 50 units/mL penicillin, 50 units/mL streptomycin, and 1 mM sodium pyruvate. The medium was aspirated, and the cells were washed in 1 mL PBS. The cells were lysed by incubating on ice for 15 min in 1 mL of lysis buffer (20 mM Tris-HCl, 50 mM NaCl, 10 μg/mL of aproptinin and leupeptin) containing 1% NP 40. A supernatant was prepared by centrifugation of the lysate and used for further experiments.

Pulse proteolysis of H-ras

Pulse proteolysis was performed as reported previously.1, 19 H-ras (31 μg/mL) was incubated at 25°C for 18–20 h in a buffer containing 20 mM Tris-HCl (pH 8.0), 50 mM NaCl, 5 mM MgCl2, 1 mM TCEP, 10 mM CaCl2, and 30 μM GDP with varying concentrations of urea. When H-ras was incubated in cell lysates, a non-hydrolyzable GDP analogue, GDP-β-S (Sigma) was used instead of GDP to prevent hydrolysis of GDP by endogenous enzymes in cell lysates. After 1-min pulse proteolysis with 0.20 mg/mL of thermolysin, the reactions were quenched by EDTA and loaded onto 15% SDS-PAGE gel, followed by staining with SYPRO Red fluorescent dye (Molecular Probes). The gel was scanned by Typhoon imaging system (GE Healthcare, Piscataway, NJ) and the band intensity was quantified with the image analysis software, ImageJ. The Cm values were determined by fitting the plots of intensity versus urea to a two-state model. Because H-ras is a ternary complex of the apo form of H-ras, GDP, and Mg2+, a rigorous description of unfolding equilibrium requires [GDP] and [Mg2+] as variables.10 However, because [GDP] and [Mg2+] do not change significantly upon unfolding, we have used a simple two-state model instead.13

The relaxation kinetic constant of H-ras was determined by pulse proteolysis as described previously.4 After 60 μg/mL H-ras was prepared in a buffer containing 4.0M urea, 20 mM Tris-HCl (pH 8.0), 50 mM NaCl, 5 mM MgCl2, 1 mM TCEP, 10 mM CaCl2, and 30 μM GDP, the reaction was dispensed into several aliquots. At designated time points, unfolded proteins in aliquots were digested by pulse proteolysis. The remaining intact proteins were quantified by SDS PAGE as described earlier. The kinetic constant was determined by fitting the band intensity change over time to a first-order rate equation.

Western blotting of H-ras

After pulse proteolysis, samples were run on 15% SDS-PAGE gel and transferred to Immobilon-FL PVDF membrane (Millipore) for 1 h at 100 V. ECL Plex system (GE Healthcare, Piscataway, NJ) was used for protein detection method because of the highest dynamic range (3.6 orders of magnitude) and linearity (R2 = 0.998). The blot was blocked in 5% dry milk for one hour and probed with anti-c-H-ras mouse monoclonal antibody (Calbiochem, Gibbstown, NJ) in 1:500 dilution overnight at 4°C. Subsequently, the blot was incubated with Cy5-linked mouse IgG antibody (GE Healthcare, Piscataway, NJ) in 1:2500 dilution for one hour at room temperature. After being washed repeatedly in Tris-buffered saline (TBS) containing 0.1% Tween 20, the fluorescent blot was scanned with Typhoon imaging system.

Acknowledgements

The authors thank Dr. Yu-Ran Na for providing an E. coli lysate. They also thank Dr. Marietta Harrison, Dr. Carol Post, and Jonathan Schlebach for helpful comments on the manuscript.

  • *

    Typically, m-values from pulse proteolysis are not so reliable due to the limited number of data points.3 To determine ΔGunf° with the Cm values from pulse proteolysis, therefore, the m-values are calculated with the number of amino acid residues in proteins based on the known statistical correlation. For H-ras, this statistical estimation of m-values does not seem to be appropriate because binding of GDP to the protein buries significant amount of solvent accessible surface area which cannot be simply estimated by the number of amino acids. Therefore, we used the m-value determined by CD to report ΔGunf°.

Ancillary