Positive Cooperative Mechanistic Binding of Proteins at Low Concentrations: A Comparison of poly (sodium N-undecanoyl sulfate) and Sodium Dodecyl Sulfate

Abstract

The interactions of the negatively charged achiral molecular micelle, poly (sodium N-undecanoyl sulfate) (poly-SUS), with four different proteins using intrinsic and extrinsic fluorescence spectroscopic probes, are studied. A comparison of poly-SUS with the conventional surfactant, sodium dodecyl sulfate (SDS), and the monomeric species, SUS, is also reported. In this work, we observed that poly-SUS preferentially binds to acidic proteins, exhibiting positive cooperativity at concentrations less than 1 mM for all proteins studied. Moreover, it appears that the hydrophobic microdomain formed through polymerization of the terminal vinyl group of the monomer, SUS, is largely responsible for the superior binding capacity of poly-SUS. From these results, we conclude that the interactions of poly-SUS with the acidic proteins are predominantly hydrophobic and postulate that poly-SUS would produce superior interactions relative to SDS at low concentrations in polyacrylamide gel electrophoresis (PAGE). As predicted, use of poly-SUS allowed separation of the His-tagged tumor suppressor protein, p53, at sample buffer concentrations as low as 0.08% w/v (2.9 mM), which is 24 times lower than required for SDS in the standard reducing PAGE protocol. This work highlights the use of poly-SUS as an effective surfactant in 1D biochemical analysis.

1. Introduction

Studies of protein-surfactant interactions are important in numerous aspects such as biochemical, industrial (food and cosmetic), pharmaceutical and in the development of analytical techniques for protein separation and detection [1–3]. Understanding such interactions definitely gives us better insight into protein structure and function [1–8]. Interactions of different classes of surfactants, such as cationic, anionic, zwitterionic, and neutral as well as surfactants with variable alkyl chain lengths and gemini surfactants, have been studied using various model proteins [9–12]. Among the various surfactants, characterization of protein interaction with sodium dodecyl sulfate (SDS) continues to be a long-standing area of ambiguity [2,3,7]. SDS-PAGE (polyacrylamide gel electrophoresis) is an established technique for protein separations [4–8]. Though this technique has been used for a wide variety of protein separations, it exhibits weak resolution for separating hydrophobic proteins or complex protein mixtures [13]. In some cases, a mixture of cationic and anionic surfactants is used to overcome this problem and zwitterionic surfactants are also sometimes considered a better alternative [13, 14]. Protein separations are primarily based on differences in binding affinities of the surfactant to the various proteins in a mixture.

In evaluating the literature on recent investigations of protein-surfactant interactions, it is evident that such studies have almost always been restricted to single or double chain monomeric species [9, 10, 15]. However, to the best of our knowledge interactions of proteins with polymerized surfactants (herein termed molecular micelles) have not been reported. Molecular micelles garner attention because of their unique physicochemical properties in comparison to those of single and double chain surfactants. Therefore, we believe this issue is of practical interest and relevance to using molecular micelles as separation reagents in 1D and 2D gel electrophoresis for hydrophobic proteins and as probes to further our understanding of their solubilization [16]. In the studies reported here, the interactions of four proteins in solution with a molecular micelle, poly (sodium N-undecanoyl sulfate) (poly-SUS), its monomeric species, SUS, and the conventional surfactant, SDS, have been examined using fluorescence spectroscopy and circular dichroism.

Poly-SUS is an amphipathic molecule with an achiral hydrophilic head group and covalently bound hydrophobic tails (Fig. 1). In general, poly-SUS is formed by polymerizing the double bond at concentrations 5 times higher than the critical micelle concentration (CMC) using γ-irradiation. This concentration ensures that a spontaneous self-assembled phase exists. Subsequently, the dynamic equilibrium between the surfactant monomer and the micelle is largely eliminated after irradiation/polymerization. Therefore, molecular micelles do not have a CMC, and their overall stability is not compromised when interacting with proteins. The irradiation process imparts a unique morphology to the molecular micelle through formation of a covalently bound highly hydrophobic micro-domain. In this study, such molecular micellar hydrophobicity has been confirmed by use of the hydrophobic probe 8-anilino-1-napthalenesulfonic acid (ANS), and it is found to be in stark contrast to conventional and second generation surfactants. Thus, it is expected that poly-SUS should have access to greater numbers of sites on proteins that are improbable as a result of the dynamic assembly and disassembly of a conventional micelle. As a result, we hypothesize that protein binding would be achieved at much lower concentrations using a molecular micelle than is achieved with a conventional surfactant such as SDS. Moreover, our group has demonstrated that interactions with molecular micelles provide superior separation schemes relative to conventional micelles due to their improved interactions [17–19]. For example, we have demonstrated that by systematically changing the concentration of poly-SUS in the running buffer, resolution of sixteen polycyclic aromatic hydrocarbons with relatively high efficiency and small k' values are achievable [17].

Figure 1.

Structures of the (a) monomer, sodium undecylenic sulfate (SUS), (b) molecular micelle, poly (sodium N-undecanoyl sulfate) (poly-SUS), and (c) the conventional surfactant, SDS.

Thus, the present work was undertaken to study the mechanism of interaction of four proteins, namely Bovine Serum albumin (BSA), ovalbumin (OVA), α-chymotrypsinogen A (aCHY) and α-lactalbumin (aLAC), with the achiral molecular micelle, poly-SUS and compare these results to interactions with the monomer SUS and the more commonly used conventional surfactant SDS. The focus of this study was to provide a basis for the development of a new analytical tool for use in biochemistry and biotechnology. Our results suggest that poly-SUS indeed exhibits stronger interactions with the proteins under consideration at much lower concentrations, as compared to the monomeric species SUS and SDS. The type of interactions as interpreted using Scatchard analysis is distinctly different, which explains the stronger interactions and higher binding constants of poly-SUS as compared to SDS and SUS. Hence, these observations suggest that due to its unique properties, poly-SUS can serve as a more favorable alternative to SDS in applications such as gel electrophoresis, protein extraction, and solubilization.

2. Materials and Methods

2.1. Materials

Serum albumin (bovine, 66 kDa, BSA, 98%), ovalbumin (egg, 45 kDa, OVA, 98%), α-chymotrypsinogen A (bovine pancreas, 26 kDa, aCHY), α-lactalbumin (bovine milk, 14.2 kDa, aLAC, 85%), p53 (43 kDa) and 8-anilino-1-napthalenesulfonic acid (ANS) were obtained at the highest purity available from Sigma-Aldrich (St. Louis, MO, USA) and used as received. Tris/Glycine buffer was used for all studies since this is the buffer traditionally used in 1D gel electrophoresis. Ultrapure water (18.2 MΩ) was obtained using an Elga PURELAB Ultra water purifier (Lowell, MA, USA). SDS (>98%), Trishydroxymethylaminomethane, and Glycine were obtained from Invitrogen Corporation (Carlsbad, CA, USA) and the Precision Plus Protein All Blue Standard marker was obtained from Bio-Rad Laboratories (Hercules, CA). Polyethyleneglycol (PEG, M.W. 200–1000) were obtained from Polysciences Inc. (Warrington, PA). All chemicals were used as received without further purification.

2.2. Methods

2.2.1. Synthesis of the Molecular Micelle, poly-SUS

Poly-SUS was synthesized according to a procedure previously reported by Warner, et al. [17] and Bergstrom [20]. The critical micelle concentrations (CMC) of monomeric SUS [~25 mM (aq)] and SDS [~8 mM (aq)] were determined by use of surface tension measurements at room temperature with a KSV Sigma 703 digital tensiometer (Fig. S1). Polymerization of SUS was achieved at a concentration of five times the CMC under γ-irradiation using a ⁶⁰Co source. ESI-MS experiments suggest the presence of different molecular weights viz. 295.115, 567.214, 381.315, 567.214, 839.309,1111.405, 1385.5055, 1657.6 corresponding to different species ranging from monomer to hexamer + Na. However, average molecular weight of the polymer as determined by use of viscosity measurement suggests a molecular weight of 534 which is close to a dimer. Viscosity measurements were performed using Aton Paar automated micro viscometer based on falling sphere method which is PC controlled by use of Visolab Filmware software (Aton Par, Austria). PEGs of different known molecular weights between 200–1000 were used as standards. The intrinsic viscosity for each polymer –solvent system was determined from the intersection of a Huggins plot (η_red vs c) and Kraemer plot lnη_rel/c vs c, where η_red is the reduced viscosity and η_rel is the relative viscosity. Then Mark-Howink equation [η]=KM^a was used to obtain K and a, the Mark-Howink constants, using the standard. Using K, a and the intrinsic viscosity ([η]) of the unknown polymer the molecular weight poly-SUS was estimated. Monomeric SUS and polymeric SUS (poly-SUS) were characterized by use of ¹H NMR (Fig. S2, supporting material) in deuterium oxide (D₂O) on a Bruker–250 MHz instrument. Complete polymerization was confirmed by observing the disappearance of the NMR chemical shift signals (5.0 – 6.0 ppm) associated with the terminal vinyl group. All poly-SUS solutions are reported using the equivalent monomer concentration, namely calculations were based on the molecular weight of the individual surfactant unit (i.e., SUS, 272 g/mol).

2.2.2. Instrumentation

Fluorescence spectra were recorded at 25 °C using a SPEX Fluorolog-3 spectrofluorimeter (Jobin Yvon, Edison, NJ) equipped with a 450-W xenon lamp and R928P photomultiplier tube (PMT) emission detector. A quartz cuvet with an optical pathlength of 1 cm was used and bandwidths for both the excitation and emission monochromators were set at 3 nm unless otherwise stated. Excitation was performed at 295 nm (Trp) and 364 nm (ANS), while emission spectra were respectively measured in the ranges of 335 – 360 nm and 460 – 525 nm. Fluorescence spectra reported herein were obtained from proteins at concentrations of 1 mg/ml in 25 mM Tris/192 mM Glycine, pH 8.4 unless otherwise indicated. Circular dichroism (CD) data were obtained using an AVIV Model 62DS (AVIV Associates, Lakewood, N.J.) spectrophotometer at 25 °C fitted with a 1 mm pathlength quartz cell. The CD spectra of native protein samples in 25 mM Tris/192 mM Glycine, pH 8.4, were acquired at concentrations that produced optimal CD signal. The Tris/Glycine buffer was filtered through a 0.45 μm nylon filter prior to sample preparation. All CD scans were conducted in triplicate in the far UV (200–240 nm) and near UV (240–320 nm) regions of the spectrum, respectively, and average spectra were recorded. All CD spectra were also corrected for background intensity of the buffer. The CD response is reported as ellipticity and displayed in units of millidegree (mdeg).

2.2.3. Determination of Binding Parameters of SDS, SUS, and polySUS to Protein

A spectrophotometric titration procedure was used to determine the characteristic binding parameters of poly-SUS, SUS, and SDS interacting with the four proteins employed in this study. The proteins (1 mg/ml) were allowed to equilibrate for thirty minutes with a range of concentrations of poly-SUS, SUS, and SDS (0 – 20 mM) in 25 mM Tris/192 mM Glycine, pH 8.4 at 25 °C. The binding isotherms, stoichiometries, and dissociation constants were determined employing Scatchard Analysis [21,22]. In biological systems where a ligand, L, binds to a receptor (macromolecule), Scatchard analysis [21,22] is typically used to determine the regions of binding in the isotherm, the binding constant for each region, and the number of ligand binding sites. The various parameters characteristic of such analyses were determined as described below:

$$\text{Fraction of surfactant bound},\alpha =(I-I_0)∕(I_m-I_0)$$ (1)

$$\text{The concentration of the bound surfactant}{S}_b=\alpha \left[\mathit{Total}\mathit{surfactant}\right]$$ (2)

where I₀ is the fluorescence intensity of the protein in the absence of poly-SUS (or SUS or SDS), I is the fluorescence intensity when the protein and poly-SUS (or SUS, SDS) are in equilibrium, and I_m is the fluorescence intensity when the protein is completely saturated with poly-SUS (or SUS, SDS). The concentration (in M) of free poly-SUS ([free poly-SUS]), was determined by 1- [bound poly-SUS]. The parameter, ν, is defined as α[Total surfactant]/[Total protein] and the concentration of free surfactant (c) was obtained from [Total surfactant](1-α). Each linear portion of a Scatchard plot (ν/c vs ν) was given a linear fit and the equilibrium binding constant (K) and number of binding sites (n) for a particular concentration region were obtained from the slope and intercept respectively.

2.2.4. Gel Electrophoresis

Instrumentation

A Bio-Rad Laboratories Mini-PROTEAN 3 Electrophoresis Module was used for PAGE separations (Hercules, CA, USA). A constant voltage of 200 V was applied for each separation by a 1000 V Bio-Rad power supply. During staining and destaining, gels were placed in plastic containers and set on a rocker (Midwest Scientific, St. Louis, MO, USA). Typical staining and destaining times for SDS-PAGE were used. The protein bands were analyzed for each gel using a Kodak Gel Logic 200 Image Analyzer (Rochester, NY, USA).

Preparation of Sample and Running Buffers

Standard 10× running buffer (RB) stock solution contained 25 mM Tris and 192 mM Glycine (pH 8.4). The RB solution was prepared by measuring an appropriate amount of poly-SUS (or SDS) into a volumetric flask, dissolving it with 50 mL of electrode buffer stock solution, and diluting it to a final volume of 500 mL with ultrapure water (18.2 MΩ). Desired pH values of the electrode buffer were achieved by the drop-wise addition of either 1 M NaOH or 1 M HCl. All poly-SUS solutions were prepared by using the equivalent monomer concentration, namely calculations were based on the molecular weight of the individual surfactant unit (i.e. SUS, 272 g/mol). Concentrations of 0.0125%, 0.025%, 0.0375%, and 0.053% w/v poly-SUS were used in the running buffer for the optimization and validation separations. The sample buffer (SB) was prepared in 1.7 mL eppendorf tubes by combining appropriate amounts of ultrapure water, 50 mM Tris HCl, pH 6.8, glycerol, bromophenol blue, and 0.078%, 0.156%, 0.313%, 0.625%, 1.25%, 2.0%, or 2.51% w/v poly-SUS (or SDS). Model proteins of equal concentration were added in the SB at a protein:SB ratio of 2:1. The reducing agent, β-mercaptoethanol, was added at 5% v/v of the SB.

Electrophoretic Separation

Running buffer totaling 325 mL was loaded into the upper and lower chambers of the Mini-PROTEAN 3 module. Each sample was heated at 95 °C for 5 minutes on a dry bath incubator from Fisher Scientific (Pittsburgh, PA, USA) unless noted otherwise. Dry bath incubator temperatures were adjusted during optimization of the poly-SUS separation protocol. Twenty microliters (20 μL) of sample was loaded into each well of the 4–20% Tris HCl gradient mini gels. When SDS was in the sample, a wide range SDS marker (6.5 – 205 kDa) from Sigma-Aldrich (St. Louis, MO, USA) was used. The migration time was less than thirty-five minutes for all separations. After each separation, gels were rinsed with ultrapure water (18.2 MΩ), stained with approximately 25 mL of Colloidal Blue Stain, and placed on a rocker. Gels were destained with ultrapure water (18.2 MΩ) until a clear background was visible.

3. Results and Discussion

3.1. Intrinsic Fluorescence Spectroscopy

Intrinsic Trp fluorescence of the protein was used to determine the relative binding properties of the three ligands, poly-SUS, SUS, and SDS, to the four proteins. The intrinsic fluorescence of a protein, contributed by the aromatic amino acids of Trp, Tyr, and Phe, is often used to study the mechanism of ligand binding to peptides and proteins [23].

All four proteins have intrinsic Trp residues in the native state, which are either buried in hydrophobic pockets or located toward the outer surfaces, as indicated in Table 1. Trp excitation at 295 nm was used to minimize excitation of tyrosine residues and subsequent heterotransfer to Trp [24]. In the absence of poly-SUS, SUS, or SDS, the intrinsic Trp fluorescence of BSA, OVA, aCHY, and aLAC displayed typical emission maxima (i.e., λ_max) of 352 nm, 346 nm, 338 nm, and 346 nm, respectively (see Fig. S3). Increasing the concentrations of poly-SUS, SUS, or SDS in the presence of the four proteins resulted in various fluorescence emission responses and emission maxima shifts (Fig. 2, Fig. S4 & S5). The variability in emission responses was expected due to heterogeneity in the number of Trp residues and their locations in each native protein. For example, BSA and aCHY have two (1 solvent accessible, 1 buried) and eight (6 solvent accessible, 2 buried) Trp residues, respectively. With increasing concentration of poly-SUS, the Trp emission of aCHY was gradually quenched which was accompanied by a red shift until a point of saturation was reached, where further increases in [poly-SUS] concentration did not result in additional quenching or shifting of emission maxima (Fig. 2A–2C) For BSA, with increasing poly-SUS (or SUS or SDS) concentration, quenching was observed in the Trp emission while an initial blue shift was followed by a continual red shift in emission wavelength maxima. Similarly, saturation of BSA with poly-SUS (or SUS or SDS) coincided with no further Trp quenching or red shifting in the emission wavelength maxima (Fig. 2D–2F). The red shift and quenching (data not shown) in the Trp emission maxima of BSA and aCHY with increasing concentration of the ligand are attributed to changes in the native conformation of the proteins [10, 25]. Such a conformational change induced by the binding of ligand to the protein suggests leads to exposure of Trp residues to a relatively hydrophilic microdomain [23]. Similar changes in the intrinsic Trp fluorescence for the other two proteins, OVA and aLAC (S4 and S5). In the presence of all three surfactants, both OVA and aLAC Trp exhibited a red shift, which is indicative of exposure to a more hydrophilic environment. The shifts were followed to understand the binding-associated conformational changes in the presence of these surfactants.

Table 1.

Physical properties of proteins studied.

Protein	Accession #	MW (kDa)	Trp*	% α-helix†	% beta sheet†	Theoretical pI‡	Theoretical Charge at pH 8.4‡
α-lactalbumin (aLAC)	P00711	14.7	4 (2/2)	43	11	5.0	−12
α-chymotrypsinogen A (aCHY)	P00766	26	8 (6/2)	14	32	8.2	−2
ovalbumin (OVA)	P01012	45	3 (1/2)	32	32	5.3	−16
albumin, bovine serum (BSA)	P02768	66	2 (1/1)	70	-	5.9	−37

^*Tryptophan (Trp) residues that are solvent accessible (sa) and buried (b) are indicated as (sa/b).

^†The secondary structure data was obtained from http://www.pdb.org

^‡The theoretical values were obtained from http://www.scripps.edu/~cdputnam/protcalc.html.

Figure 2.

Fluorescence wavelength maxima shifts of Trp_aCHY and Trp_BSA in the presence of increasing monomeric concentration (0 – 20 mM) of SDS (open circles), poly-SUS (open diamonds), and SUS (open triangles) in association with aCHY (38 μM, A–C) and BSA (15 μM, D–F), respectively, determined by steady state fluorescence (λ_ex = 295 nm, 25°C). The lines have been included to guide the eye.

3.2. Binding Studies and Scatchard Analysis

Analyses of Scatchard plots reveals the type of binding, particularly when multi-site ligand binding is suspected [26]. Generally, the binding isotherm displays four characteristic regions with increasing surfactant concentration: (1) specific binding to high energy sites on the protein, which are believed to be electrostatic, (2) noncooperative association, (3) cooperative binding as evidenced by a marked increase in binding and where protein unfolding is believed to occur, and (4) saturation in which no further binding occurs and micelles co-exist with the saturated protein [8,27]. Moreover, distinct differences in binding, namely positive or negative cooperative binding of the ligand L (or surfactant) to the target, are obtainable [28].

With regard to the binding studies reported here, it was observed that binding of poly-SUS to BSA occurred in a significantly lower concentration regime as compared to SUS and SDS. Interestingly, for BSA at concentrations between 0 and 0.8 mM, the slope of the fraction bound curve (see inset of Fig. 3) for poly-SUS is observed to be 8.4 (≈ 2³) and 16.6 (≈ 2⁴) times that for SDS and SUS, respectively, suggesting that poly-SUS binding to BSA is highly cooperative. Thus, a surfactant with one more methylene group (SDS) binds BSA twice as fast as a surfactant with one less carbon (SUS), while overall binding is exponentially greater when the terminal double bond of the monomer is polymerized (poly-SUS). At concentrations as low as 0.8 mM, the fraction of poly-SUS bound to BSA reached 90% while it required five times more SDS and twelve times more SUS to reach the same bound fraction (Fig. 3).

Figure 3.

Fraction bound of poly-SUS (solid diamonds), SUS (open triangles), and SDS (open circles) to BSA (15 μM) with increasing monomeric surfactant concentration (0 – 20 mM).

A similar trend was observed for aCHY, where the slope of the fraction bound curve at low concentrations was greater for poly-SUS ([poly-SUS] < 1.0 mM) than SDS or SUS. However, the corresponding slope for aLAC was similar for poly-SUS and SDS and the slope for OVA was greater for SDS than for poly-SUS (data not shown). It is evident from Fig. 2 and 3 that the binding of poly-SUS to BSA and aCHY are significantly better as compared to SDS and SUS. These results suggest that the saturation binding points were attained at much lower concentrations (~ 1 mM) with poly-SUS, while saturation for the other two surfactants was attained at remarkably higher concentration (~4.7 mM for SDS and ~10 mM for SUS) which corresponds to their CMC in this medium (determined experimentally by tensiometry) [29]. Furthermore, we believe that the better binding performance of poly-SUS in the presence of BSA and aCHY as compared to the other two surfactants is due to the absence of any transition from a monomeric species to a low-aggregated state followed by a full micellar state to saturate the proteins [10]. Similar binding behavior was observed with aLAC. However, with OVA, the binding was relatively weaker for SDS as compared to poly-SUS. This is attributed to the significantly lower surface hydrophobicity of OVA [30]. This observation again suggests that the binding of poly-SUS with proteins are assisted primarily through hydrophobic interactions.

Examination of the Scatchard plots for binding of poly-SUS, SDS, and SUS clearly suggests that the binding mechanism of these three surfactants to the various proteins studied is significantly different from each other. Different characteristics of the Scatchard plots in different concentration region suggest that the binding of these surfactants to the four different proteins follows separate mechanisms in various concentration regions (Table 2).

Table 2.

Cooperative binding type for poly-SUS, SUS and SDS in the low concentration region (< 1.0 mM).

Surfactant		BSA	OVA	aCHY	aLAC
pSUS	Region1	sp, +	sp, +	sp,+	sp,+
	Region2	+	−	−	−
	Region3	−	+	−	−
	Region4	n.sp	n.sp	n.sp	n.sp
SUS	Region1	sp, +	sp	sp	sp, −
	Region2	−	−	+	−
	Region3	n.sp	n.sp	−	−
	Region4	n.sp	n.sp	n.sp	−
SDS	Region1	sp	sp, −	sp,+	sp,−
	Region2	−	−	−	−
	Region3	+	−	+	n.sp
	Region4	n.sp	n.sp	n.sp	n.sp

Sp=specific binding, + = positive cooperative binding, − = negativecooperative binding, n.sp= non specific binding

The analysis (Fig. 4 and 5, Tables 2 and 3) revealed that poly-SUS followed a highly cooperative binding mechanism especially in the low concentration regions, wherein it was either unbound or fully bound over small changes in concentration spending little time in partially bound intermediate states [31]. The observed co-operativity may also be attributed to raveling and unraveling capabilities of the molecular micelle depending on the environment (Scheme S1). This was presumably due to polymerization of the terminal double bond. According to these studies, the binding mechanism of poly-SUS to BSA is in direct opposition to what has been observed for the conventional anionic surfactant SDS. The concave downwards nature of the Scatchard plot with very high binding constant suggested positive cooperative binding of poly-SUS in the low concentration regime (0.2 < [poly-SUS] < 0.8 mM). The binding constant was found to be two orders of magnitude greater than that observed for SDS in the same concentration range. Thus, below 2.7 mM the binding of poly-SUS to one site on BSA increases the poly-SUS binding affinity to subsequent sites of the protein. The negative cooperative binding of SDS to BSA in the similar low concentration region (0.2 < [SDS] < 0.8 mM) was evident from the concave upwards nature of the Scatchard plot. Typically observed in this low concentration region, negative cooperative binding occurs when the surfactant monomer binds to a site on the protein with no allosteric effects present to facilitate additional ligand binding [32]. Thus, SDS binding is highly specific in the low concentration regime (0.2 < [SDS] < 0.8 mM). Above 2.7 mM, poly-SUS exhibited a negative cooperative binding suggesting that in this concentration region, BSA was already saturated with poly-SUS and thus binding at one site lowers the binding affinity at adjacent sites. However, at higher concentrations (3.4 – 5.4 mM), positive cooperative binding was observed for SDS followed by a linear region at higher concentrations characteristic of non-specific binding [21]. The highest binding constant for SDS was calculated to be 4.5 × 10⁴ M⁻¹ in region III (4.1 – 5.4 mM) where the onset of SDS micellization occurs. The binding of SDS agreed with the isotherm observed by Takeda et al. (1981) who used conductometric and chromatographic methods to study the binding events for SDS associating with BSA [33]. The high n values observed in certain regions especially for poly-SUS binding to proteins suggests substantial unfolding of the protein due to the already bound surfactants leading to exposure of additional binding sites. However relatively low binding constants in these regions may be attributed to steric factors and other thermodynamic factors associated with the binding [34].

Figure 4.

Scatchard Plots of BSA with (A) SDS and (B) poly- SUS

Figure 5.

Scatchard Plots of OVA with (left) SDS and (right) poly- SUS. The inset expands the low concentration regions of the corresponding plots.

Table 3.

Scatchard Analysis data for the interaction of the four proteins with the surfactants.

Protein	poly-SUS			SITS			SDS
	Concentration Range (mM)	K (M−1)	n	Concentration Range (mM)	K (M−1)	n	Conce n-tration Range (mM)	K (M−1)	n
BSA	0.2–0.8	1.30×104	0.4	0.2–0.8	3.65×102	227.7	0.2–0.8	6.80×102	67.3
	1.0–2.7	4.00×104	23.7	1.0–4.4	5.90×101	9950	1.0–3.4	2.40×103	0.63
	3.4–4.7	8.00×102	1997.0	4.7–8.0	2.73×102	74.9	4.1 –5.4	4.50×104	5.3
	5.0–12.0	2.80×103	21.9	10.0–16.0	1.19×103	481.1	6.0–16.0	3.71×103	81.7
OVA	0.2–1.0	1.42×103	0.4	0.7–1.0	4.01×102	8.5	0.2–0.5	2.36×103	0.1
	1.7–2.7	1.07×103	2.2	1.7–3.4	2.60×102	17.4	0.7–1.0	1.64×103	0.1
	3.4–8.0	2.85×103	28.5	4.1–5.1	4.24×102	29.3	1.7–2.7	8.58×104	0.5
	10.0–18.0	8.14×103	43.7	5.4–7.0	1.28×103	76.1	3.4–4.4	2.65×104	3.97
							5.1–12.0	6.63×104	3.09
aCHY	0.2–0.5	2.31×103	1.3	0.2–0.5	2.01×103	0.7	0.2–0.5	2.20×103	0.2
	0.7–0.1	2.93 ×103	0.3	0.7–1.0	1.16×103	2.9	0.7–1.0	1.07×103	3.2
	1.7–3.4	5.38×102	173.9	1.7–3.4	7.27×102	11.2	1.7–3.4	9.61×102	1.0
	4.1–6.0	8.45×102	41.6	4.1–5.1	6.35x×102	28.6	4.1–6.0	2.57×103	20.6
	8.0–12.0	1.75×103	43.6	5.4–8.0	5.74×102	120.3
aLAC	0.5–1.0	2.24×103	0.7	0.2–1.0	1.24×100	0.31	0.2–0.5	2.44×102	1.7
	1.7–4.4	6.81×102	148.7	1.7–3.4	8.43×100	13.4	0.7–2.2	4.80×101	7.2
	4.7–8.0	1.35xlO3	53.3	4.1–5.1	3.08×101	36.4	2.7–4.1	1.06×100	2.7
	10.0–16.0	2.15xlO3	113.4	5.4–8.0	1.58×101	95.0	5.4–8.0	2.30×103	0.2

By comparison of the cooperativity profiles for all of the proteins studied (Table 2), it is apparent that positive cooperativity dominated for poly-SUS in the low concentration region (0 < [poly-SUS] < 1.0 mM). This observation suggests maximum molecular micelle complexation with each protein at lower concentrations. The positive cooperative binding to all four proteins (Figures 5 and S5) in this concentration range suggests that poly-SUS binding to the proteins studied is independent of protein size and charge. We attribute this phenomenon to the specificity (absence of micelle assembly and disassembly due to the covalently bound micellar structure), flexibility (ability to adopt different conformations), and hydrophobicity (large hydrophobic microdomain) that exists in poly-SUS, giving it access to more binding sites and thus promotion of positive allosteric modulation [32]. Also, the observed positive cooperativity at such low concentrations may be attributable to the greater number of anionic head groups on the polymerized molecular micelle as compared to its monomeric species. Summarizing the results from the intrinsic fluorescence data, poly-SUS binding to the proteins studied is (1) more cooperative than SDS and SUS, (2) occurs at much lower concentrations (<1.0 mM) than SDS (> 4.1 mM) and SUS (> 10 mM) for two of the proteins, and (3) occurs through a primarily hydrophobic interaction under the experimental conditions explored.

For the number of binding sites, values of 24 and 5 were found on BSA for poly-SUS and SDS, respectively. Although the number of SDS binding sites is in agreement with previous reports [34,35], it is significantly lower than the number of binding sites found for poly-SUS. The large number of poly-SUS binding sites on BSA offers an explanation for the efficiency with which poly-SUS was observed to interact with BSA (at concentrations as low as 0.8 mM). In addition, the amount of poly-SUS bound to BSA as compared to SDS at concentrations < 1.0 mM may be explained by the large number of sites made available for binding by poly-SUS-induced conformational changes to BSA (see section 3.4.).

3.3. Extrinsic Fluorescence Spectroscopy

In addition to intrinsic fluorescence spectroscopy, the extrinsic probe, 8-anilino-1-naphthalene-sulfonate (ANS), has been used to monitor surfactant saturation concentrations and surfactant-induced unfolding of the four proteins [34]. ANS is a well known hydrophobicity probe whose fluorescence is extremely weak in hydrophilic environment and increases abruptly as it migrates to a hydrophobic environment such as the hydrophobic patches within proteins. These changes in fluorescence of the external probe are used as a tool to understand the surfactant binding to the proteins.

In this study, fluorescence data were collected for excitation wavelengths at 295 nm and 364 nm representing Trp and ANS, respectively. ANS is both hydrophobic, i.e. containing a naphthalene moiety, and hydrophilic, i.e. possessing an anionic sulfonate head group. An increase in ANS emission accompanied with blue shift was observed in the presence of each protein due to hydrophobic interaction and subsequent energy transfer from Trp. ANS is known to bind to the hydrophobic sites of proteins and exhibit enhanced fluorescence emission due to Forster Resonance Energy Transfer (FRET) from Trp residues of proteins [34]. The extent of FRET depends on the proximity of ANS to the Trp residues. The greatest intensity increase resulted from the interaction with BSA (see Fig. S6). This was attributed to efficient FRET from Trp residues in BSA to bound ANS suggesting that these two molecules were in close proximity (< 10 nm). However, little energy transfer was observed from Trp_aCHY to ANS which suggested the inaccessibility of the ANS to the Trp.

As poly-SUS (or SDS or SUS) was titrated into the cuvet containing 1 mg/ml of protein and 20 μM ANS and allowed to equilibrate, we envisaged ANS amphipathic contacts would be replaced by poly-SUS (or SDS or SUS) amphipathic contacts on the proteins. This explanation was complementary to the observation that the ANS emission dropped on addition of polySUS (or SDS or SUS) as it was replaced by polySUS (or SDS or SUS) from the hydrophobic patches of the protein due to stronger hydrophobic interactions of the later. In addition, the ANS emission maximum red shifted indicating ANS was moving to a more hydrophilic environment. After addition of a certain amount of poly-SUS (or SDS or SUS), ANS fluorescence was again found to increase accompanied with a blue shift. When the protein-surfactant interaction was complete, we suppose that ANS would be found either in the hydrophobic core of the molecular micelle (poly-SUS), in the bulk aqueous environment due to its hydrophilicity, or sterically situated between the two environments. An increase in the ANS emission after the saturation point (concentration of surfactant for which the protein binding sites are saturated) suggested that ANS once again migrates to a relatively hydrophobic environment. This signature in ANS fluorescence emission is used to obtain the saturation binding concentrations of a surfactant to a protein. The concentration of poly-SUS (or SDS or SUS) coinciding with complete ANS displacement was believed to be the saturation concentration of poly-SUS (or SDS or SUS), interacting with the protein since excess poly-SUS (or SDS or SUS) was then available for hydrophobic interaction with ANS. The saturation concentrations determined by extrinsic fluorescence spectroscopy (Fig. 6) were consistent with the values obtained from intrinsic Trp fluorescence of the proteins studied. For example, the saturation concentration of poly-SUS, SUS, and SDS for interaction with aLAC was observed to be 1.1 mM, 2.9 mM, and 4.1 mM, respectively using extrinsic fluorescence which is close to the values obtained using intrinsic fluorescence (1.0 mM, 2.9 mM, and 4.4 mM, for poly-SUS, SUS, and SDS respectively). Thus, it was confirmed that poly-SUS binds to aLAC at concentrations at least four times less than SDS. At pH 8.4, all of the proteins should have a net negative charge (Table 1). Therefore, the interaction of ANS with the four acidic proteins was expected to be primarily hydrophobic, although electrostatic contacts (e.g., with lysine, arginine, or histidine) are possible, but to a lesser extent. When taken in aggregate, analyses of the results using the extrinsic reporter molecule indicate binding of poly-SUS to the proteins studied is primarily hydrophobic.

Figure 6.

Comparison of saturation concentrations (mM) for poly-SUS, SUS, and SDS associating with the proteins, BSA, OVA, aCHY, and aLAC (1 mg/ml), determined from intrinsic and extrinsic fluorescence spectroscopy. Saturation was not attained for SUS/BSA using the extrinsic probe; therefore, no concentration is indicated.

Summarizing the results thus far, we conclude that the interaction between the acidic proteins and poly-SUS, SDS, and SUS would be primarily hydrophobic due to the hydrophobicity of the core of poly-SUS (or hydrophobicity of the aliphatic chain in the conventional micelle). Another noticeable observation from these studies was a greater increase in ANS emission in the presence of poly-SUS over SDS, which suggests that poly-SUS is significantly more hydrophobic than SDS (Fig. S7).

3.4.Effect of poly-SUS on Protein Denaturation: Circular Dichroism Studies

To evaluate the effect of poly-SUS on protein conformation as compared to the conventional surfactant SDS, circular dichroism measurements were used to monitor changes in the secondary structure. Far UV CD spectroscopy is the characteristic region of the electromagnetic spectrum where secondary structure transitions in proteins are gauged. In this region, the peptide backbone is the chromophore; a protein that is primarily α-helical in structure exhibits two negative bands at 208 nm and 222 nm. Alterations in the far UV spectra for aLAC in the absence and presence of poly-SUS and SDS are shown in Fig. S8. A greater predominant shift in the minimum at 208 nm is observed for poly-SUS relative to SDS in the presence of aLAC at a concentration of ≈ 4 mM (or 0.12 %w/v). No appreciable differences in CD spectra were observed for poly-SUS and SDS at ≈ 2.9 mM (or 0.08 %w/v). In principle, the CD spectrum of a protein is the sum of percentages of all possible secondary structural motifs (i.e., α-helix, β-sheet, β-turn, and random coil). Thus, as the 208 nm minimum (−θ_208nm) increases with increasing concentration of poly-SUS, aLAC begins to unfold. The shift in the CD signal for the 208 nm minimum at ≈ 4 mM (or 0.12 %w/v) suggests a transition in aLAC conformation toward random coil in the presence of poly-SUS [36]. Additional evidence for the poly-SUS-induced secondary structural change in aLAC was observed upon addition of greater amounts of poly-SUS, resulting in dramatic shifts in the minimum at 208 nm relative to SDS (data not shown). In addition, when binding to BSA in either the native or reduced states, poly-SUS perturbs the secondary structure causing significant unfolding (Fig. 7, A–D) as compared to SDS. To further probe the concentration effect of poly-SUS on the secondary structure change in BSA, we have monitored the change in ellipticity at 208 nm. In the concentration range from 0 – 25 mM the ellipticity at 208 nm (−θ_208nm) for poly-SUS increased dramatically compared to SDS (see Fig. S9) suggesting decreased α-helical content and significant unfolding of BSA. Though an increase in ellipticity was observed at 222 nm, the trend was not as pronounced as at 208 nm. Over the concentration range studied, SDS did not appreciably change the conformation of BSA. From these data, one can deduce that poly-SUS overcomes the low dielectric constant of the hydrophobic interior³⁷ to induce major conformational changes. This exponentially increases the capacity of poly-SUS to bind to the interior sites complementing its ability to bind the solvent accessible hydrophobic regions. This is also complementary to the Scatchard analysis data where we find that the binding constant of poly-SUS with aLAC is orders of magnitude higher than SUS and SDS, suggesting stronger binding of poly-SUS at significantly lower concentration as compared to the conventional surfactants.

Figure 7.

Changes in secondary structure of BSA in the native (A) and (C) and reduced (B) & (D) states in the presence of 0.156 % w/v poly-SUS and 0.156 % w/v SDS as indicated. The buffer was 25 mM Tris/192 mM Glycine at pH 8.4 and 25° C. A 1 mm pathlength quartz cuvette was used.

In summary, the present work emphasizes the critical interactions that contribute to the binding mechanism involving a molecular micelle and four proteins with the aim of contributing to the growing need for new and enhanced analytical tools for biochemistry and biotechnology. By use of intrinsic and extrinsic fluorescence spectroscopy, we have determined that the molecular micelle, poly-SUS, exhibits a positive cooperative binding mechanism for all proteins studied and binds two of the proteins at lower concentrations than SUS and SDS. Further, our circular dichroism results show that poly-SUS disrupts the secondary structure of BSA and aLAC at low concentrations, which is in stark contrast to SDS.

3.5. Comparison of poly-SUS and SDS in Polyacrylamide Gel Electrophoresis

The fluorescence and CD studies clearly suggest the distinctive binding property of poly-SUS with the proteins studied, irrespective of their size. We initially postulated that poly-SUS would perform superior to SDS at low concentrations in polyacrylamide gel electrophoresis due to greater hydrophobic interactions and greater binding constants observed from our spectroscopic data. When studied in 1D-gels, poly-SUS indeed produced separations at significantly lower concentration as compared to SDS. To verify this unique behavior of poly-SUS, a His-tagged protein, p53, containing long regions of unordered structure [38] was selected as the analyte of interest. This protein was selected because it is a major target for anticancer therapy due to mutations that alter its ability to contribute to tumor suppression [39, 40]. Facile detection of poly-SUS, SUS, and SDS binding efficiency to p53 was evaluated in the porous gel medium. Representative PAGE-SDS and PAGE-poly-SUS separations are shown in Fig. 8. Migration of p53 through the gel was influenced by the ligand binding efficiency as judged by the location of the band in the gel. In a gradient gel, which was used here, the stacking region of the gel contains large pores that are capable of accommodating large aggregates whereas the resolving region of the gel contains progressively smaller pores as the gel is traversed. It was observed (Fig. 8A) that at a loading of 0.08% w/v (2.9 mM) SDS in the sample buffer a trail of p53 was essentially smeared down the lane presumably due to a lack of binding with SDS. We propose that the water molecules that solvate the hydrophilic residues on the protein surface began interacting with the hydrophilic head groups of the SDS monomers and micelles, causing a reduction in protein-protein repulsive forces and subsequent aggregation or precipitation of p53. The SUS monomer separation exhibited bands that were diffuse (see Fig. S10). Examination of this gel suggests lower protein binding with the addition of one degree of unsaturation in the molecule, which is consistent with the intrinsic fluorescence results. From the work by Sprague et al. [41], we know that the presence of unsaturation originating from the terminal double bond in SUS increases the CMC by a factor of two over its saturated counterpart. Thus, micellar species with widely differing aggregation numbers interacting with the protein may contribute to the diffuse bands and streaking. Conversely, poly-SUS binding to p53 is highly efficient down to concentrations in the sample buffer as low as 0.08% w/v (2.9 mM) as observed in Fig. 8B. Thus, it was concluded from these data, as predicted by results from our fluorescence spectroscopy studies, that poly-SUS has robust association with water soluble proteins like p53 that is attainable at concentrations as high as 2% w/v and as low as 0.08% w/v. This robustness is attributed to polymerization of poly-SUS, having conformational flexibility, and having highly hydrophobic microdomains providing it with access to sites on the protein that are not attainable with a conventional micelle. Collectively, these PAGE separations are consistent with our hypothesis that poly-SUS at low concentrations would exhibit greater binding efficiency and separation effectiveness in PAGE as compared with SDS.

Figure 8.

PAGE of p53 on a 4–20% gradient gel using (A) the conventional surfactant, SDS, and (B) the molecular micelle, poly-SUS. (A) SDS: lane 1, precision plus molecular weight marker (MWM, 10–250 kDa); lane 2, 2% w/v SDS; lane 3, 0.08% w/v SDS. (B) poly-SUS: lane 1, 2% w/v poly-SUS; lane 2, 0.08% w/v poly-SUS; lane 3, BSA standard and 2% w/v poly-SUS; lane 4, OVA standard and 2% w/v poly-SUS. Lanes 3 and 4 were used as markers to identify the approximate molecular weight of p53 in the poly-SUS separation.

4. Conclusions

Detailed examination of the data reported herein show that poly-SUS is a robust surfactant that is facile, flexible, and hydrophobic with high affinity for globular proteins over an appreciably low concentration range. The salient positive cooperative binding mechanisms of the four proteins with poly-SUS as compared to the mostly negative co-operative binding mode with SDS and their poly-SUS binding affinities at low concentration are striking outcomes of the present study. These properties were understood through coupling of data from intrinsic and extrinsic fluorescence spectroscopy, circular dichroism, and polyacrylamide gel electrophoresis, thereby revealing potential applications of molecular micelles in protein separations over the conventional surfactants.^{9–12, 43–45} Extrapolating from our data, we propose molecular micelles as solubilizing agents for hydrophobic proteins. The extraordinary hydrophobicity and other structural features such as specificity achieved through the molecular micelle formation of the monomeric surfactant molecules as compared to the dynamic micellar self-assembly makes them highly competent candidates in proteomics. Moreover, results from the data presented here suggest that water soluble, hydrophobic, amino acid-based molecular micelles may exhibit greater recognition and higher binding affinity of proteins, which may facilitate enhanced solubilization, separation, and identification in proteomics studies. Preliminary gel electrophoresis studies also reveal better performance of poly-SUS as compared to SDS at low concentrations. Further studies are ongoing in our laboratory to exploit the application of this and other molecular micelles in 1D and 2D gels as a significantly better alternative to existing surfactants.^{9–12, 43–45}