University of Southern California Dissertations and Theses

Controlling the form-dynamics-function relationship of proteins with light illumination

(USC Thesis Other)

Controlling the form-dynamics-function relationship of proteins with light illumination

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content CONTROLLING THE FORM-DYNAMICS-FUNCTION RELATIONSHIP OF PROTEINS WITH LIGHT ILLUMINATION by Shao-Chun Wang ____________________________________________________________________ A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (CHEMICAL ENGINEERING) May 2008 Copyright 2008 Shao-Chun Wang ii Dedication The work is dedicated to my beloved family. iii Acknowledgments I would like to thank my advisor Dr. C Ted Lee for his support and encouragement. I also thank Dr. Richard Roberts, Dr. Katherine Shing, Dr. Philip Stephens, and Dr. Ping Wang for their insightful thoughts, discussions and supports in my research. Many thanks go to Dr. Antonio Faraone and Dr. Boualem Hammouda at NCNR, who gave precious advices and instructions on protein dynamics measurements and analysis. I thank Tao Wei for his help on protein dynamics analysis. Next I would like to thank my colleagues and also friends, Anne Laure Le Ny, Andrea Hamill, and Jing (Kiki) Zhang, who always gave me great support by being with me and sharing their thoughts. I would like to thank my dearest parents for their love, support, and guidance over the years. Finally, I would like to thank my boy friend, Fu-Hsuan (Sean) Chiu for giving his support and encouragement constantly. iv Table of Contents Dedication Acknowledgements List of Tables List of Figures Abstract Chapter 1 Introduction 1.1 Protein Form, Function and Dynamics 1.2 Photoresponsive Surfactant 1.3 Bovine Serum Albumin (BSA) 1.4 Lysozyme Chapter 2 Protein Secondary Structure Controlled with Light and Photoresponsive Surfactants 2.1 Abstract 2.2 Introduction 2.3 Experimental Details 2.4 Results and Discussion 2.5 Conclusion 2.6 Acknowledgments Chapter 3 Enhanced Enzymatic Activity through Photoreversible Conformational Changes 3.1 Abstract 3.2 Introduction 3.3 Experimental Details 3.4 Results and Discussion 3.5 Conclusion 3.6 Acknowledgments Chapter 4 Light Induced Protein Dynamics Observed with Neutron Spin Echo 4.1 Abstract 4.2 Introduction 4.3 Experimental Details ii iii vi vii xi 1 1 10 11 13 16 16 17 20 22 36 38 39 39 40 44 50 67 68 69 69 70 73 v 4.4 Results and Discussion 4.5 Conclusion 4.6 Acknowledgments Chapter 5 Conclusion and Future Work 5.1 Conclusion 5.2 Future Work 5.2.1 Effect of azoTAB on Ribonuclease A structure 5.2.2 Effect of azoTAB on dynamics of Ribonuclease A 5.2.3 Effect of azoTAB on function of Ribonuclease A Bibliography 83 100 101 102 102 103 103 106 108 112 vi List of Tables Table 3.1: Values of the radius of gyration determined from Guinier analysis of the SANS data in Figure 3.1 Table 3.2: Effect of azoTAB on the kinetic parameters of lysozyme [lysozyme] = 0.002 mg/mL, pH 5.0 Table 4.1: Analysis of H/D exchange kinetics of lysozyme in the presence of azoTAB surfactant 52 63 96 vii List of Figures Figure 1.1: General structure of amino acid Figure 1.2: Classification of protein structure Figure 1.3: AzoTAB chemical structure and isomerization Figure 1.4: UV-vis absorption spectrum of azoTAB ([azoTAB]=0.5 mM, path length = 2 mm) Figure 1.5: Photocontrol of BSA unfolding Figure 1.6: (a) Crystal structure of hen egg-white lysozyme (PDB code 6LYZ). The arrow points to the cleft of active site. (b) lysozyme- azoTAB (8.5 mM) unfolding/refolding cycle observed with SANS Figure 2.1: Deconvolved FT-IR spectra of BSA in solution with varying azoTAB surfactant concentration under (a) visible and (b) UV light, respectively. The maximum of the amide I band is shifting to lower wavenumber with increasing surfactant concentration. Vertical dashed lines in (a) and (b) indicate the maximum of the amide I band without surfactant (− −) and with 19.85 mM azoTAB surfactant (−⋅−), respectively. Dashed curves in (a) show the individual Gaussian band for each secondary structure element of pure BSA (see text). Also shown are the second derivatives of the FT-IR spectra under (c) visible and (d) UV light, respectively Figure 2.2: Concentration dependences of the contribution to the secondary structure resulting from (a) α-helix (1653 cm -1 ), (b) β-strand (1630 cm -1 ), (c) unordered (1645 cm -1 ), and (d) β-turn (1677 cm -1 ) conformations under visible (•) and UV ( ○) light illumination, respectively. The vertical dashed lines in (a) - (d) indicate the tertiary structure boundaries with respect to surfactant concentration obtained from SANS data. The remaining peak from side-chain vibrations was found to slightly increase from 0.9 % to 2.5% of the total secondary structure with azoTAB addition. Also shown in (e) are the tertiary structures of BSA under visible or UV light illumination from ref 1 as a function azoTAB surfactant concentration. N: native heart-shaped structure; N d : distorted heart-shaped structure; F: partially unfolded; E: Elongated 2 3 10 11 12 14 23 27 viii Figure 2.3: Difference spectra of BSA defined as ΔA = spectra collected under UV light – spectra collected under visible light at a constant surfactant concentration in the amide I region Figure 2.4: FT-IR spectra of a BSA solution containing 5.02 mM azoTAB under repeated light cycles (visible → UV → visible → UV → visible → UV) in the amide I Region Figure 3.1: SANS data of lysozyme-azoTAB solutions as a function of surfactant concentration under visible (closed symbols) and UV (open symbols) light. Pure lysozyme (•), 3.6 mM azoTAB ( S , U ), 8.5 mM azoTAB ( , ), 12.0 mM azoTAB ( ¡ , ), and 18.6 mM azoTAB ( T , V ). [Lysozyme] = 10 mg/mL in pH 5.0 buffer Figure 3.2: PDDFs of lysozyme-azoTAB solutions as a function of surfactant concentration under visible (black lines) and UV (gray lines) light. [Lysozyme] = 10 mg/mL in pH 5.0 buffer Figure 3.3: In vitro conformations of lysozyme in solution determined from shape-reconstruction analysis of the SANS data in Figure 3.1. Best- fit structures are shown in blue and consensus envelopes are shown in red. The crystal structure of lysozyme (PDB code 6LYZ, space-filling and ribbon) is shown for comparison with arrows pointing to the active- site cleft between the α and β domains Figure 3.4: (a) Effect of azoTAB concentration on lysozyme activity against Micrococcus luteus-RBB under visible (•) and UV ( ○) illumination. [lysozyme] = 0.002 mg/mL; [Micrococcus luteus-RBB] = 0.8 mg/mL. (b) Optical micrographs of Micrococcus luteus-RBB cells (0.8 mg/mL) as a function of azoTAB concentration and light conditions Figure 3.5: Lysozyme activity against glycolchitin-RBB as a function of azoTAB concentration under visible (•) and UV ( ○) illumination, respectively. [lysozyme] = 0.002 mg/mL; [glycolchitin-RBB] = 1.5 mg/mL; pH 5.0; 37°C Figure 3.6: (a) Initial velocity profile of lysozyme against glycolchitin- RBB without ( ) and with 0.2 mM trans azoTAB (•) and cis azoTAB ( ○). (b) Linweaver-Burk plot of native lysozyme ( ) and lysozyme with 0.2 mM trans azoTAB (•). [lysozyme] = 0.002mg/mL; pH 5.0; 37 °C 33 35 51 ● 53 ● 55 ● ● ● ● ● ● 58 60 63 ix Figure 3.7: Photo-regulation of lysozyme activity against glycolchitin- RBB. (a) Reaction initiated with trans azoTAB ( ), followed by UV illumination to photoisomerize azoTAB to the cis state (---). (b) Reaction initiated with cis azoTAB (---), followed by visible illumination to photoisomerize azoTAB to the trans state ( ). [lysozyme] = 0.002 mg/mL, [glycolchitin-RBB] = 1.5 mg/mL, [azoTAB] = 0.2 mM, pH 5.0; 37 °C Figure 4.1: (a) SANS data and (b) PDDFs of lysozyme-azoTAB solutions as a function of surfactant concentration and light illumination (closed symbols for trans and open symbols for cis azoTAB). Pure lysozyme (•), 5.0 mM azoTAB ( ¡ , ), 8.0 mM azoTAB ( S , U ), and 12.0 mM azoTAB ( T , V ). [Lysozyme] = 10 mg/mL in pH 7.2 buffer Figure 4.2: NSE data of azo-TAB-lysozyme as a function of time t at representative Q values Figure 4.3: NSE data of lysozyme effective diffusion coefficient with (a) 0, (b) 5.0 (trans), (c) 5.0 (cis), (d) 8.0 (trans), (e) 8.0 (cis), (f) 12.0 (trans), and (g) 12.0 (cis) mM azoTAB. The ribbon (PDB 6lyz) and space filled structure next to each plot represents the protein conformation at the corresponding azoTAB concentration. In each plot, the horizontal red dashed line represents the center-of-mass translational diffusion coefficient; the solid line represents rigid body calculations using SANS data (−) and crystal structure (6lyz) (−). The dashed lines in (d) and (g) represent calculation results of domain models using soft- linker-domain model assuming two-(---) and three-(---) domains and freely-jointed-domain model (---) Figure 4.4: (a) Representative lysozyme H/D exchange spectra in the time range of 5 min to 10 hr at 20 °C. Arrows in the spectra indicate direction of peak intensity changes. [azoTAB] = 8.0 mM. (b) H/D exchange kinetics of lysozyme as a function of azoTAB concentration and light illumination (closed symbols for trans and open symbols for cis azoTAB). Pure lysozyme (•), 5.0 mM azoTAB ( ¡ , ), 8.0 mM azoTAB ( S , U ), and 12.0 mM azoTAB ( T , V ). [Lysozyme] = 10 mg/mL in pH 7.2 buffer Figure 4.5: (a) Ensemble-averaged fluorescence spectra of Alexa 532/Alexa 594 labeled T4 lysozyme with varying concentration of azoTAB. The inserted picture represents the structure of T4 lysozyme (PDB code 6LZM) with the red and blue dots locating cys 54 and cys 97 where the dye pair binds, respectively. (b) Ensemble-averaged FRET efficiency as a function of azoTAB concentration 67 84 86 88 94 98 x Figure 5.1: SANS data of azoTAB-ribonuclease A solutions as a function of surfactant concentration and light illumination. [RNase A] = 10 mg/mL Figure 5.2: FT-IR FSD spectra of RNase A without varying concentration of azoTAB. [RNase A] = 10 mg/mL, pH 7.2 Figure 5.3: NSE data of the effective diffusion coefficient of RNase A with (a) 0, (b) 10 (trans), (c) 10 (cis) mM azoTAB. The space filled structure next to each plot represents the protein conformation at the corresponding azoTAB concentration obtained from the SANS data in Figure 5.1 using GA_STRUCT Figure 5.4: (a) Relative activity of RNase A against yeast RNA as a function of azoTAB concentration (•) and cis azoTAB ( ○). (b) Linweaver-Burk plot of native RNase A (•) and RNase A with 2.5 mM azoTAB (trans ( S ) and cis ( T )) Figure 5.5: (a) Chemical structure of the fluorescence based substrate, 6- FAM-dArU(dA) 2 -6TAMRA. (b) Illustration of the basic idea of th fluorescence based substrate Figure 5.6: (a) Emission spectra of 6-FAM~IArU(dA)2~5-TAMRA (solid line) and its cleavage product (dashed line) on excitation at 490 nm. (b) Exponential rise analysis of the cleavage of 6-FAM-dArU(dA)2- 6-TAMRA by ribonuclease A. "CPS" refers to photon counts per second 104 106 107 109 110 111 xi Abstract The interaction of a light-responsive azobenzene-based surfactant (“azoTAB”) with proteins has been investigated as a means to photoreversibly control protein structure, dynamics, and function. AzoTAB undergoes a reversible photoisomeriztion upon exposure to appropriate wavelength of light, with the visible-light, trans isomer being more hydrophobic and, thus, inducing a greater degree of protein unfolding than the UV-light, cis form. AzoTAB is found to induce triggered and localized protein unfolding, measured directly in solution with small- angle neutron scattering (SANS) experiments, and to further influence the biological function and dynamics of proteins. For example, the relationship between photoreversible changes in secondary and tertiary structure of BSA, found to exist as one of three discrete forms depending on the azoTAB concentration, has been examined. Similarly, photo-control of the form-function relationship of lysozyme has been studied. With azoTAB in the trans form under visible light a partially- unfolded intermediate conformation of lysozyme with an exposed active site was found, while lysozyme was observed to refolded to a native-like structure upon UV illumination. In addition, enhanced dynamics within the partially-unfolded form of lysozyme were observed with neutron spin echo (NSE) measurements and thought to contribute to a nearly 8-fold enhancement in the enzyme activity compared to the native state. Combined, these results provide insight into a unique light-based xii method of controlling the complete structure-dynamics-function relationship of proteins. 1 Chapter 1 Introduction 1.1 Protein Form, Function and Dynamics Proteins are organic macromolecules which exist everywhere in living organisms. They are catalysts that transport and store other molecules, provide mechanical support, give immune protection, generate movement, transmit nerve impulses, and control growth and differentiation. Basically, proteins are responsible for nearly every cell function. With such a heavy duty, protein malfunction can often lead to fatal diseases. For instance, Alzheimer’s disease, mad cow disease and Huntington’s disease all stem from alteration of protein conformation, leading to the formation of polymers (e.g., amyloids) resulting from aggregation of misfolded proteins. On the other hand, proteins formed by specific arrangements of amino acids varying in size, shape, charge, hydrogen bonding capacity, hydrophobic character and chemical reactivity are useful in many applications including therapeutic agents, catalysts, and materials. Thus, understanding the protein form-function relationship is beneficial both academically and industrially. Proteins are polymers formed by amino acids. There are 20 naturally occurring amino acids involved in protein composition; each has the general structure as shown in Figure 1.1. The center of an amino acid is the alpha carbon (C α ) which the amine group, 2 Figure 1.1: General structure of amino acid. carboxyl group, a hydrogen atom and a side chain (R group) are covalently bound to. The properties of amino acids are determined by the side chain, and depending on the side chain amino acids can be classified as either acidic, basic, polar or nonploar. The structure of proteins is usually discussed in terms of four levels as shown in Figure 1.2. Through the translation process catalyzed by ribosome, a long chain polymer of amino acid residues is produced by condensation reactions that repeatedly link two amino acid residues by a peptide bond to form the main chain/backbone of a protein. The primary structure of the protein is defined by the resulted specific sequential arrangement of the amino acid residues. Under suitable conditions (solvent, temperature, slat, etc.), the one dimensional amino acid sequence spontaneously folds into a three dimensional structure, native state, such that the free energy is minimized, preferentially with hydrophobic amino acids in C C N O O H H H H R 3 Figure 1.2: Classification of protein structure. Secondary Structure Tertiary Structure Sheet Quaternary Structure Primary Structure Helix C α C N O R i C α C N O R i C α C N O R i Residue i Residue i-1 Residue i+1 4 the core and hydrophilic (polar, or charged) amino acids on the surface of the protein depending on the differing degrees of polarity and water affinity of amino acids side chains. This folding is manifested by the formation of secondary structure elements that are stable due to simultaneously minimizing the strain on the relatively rigid amide linkages, while at the same time satisfying the hydrogen-bonding potential of the main- chain N−H and C=O groups. The first successfully predicted secondary structure elements were α-helices and β-sheets. 119 α-helices have a spring-like conformation locally formed by a single consecutive set of residues in a amino acid sequence. β-sheets are composed of several independent sets of amino acid residues in a protein. Each consecutive set (typically 5−10 residues) refers to a local structure, β-strands. The non- local structures, β-sheets, are formed by laterally connecting β-strands with hydrogen bonds. Since then, other secondary structure elements such as loops, which connecting helices and strands and other forms of helices were discovered. Ultimately, these secondary structure elements arrange into the overall shape or tertiary structure of the protein, which is largely driven by hydrophobic interactions in the protein. Other interactions also stabilize the tertiary structure of a protein including disulfide bonds, hydrogen bonding, and ionic interaction. Some proteins exist as oligomers that are composed of several identical subunits or a set of different subunits that are not covalently linked. The three dimensional organization of these subunits is the quaternary structure of the protein. These folding events, which lead to the formation of the native (or active) state, are essential for biological function of proteins. Taking enzymes, (i.e., proteins that 5 catalyze chemical reactions,) as an example, residues involved in chemical reactions can be distantly separate pieces in the amino acid sequence. In the native state, an enzyme folds to form a cavity of active residues in which substrates can be selectively recognized and efficiently bound with scaffolding support from other residues. However, proteins are not static entities and instead regularly undergo different types of conformational changes during the course of activity. Although the static conformation of proteins can be obtained from high resolution X-ray crystallography, the native structure of a protein is actually a collection of slightly different structures, each corresponding to a local minima in the potential energy surface of the system and forming an equilibrium of structures. 172 This structure equilibrium may be perturbed by various factors such as heat or ligand binding. For example, substrate binding can cause different levels of structural changes, from localized conformational changes to hinge bending motions reorientating entire protein domains. Protein motions resulting from continuous conformational changes, which require suitable structural flexibility, lead to protein dynamics including subpicosecond atomic vibrations, pico- to nanosecond backbone and side-chain fluctuations, and milli- to second conformational rearrangement and breathing motions. 11 These protein motions are functionally important even though mechanisms of protein function are not fully understood. For example, the backbone and side-chain fluctuations may be involved in molecular recognition and loop motions may be required to exclude water or for repositioning of catalytic residues. 11, 60 6 To reveal how proteins function, numerous researchers have investigated the form-function relationship of proteins to examine the importance of protein conformation. On the other hand, protein flexibility and dynamics are considered equally important to protein function. 4, 16, 35, 107, 115, 172 However, only a few studies have reported a direct relationship between dynamics and biological function, leaving much as yet undiscovered about the broader form-dynamics-function relationship of proteins. 33, 78, 169 As such, precise knowledge of how a protein unfolds/refolds and fluctuates between structural intermediates in response to various stimuli is required. To achieve this goal, numerous factors that can cause protein unfolding have been examined to probe conformational changes in proteins, including pH, 113 temperature, 47, 70 pressure, 139, 163 and chemical denaturants (urea, GdmCl or surfactants). 39, 58, 97, 105, 161 Among these denaturants, surfactants (i.e., molecules containing both a hydrophobic tail and a hydrophilic, typically charged, head group) can provide specific insight into protein folding processes, particularly when considering that proteins routinely contact a variety of amphiphilic molecules (e.g., cell membranes, surfactant-assisted solubilization and crystallization, etc.). Surfactants bound to proteins generally shield the unfavorable interaction of the nonpolar amino acids in the protein core with water, allowing the protein to unfold. 3, 152 In the dissertation, azobenzene-based photoresponsive surfactants (azoTAB) will be utilized as a means to reversibly controlling protein conformation, function and dynamics with simple light illumination. 58, 97 The photoresponsive surfactants, undergo a reversible trans ↔ cis photoisomerization (see Figure 1.3, below) upon exposure to 7 visible or UV light, respectively. 42-44, 66, 137 Enhanced protein-surfactant interactions occur under visible light as a consequence of the relative hydrophobicity of the planar trans isomer compared to the bent cis conformation. This leads to protein unfolding that can be reversed with exposure to UV light, and further allows photo-control of the biological function and dynamic fluctuations of proteins. 58, 97 Among the numerous techniques of protein structure determination, X-ray crystallography has been the principle technique, providing unparalleled atomic resolution of protein structures. By irradiating a protein crystal with a narrow X-ray beam, which has a wavelength on the order of a covalent bond, the X-ray scattering pattern by the protein crystal can be resolved into abundant information of protein structure. More than ten thousands native protein structures have been successfully elucidated by applying this technique. However, as mentioned above the protein structures are not static in nature and instead proteins usually adopt partially-folded conformation when performing certain functions. In these cases the application of X-ray crystallography is limited. Therefore, other techniques must be employed to observe protein conformations in more complex solution environments, such as the small angle- neutron scattering (SANS) measurements in Chapter 3 and Chapter 4. SANS is a scattering technique that measures the deflection of a neutron beam by a sample at small angles. Several data analysis techniques allow one to obtain structural information from a protein scattering pattern, including Guinier analysis or calculation of pair-distance distribution function (which give low-resolution structural information of protein size and 8 shape) or reliable ab initio shape reconstruction techniques such as GA_STRUCT or GASBOR (which enable one to visually “observe” protein conformation in solution.) For protein secondary structure determination, Fourier transform infrared spectroscopy (FT-IR) and circular dichroism spectroscopy (CD) are the most popular techniques. CD measures the asymmetry of an optically active material by observing the differences in absorption of left-handed and right-handed polarized light. The CD spectrum of a protein in the far-UV region (190-250nm) contains secondary structure information from which detailed information can be obtained by decomposing the overall spectrum into characteristic spectra of individual secondary structure elements. On the other hand, FT-IR measures sample absorption of infrared light that is energetic enough to excite molecular vibrations. The vibrations of peptides in a protein result in nine characteristic bands (amide A, B, I, II…). The best studied band, namely amide I which ranges from 1700−1600 cm -1 , arises mainly from C=O stretching of the peptide and is directly related to backbone conformation. Thus, secondary structure information of proteins can be determined from amide I with the aid of resolution enhancement techniques. Both CD and FT-IR techniques are extensively used for studying protein folding/unfolding upon varying temperature, pH, or denaturants. However, neither of these techniques is perfect for secondary structure determination. For instance, although CD spectra usually give more accurate estimates of the portion of α helices than FT-IR, it also usually results in a much lower percentage of β structure compared to X-ray crystallography. In contrast, FT-IR has been shown to allow accurate β structure estimation as well as acceptable α helices percentages. 32, 123 The strong absorbance of 9 azoTAB makes secondary structure analysis from CD spectra (between 260 and 180 nm) very difficult, while with IR spectra the sharp azoTAB peak at 1600 cm -1 can be easily subtracted. Thus, FT-IR is a better technique than CD for studying the effect of azoTAB on protein secondary structure. In the following sections, an introduction of the photoresponsive azoTAB surfactant will be given, followed by a brief introduction into the application of azoTAB to photocontrol of the folding of bovine serum albumin (BSA) and lysozyme as will be discussed in greater detail in Chapter 2, 3 and 4. In Chapter 2 the ability to control BSA conformation by using photoresponsive azoTAB surfactant and light (UV or visible light) will be examined through a secondary structure investigation using FT-IR, along with a careful comparison with tertiary structure observed from SANS. The ability to photo- control the form-function relationship of lysozyme will then be discussed in Chapter 3, with enzyme conformation and enzyme function directly related by comparing protein conformation obtained from SANS and the activity profile using various protein assays. In Chapter 4, enhanced lysozyme dynamics induced by azoTAB surfactant and light will be observed with neutron spin echo spectroscopy combined with tertiary structure obtained from SANS to study the protein structure-dynamics-function relationship. Finally, a strategy to study the relationship between protein dynamics and function for ribonuclease A will be discussed in Chapter 5. 10 1.2 Photoresponsive Surfactant A cationic azobenzene trimethylammonium bromide surfactant (azoTAB) of the form in Figure 1.3 was synthesized according to published procedures. 66, 136 Briefly, 4- ethlyalaniline was azocoupled with phenol followed by alkylation with 1,4- dibromobutylane and quaternization with trimethylamine. The structure of the product was determined by 1 H-NMR. N N CH 3 CH 2 O(CH 2 ) 4 N + (CH 3 ) 3 Br N + (CH 3 ) 3 Br N N O(CH 2 ) 4 CH 2 CH 3 UV Light Visible light UV Light Visible light Figure 1.3: AzoTAB chemical structure and isomerization. Generally, azoTAB exists as the trans isomer in solution in the dark or under visible light, with a maximum absorption at 350 nm and dipole moment across the – N=N– bond of ~0.5D. When illuminated with UV light, photoisomerization to the cis isomer occurs, which induces an increase of the dipole moment to ~3.1 D across the – N=N– bond and results in a decrease of the absorbed band at 350 nm and increase in the band at 434 nm, as shown in Figure 1.4. trans cis 11 Given the difference in dipole moments upon illumination with UV or visible light, the photoresponsive surfactant possess light tunable physical properties, such as surface tension, electric conductivity, hydrophobicity. 44 Taking advantages of this property, photoresponsive surfactants have been applied in our group to photocontrol protein structure including secondary, 161 tertiary 58 and quaternary structure, 59 along with DNA condensation. 96 0 0.5 1 1.5 2 200 250 300 350 400 450 500 550 Absorbance Wavelength (nm) trans cis Figure 1.4: UV-vis absorption spectrum of azoTAB ([azoTAB] = 0.5 mM, path length = 2 mm). 1.3 Bovine Serum Albumin (BSA) Serum albumin is the most abundant protein in the circulatory system of mammals, responsible for regulation of the osmotic pressure for proper distribution of body fluids between vascular and other tissues. The protein has an isoelectric point of 4.7, which leads to a net negative charge of -18 at physiological conditions. The structure of BSA is known to be dominated by nine helical loops connected with 17 disulfide 12 bridges, giving rise to three protein domains each containing one small and two large helical loops. This leads to a secondary structure of BSA with approximately 67% helical content, while the remainder consists of ~23% extended conformations (β-strands) and ~10% β-turns. 15, 24, 121 Upon acid denaturation, BSA exhibits three discrete conformations, including a heart-shaped native N form (the well-known X-ray crystallographic structure in the native state), a partially unfolded F form, and a highly elongated E form 24 as shown in Figure 1.5. Figure 1.5: Photocontrol of BSA unfolding. 97 In Chapter 2, similar conformational changes of N ↔ F and F ↔ E transitions are observed with addition of azoTAB and light illumination. The secondary structure rearrangements of BSA in response to photoresponsive surfactant and light illumination will be utilized as a means to fully characterizing the in vitro conformations obtained vis 0 5 10 mM [S] UV visible UV visible UV E α-helices N F N terminus Increasing surfactant concentration 13 from SANS, as well as to answer questions about the local reversibility of photo- controlled protein folding. By using Fourier transform infrared (FT-IR) spectroscopy, quantitative information on the changes in BSA secondary structure induced by azoTAB and light illumination will be obtained. Directly comparing changes in secondary (FT- IR) and tertiary (SANS) structure observed during the N → F and F → E transitions, while at the same time reversibly controlling these transitions with light, provides unprecedented insight in the mechanisms of BSA folding. 1.4 Lysozyme Lysozyme, a bacteriolytic enzyme which is present at high concentration in various tissues and secretions, was first discovered by Alexander Fleming in 1922. The enzyme catalyzes the hydrolysis of β(1→4) linked copolymers of N-acetylglucosamine and N-acetylmuramic acid in certain bacteria cell walls or homopolymers of N- acetylglucosamine, chitin. 77 As shown in Figure 1.6(a), the native structure of lysozyme can be roughly divided into two domains: a structured α-domain that is folded around a central core of hydrophobic residues and has four α-helices and one 3 10 helix, and a sheet-like β-domain consisting mainly of hydrophilic residues either on the outer surface of the molecule or lining the cleft forming an antiparallel β-sheet and a long loop. The active site of lysozyme, which is located in the cleft between the α and β- domains, can be divided into 6 sites (sites A to F). The cleft of the active site is partially lined with nonpolar side chains of amino acids for binding of the nonpolar regions of substrates, along with a hydrogen bonding site for acylamino and hydroxyl groups. With 14 UV visible some degree of flexibility, lysozyme has been shown to exhibit “hinge-bending motions” between the two separate domains upon substrate binding. 4, 16, 31, 106, 107, 172 Due to these hinge-bending motions, the active site is “opened” to initiate the formation of the enzyme-substrate complex and “closed” upon substrate binding. Figure 1.6: (a) Crystal structure of hen egg-white lysozyme (PDB code 6LYZ). The arrow points to the cleft of active site. (b) lysozyme-azoTAB (8.5 mM) unfolding/refolding cycle observed with SANS. 160 Although surfactants are usually considered as d0enaturants that interact and unfold the active site and decrease the enzymatic activity, with the presence of the nonionic surfactant polyoxyethylenesorbitan (Tween 20) lysozyme exhibits enhanced activity toward its natural substrate, Micrococcus luteus. 10, 64, 65, 83, 84 In Chapter 3, the photo-control of enzyme structure and activity by using photoresponsive surfactants will be demonstrated. AzoTAB is found to bind and swell parts of the protein away from the active site, leading to enhanced exposure of the active site in the lysozyme-azoTAB complexes as determined from SANS displayed in Figure 1.6(b). Upon assay, the activity of lysozyme was enhanced up to ~7 fold with the addition of azoTAB, possibly α-domain β-domain Active site (a) (b) 15 due to a higher accessibility of substrate binding with a more exposed active site. It is possible that with azoTAB binding the flexibility of the protein is enhanced, which allows the protein to perform increased hinge-bending motions. With this enhanced flexibility, the substrate binding efficiency may be increased. In Chapter 4, the effect of internal dynamics on lysozyme induced by azoTAB and light will be examined using neutron spin echo measurement. The azoTAB-lysozyme complex is found to perform enhanced internal motions with increased flexibility observed from hydrogen/deuterium exchange kinetics. 16 Chapter 2 Protein Secondary Structure Controlled with Light and Photoresponsive Surfactants * 2.1 Abstract The interaction of a light-responsive azobenzene surfactant with bovine serum albumin (BSA) has been investigated as a means to examine photoreversible changes in protein secondary structure. The cationic azobenzene surfactant undergoes a reversible photoisomeriztion upon exposure to the appropriate wavelength of light, with the visible- light (trans) form being more hydrophobic and, thus, inducing a greater degree of protein unfolding than the UV-light (cis) form. Fourier transform infrared (FT-IR) spectroscopy is used to provide quantitative information on the secondary structure elements in the protein (α-helices, β-strands, β-turns, and unordered domains). Comparing the secondary structure changes induced by light illumination in the presence of the photoresponsive surfactant with previous measurements of the tertiary structure of BSA obtained from small-angle neutron scattering (SANS) allows the three discrete conformation changes in BSA to be fully characterized. At low surfactant concentrations an α-helix → β-structure rearrangement is observed as the tertiary structure of BSA changes from a heart-shaped * The Chapter is a published article (Journal of Physical Chemistry B 110, 16117-16123, 2006.) 17 to a distorted heart-shaped conformation. Intermediate surfactant concentrations lead to a dramatic decrease in the α-helix fraction in favor of unordered structures, which is accompanied by an unfolding of the C-terminal portion of protein as evidenced from SANS. Further increases in photosurfactant concentration lead to a β → unordered transition with the protein adopting a highly elongated conformation in solution. Each of these protein conformational changes can be precisely and reversibly controlled with light illumination, as revealed through FT-IR spectra collected during repeated visible ↔ UV light cycles. 2.2 Introduction Proteins are the workhorses of biology. Beginning as polyelectrolyte chains formed from a specific sequence of the 20 naturally occurring amino acids (this sequence is termed the primary structure of the protein), proteins fold in response to differing degrees of water affinity of these amino acids such that the free energy is minimized, preferentially with hydrophobic amino acids in the core and hydrophilic amino acids on the surface of the protein. This is manifested locally by the formation of secondary structure elements such as α-helices and β-sheets that are stable due to simultaneously minimizing the strain on the relatively rigid amide linkages, while at the same time maximizing the formation of hydrogen bonds. Ultimately, these elements arrange into the overall shape or tertiary structure of the protein. These folding events, which lead to the formation of the native state, give rise to a functional protein. However, proteins are not static entities and instead regularly undergo conformational changes to 18 intermediately-folded states during the course of activity, particularly upon interaction with ligands or substrates. As such, information on the native state alone will not provide complete understanding of the form-function relationship of a protein. Instead, precise knowledge of how a protein unfolds/refolds in response to various stimuli is required. To achieve this goal, photoresponsive surfactants can be utilized as a means of reversibly controlling protein conformation with simple light illumination. 58, 97 Since protein folding results from an interplay between interactions among the protein amino acids and the solvent (i.e., hydrophobic, electrostatic, van der Waals, and hydrogen bonding), alteration of these intramolecular forces responsible for maintaining the secondary and tertiary structure results in changes in protein conformation. 118, 128 Indeed, based on the importance of protein folding in biological, chemical, and industrial applications, extensive efforts have been devoted to probe conformational changes in proteins with pH, temperature, pressure, and chemical denaturants. Among these denaturants, surfactants (i.e., molecules containing both a hydrophobic tail and a hydrophilic, typically charged, head group) can provide specific insight into protein folding processes, particularly when considering that proteins routinely come in contact with a variety of amphiphilic molecules (e.g., cell membranes, surfactant-assisted solubilization, dispersion, and crystallization, etc.). Surfactants bound to a protein generally shield the unfavorable interaction of the nonpolar amino acids of the protein with water, allowing the protein to unfold. 3, 152 In the case of azobenzene-based photoresponsive surfactants, which as shown in Figure 1.3 undergo a reversible trans ↔ cis photoisomerization upon exposure to visible 19 and UV light, respectively, 42, 43, 66, 137 enhanced protein-surfactant interactions occur under visible light as a consequence of the relative hydrophobicity of the planar trans isomer compared to the bent cis conformation, leading to protein unfolding that can be reversed with exposure to UV light. 58, 97 Direct photo-control of the folding of bovine serum albumin (BSA) was demonstrated with in vitro tertiary structures obtained by applying shape-reconstruction techniques to small-angle neutron scattering (SANS) data. Three discrete folding forms of BSA were detected, including a heart-shaped N form at low surfactant concentrations (similar to the well-known X-ray crystallographic structure in the native state 68 ), a partially unfolded F form at intermediate surfactant concentrations (with the C-terminal portion separated from the remainder of the protein), and a highly elongated E form at high surfactant concentrations (although some residual folding, i.e., “kinks”, were evident in the shape-reconstructed conformations), while N ↔ F and F ↔ E transitions could be induced with light. The structure of BSA is known to be dominated by nine helical loops connected with 17 disulfide bridges, giving rise to three protein domains each containing one small and two large helical loops. This leads to a secondary structure of BSA with approximately 67% helical content, while the remainder consists of ~23% extended conformations (β-strands) and ~10% β-turns. 15, 24, 121 Upon unfolding of BSA under acid conditions 88 or through the addition of surfactants, 146, 149 the helicity decreases but still represents approximately 50% of the secondary structure, suggesting that the six large helical loops in BSA are not disrupted. 147 Thus, it is possible that the kinked structures observed in the photosurfactant-unfolded E form with SANS represent these remaining helical loops. 20 In the present study, we utilize measurements of the secondary structure of BSA in response to photoresponsive surfactant and light illumination as a means of fully characterizing the in vitro conformations obtained with SANS, as well as answering question about the local reversibility of photo-controlled protein folding. By using Fourier transform infrared (FT-IR) spectroscopy, quantitative information on the changes in BSA secondary structure induced by photosurfactant and light illumination are obtained. Directly comparing changes in secondary (FT-IR) and tertiary (SANS) structure observed during the N → F and F → E transitions, while at the same time being reversibly controlling these transitions with light, provides unprecedented insight in the mechanisms of BSA folding. 2.3 Experimental Details An azobenzene trimethylammonium bromide surfactant (azoTAB) of the form N N CH 3 CH 2 O(CH 2 ) 4 N + (CH 3 ) 3 Br was synthesized according to published procedures. 66, 136 Conversion from the trans form to the cis isomer was achieved by illumination with a 200 W mercury arc lamp (Oriel, model no. 6283) equipped with a 320 nm band-pass filter (Oriel, model no. 59800) in combination with a heat-absorbing IR filter (Oriel, model no. 59060) to isolate the 365 nm line. For conversion from the cis form to the trans form, a 400 nm long-pass filter (Oriel, model no. 59472) was used to illuminate the sample with the 436 nm line from the 21 mercury lamp. Absorption measurements indicate that under visible-light the surfactant exhibits an approximately 75/25 trans/cis equilibrium, while under UV light the surfactant is primarily in the cis form (> 90% cis). 96 Highest quality BSA was purchased from Roche and used as received. Samples were prepared by dissolving 10 mg/mL BSA and varying amounts of crystallized surfactant into a D 2 O phosphate buffer solution (8.3 mM, pD = 7.30 measured with a standard pH electrode and corrected according to pD = pH + 0.4 for deuterium isotope effects). Samples were prepared 48 hours prior to FT-IR measurements to ensure complete H-D exchange. Infrared spectra were measured with a Genesis II FT-IR spectrometer (Mattson Instruments). Solutions were loaded between a pair of CaF 2 windows using a 50 μm Teflon spacer contained in a demountable liquid cell equipped with a circulating water jacket (T = 20 °C). A liquid light guide (Oriel, model no. 77557) was oriented within the spectrometer to directly illuminate the sample with UV or visible light for 90 mins prior to and during data collection. The sample chamber was continuously purged with dry air to eliminate water vapor. For each spectrum, a 250-scan interferogram was collected with a 4 cm -1 resolution. The absorbance due only to the protein secondary structure was obtained by subtracting the spectra measured for a pure surfactant solution under otherwise identical conditions, allowing the relatively sharp surfactant peaks at ~ 1600 cm -1 to be removed. The resulting corrected spectra were flat in the region between 2000 and 1750 cm -1 , and were used for further analysis. The technique of Fourier self- deconvolution (FSD) was applied to the original spectra to resolve the overlapping bands 22 in the Amide I region using a band-narrowing factor k = 2.0 and a full width at half height of 25.15 cm -1 . Second derivative spectra were obtained with the Savitsky-Golay function for a 3 rd order polynomial, using a five data point window. Difference spectra at varying surfactant concentration were obtained by subtracting the spectra collecting under visible light from the spectra collecting under UV light illumination. For pure BSA solutions, the difference spectra showed no significant absorbance (< 1% throughout the amide I region). Curve fitting was carried out on the deconvolved spectra by assuming Gaussian band profiles. The number of peaks and initial values for the peak positions, intensities, widths, and heights were estimated from the second derivative and deconvolved spectra. The percentages of each structural component were estimated from the relative areas of each curve. 2.4 Results and Discussion FT-IR Spectra of BSA-azoTAB mixtures. Figures 2.1(a)-(d) show the FSD and second derivative spectra of BSA solutions in the amide I region of 1700-1600 cm -1 with varying azoTAB surfactant concentration under visible (trans form) and UV (cis form) light. In the absence of surfactant, the spectrum for pure BSA can be resolved into four major bands as shown in Figure 2.1(a), which can be assigned according to the literature 14, 19, 47, 63, 112, 142 as 1677 cm -1 (β-turns), 1653 cm -1 (α-helices), 1630 cm -1 (β- strands, short segments connecting helical structures) and 1610 cm -1 (vibrations of the aromatic side chains). In addition, a minor peak (~ 2% of the total structure) is found at approximately 1641 cm -1 , which will be discussed below. 23 UV light 1600 1620 1640 1660 1680 1700 2nd derivative Wavenumber (cm -1 ) Visible light 19.85 mM 17.95 mM 15.97 mM 10.04 mM 6.57 mM 5.02 mM 2.98 mM 1.99 mM 0.99 mM 0.00 mM 1600 1620 1640 1660 1680 1700 Wavenumber (cm -1 ) UV light 19.85 mM 17.95 mM 15.97 mM 10.04 mM 6.57 mM 5.02 mM 2.98 mM 1.99 mM 0.99 mM 0.00 mM [azoTAB] Visible light Figure 2.1: Deconvolved FT-IR spectra of BSA in solution with varying azoTAB surfactant concentration under (a) visible and (b) UV light, respectively. The maximum of the amide I band is shifting to lower wavenumber with increasing surfactant concentration. Vertical dashed lines in (a) and (b) indicate the maximum of the amide I band without surfactant (− −) and with 19.85 mM azoTAB surfactant (−⋅−), respectively. Dashed curves in (a) show the individual Gaussian band for each secondary structure element of pure BSA (see text). Also shown are the second derivatives of the FT-IR spectra under (c) visible and (d) UV light, respectively. (d) (c) (b) (a) 24 As the surfactant concentration is increased under either visible or UV light, a broad band develops in the region of ~ 1645 cm -1 (particularly evident under visible light, see Figures 2.1(a) and (c)), which is a typical location for unordered structures. At the same time, the influence of the α-helix peak on the spectra decreases, resulting in the maximum in the amide I band shifting to lower wavenumbers and indicating a loss of secondary structure (primarily a helix-to-unordered transition) as BSA unfolds. The second derivative spectra (Figures 2.1(c) and (d)) also indicate a gradual decreases in the α-helical band accompanied by an increase in unordered structures with increasing surfactant concentration. While similar trends are seen under both visible and UV light, the effects are less obvious and require higher surfactant concentrations with UV illumination, as the more hydrophobic trans form of the surfactant results in a greater degree of protein unfolding compared to the cis isomer. As discussed below, the appearance of the unordered peak in Figure 2.1 coincides with the discrete unfolding of BSA to the F form observed with SANS, 97 while the continual shift in the peak maximum is consistent with the F → E transition observed at higher surfactant concentrations. The minor peak found in the spectrum for pure BSA at 1641 cm -1 is difficult to unequivocally define. Although this region is typically assigned to random or unordered structures, as was done above for the broad peak that develops at 1645 cm -1 , comparing the X-ray crystallographic structure of serum albumin to various spectroscopic techniques and sequence-based predictive methods generally indicates that native BSA contains no unordered domains. 24, 121 A “hydrated peak” at ~1640 cm -1 has also been attributed to either β-turn or random conformations within the protein core that contain bound water 25 molecules, 12, 13 perhaps forming hydrogen bonds within the loops of the β-turns and thereby disrupting the turn structure, 38 consistent with the idea that BSA, even in the crystalline state, is known to contain a large amount of bound water. 24, 121 As seen from the second derivative spectra in Figures 2.1(c) and (d), the 1641 cm -1 peak is evident only at low surfactant concentrations (below 1 mM and 3 mM azoTAB under visible and UV light, respectively, while at 2 mM azoTAB this peak could be reversibly disrupted and reformed with cycles of visible and UV light exposure), supporting the assignment to a hydrated peak with surfactant binding potentially replacing bound water molecules. Indeed, the core of BSA becomes accessible to the hydrophobic probe Nile red at an azoTAB concentration of ~1 mM under visible light, while at the same time the heart- shaped structure of BSA becomes somewhat distorted. 97 Despite this evidence, it is beyond the resolution of FT-IR to definitively assign the origin of this peak, thus, the simplified approach of assigning all peaks in the region of 1645-1640 cm -1 to unordered structures is utilized. Quantitative analysis of the FT-IR Spectra. In order to determine the individual contributions to the overall protein secondary structure, the FSD spectra were deconvoluted into five peaks (β-turns at 1677 cm -1 , α-helices at 1653 cm -1 , unordered structures at 1640 cm -1 , β-strands at 1630 cm -1 , and side-chain vibrations at 1610 cm -1 ) as shown in Figure 2.2. The results for pure BSA indicate 65% α-helix, 25% β-strand, 6% β-turn, and 2% unordered structures, with ~1% of the structure occurring in the side-chin region. The α-helix percentage of has generally been reported to range from 55 − 60% using FT-IR 14, 86, 125, 167 or Raman 29 spectroscopy to 68% using CD 127, 149 spectroscopy, 26 while a value of 65% has been estimated from the sequences of several albumins. 120 Values in the literature for the percentage of β structures vary more widely, however, ranging from 35% − 44% using FT-IR (22% − 30% of β-strand and 11% − 16% β-turn) 14, 86, 125, 167 and 3% − 30% with CD (β-strand), while X-ray crystallography indicates that ~33% of the protein exists as β structures. 69 As shown in Figure 2.2, each secondary structure component changes in a nonmonotonic manner in response to surfactant concentration and light illumination. The α-helix percentage (Figure 2.2(a)) decreases to an intermediate plateau value at low surfactant concentrations, followed by a second depression and eventually reaching a constant value of ~ 50%. In addition, less α-helical character is consistently observed when illuminating with visible compared to UV light. At the same time, the percentage of unordered structures (Figure 2.2(c)) is seen to increase with increasing surfactant concentration, with a greater unordered component found under visible light. The β-turn and β-strand structures in Figures 2.2(b) and (d), respectively, initially increase with surfactant addition and then gradually decrease at higher surfactant concentrations (especially the β-strand content), with the loss of β structure again occurring more rapidly under visible light. Combined, these results are consistent with the more-hydrophobic trans form of the surfactant causing a greater degree of unfolding and, hence, more extensive loss of secondary structure in BSA. In the sections that follow, we demonstrate that this apparent complex behavior of the secondary structure elements can in fact be directly related to changes in the protein tertiary structure that were observed with SANS. 27 Figure 2.2: Concentration dependences of the contribution to the secondary structure resulting from (a) α-helix (1653 cm -1 ), (b) β-strand (1630 cm -1 ), (c) unordered (1645 cm -1 ), and (d) β-turn (1677 cm -1 ) conformations under visible (•) and UV ( ○) light illumination, respectively. The vertical dashed lines in (a) - (d) indicate the tertiary structure boundaries with respect to surfactant concentration obtained from SANS data. The remaining peak from side-chain vibrations was found to slightly increase from 0.9 % to 2.5% of the total secondary structure with azoTAB addition. Also shown in (e) are the tertiary structures of BSA under visible or UV light illumination from ref 1 as a function azoTAB surfactant concentration. N: native heart-shaped structure; N d : distorted heart- shaped structure; F: partially unfolded; E: elongated. 5 10 15 0 5 10 15 20 Percentage (%) [S] (mM) β-turn 45 50 55 60 65 70 0 5 10 15 20 α-helix Percentage (%) [S] (mM) FE N F E N 22 24 26 28 30 32 0 5 10 15 20 Percentage (%) [S] (mM) β -strand 0 5 10 15 20 0 5 10 15 20 Percentage (%) [S] (mM) Unordered (b) (d) (c) (a) (e) 28 Pre-transition. The in vitro tertiary structures of BSA obtained through SANS 97 (Figure 2.2(e)) revealed that the pure protein exists as a heart-shaped structure in solution (the so-called N form), similar to the X-ray crystallographic structure. With the addition of a small amount of azoTAB surfactant (0.55 mM and 1.60 mM under visible and UV light, respectively), BSA was found to adopt a slightly distorted heart-shaped conformation, termed here the N d form, with the left side of the protein (the C-terminus, or domain III) possibly indicating some degree of swelling. However, from the SANS data alone it was not possible to determine whether this “pre-transition” was indeed a structural rearrangement in BSA or simply a consequence of surfactant accumulating on this part of the protein, since both the protein and bound surfactant molecules scatter neutrons. As shown in Figure 2.2, however, over the same surfactant concentrations where these subtle differences in BSA tertiary structure were observed, the secondary structure of BSA exhibits dramatic changes. For example, at 0.63 mM azoTAB under visible light (surfactant-to-protein molar ratio, S/P = 4.2) and 1.49 mM azoTAB under UV light (S/P = 10), the helicity of BSA has dropped from 65% to 57.8% and 58.5%, respectively, coinciding with an apparent plateau in the α-helix percentage. Interestingly, this drop of ~7% in helicity is not accompanied by a similar increase in unordered domains, instead an increase in β structures is found (from 31.7% for pure BSA to either 37.2% at 0.63 mM azoTAB under visible light or 38.8% at 1.49 mM azoTAB under UV light). Thus, it appears that the N → N d pre-transition at low surfactant concentrations is a result of an α → β rearrangement in the protein secondary structure, likely a result of 29 the repulsion of positive charges that develop on the protein upon surfactant binding, which would favor the formation of extended conformations. 14 Similar α → β transitions have been observed with FT-IR upon the interaction of serum albumin with acid, 14 or small organic molecules, 114, 124, 125, 167 while the β-sheet percentage of BSA begins to increase at sodium dodecyl sulfate (SDS) concentrations as low as 0.25 mM (S/P = 26). 149 Due to the relatively low surfactant concentrations required to initiate the N → N d conformational change shown in Figure 2.2, it is likely that this pre-transition is a result of surfactant molecules interacting with high-affinity binding sites on the protein. For comparison, the five binding sites for myristic acid with BSA have been located with X- ray crystallography to be predominantly within hydrophobic pockets of serum albumin (one in subdomain IB, one between IA and IIA, two in IIIA and one in IIIB). 34 Thus, the FT-IR and SANS results support the idea of the N → N d pre-transition being a result of an α → β rearrangement in the protein secondary structure, possibly in domain III. N → F transition. Following the pre-transition, analysis of the SANS data revealed that BSA unfolded to a partially-unfolded F form at surfactant concentrations of approximately 0.7 mM and 1.9 mM under visible and UV light, respectively. As seen in Figure 2.2, as this N → F transition is crossed the α-helix percentage of BSA undergoes a second decrease of ~ 10%, now accompanied by a similar increase in unordered structures. At the same time, the percentage of total β structures remains largely unchanged, with the β-turn proportion increase being offset by the β-strand decrease. Thus, the N → F transition appears to be primarily a helical-to-unordered structural transition of the protein. Similar helical-to-unordered transitions have been observed in 30 many proteins with unfolding, and in the case of BSA both thermal 111, 112 and chemical (urea or guanidine hydrochloride) denaturantion 148 have been shown to disrupt α-helices in favor of unordered structures. Furthermore, ionic surfactants (e.g., SDS, and dodecyltrimethylammonium bromide (DTAB)) have also been shown to induce helical- to-unordered transitions in BSA, in some cases accompanied with a slight increase in β structures. 149 Interestingly, within the F-form region at higher surfactant concentrations, the α- helix content gradually reaches a constant value with no further decrease observed at higher surfactant concentrations. The fact that the remaining helical content is relatively large (~47% under visible light and ~50% under UV light) agrees with the suggestion proposed by Takeda et al. that surfactants (e.g. sodium n-alkyl sulfates and n-alkyl trimethylammonium bromides) do not disrupt the helical structure in the six large loops in BSA. 149 Similar values of this residual helical content in “unfolded” BSA have been reported with thermal (44 % at 65 °C), 111 acidic (~42%), 113 or surfactant (47% − 53%) 145, 149 denaturation. However, judging BSA unfolding solely from the helical content in Figure 2.2 would result in erroneous estimates of the degree of denaturation, since the most dramatic unfolding (the F → E transition discussed below) occurs independent of loss of helical structure. This illustrates the importance of combined secondary and tertiary structure measurements in protein folding studies. F → E transition. Increasing the azoTAB concentration into the region where BSA adopts the aforementioned elongated E form, while not demonstrating any significant changes in the helical structure (indeed, the “kinks” observed in the E form 31 from SANS are consistent with these unfolded α-helical segments), does result in an increase unordered structures with a decrease in β structures. Similar β-to-unordered transitions induced by surfactant has been observed in immunoglobulin G (IgG) which contains a large amount of β structure. 155 In the presence of SDS/Tween 20 at 20°C, a significant amount of IgG β-sheets are transformed to unordered structures due to the destabilizing effect of the hydrophobic surfactant tails penetrating the hydrophobic domains of the protein and weakening the intramolecular hydrophobic interactions within the protein. Overall, the secondary structure profiles indicate that with visible-light illumination (azoTAB in the relatively hydrophobic trans state) BSA is more unfolded, exhibits less α-helical and β structures, and has more unordered domains compared to UV-light illumination (hydrophilic cis form). This observation is consistent with studies of BSA unfolding induced by a variety of sodium n-alkyl sulfate and n-alkyl trimethylammonium bromides surfactants, where it was shown that increasing the length of the alkyl tail (resulting in a more hydrophobic surfactant) lowered the surfactant concentration required to unfold the protein. 147 In addition, there are some more subtle differences between the effects of the trans and cis isomers of azoTAB on BSA unfolding that cannot be explained by simple differences in hydrophobicity. Under visible light, the final values of the α-helical, β-sheet, and random structures at high azoTAB concentrations all differ from the asymptotic values under UV light, implying a fundamental difference in the mechanisms of interaction of the planar trans and bent cis isomers of azoTAB with the protein. Indeed, it would be expected that the trans 32 conformation of azoTAB would present less steric hindrance for the hydrophobic interaction of the benzene rings of the surfactant with the protein. Thus, some moieties of the protein could be inaccessible to the relatively bulky cis form of azoTAB, thereby resulting in a slightly smaller degree of BSA unfolding. Difference spectroscopy. To compare the effects of visible and UV light illumination on the secondary structure of BSA in azoTAB solutions, difference spectra were calculated by subtracting the protein spectrum obtained under visible light from the spectrum under UV light at a given surfactant concentration, as shown in Figure 2.3. Positive peaks in the difference spectra then correspond to an increase in absorbance upon illumination with UV light, while negative values indicate that the absorbance is greater under visible light. At low surfactant concentrations (< 1.99 mM) a broad negative peak at ~1650 cm -1 is observed, which could be resolved through Fourier self- deconvolution (data not shown) into a positive peak at 1656 cm -1 , a negative peak at 1651cm -1 , and a broad negative peak at 1645 cm -1 . The negative peak at 1645 cm -1 indicates that unordered structures appear at the early stages of azoTAB-initiated protein unfolding under visible light (e.g., this negative peak was seen at [azoTAB] as low as 0.99 mM under visible light, see also Figure 2.2(c)). The peaks at 1656 and 1651 cm -1 are likely due to a shift of the α-helix band to lower wavenumbers caused by a conformational or environmental change upon protein unfolding. 63, 81 Specifically, solvent (D 2 O) would have a higher accessibility to the protein under visible light (BSA exists as the N form under UV light and the F form under visible light at 1 mM 33 -0.004 -0.002 0 0.002 0.004 0.006 0.008 0.01 0.012 1600 1620 1640 1660 1680 1700 0.99 mM 1.99 mM 2.98 mM 4.05 mM 5.02 mM 6.57 mM 8.97 mM 10.04 mM 13.18 mM Δ Absorbance Wavenumber (cm -1 ) Figure 2.3: Difference spectra of BSA defined as ΔA = spectra collected under UV light – spectra collected under visible light at a constant surfactant concentration in the amide I region. azoTAB), leading to an increase in hydrogen (deuterium) bonding along the polypeptide backbone, well known to cause a shift of the α-helix band to lower wavernumbers. 81 The difference spectra at low azoTAB concentration also display an upward-trending peak at ~1640 cm -1 , agreeing with the previous observations in Figure 2.1 that the “hydrated peak” associated with bound water molecules is less prevalent under visible versus UV light. As the surfactant concentration is increased from 2 to 3 mM, the difference spectra in the helical region inverts to positive values, indicating that the mismatch between the α-helix peak positions under visible and UV light no longer exists. Note that this is similar to the azoTAB concentration where BSA exists as the F conformation under both 34 visible and UV light (2.21 mM azoTAB), thus, the α-helical segments have approximately the same exposure to solvent, independent of the light conditions. With further increases in the surfactant concentration an increasingly positive α- helix peak develops in the difference spectra, a result of two separate effects. The relatively hydrophobic trans conformation of azoTAB leads to a greater loss of helical structure in Figure 2.2(a) than the cis isomer, which would give the observed positive values in the difference spectra in the helical region in Figure 2.3. In addition, the width of the α-helix band is observed to increase on going from UV to visible light illumination, in agreement with the general correlation between an increase in bandwidth accompanying greater protein flexibility, 6 again related to the fact that the protein is less compact under visible light. The combined effects of a decrease in absorbance along with a broadening of the α-helical band as the surfactant concentration is increased lead to increasingly sharper peaks in the difference spectra. A negative peak in the difference spectra at ~1635 cm -1 becomes evident for azoTAB concentrations greater than 4 mM, which could be deconvoluted into a negative peak at 1645 cm -1 (unordered structures, not directly visibly from the difference spectra due to overlap with the strong positive peak at 1651 cm -1 ) and a relatively small upward- trending peak at ~1630 cm -1 (β-strands). Taken as a whole, the difference spectra support the trends observed in the quantitative analysis of the protein secondary structure in Figure 2.3, and confirm the suggestion that the trans azoTAB surfactant induces more protein unfolding and unordered structures than the cis isomer. 35 Photoreversible protein folding observed with FT-IR. A primary objective of this study was to demonstrate control of protein conformation in a photoreversible manner. While in the previous SANS study it was demonstrated that the tertiary structure of BSA could be recovered with light illumination, 97 it was not possible to detail the precise changes in the local folding state. As shown in Figure 2.4, precise photoreversible control of BSA secondary structure is indeed observed through repeated visible ↔ UV light cycles. Similar cycles were applied at all of the surfactant concentrations utilized in Figure 2.2 (data not shown) with similar results (i.e., reversibility within the resolution of the FT-IR technique). 0 0.05 0.1 0.15 0.2 0.25 0.3 1600 1620 1640 1660 1680 1700 vis UV vis UV vis UV Absorbance Wavenumber (cm -1 ) Figure 2.4: FT-IR spectra of a BSA solution containing 5.02 mM azoTAB under repeated light cycles (visible → UV → visible → UV → visible → UV) in the amide I region. 36 While conformational changes in proteins can be induced by the addition of a variety of chemical denaturants or through changes in temperature or pressure, in many of these cases the changes in protein structure are not reversible, primarily due to exposure of the hydrophobic domains of the protein upon unfolding. For example, while the structural changes in BSA with thermal denaturation are reversible below 50 °C, higher temperatures lead to the formation of intermolecular β-sheets and irreversible aggregation of the protein. 99, 111, 112 Thus, to achieve truly reversible unfolding typically requires that the hydrophobic moieties of the protein be protected in some manner. For this purpose, surfactants have demonstrated great utility in protein folding studies, where binding of the surfactant to hydrophobic groups inherently protects the unfolded protein from aggregation. In fact, a surfactant-based system was used in one of the few successful examples of reversible unfolding of a membrane protein, where exchanging from a “harsh” to a “mild” detergent through dialysis or rapid mixing permitted the refolding of bacteriorhodopsin. 27, 75, 101, 103 In the case of azobenzene-based photoresponsive surfactants, exchanging the nature of the surfactant can be achieved with light illumination, providing a simple and fast technique to reversibly control the conformation of proteins. 2.5 Conclusion The ability to control protein conformation at the secondary structure level with light illumination using photoresponsive surfactants has been demonstrated in this study. The visible-light (trans) form of the azobenzene surfactant is more hydrophobic and 37 exhibits more interaction with the protein than the UV-light (cis) form of the surfactant. As a consequence, exposing the surfactant-protein solution to visible light results in a greater degree of protein unfolding, while exposing to UV-light causes surfactants molecules to dissociate from the protein and the protein to refold. The FT-IR spectroscopic data revealed three secondary structure transitions of BSA induced by photoresponsive surfactant and light, related to tertiary structure transitions observed with previous small-angle neutron scattering experiments. At low surfactant concentrations (0.55 mM and 1.60 mM under visible and UV light, respectively), the secondary structure of the protein undergoes an α-helix → β-structure transition, occurring in the same region where the heart-shaped native state of BSA in solution becomes slightly distorted (N → N d transition). Intermediate surfactant concentrations result in an α-helix → unordered structural transition as the tertiary structure of the protein was found to unfold from the native to a partially unfolded conformation (F form). At higher surfactant concentrations, a β → unordered transition is observed with the protein adopting a highly-elongated shape in solution (E form). As a result of these combined secondary and tertiary structure measurements, new insight into the nature of BSA unfolding was obtained. Finally, the use of photoresponsive surfactants was demonstrated to provide a means to precisely and reversibly unfold/refold proteins with simple light illumination, leading to a new tool in the study of protein folding phenomena. 38 2.6 Acknowledgements We acknowledge the Charles Lee Powell Foundation and the James H. Zumberge Faculty Research and Innovation Fund for support of this research. 39 Chapter 3 Enhanced Enzymatic Activity through Photoreversible Conformational Changes * 3.1 Abstract The interaction of a light-responsive surfactant with lysozyme at pH 5.0 has been investigated as a means to control protein structure and enzymatic activity with light illumination. The cationic azobenzene surfactant undergoes a reversible photoisomeriztion upon exposure to the appropriate wavelength of light, with the visible- light (trans) form being more hydrophobic and, thus, inducing a greater degree of protein unfolding than the UV-light (cis) form. Conformational changes as a function of photoresponsive surfactant concentration and light illumination were measured through shape-reconstruction analysis of small-angle neutron scattering (SANS) data. The SANS-based in vitro structures indicate that lysozyme transitions from a native-like structure at low surfactant concentration to a partially-unfolded conformation at higher surfactant concentrations under visible light illumination, while UV-light illumination causes the protein to refold to a near-native structure. Protein swelling occurs principally away from the active site near the hinge region connecting the α and β domains, leading * The Chapter is a published article (Biochemistry 46, 14557-14566, 2007.) 40 to an increase in the observed separation distance of the α and β domains in the ensemble SANS measurements, a likely result of enhanced domain motions and increased flexibility within the protein. This swelling of the hinge region is accompanied by an 8- fold increase in enzymatic activity relative to the native state. Both enzyme swelling and superactivity observed under visible light can be reversed to native-like conditions upon exposure to UV light, leading to complete photoreversible control of the structure and function of lysozyme. 3.2 Introduction Azobenzene-based photoresponsive surfactants have recently been utilized to induce reversible changes in protein conformation with light illumination, with relatively high resolution in vitro protein structures during the structural transitions determined with small-angle neutron scattering (SANS). 58, 59, 97, 161 A photoisomerization between the trans (relatively hydrophobic) and cis (relatively hydrophilic) forms of the azobenezene moiety allows photocontrol of a wide range of surfactant properties, 44 including interaction with various protein domains. For bovine serum albumin (BSA) in the presence of the photosurfactant, the initial unfolding events were localized to the hydrophobic α-helical segments in the C-terminal portion of the protein. 97, 161 Furthermore, through changes in light illumination, reversible transitions between intermediately-folded conformations were achieved. In contrast, for α-chymotrypsin with a primarily β structure, only small changes in the overall size of the protein were observed (~7% increase in the radius of gyration), however, this subtle structural 41 rearrangement was sufficient to convert intramolecular β structures into intermolecular β sheets and lead to eventual amyloid fibril formation. From the pre-amyloid oligomer structures determined with SANS, photoreversible transitions from corkscrew-like hexamers to rope-like dodecamers were observed, potentially capturing the initial stages of fibril formation. For lysozyme containing both α and β domains, SANS data indicated that the photosurfactant primarily swelled the α-domain of the protein and particularly helix A, while the β-domain and the active-site cleft remained relatively intact. 58 From these structural studies in lysozyme, the question remains as to what effect, if any, would the reversible changes in protein conformation have on enzymatic activity. Protein function is, to as large extent, determined by protein conformation, particularly in the case of enzymes where folding results in an active site that allows for selective binding of substrates. However, the static form-function relationship of the classic “lock-and-key” mechanism has been replaced in modern enzymology with the view that enzyme dynamics can have an equally important role in catalysis. Thus, various hypotheses have been proposed with this view in mind, 9, 17, 22, 116, 171 all with the underlying theme that significant conformational flexibility is required during the course of the reaction. 8 Thus, it is generally viewed that for the enzyme-substrate complex to surpass the activation-energy barrier requires “conformational sampling” 9 or “dynamic excursions” 135 along the reaction pathway towards the formation of transition-state conformations. 61, 153 Interestingly, this flexibility is not necessarily proximal to the active site, and instead may be in distal regions away from the active site. 9, 151 This has been demonstrated through various mutation studies where replacement of distal amino acids 42 has led to concurrent changes in reaction rates and protein flexibility, potentially by increasing the probability of sampling transition-state conformations. 21, 62, 151 These results seem to suggest a general procedure by which enzymatic activity could be increased through enhancements in enzyme flexibility, provided that these enhancements are not achieved at the expense of denaturing the active site. Lysozyme catalyzes the hydrolysis of β(1→4) linked polysaccharide copolymers of N-acetylglucosamine and N-acetylmuramic acid found in certain bacteria cell walls, as well as homopolymers of N-acetylglucosamine (i.e., chitin). The enzyme active site, divided into 6 subsites A through F, resides in a cleft between the two domains of lysozyme: the α-domain folded around a central hydrophobic core containing four α- helices and one 3 10 helix, and a sheet-like β-domain consisting mainly of hydrophilic residues either on the outer surface of the molecule or lining the cleft. The cleavage of β(1→4) linkages occurs between sites D and E close to the catalytic residues Glu35 and Asp52, 162 leading to formation of an oxocarbenium intermediate that is electrostatically stabilized by Asp52. This intermediate can then be hydrolyzed upon direct attack by a nucleophilic water molecule (hydrolysis) completing the reaction, or a β(1→4) linkage can be regenerated upon reaction with a second substrate molecule that becomes bound to the vacant subsites E and F (transglycolsylation). 49, 77, 82 Enhanced lysozyme activity has been observed upon modification of specific amino acids within the protein. For example, a mutant lysozyme with deleted Arg14 and His15 residues (both located in helix A of the α domain distal to the active site) was observed to exhibit increased activity (~140 %) attributed to enhanced mobility of the 43 residues near or at the active site. 78, 109 Conversely, the presence of ionic surfactants such as sodium n-alkyl sulfates and n-alkyl trimethylammonium bromides usually deactivate the enzyme by interacting directly with the active site or indiscriminately denaturing the protein. 64, 65, 84, 100 These alkyl-based surfactants unfold all regions of the protein through non-specific interactions with protein hydrophobic domains. In contrast, the “localized swelling” of lysozyme observed in the presence of the azobenzene-based photosurfactant in regions away from the active site suggest that the effect of the photosurfactant on lysozyme activity could be unique from traditional surfactants, potentially increasing reactivity through enhancements in protein flexibility. In the present work, the conformation and activity of lysozyme are controlled through the use of a photoresponsive surfactant and light illumination. Shape-reconstruction analysis 72, 144 of small-angle neutron scattering data is used to provide relatively high- resolution information on the location of protein unfolding. The observed conformational changes are correlated with enzyme activity measured through two different assays, Micrococcus Luteus and glycolchitin. The photosurfactant is found to swell the hinge region connecting the α and β domains, leading to a more flexible protein and resulting in dramatic increases in enzyme activity. Moreover, the observed light-induced superactivity can be photoreversibly controlled through surfactant isomerization. 44 3.3 Experimental Details An azobenzene trimethylammonium bromide surfactant (azoTAB) of the form N N CH 3 CH 2 O(CH 2 ) 4 N + (CH 3 ) 3 Br was synthesized according to published procedures. 66, 136 The surfactant undergoes a reversible photoisomerization upon exposure to the appropriate wavelength of light with the trans isomer (434-nm visible light) exhibiting a lower dipole moment and, hence, being more hydrophobic than the cis isomer (350-nm UV light). 136 To eliminate the potential of UV deactivation of the enzyme, 40, 41, 138 cis surfactant solutions were pre-converted under UV-light from an 84-W long wave UV lamp, 365 nm (Spectroline, model number XX-15A) for at least 30 min prior to the addition of an enzyme stock solution. The combined solutions were then maintained in the dark during the entire reaction period of 1-3 hours, with absorbance spectra measured after each experiment to ensure the surfactant remained in the cis form (the half-life of dark conversion from the cis to the trans form is ~24 hours). 96 However, control experiments demonstrated that activity was unaffected by direct UV illumination of the enzyme in the presence of the photosurfactant. Thus, azoTAB appears to offer similar protective properties as other UV scavengers such as ascorbate 40, 41 due to the strong absorbance of the surfactant in the UV region. In contrast, pure enzyme showed an ~70% decrease in activity upon exposure to UV light. 45 For the dynamic photo-response assays and optical microscopy, conversion to the cis form was achieved by illuminating the enzyme-surfactant solutions with a liquid light guide (Oriel, model number 77557) attached to a 200-W mercury arc lamp (Oriel, model no. 6283) equipped with a 320-nm band-pass filter (Oriel, model no. 59800) in combination with a heat-absorbing filter (Oriel, model no. 59060), effectively isolating the 365-nm mercury line (UV-A). Conversion back to the trans form was achieved with a 400-nm long-pass filter (Oriel, model no. 59472) to isolate the 436-nm mercury line. Enzymatic Assays Highly-purified lysozyme from hen egg white (L7651), lyophilized Micrococcus luteus (M3770), glycolchitosan (G7753), Remazol brilliant blue R (R8001) and phosphate buffer (8.3 mM) were purchased from Sigma and used as received. The buffer was adjusted to pH 5.0 with the addition of HCl. The standard lysozyme assay of monitoring the decrease in optical density during lysis of Micrococcus leteus cells walls was deemed inappropriate in the presence of azoTAB due to potential surfactant-induced cell aggregation (see below). Hence, the alternative assays described below were utilized. Preparation of Micrococcus luteus conjugated with Remazol brilliant blue (Ml- RBB). Ml-RBB was prepared as described by Ito et al. 80 To 40 mL of a suspension of Micrococcus luteus cells (15 mg/mL), a solution of 400 mg of Remazol brilliant blue R (RBB-R) in 40 mL distilled water was slowly added under constant stirring at 50 °C. Subsequently, 8 g of sodium sulfate was added over the course of 30 mins. A solution of 46 400 mg trisodium phosphate in 4 mL distilled was then added and the mixture was stirred at 50 °C for another 30 mins. The mixture was centrifuged at 2600 rpm for 10 mins and the supernatant was discarded. The pellet was washed with 40-mL phosphate buffer (50 mM) until the supernatant was colorless, followed by washing twice with distilled water. Ml-RBB was then lyophilized and stored at –20 °C. Preparation of glycolchitin conjugated with Remazol brilliant blue (glycolchitin- RBB). Glycolchitin-RBB was obtained by acetylation of glycolchitosan 129 followed by coloration with RBB-R. 166, 168 Briefly, 1.5 g of glycolchitosan, a water-soluble derivative of chitosan, was dissolved in 150 mL sodium tetraborate solution (100 mM). Acetic anhydride was then slowly added under constant stirring until an acetic acid concentration of 2 wt % was reached, followed by adjusting to pH 9 with NaOH. After 30 mins, glycolchitin was precipitated from the mixture by addition of acetonitrile, repeated several times until the pH of the glycolchitin solution was neutral. The product obtained was then dissolved in 75 mL water and gently heated to 50 °C, with a RBB-R solution (150 mg/mL) slowly added under constant stirring. After an hour, 3 g of sodium sulfate was added in several aliquots, with 0.3 g of trisodium phosphate subsequently added with the reaction continued for another 75 mins at 50 °C. Glycolchitin-RBB was then dialyzed against water for two days to remove excess dye, salts, and low molecular weight product. The final product was lyophilized and stored at –20 °C. Lysozyme activity against Ml-RBB. 1.6 mg/mL of Ml-RBB was suspended in an 8.3 mM, pH 5.02 phosphate buffer. A 1.0 mL lysozyme solution (0.008 mg/mL), 1.0 mL buffer solution, and the desired amount of a stock surfactant solution were then added to 47 2.0 mL of the Ml-RBB suspension. The reaction mixture was then incubated at 37 °C with continuous, gentle stirring. At suitable time intervals, 400 μL of the reaction volume were withdrawn and immediately vortexed for 10 sec to quench the reaction due to unfolding of the enzyme induced by exposure to the air-liquid interface, 45, 133 followed by centrifugation at 15,000 rpm for 5 min to remove the insoluble cell walls, leaving the hydrolyzed reaction product remaining in the supernatant. The absorbance of the supernatant at 600 nm was measured and found to increase linearly with time for two hours. Lysozyme activity at different conditions was determined from the initial rate of increase of the absorbance at 600 nm due to the dyed product, expressed as a percentage relative to pure lysozyme. Activity of a lysozyme solution was measured after vortexing the enzyme solution for 10 sec confirming loss of activity due to unfolding. Lysozyme activity against glycolchitin-RBB. 1 mL of glycolchitin-RBB (2 mg/mL) was mixed with 0.5 mL of an azoTAB solution at appropriate concentrations and 0.5 mL of a lysozyme solution (0.008 mg/mL) in an 8.3 mM, pH 5.02 phosphate buffer to give a final concentration of 1 mg/mL glycolchitin-RBB and 0.002 mg/mL lysozyme. The mixture was incubated at 37 °C under gentle stirring. At suitable time intervals 200 μL of the reaction volume was withdrawn and mixed with 200 μL of acetoniltrile to quench the reaction and precipitate non-reduced glycolchitin. The mixture was then cooled on ice and centrifuged at 15,000 rpm for 5 min at room temperature. Lysozyme activity was determined from the initial rate of increase of the absorbance at 600 nm over the range of 1 hour, expressed relative to the rate of pure lysozyme. 48 Kinetic parameters of lysozyme with and without the presence of azoTAB were obtained using substrate concentrations ranging from 0.01–0.44 wt % of glycolchitin- RBB. The reaction mixture was incubated at 37 °C for 1 hour. The maximum initial velocity V m and the apparent Michaelis constant K M of the enzyme were determined from linear-regression analysis of double-reciprocal Lineweaver-Burk plots. Data that demonstrated inhibition due to high substrate concentration and, thus, presented an upward trend in the Lineweaver-Burk plots were excluded from the analysis of kinetic parameters. According to the Lineweaver-Bruk equation (1/v = 1/V m + K M /V m [S], where v and [S] represent the initial velocity and substrate concentration, respectively), K M was obtained from the x-axis intercept of −1/K M , while V m was determined from the y-axis intercept of 1/V m . Due to the inability to determine the accurate molecular concentration of the polymer substrate in the reaction mixture, V m is presented as the change of absorbance at 600 nm per minute. Optical microscopy of ML-RBB cells. 2.0 mL of a stock azoTAB surfactant solution were added to 2.0 mL of a 1.6 mg/mL Ml-RBB suspension in an 8.3 mM, pH 5.02 phosphate buffer. The samples were then observed with an Olympus IX71 inverted microscope equipped with a 40× objective lens (SLCplanFl) and a 1.6× magnification changer resulting in 64× total magnification, and recorded with a CCD digital camera (Hamamatsu, model no. C4742-95). At each azoTAB concentration, the same solution was used to obtain images under both visible and UV light with the samples exposed to UV light for at least 30 mins to convert the surfactant to the cis form. 49 Small-Angle Neutron Scattering The neutron scattering data were collected on the 30-m NG3 SANS instrument at NIST. 53 Two sample-detector distance were used (1.33 and 7.0 m) combined with a 25- cm detector offset to give a Q-range of 0.0048–0.46 Å -1 , where Q = 4πλ -1 sin(θ/2) and θ is the scattering angle. The net intensities were corrected for the background and empty cell (pure D 2 O), followed by accounting for the detector efficiency using the scattering from an isotropic scatterer (Plexiglass), and then converted to an absolute differential cross section per unit sample volume (in units of cm -1 ) using an attenuated empty beam. The coherent scattering intensities of the sample were obtained by subtracting the incoherent contribution from the hydrogen atoms in lysozyme (0.0004 cm -1 ) and the surfactant (0−0.0012 cm -1 ). The SANS data were analyzed using three complementary techniques: Guinier analysis, calculation of the pair distance distribution functions (PDDFs), and a shape- reconstruction algorithm. The PDDFs were calculated assuming a monodisperse system using GNOM 144 over a Q-range of ca. 0.02–0.3 Å -1 to exclude protein intermolecular interactions at low Q. 58, 59 The maximum particle diameter (D max ) was selected to give a smooth return of the PDDF to zero at D max . The shape reconstructions were performed by approximating lysozyme as containing 1000 scattering centers in the program GA_STRUCT 72 over a Q-range of 0.01–0.3 Å -1 , again to exclude intermolecular interactions and to avoid length scales too small for protein continuity at high Q. Briefly, GA_STRUCT utilizes a genetic algorithm to optimize the positions of the 1000 scattering centers until the calculated scattering data best fit the experimental data. Ten 50 independent runs are performed, with the individual protein shapes from each run averaged to give the consensus envelope. 72 3.4 Results and Discussion The ability to control lysozyme conformation with azoTAB surfactant at pH 5.0 is shown in the SANS data in Figure 3.1 as a function of azoTAB concentration and light conditions. Under visible light at even the lowest azoTAB concentration studied (3.6 mM), the scattering curves begin to deviation from pure lysozyme at Q ~ 0.2 Å -1 or length scales (L = 2π/Q) of approximately 31 Å, similar to the diameter of lysozyme (36 Å). 140 This suggests that lysozyme swells with increasing trans azoTAB concentration. Under UV light, however, with the surfactant converted to the cis state, evidence of swelling is not observed until 12.0 mM azoTAB, or about 3–4 times the concentration under visible light. Thus, over a wide concentration region photoreversible protein folding can be achieved, similar to previous results obtained at pH 7. 58 From the data in Figure 3.1, radii of gyration (R g ) were calculated from the Guinier approximation I(Q) = I(0)exp(−Q 2 R g 2 /3), valid in the region QR g < 1.3. As seen in Table 3.1, azoTAB under both visible and UV light increases the values of R g relative to the pure lysozyme (R g = 12.9 Å, in good agreement with published values of the native state of 13.3 and 13.5 Å 58, 141 ), with again the deviation from the native state greater under visible comparing to UV light. Note, however, that even the largest value of R g = 19.2 Å in Table 3.1 is relatively low compared to values reported for denatured lysozyme 51 in urea (R g = 28.7 Å) and alcohol (R g = 24.9 Å), 85 suggesting relative mild swelling of the protein with azoTAB. 0.001 0.01 0.1 1 0.01 0.1 I(Q) / cm -1 Figure 3.1: SANS data of lysozyme-azoTAB solutions as a function of surfactant concentration under visible (closed symbols) and UV (open symbols) light. Pure lysozyme ( ●), 3.6 mM azoTAB ( S , U ), 8.5 mM azoTAB ( , ), 12.0 mM azoTAB ( ¡ , ), and 18.6 mM azoTAB ( T , V ). [Lysozyme] = 10 mg/mL in pH 5.0 buffer. Q / Å -1 52 Table 3.1: Values of the radius of gyration determined from Guinier analysis of the SANS data in Figure 3.1. Guinier analysis [azoTAB] (mM) R g (Å) 0.0 12.9 visible light 3.6 13.9 8.5 17.0 12.0 17.8 18.6 19.2 UV light 3.6 12.8 8.5 12.8 12.0 13.2 18.6 14.2 Pair distance distribution functions (PDDFs) shown in Figure 3.2, related to the probability P(r) of finding two scattering within the protein centers a distance r apart, were calculated from the SANS data in Figure 3.1. For a globular protein, the PDDF is expected to have a symmetric, inverse parabolic shape with a peak position and maximum dimension (D max ) approximately given by the protein radius and diameter, respectively. As seen in Figure 3.2, increasing azoTAB concentration under visible illumination results in D max increasing from 42 Å to 57 Å, while that maxima shift from ~15 Å to 22 Å. Under UV light less effect is again observed, consistent will a smaller degree of unfolding with the cis surfactant. For comparison, D max increases from 42 Å to 75 Å in lysozyme denatured with urea. 28 53 0 0.0005 0.001 0.0015 0.002 0.0025 0.003 0 1020 3040 5060 P(r) r / ? 18.6 mM 18.6 mM 12.0 mM 8.5 mM 3.6 12.0 0, 8.5, 3.6 mM Figure 3.2: PDDFs of lysozyme-azoTAB solutions as a function of surfactant concentration under visible (black lines) and UV (gray lines) light. [Lysozyme] = 10 mg/mL in pH 5.0 buffer. To gain more precise information on the nature of protein unfolding with azoTAB, a shape-reconstruction algorithm was applied to the SANS data. As previously described, 58 the protein is approximated as a collection of scattering centers whose positions are adjusted to fit the experimental scattering curve. The results of this shape reconstruction analysis are shown in Figure 3.3 along with the X-ray crystallographic structure of lysozyme (PDB code 6LYZ). The structure from the run best fitting the data are shown in blue, while the “consensus envelops” obtained by averaging the 10 independent runs for each data set are displayed in red to demonstrate consistency of the fits. Although the resolution of the SANS technique (~2π/Q max ) 144 is reduced compared r / Å -1 54 to X-ray crystallography, these and similar 72, 143 structures have demonstrated the ability of SANS to determine precise structural detail of proteins in vitro, such as location of the active site cleft of lysozyme between the α and β domains, 58 the mechanism of BSA unfolding, 97, 161 and the rope-like structures of pre-amyloid oligomers. 59 As shown in Figure 3.3, SANS can determine the conformation of partially- unfolded conformations unattainable with traditional crystallography. Thus, while at low azoTAB concentrations (3.6 mM under visible light and up to 12 mM under UV light) the SANS-based structures are similar to the crystal structure of pure lysozyme, the true utility of SANS is seen at elevated azoTAB concentrations were the degree, and location, of protein swelling can be determined. From examination of Figure 3.3, lysozyme is observed to swell primarily in the lower portion of the molecule away from the active site in the so-called hinge region, which gives rise to a progressively open active-site cleft. This unfolding mechanism in similar to previous SANS and FT-IR measurements at pH 7, which demonstrated swelling induced by a similar azoTAB derivative was also in the hinge region and accompanied by a loss of α-helical content. 58 The similarity in the unfolding mechanism observed in Figure 3.3 at pH 5 and in previous work at pH 7 illustrates the robust nature of shape-reconstruction of SANS data to examine the structure of partially-unfolded proteins. 55 Figure 3.3: In vitro conformations of lysozyme in solution determined from shape- reconstruction analysis of the SANS data in Figure 3.1. Best-fit structures are shown in blue and consensus envelopes are shown in red. The crystal structure of lysozyme (PDB code 6LYZ, space-filling and ribbon) is shown for comparison with arrows pointing to the active-site cleft between the α and β domains. visible light UV light 3.6 mM 8.5 mM 12.0 mM 18.6 mM 3.6 mM 8.5 mM 12.0 mM 18.6 mM pure lysozyme (6LYZ) 56 Note that the appearance of swollen regions in Figure 3.3 cannot be due to simple surfactant aggregation on the protein. This alternative mechanism, common with proteins in the presence of high concentrations of sodium dodecyl sulfate (SDS), is often termed the necklace-and-beads model, with SDS micelles (the “beads”) aggregating along the unfolded protein chain (the “necklace”). 30 In contrast, the SANS measurements in Figure 3.1 at 10 mg/mL (0.69 mM) lysozyme would give at most from 5 to 26 surfactant molecules bound per protein, assuming complete binding. These values are much smaller than typical aggregation numbers of even a single micelle “bead” ranging from 40–100 SDS molecules per micelle. 76, 152 Nevertheless, a control experiment was performed at the protein contrast-matching point (60/40 H 2 O/D 2 O) 7 to render the protein “invisible” to the neutron beam, giving a SANS spectrum (not shown) of essentially zero scattering from azoTAB alone. This conclusively demonstrates that azoTAB micellar aggregation on the protein is not contributing to an artificial increase in the scattering intensity; thus, the swollen regions in Figure 3.3 are due to protein unfolding. From these SANS-based in vitro conformations, the question remains as to what effect, if any, the increased exposure of the active-site cleft would have on lysozyme activity. Based on the conformations in Figure 3.3, however, any effect of azoTAB on lysozyme activity is expected be photoreversible since the trans, visible-light form of the surfactant induces a greater degree of active-site exposure compared to the cis, UV-light conformation. Lysozyme activity determined by Ml-RBB. To determine the effect of surfactant and photo-initiated conformational changes on protein function, hydrolysis of dye-labeled 57 Micrococcus luteus (Ml-RBB) with lysozyme was measured as a function of azoTAB concentration and light condition, as shown in Figure 3.4(a). At low azoTAB concentrations, lysozyme activity is enhanced with surfactant under both UV and visible light, while elevated azoTAB concentrations result in a maximum in the activity curves and eventual deactivation of the enzyme. Interestingly, both the onset of “superactivity” and enzyme deactivation are observed at lower azoTAB concentrations under visible compared to UV light. While an enhancement of activity relative to the native state could be consistent with the increase of active-site exposure in the SANS-based in vitro structures above (discussed further below), the origin of the deactivation step remains unclear. At least two possible factors could be responsible for this latter effect. First, azoTAB concentrations greater than 0.5 mM could induce complete protein denaturation as opposed to swelling in Figure 3.3, with the relatively-hydrophobic trans form of the surfactant resulting in greater unfolding compared to the relatively-hydrophilic cis isomer. Note that while 0.5 mM azoTAB is low compared to the surfactant concentrations in Figure 3.3, the significant difference in lysozyme concentrations required for the activity (0.002 mg/mL) and SANS (10 mg/mL) measurements do not allow ruling out this phenomenon. 58 0 100 200 300 400 500 00.511.5 22.5 3 Relative Activity (%) [azoTAB] (mM) Figure 3.4: (a) Effect of azoTAB concentration on lysozyme activity against Micrococcus luteus-RBB under visible (•) and UV ( ○) illumination. [lysozyme] = 0.002 mg/mL; [Micrococcus luteus-RBB] = 0.8 mg/mL. (b) Optical micrographs of Micrococcus luteus-RBB cells (0.8 mg/mL) as a function of azoTAB concentration and light conditions. (b) (a) UV light Visible [azoTAB] (mM) 1 2 59 Conversely, the decrease in activity at elevated azoTAB concentrations could result from surfactant interacting directly with the substrate, resulting in an effective loss of activity without changing the enzyme conformation. M. luteus is a gram-positive bacterium with a negatively-charge cell wall consisting of a rigid layer of highly cross- linked peptidoglycan embedded with teichuronic acids. At physiological pH, lysozyme exhibits a net positive charge (pI = 11.0), thus, the negative charge on the cell wall has been determined to be an important feature during hydrolysis. 129, 150 Thus, the presence of cationic azoTAB could neutralize the cell walls, leading to an effective decrease in enzyme reactivity. To examine this effect, optical micrographs of Ml-RBB cells were obtained under varying surfactant concentrations and light conditions, as shown in Figure 3.4(b). At low azoTAB concentrations (< 0.5 mM under visible [not shown] and 1 mM under UV light), the bacteria cells remain well dispersed in solution, indicating that much of the negative charge of the cells responsible for dispersion remains intact. With increased surfactant concentration, however, the M. luteus cells exhibit enhanced aggregation, becoming particularly pronounced beyond ca. 1 mM and 2 mM azoTAB under visible and UV light, respectively, consistent with regions of diminished activity in Figure 3.4(a). Apparently, enhanced cell aggregation causes the accessibility of lysozyme to the peptidoglycin substrate to be substantially reduced, explaining the loss of activity observed above. The hydrophobic trans isomer results in a higher degree of cell aggregation compared to the cis isomer. Interestingly, cell aggregation and return to the well-dispersed state could be repeatedly and reversibly initiated with visible ↔ UV light cycles and the appropriate azoTAB concentration (not shown). 60 0 100 200 300 400 500 600 700 800 0 0.2 0.4 0.6 0.8 1 Relative Activity (%) [azoTAB] (mM) Lysozyme activity determined by glycolchitin-RBB. Based on the above results, the standard Ml-RBB assay cannot properly assess the effect of azoTAB on lysozyme activity. Thus, a glycolchitin-RBB assay was used to minimize electrostatic interactions between the substrate and azoTAB. Glycolchitin is a neutral polymer composed of N- acetyl-D-glucosamine, and even when reacted with the anionic dye Remazol brilliant blue (degree of substitution ~ 2%) the net negative charge of glycolchitin-RBB is significantly lower than the M. luteus substrate. Lysozyme activity against glycolchitin- RBB as a function of azoTAB concentration and light conditions are shown in Figure 3.5. Figure 3.5: Lysozyme activity against glycolchitin-RBB as a function of azoTAB concentration under visible (•) and UV ( ○) illumination, respectively. [lysozyme] = 0.002 mg/mL; [glycolchitin-RBB] = 1.5 mg/mL; pH 5.0; 37 °C. 61 As was the case with the M. luteus substrate, enhanced activity over the native state is observed with increased azoTAB concentration. In contrast to Figure 3.4, however, the respective activities under visible and UV light level off at about 0.3 mM trans azoTAB and 0.6 mM cis azoTAB and remain essentially constant up to 12 mM azoTAB (data not shown). Thus, the decrease in activity at elevated azoTAB concentrations in Figure 3.4 does indeed appear to be an artifact of the cell aggregation due to neutralization. With the effects of cell wall aggregation removed, 500–600% superactivity is observed with as little as 0.1 mM azoTAB under visible light and 600–700% superactivity occurs at ~ 0.3 mM surfactant under UV light. While a small degree of photocontrolled activity was observed with Ml-RBB at low surfactant concentrations ([azoTAB] < 0.3 mM), in Figure 3.5 without the depressing effect of cell aggregation the differences in activity between the visible and UV states are better resolved. Thus, it appears that at low surfactant concentrations lysozyme exhibits higher activity in the presence of the hydrophobic trans isomer compared to systems containing cis azoTAB as well as the native state. These results correlate with the shape-reconstruction analysis of the SANS data in Figure 3.3, which indicate that the trans form of azoTAB induces a higher degree of protein swelling than the hydrophilic cis form. To determine enzyme kinetic parameters, initial velocity verses substrate concentration profiles and double-reciprocal Lineweaver-Burk plots of lysozyme against glycolchitin-RBB were generated, as shown in Figure 3.6. Interestingly, the initial- velocity profiles all go through a maximum in substrate concentration, suggesting eventual substrate inhibition of the enzyme. From the Lineweaver-Burk plots in Figure 62 3.6(b), substrate inhibition is evident at low values of inverse substrate concentration; thus, the kinetic parameters in Table 3.2 were obtained using the linear portion of the data from a slope (K M /V max ) and x-intercept (–1/K M ). Due to a very limited linear region for the cis azoTAB data (not shown), apparent kinetic parameters were only estimated for pure lysozyme and lysozyme in the presence of trans azoTAB. For native lysozyme activity against M. luteus, a similar substrate-inhibition effect has been observed, attributed to the strong electrostatic attraction between substrate and enzyme causing multiple attachments of substrates to the enzyme and hindering the substrate entering enzyme active site. 154 However, the inhibition effect observed in Figure 3.6 is less likely to be a result of electrostatic attraction considering the low charge of glycolchitin-RBB compared to M. luteus. Thus, the major inhibition effect may be the competing transglycolsylation reaction. 5 As mentioned above, the active site of lysozyme is divided into six subsites A-F. The scissile bond locates between subsites D and E, with cleavage of the β(1→4) linkage of the polysaccharide leading to the formation of a positively- charged oxocarbenium intermediate bound to site A through D and stabilized by Asp52. During the normal reaction pathway, a nucleophilic water molecule hydrolyzes this intermediate, forming a reduced-sugar product that is released from the active site. However, if a second substrate occupies the vacant sites E and F prior to hydrolysis, a β(1→4) linkage between the intermediate and this second substrate can occur, resulting in transglycosylation. 49, 77, 82 Thus, elevated substrate concentrations can increase the occurrence of the competing transglycosylation reaction, the likely origin of the substrate inhibition observed in Figure 3.6. 63 Figure 3.6: (a) Initial velocity profile of lysozyme against glycolchitin-RBB without ( ■) and with 0.2 mM trans azoTAB (•) and cis azoTAB ( ○). (b) Linweaver-Burk plot of native lysozyme ( ■) and lysozyme with 0.2 mM trans azoTAB (•). [lysozyme] = 0.002mg/mL; pH 5.0; 37 °C. Table 3.2: Effect of azoTAB on the kinetic parameters of lysozyme. [lysozyme] = 0.002 mg/mL, pH 5.0. azoTAB K M 10 4 ⋅V m (mg⋅mL -1 ) (A 600 ⋅min -1 ) 0 mM 0.34 2.06 0.2 mM (trans) 1.72 16.3 -40 -20 0 20406080 1/[substrate] (a) 0 0.0001 0.0002 0.0003 0.0004 0.0005 0.0006 0.0007 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 Initial Velocity (ΔA600/min) [substrate] (wt %) pure lysozyme cis azoTAB trans azoTAB 5000 1 10 4 1.5 10 4 2 10 4 1/V m (b) 64 As shown in Table 3.2, both of the Michaelis constant K M and the maximum velocity V max increase in the presence of 0.2 mM trans azoTAB. Assuming that the rate- limiting step is product formation, K M represents the dissociation constant between enzyme and substrate; thus, the increase of K M suggests a decrease in substrate binding affinity towards the slightly-unfolded form of lysozyme in the presence of trans azoTAB compared to the native state. In contrast, the 8-fold increase in the maximum velocity V max , which can be related to an increase in the enzyme turnover number k cat through the equation k cat = V m /[E] 0 , suggests that the swollen structures of lysozyme in the presence of trans azoTAB have the effect of increasing the overall reaction rate. Recalling the SANS-based in vitro structures can provide insight to these two effects. With the net increase in the separation distance of the two domains, likely a result of enhanced domain motions in the presence of trans azoTAB, the substrate binding affinity has been reduced due to a slight perturbation of the active site, yet the reaction is enhanced due to an increase in flexibility of the enzyme. The later occurrence of substrate inhibition with trans azoTAB may also be a result of enhanced flexibility favoring hydrolysis (increase in V max ) over transglycosylation (decrease in substrate binding affinity). In order to form the enzyme-substrate complex, it has been reported that the active-site cleft of lysozyme has to first open (to allowing the substrate to enter the active site) and then close (to return the enzyme to a state similar to the native conformation) through hinge-bending motions of the α and β domains. 31, 46, 54, 94, 104, 158, 172 Thus, it may be expected that domain motions play an important role in the enzyme catalytic process. Indeed, a number of studies support this relationship between conformational flexibility 65 and enzyme activity, 56, 122, 131, 170 with increases in flexibility and internal fluctuation leading to enhanced enzyme activity. 35, 78, 109, 115, 169 For the specific case of lysozyme, a mutant with residues Arg14 and His15 deleted has exhibited increased internal motions upon inhibitor binding and higher activity against glycholchitin comparing to the wild- type enzyme, despite the fact that both of these residues are distal the active-site cleft. 78, 109 Similarly, replacing the bulky tryptophan residue at subsite B of the active site with smaller tyrosine or phenylalanine residues gave looser binding of substrates yet enhanced activity by up to 200%. 93, 102 Conversely, when residues Met12 (α-helix) and Leu56 (β- sheet), which face each other across the cavity of the hydrophobic core in the α domain, were replaced with more hydrophobic residues, the mutant exhibited enhanced stability and rigidity and a reduction in activity, possibly due to restricted internal motions. 115 The rates of enzymatic reactions are controlled by the height of the activation energy barrier, or equivalently the probability of sampling transition-state conformations. Thus, to induce superactivity requires that enzyme flexibility be increased in such a way that these transition-state structures are preferentially sampled, with flexibility promoting conformational changes along the reaction pathway. 9, 61, 135, 153 With this view, the increase in activity observed under visible light could be a result of the higher degree of swelling localized near the hinge region, which would be expected to lead to enhanced hinge-bending motions and a net increase in the separation distance of the α and β domains. At the ensemble level as in Figure 3.3, this is manifested by an overall broadening of the active-site cleft, with the SANS structures representing the z-average of all conformations in solution, 59 analogous to regions with high temperature factors (B 66 factors) in X-ray crystallographic structures that are often associated with regions undergoing large thermal motions. Thus, azoTAB appears to induce superactivity in lysozyme by binding at a location removed from the active site and resulting in a more flexible enzyme undergoing fluctuating conformational changes. In situ photocontrol of enzyme activity. One potential application of the photoresponsive azoTAB surfactant is to be act as a photo-regulator in biocatalytic systems, where simple light illumination can be used for in situ control of enzyme activity. Figure 3.7 demonstrates this photoregulation of lysozyme activity against glycolchitin-RBB. In Figure 3.7(a) the reaction begins with 0.2 mM trans azoTAB in the cis state (~550% superactivity from Figure 3.5), followed by UV illumination to convert azoTAB to the cis state, immediately leading to a decrease in reaction rate (activity ~120%), a result of enhanced protein swelling under visible versus UV light. While similar photo-switching of biocatalytic activity has been obtained through covalent attachment of photoresponsive groups to an enzyme, 79, 164, 165 this enzyme modification process is relatively complex and time consuming compared to simple mixing of enzymes with the azoTAB surfactant. Furthermore, with covalent linkages enzyme activity is typically decreased slightly compared to the native state even in the “on” state. 79, 164, 165 In contrast, azoTAB offers a unique method to induce superactivity through interacting with hydrophobic, often α-helical regions of the protein removed from the active site. 67 0.16 0.18 0.2 0.22 0.24 0.26 0 10203040 506070 Absorbance Time (min) Figure 3.7: Photo-regulation of lysozyme activity against glycolchitin-RBB. (a) Reaction initiated with trans azoTAB ( ), followed by UV illumination to photoisomerize azoTAB to the cis state (---). (b) Reaction initiated with cis azoTAB (---), followed by visible illumination to photoisomerize azoTAB to the trans state ( ). [lysozyme] = 0.002 mg/mL, [glycolchitin-RBB] = 1.5 mg/mL, [azoTAB] = 0.2 mM, pH 5.0; 37 °C. 3.5 Conclusions The ability to photo-reversibly control the form-function relationship of lysozyme using the photoresponsive azoTAB surfactant has been demonstrated through measurements of enzyme activity combined with in vitro protein conformations determined with small-angle neutron scattering. Shape-reconstruction analysis applied to the SANS data indicates that the addition of azoTAB causes lysozyme to partially unfold relative to the native state, with swelling observed primarily away from the active site in the hinge region connecting the α and β domains. As a result, enhanced domain motions are detected as an increase in the average distance between the α and β domains, thereby suggesting that the active site become increasingly accessible upon protein swelling. (b) 0.08 0.1 0.12 0.14 0.16 0.18 0.2 0.22 0 10203040 506070 Time (min) Absorbance (a) trans trans cis cis 68 These results are consistent with activity assays that demonstrate that a 5– fold to 7–fold increase in lysozyme activity in the presence of azoTAB. As a result, in situ photo- control of enzyme activity was also demonstrated. Combined, these results provide for direct evidence of the connection between protein form and protein function, both of which can be photo-reversibly controlled. 3.6 Acknowledgements This material is based upon work supported by the National Science Foundation under Grant No. 0554115. We also thank W.T. Heller for graciously supplying the GA_STRUCT program. We acknowledge the support of the National Institute of Standards and Technology, U.S. Department of Commerce, in providing the neutron research facilities used in this work. 69 Chapter 4 Light-Induced Protein Dynamics Observed with Neutron Spin Echo Measurements 4.1 Abstract A light-responsive surfactant has been used as a means to control protein structure and dynamics with light illumination. The cationic azobenzene surfactant, which undergoes a reversible photoisomerization upon exposure to the appropriate wavelength of light, adopts a relatively hydrophobic, trans structure under visible light illumination and a relatively hydrophilic cis structure under UV light illumination. Therefore, the visible-light trans form induces a higher degree of protein unfolding than UV-light cis form. Small-angle neutron scattering (SANS) and neutron spin echo (NSE) spectroscopy were used to measure the tertiary structure and internal dynamics of lysozyme. The SANS-based in vitro structures indicate that under visible light the photosurfactant induces a partially-unfolded structure of lysozyme with unfolding occurring principally away from the active site near the hinge region connecting the α and β domains. Upon UV exposure, however, the protein refolds to a native-like structure. At the same time, enhanced internal dynamics of lysozyme were detected with the surfactant in the trans form through NSE measurements of the Q-dependent effective diffusion coefficient (D eff ) of the protein, which were found to deviate from rigid body translational and rotational 70 motions, indicating enhanced internal dynamics with the protein in the partially-unfolded form. In contrast, the D eff values of lysozyme in the presence of cis azoTAB largely agree with the rigid body calculation as well as those measured for pure lysozyme, suggesting that the native protein is dormant on the nanosecond time and nanometer length scales. Lysozyme internal motions were modeled by assuming a protein of two (α and β domains) or three (α and β domains plus the hinge region) domains connect by either soft linkers or rigid, freely-rotating bonds. Protein dynamics were also tracked with FT-IR through hydrogen/deuterium exchange kinetics, which also demonstrated enhanced protein flexibility induced by the trans from of the surfactant relative to the native protein. Ensemble-averaged intramolecular fluorescent resonance energy transfer (FRET) measurements further demonstrated the enhanced effect of the trans form of the photosurfactant on the dynamics of lysozyme. Combined, these results provide insight into a unique light-based method of controlling protein dynamics and function. 4.2 Introduction Proteins are the building block of life. They are responsible for a variety of functions, such as catalyst, transport, etc. To function appropriately, proteins need to adopt certain structures. For example, some enzymes exhibit a structure of a closed active site upon ligand/substrate binding and an open active site upon product release. 7, 172 Static images of proteins bound with ligand began to emerge in the past decade, providing insight into the changes in protein conformation and dynamics required for protein function. 82 To examine protein dynamics, which connects protein structure and 71 function, a wide range of techniques are required to cover the broad range of length scales and time scales of protein internal motions, from fast local fluctuations of an atom to slow collective structural rearrangements of molecular domain motions. 1, 35 Static crystallographic structures provide protein dynamics averaged over hours. 108 In contrast, neutron spectroscopy can probe motions on time scales ranging from 10 -12 to 10 -8 s. 35, 50 Computer simulations such as molecular dynamics simulation and normal mode analysis have also been used to predict large scale domain motions at nano/picosecond time scales. 4, 18, 35, 67 In addition, conformational flexibility required for transition from one conformational state to another have been examined by hydrogen-deuterium exchange and connected to enzyme activity. 78, 109 Single-molecule measurements provide real time observation of hidden dynamic events upon protein functioning. 31 Each of these techniques provides important insight to protein dynamics, thereby making the protein structure-dynamics-function relationship better understood. Among these techniques, neutron scattering has the advantage of providing a method to measure dynamics over a range of time and length scales. 50 The fact that neutrons are electrically neutral with wavelengths of ~Å and energy levels of ~meV has made neutron spectroscopy an important tool for studying condensed matter. For example, neutron spin echo (NSE), an quasielastic method, has been successfully applied to study the dynamics of microemulsions (mixture of surfactant, water and oil), revealing the dynamics of shape fluctuations of surfactant interfaces. 48, 73, 91, 110 Polymer dynamics from short linear polymer chains to complex polymer mixtures and block copolymers have also examined with NSE and compared to polymer behavior calculated from 72 different dynamic models. 110, 130 In terms of protein dynamics, the application of neutron spectroscopy initially involved protein powders with low water content where translational and rotational motions can be neglected. Since then hydration effects and the influence of solvents on protein dynamics have been studied, and the application of neutron spectroscopy to the study proteins in solution has gradually advanced. 50 Recently, protein domain motions of a DNA polymerase in D 2 O solution have been successfully observed with NSE measurements revealing the dynamics associated with interdomain motions that are functionally important to the catalytic behavior of the enzyme. 18 From the NSE data, the effective diffusion coefficient of Taq polymerase was obtained and fit with a domain model characterized by both intradomain and interdomain motions with the assumption that the motion of each individual domain being independent of others. In previous studies, an azobenzene-based surfactant (azoTAB), which is in the relatively hydrophobic, trans structure under visible light illumination and the relatively hydrophilic cis structure under UV light illumination, was used to investigate the structure-function relationship of lysozyme. 160 The ensemble SANS measurements indicated that in the presence of trans azoTAB the tertiary structure of lysozyme exhibited a higher degree of swelling primarily in the hinge region connecting the α and β domains, leading to an increase in the observed separation distance of the two domains relative to the protein in the presence of cis azoTAB-lysozyme, which exhibited a compact, native like structure. The swelling of the tertiary structure was accompanied by an approximately 7-fold enhancement in activity comparing to the native protein as well 73 as the cis-azoTAB-protein complex, while increased protein flexibility accompanying internal dynamics induced by azoTAB under visible light was considered as a possible cause of enhanced activity. The goal of this Chapter is to examine the effect of azoTAB and light illumination on the structure-dynamics relationship of lysozyme. Neutron spin-echo (NSE) spectroscopy is used to observe the internal motion of the protein. NSE is a quasi-elastic scattering technique that provides time resolution of 0.01-100 ns and spatial resolution of 2-200 Å. In this study, NSE experiments are used to observe the time-dependent protein internal dynamics, along with small-angle neutron scattering (SANS) measurements providing the ensemble-averaged tertiary structure of lysozyme-azoTAB complexes. In addition, H/D exchange kinetics of lysozyme-azoTAB complexes are collected by FT-IR spectroscopy to detect flexibility in the protein. Ensemble-averaged fluorescent resonance energy transfer (FRET) between a donor and acceptor dye located on opposite protein domains is also examined to observe the effect of azoTAB on the tertiary structure and dynamics of lysozyme. 4.3 Experimental Details Materials. Azobenzene trimethylammonium bromide surfactants (azoTAB) of the form N N CH 3 CH 2 O N + (CH 3 ) 3 Br CH 2 n 74 were synthesized according to published procedures. 66 For H/D exchange observed with FT-IR, a surfactant with an n = 2 was used. A surfactant with am n = 4 was used in all other experiments. The longer the alkyl chain, the more hydrophobic the molecule. Therefore, the surfactant of n = 4 (CMC = 4.6 and 10.5 mM for trans and cis, respectively) 66 tends to bind more to a protein and induce a higher degree of protein unfolding through hydrophobic interaction as compared to the surfactant of n = 2 (trans/cis CMC = 9.5/11.5 mM) 66 . The surfactants undergo a reversible photoisomerization upon exposure to the appropriate wavelength of light. Photoisomerization of the surfactant with light illumination changes the dipole moment across the –N=N– bond. When existing as the trans isomer the surfactant has a lower dipole moment (planar structure) than the cis isomer (bent structure). Therefore, the trans isomer is more hydrophobic than cis isomer. Under visible-light illumination the surfactant exhibits a 75/25 trans/cis photostationary state, whereas with UV light illumination the surfactant exists primarily in the cis form (> 90% cis). To collect FT-IR and fluorescent spectra, an azoTAB solution without protein was preconverted to the cis from using an 84 W long wave UV lamp, 365 nm (Spectroline, Westbury, NY, model number XX-15A) prior to addition of the enzyme. For the neutron spin echo experiments, the protein-surfactant solution was illuminated with a 200 W mercury arc lamp (Oriel, model no. 6283) equipped with a 320 nm band-pass filter (Oriel, model no. 59800) in combination with a heat-absorbing filter (Oriel, model no. 59060), effectively isolating the 365 nm line (UV-A) to convert azoTAB to the cis form. During the period of FT-IR and NSE data collection, a liquid light guide (Oriel, model 75 number 77 557) was attached to the arc lamp to continuously illuminate the sample with UV light in order to maintain azoTAB in the cis state. Highly-purified lysozyme from hen-egg-white (L7651) and phosphate buffer (8.3 mM) were purchased from Sigma and used as received. Small-Angle Neutron Scattering. The neutron scattering data were collected on the 30-m NG-3 SANS instruments at NIST. 53 Two sample-detector distance were used (1.33 and 7.0 m), combined with a 25 cm offset of the detector, to give a Q range of 0.0048 - 0.46 Å -1 , where Q = 4πλ -1 sin(θ/2) and θ is the scattering angle and λ is 6 Å. The net intensities were corrected for the background (pure D 2 O) and empty cell, followed by accounting for the detector efficiency using the scattering from an isotropic scatterer (Plexiglass), and then converted to an absolute differential cross section per unit sample volume (in units of cm -1 ) using an attenuated empty beam. The coherent scattering intensities of the sample were obtained by subtracting the incoherent contribution from the hydrogen atoms in lysozyme (0.004 cm -1 ) and the surfactant (0.002 cm -1 ). The SANS data were analyzed by calculation of the pair distance distribution functions (PDDFs) and a shape-reconstruction algorithm. The PDDFs were calculated assuming a monodisperse system using GNOM 144 over a Q range of ca. 0.02 − 0.3 Å -1 , to exclude effects of protein interactions at low Q. The maximum particle diameter (D max ) was selected to give a smooth return of the PDDF to zero at D max . The shape reconstructions were performed using GA_STRUCT supplied by Dr. William Heller. 72 The Q range of 0.01 − 0.3 Å -1 was employed with 1000 scattering centers used to 76 represent the protein. This range was used to exclude potential protein interactions that would be exhibited at low Q and to avoid length scales too small for protein continuity at high Q. Neutron Spin Echo Experiments. NSE experiments were performed on the NG-5 neutron spin echo spectrometer at NIST 132 using a quartz sample cell with a 4 mm path length and the neutron wavelength of 8 Å at 25 °C. The data were collected over a Q range of 0.046 − 0.246 Å -1 . All samples were prepared with buffered D 2 O solution (8.3 mM sodium phosphate, pH 7.2) to achieve a protein concentration of 10 mg/mL at varying surfactant concentrations. The data were corrected with the solvent (D 2 O buffer), the empty cell, and the instrumental resolution collected from a set of elastic reference samples to obtain normalized intermediate scattering functions I(Q,t)/I(Q,0) of lysozyme at several wave vectors. The DAVE software package was used for elements of the data reduction and analysis. 74 To obtain the effective diffusion coefficient (D eff ) of the protein, the data were fit with the single-exponential decay function 2 (,) exp( ( ) ) (,0) eff IQ t D QQt IQ =− . (1) Three models were employed to calculate the Q-dependence of D eff , including rigid body analysis, a soft-linker-domain model proposed by Bu et al, 18 and a freely- jointed-domain model proposed by Akcasu et al. 2 The rigid body model calculates the contributions of translation and rotational motions of rigid lysozyme to the Q dependent oscillation of D eff . The soft-linker-domain model assumes the protein to be composed of rigid domains connected by soft spring linkers where the first cumulant, the initial slope of the dynamic scattering function, Ω(Q) = −∂ln[I(Q,t)/I(Q,0)]/∂t, is explicitly 77 independent of the interdomain spring constant. The freely-jointed-domain model treats the protein domains as polymer segments along a freely jointed chain consisting of n identical beads connected by freely rotating rigid bonds of length ζ. Rigid body calculations were applied to the crystal structure of lysozyme (PDB code 6LYZ) and the solution structures of lysozyme-azoTAB complexes obtained from SANS by GA_STRUCT to account for translational and rotational motions through the equation 18 (( ) 2 () (()()) () jl jl iQ r r TR jl jl B eff iQ r r jl jl bb QH Q L j H L l e kT DQ Q bbe − − 〈 +〉 = 〈〉 ∑ ∑ , (2) where b j and b l are neutron scattering lengths of effective scattering centers j and l, respectively. For the lysozyme crystal structure, the scattering centers were taken as each amino acid residue with the center of each residue being the average coordinate and the neutron scattering length of each residue being the sum of the neutron scattering lengths of all atoms in the residue. For the solution structure of lysozyme obtained by GA_STRUCT, b values were assumed identical for all scattering centers. L(j) = Q × r j is the angular momentum vector, while H T and H R are the translational and rotational mobility tensors, in which the three principle-axis coefficients (D T x , D T y , D T z in H T and D T x , D T y , D T z in H R ) are calculated by the program HYDROPRO. 23, 37 In the soft-linker-domain model, the lysozyme molecule was divided into n rigid domains (n = 2 for a two-domain model, namely the α and β domains; or n = 3 for a 78 three-domain model of the α and β domains and the hinge region). D eff was calculated using the equation 18 1 () n ii i eff lys DS DQ S = = ∑ , (3) where D i = (k b T)/ξ i with ξ i being the friction constant of the i th domain. S i and S lys were obtained by the equation , sin () i N mn i mn i mn Qr r SQ Qr r ∈ ⎡ −⎤ ⎣ ⎦ = − ∑ , (4) which is the rotationally-averaged static form factor of the i th domain with N i being the number of scattering centers in the i th domain. S lys was calculated using N = 1000, namely the total scattering centers in lysozyme. The freely-jointed-domain model treats the protein as n rigid domains connected by (n-1) bonds of length ζ, with D eff calculated from the equation 2 () () () eff b Q DQ kT SQ μ = , (5) where S(Q) is the static scattering function given by 1 00 0 00 () ()1 () 1 () 1 2 1 1( ) 1 ( ) 1 n jQ jQ j Q n SQ n nj Q j Q n ζζ ζ ζζ − ⎡⎤ ⎡ ⎤ − − =⋅ + − ⎢⎥ ⎢ ⎥ −− − ⎣ ⎦ ⎣⎦ (6) and j 0 is the zeroth order spherical Bessel function. The mobility tensor μ(Q) is given by 12 () () () QQ Q μ μμ = − , (7) where 79 10 () ( ) Qj Q H μν μν μζ − = (8) and [] () 1 2 2 2 ,1 1 () 2(1 ) nn iQ R T j j j j QHCeQCH μν μ ν μν ξ μζ ζτ − ⋅ == ⎡ ⎤ =⋅ ⎣ ⎦ − ∑∑ . (9) In these equations H is the preaveraged hydrodynamic interaction matrix given by 0 0 12 (1 ) ( ) Hdxjx μν μν μν μν δδτ ξπ ∞ − ⎡ ⎤ =+− ⎢ ⎥ ⎣ ⎦ ∫ (10) and τ = ξ/6πζη is the draining parameter determined by the friction constantξ, the solvent viscosity η, and bond length ζ. C is an (n-1) × n matrix defined by ,1 jj j C μμ μ δ δ − = − , (11) where δ is the Kronecker delta. The indices μ andν runs over all protein domains (from 1 to n), while the index j runs over all bonds (from 1 to (n-1)). The ensemble average is approximated as () () () () 2 1 1 2 00 2 iQ R j jQ eQ jQ jQ Q μν μν ζ ζζ ζ ζ ζ −− ⋅ ⋅ ⎡ ⎤ ⋅= ⋅ ⋅− ⎢ ⎥ ⋅ ⎣ ⎦ (12) when ⏐μ−ν⏐= 1 and μ = j+1 or ν = j+1, otherwise () () 2 2 0 3 iQ R j eQ jQ μν μ ν ζ ζζ − ⋅ ⋅= ⋅ . (13) To apply the freely-jointed-domain model lysozyme was divided into n = 3 domains, representing the α and β domains and the hinge region, connected by (n -1) = 2 bonds. The approximate center-of-mass of each domain was visually selected from the PDB file produced by GA_STRUCT, while each scattering center was associated with 80 the closest domain. The bond length ζ was taken as the average distance between the centers-of-mass of the three domains. Kinetics of Hydrogen-Deuterium exchange measured with FT-IR. The amide I (1700−1600 cm -1 ) and II (1600−1500 cm -1 ) regions in protein IR spectra result from C=O stretching and N-H bending coupled with C-N stretching, respectively. Upon exposure to D 2 O, the amide II peak dramatically shifts to lower wavenumbers (from ~1550 to ~1450 cm -1 ), while the amide I peak shifts only slightly by 2−10 cm -1 . The Amide II band is known to reflect changes in conformation, stability and flexibility of proteins upon H/D exchange. Therefore, observing the decay of the amide II intensity scaled by amide I intensity allows monitoring of the global H/D exchange rate of lysozyme at different azoTAB concentrations and light conditions. All spectra were measured with a Genesis II FTIR spectrometer (Mattson Instruments) and a demountable liquid cell with a water circulated jacket to control the temperature at 20 °C. For each time point, 100-scan interferograms were collected and averaged at a resolution of 4 cm -1 . The exchange was started by injecting buffered D 2 O solutions (8.3 mM sodium phosphate, pH 7.2) at an appropriate azoTAB surfactant concentration (0 − 12mM) into a vial containing an appropriate amount of lyophilized lysozyme to give a protein concentration of 10 mg/mL, followed by brief gentle mixing. The mixture was then immediately transferred to a demountable FT-IR liquid cell consisted of a pair of CaF 2 windows and a Teflon spacer (100 μm). Data collection started after a delay time of ~2.5 min. A liquid light guide inserted into the FT-IR sample chamber was focused on the sample area to introduce direct UV illumination throughout 81 the data collection process of cis-azoTAB-lysozyme complexes. Dried air was continuously purged into the chamber to eliminate the effect of water vapor. A set of reference spectra (buffered D 2 O and azoTAB without protein) were collected at identical conditions for background subtraction. The protein peaks in the amide I and amide II regions were obtained by subtracting the corresponding reference spectra from the spectra of the protein solution. The fraction (F) of unexchanged hydrogen atoms in the protein was then determined by () ( ) (0) ( ) t F ω ω ωω − ∞ = − ∞ , (14) where ω(t) = A II (t)/A I (t) with A II (t) and A I (t) representing the absorbance of amide II at ~1542 cm -1 and Amide I at ~1649 cm -1 , respectively, at time t. The value of ω(0) was obtained from a spectrum of undeuterated lyophilized lysozyme in KBr pellets. ω(∞) was determined from protein samples aged for 10 days in D 2 O at room temperature to ensure complete exchange. The protein sample without azoTAB required incubation at 37 °C to completely remove the amide II peak. The H/D exchange kinetics of lysozyme at different azoTAB concentrations was analyzed with a multi-exponential equation for i amide regions ( ) exp( ) ii i Ft a kt =− ∑ , (15) where k i represents the rate constant of different amide groups with similar exchange rates and a i is proportional to the relative fraction of the i th region in the protein. The mean kinetic rate constant 〈k〉 was defined as 71 82 1 / ii i i i ak k a − = ∑ ∑ . (16) Intramolecular fluorescence resonance energy transfer (FRET) spectroscopy. Wildtype T4 lysozyme was used to study of the degree of FRET in the protein with changing azoTAB concentrations. The construct of the protein was kindly provided by Dr. Brian Matthews at the University of Oregon. Labeling of the donor (Alexa Fluor 532 maleimide) and acceptor (Alexa Fluor 594 maleimide) dyes to the two cysteine residues in the protein (cys 54 and cys 97, located on β and α domains, respectively) was achieved by thiolation using the standard procedure was provided by Molecular Probes. Briefly, the donor dye was slowly added into the protein solution (1 mg/mL) in phosphate buffer (pH 7.2, 20 mM) at a dye-to-protein ratio of 1:1. The mixture was allowed to react overnight at 4 °C. Unreacted dye was then removed by extensive dialysis. The percentage of labeling was kept low to prevent double-labeling of the same dye on one protein molecule. Donor dye labeled protein was then separated from unlabeled protein by hydrophobic interaction chromatography (phenyl sepharose high performance from GE Healthcare) and further labeled with the acceptor dye with similar procedures. About 90% and 70% of the product was labeled with the donor and the acceptor, respectively, estimated from UV-vis spectroscopy. The ensemble-averaged fluorescence of the protein was obtained by excitation at 510 nm. The emission spectra exhibited peaks of both donor and acceptor at 544 nm and 617 nm, respectively, further confirming the success of the dye pair labeling. To observe 83 the effect of azoTAB surfactant on the FRET, 0.05 mL of a 0.03 mg/mL protein solution and varying amounts of concentrated surfactant solution were added into a 0.9 mL phosphate buffer solution (8.3mM, pH 7.2) resulting in a solution of 1.6 μg/mL protein and the desired surfactant concentration (0 − 0.14 mM). The relative ensemble-averaged FRET efficiency was evaluated using the equation E ET = I A /(I A +I D ), where I A and I D correspond to the fluorescence intensity of acceptor at 544 nm and donor at 617 nm, respectively, normalized by the efficiency of labeled protein without the presence of surfactant. 159 4.4 Results and Discussion SANS data of lysozyme-azoTAB complexes as a function of azoTAB concentration and light conditions are shown in Figure 4.1. Similar to lysozyme-azoTAB complexes at pH 5.0, 160 with increasing azoTAB concentration the scattering curves of azoTAB-lysozyme complexes under visible light shown in Figure 4.1(a) begin to deviate from pure lysozyme at Q ~ 0.2 Å -1 , which corresponds to a length scale L (= 2π/Q) of 31 Å. Compared to the published value for the diameter of lysozyme (~35 Å), 58, 140 this deviation suggests that azoTAB induces swelling of the lysozyme tertiary structure. While a marked deviation of the scattering curves from that of pure lysozyme is induced by trans azoTAB, the scattering curves for cis azoTAB/lysozyme complexes are almost identical to that of the pure protein, indicating the structure of lysozyme is not significantly influenced by cis azoTAB. 84 Radii of gyration (R g ) of lysozyme at different azoTAB concentrations and light conditions were also calculated from the SANS data by Guinier analysis using the equation I(Q) = I(0)exp(−Q 2 R g 2 /3), where I(0) is the extrapolated scattering intensity at Q = 0. The R g values were observed to steadily increase with increasing azoTAB concentration under visible light (R g = 15.2, 16.9, and 18.3 Å with 5.0, 8.0, 12.0 mM azoTAB, respectively). Upon illumination with UV light, however, lysozyme/azoTAB complexes apparently refold to dimensions similar to native lysozyme (R g = 13.5, 12.5, 12.9 and 13.5 Å at azoTAB concentrations of 0, 5.0, 8.0, 12.0 mM, respectively), supporting the notion that lysozyme is swollen by trans azoTAB and refolds to a native- like conformation upon conversion to cis azoTAB. 0.01 0.1 1 0.01 0.1 I(Q)/cm -1 Q/A -1 o 0 0.4 0.8 1.2 1.6 0 102030405060 r (A) P(r) (x10 3 ) o Figure 4.1: (a) SANS data and (b) PDDFs of lysozyme-azoTAB solutions as a function of surfactant concentration and light illumination (closed symbols for trans and open symbols for cis azoTAB). Pure lysozyme (•), 5.0 mM azoTAB ( ¡ , ), 8.0 mM azoTAB ( S , U ), and 12.0 mM azoTAB ( T , V ). [Lysozyme] = 10 mg/mL in pH 7.2 buffer. (a) (b) 85 The pair distance distribution function (PDDFs) were calculated from the SANS data, as shown in Figure 4.1(b). PDDFs calculate the probability of finding two scattering centers a distance r apart within the protein. For globular proteins, the most probable distance between scattering centers (i.e., the maxima of the curves in Figure 4.1(b)) would then represent the radius of a protein, while the r value were the PDDFs return to zero would indicate the maximum dimension within the protein. With increasing trans azoTAB concentration in Figure 4.1(b), both the protein radius and the maximum dimension shift to higher values, again indicating that trans azoTAB induces protein swelling. Upon conversion of azoTAB to the cis form with UV illumination, however, the protein refolds such that the radius and maximum dimension return to values similar to that obtained for pure lysozyme. NSE data of the normalized intermediate scattering functions I(Q,t)/I(Q,0) of lysozyme with different azoTAB concentrations and light illuminations at representative Q values are shown in Figure 4.2. The data can be fit with single exponential decay function to obtain the Q-dependent effective diffusion coefficients D eff of lysozyme at the corresponding surfactant concentration and light condition, as shown in Figure 4.3(a)-(g). The intramolecular interference of a rigid body is known to cause oscillations of the diffusion coefficient as a function of Q, therefore the contributions of translational and rotational diffusion to the D eff (Q) was calculated. 18, 51 These rigid-body calculations for lysozyme solution structures determined by shape reconstruction of the SANS data are plotted as blue solid lines at each azoTAB concentration, while the rigid-body calculations form the lysozyme crystal structure (6LYZ) is a solid red line. The NSE data 86 Figure 4.2: NSE data of azo-TAB-lysozyme as a function of time t at representative Q values. 0.01 0.1 1 5mM cis Q=0.061 Q=0.080 Q=0.098 Q=0.122 Q=0.160 Q=0.196 0 06090 I(Q,t)/I(Q,0) 0.01 0.1 1 02468 10 12 14 12mM cis Q=0.061 Q=0.080 Q=0.098 Q=0.122 Q=0.160 Q=0.196 0 06090 t (ns) I(Q,t)/I(Q,0) 0.01 0.1 1 8mM cis Q=0.085 Q=0.104 Q=0.122 Q=0.197 Q=0.209 I(Q,t)/I(Q,0) 0.01 0.1 1 02468 10 12 14 12mM trans Q=0.061 Q=0.080 Q=0.098 Q=0.122 Q=0.160 Q=0.196 C t (ns) I(Q,t)/I(Q,0) 0.1 1 5mM trans Q=0.061 Q=0.080 Q=0.098 Q=0.122 Q=0.160 Q=0.196 0.06090 I(Q,t)/I(Q,0) 0.1 1 8mM trans Q=0.085 Q=0.104 Q=0.122 Q=0.160 Q=0.209 I(Q,t)/I(Q,0) 0.01 0.1 1 02 46 8 10 12 14 Q=0.085 Q=0.104 Q=0.122 Q=0.156 Q=0.209 I(Q,t)/I(Q,0) t (ns) pure lysozyme 87 of pure lysozyme shown in Figure 4.3(a) indicate that the effective diffusion coefficient oscillates around the low-Q translational diffusion coefficient collected by dynamic light scattering (red dashed line), suggesting that the protein exhibits as a rigid structure on the nanosecond time scale and nanometer length scale in the native form. The fact that rigid- body calculations for pure lysozyme from both the crystal and the solution structures are similar and agree with the experimental data further confirm that the native protein in solution behaves as a rigid body exhibiting little variation from a static structure. In the presence of azoTAB, however, the NSE data shown in Figures 4.3(b)-(g) exhibit differing degrees of deviation from the rigid-body calculations at higher Q values, implying that the protein is undergoing dynamic motions beyond intramolecular translational and rotational interference of a rigid body. Furthermore, trans azoTAB seems to induce deviations at lower Q-values (at Q ~ 0.15Å -1 , or a length scale of ~42Å) than cis azoTAB (at Q ~ 0.2Å -1 , or a length scale less than ~31Å). These deviations in the D eff of lysozyme from a rigid body calculation with cis azoTAB at length scale smaller than the diameter of lysozyme (~35 Å) suggest that the protein performs small scale internal motion. Radii of gyration obtained from the SANS data and the maximum dimensions from PDDF analysis indicate cis-azoTAB-lysozyme complexes have a native-like conformation. Furthermore, lysozyme conformation obtained from SANS data by GA_STRUCT (red features in Figure 4.3) confirm the native-like structure of cis- azoTAB-lysozyme complexes. However, the fluorescence of lysozyme-azoTAB complexes using Nile red as a probe suggests that in the presence of cis azoTAB the protein is not as compact as native lysozyme. 58 The nonionic hydrophobic probe, Nile 88 8 9 10 11 12 13 14 15 16 0.05 0.1 0.15 0.2 Pure lysozyme Q (A -1 ) o D eff (A 2 /ns) o 8 9 10 11 12 13 14 15 16 0.05 0.1 0.15 0.2 5mM trans Q (A -1 ) o D eff (A 2 /ns) o 8 9 10 11 12 13 14 15 16 0.05 0.1 0.15 0.2 5mM cis Q (A -1 ) o D eff (A 2 /ns) o Figure 4.3: NSE data of lysozyme effective diffusion coefficient with (a) 0, (b) 5.0 (trans), (c) 5.0 (cis), (d) 8.0 (trans), (e) 8.0 (cis), (f) 12.0 (trans), and (g) 12.0 (cis) mM azoTAB. The ribbon (PDB 6lyz) and space filled structure next to each plot represents the protein conformation at the corresponding azoTAB concentration. In each plot, the horizontal red dashed line represents the center-of-mass translational diffusion coefficient; the solid line represents rigid body calculations using SANS data (−) and crystal structure (6lyz) (−). The dashed lines in (d) and (g) represent calculation results of domain models using soft-linker-domain model assuming two- (---) and three-(---) domains and freely- jointed-domain model (---). (a) (b) (c) 89 Figure 4.3, Continued. 8 9 10 11 12 13 14 15 16 0.05 0.1 0.15 0.2 8mM trans Q (A -1 ) o D eff (A 2 /ns) o 8 9 10 11 12 13 14 15 16 0.05 0.1 0.15 0.2 8mM cis Q (A -1 ) o D eff (A 2 /ns) o 8 9 10 11 12 13 14 15 16 0.05 0.1 0.15 0.2 Q (A -1 ) o D eff (A 2 /ns) o 12mM trans 8 9 10 11 12 13 14 15 16 0.05 0.1 0.15 0.2 D eff (A 2 /ns) Q (A -1 ) o o 12mM cis (d) (e) (f) (g) Å 90 Red, exhibits intensive increases in fluorescence when partitioned into nonpolar environments, such as the hydrophobic core of protein. The subtle increase of Nile red fluorescence in lysozyme solution with high concentration cis azoTAB indicates that the protein is relaxed/loosened although not as unfold as with trans azoTAB. With a relaxed native-like conformation, the deviation of the protein dynamics from rigid body is possibly due to small scale fluctuations induced by azoTAB. The pronounced deviation of D eff of lysozyme at lower Q values with higher concentrations of trans azoTAB indicates that the protein in these cases performs large scale motions, possibly domain motion such as hinge bending motions or shear. The active site of lysozyme is located in a cleft between the two protein domains: an α domain consisting mainly of α helical segments and a β domain composed primarily of β sheet structures. Connecting the two domains is a hinge region identified by theoretical simulations with experimental structure data to be comprised of helix C, residues Glu35 and Ser36 on the loop succeeding helix B, and Ile55 and Leu56 at the turn between strands II and III. 57 To form a substrate-enzyme complex, it is well accepted that the protein has to perform hinge-bending motions with the protein active site first opened (substrate-free state) and then closed to bind with the substrate (substrate-bound state). X-ray crystallographic data of substrate-free T4 lysozyme exhibited a more opened active site than the substrate-bound form. 95 Intramolecular single-molecule FRET of T4 lysozyme also observed events resulted from conformational changes related to protein function in real time. 31 From shape-reconstruction of the SANS data in Figure 4.1, the active site cleft can be observed from the solution structures of lysozyme in the 91 upper right of the molecule, as shown in Figure 4.3. Upon increasing the concentration of trans azoTAB, the protein is observed to swell primarily in hinge region at bottom of the molecule, progressively resulting in a relatively open active site cleft, consistent with previous measurments. 58, 160 Due to the deviation from the rigid-body calculations of the measured D eff values for the lysozyme with trans azoTAB, especially at 8.0 and 12.0 mM azoTAB where the protein shows the most significant degree of unfolding, two models are applied to estimate the contributions of protein domain motions to the Q-dependent effective diffusion coefficient. The first model, proposed by Bu et al., treats a protein as separate rigid objects connected by soft spring linkers. 18 Considering the solution structure of lysozyme obtained from SANS along with the location of the active site and the known dynamics required for enzyme function, the protein molecule was divided into either two domains (α and β domain) or three domains (α and β domain plus the hinge region). When applying this soft-linker-domain model to lysozyme with 8.0 and 12.0 mM trans azoTAB, the results obtained by assuming two (blue dashed line) or three (green dashed line) rdomains exhibit a peak at Q ~ 0.17 Å -1 as shown in Figures 4.3(d) and (f). To obtain these results, it was necessary to increase the friction constant calculated from Kirkwood’s formula 89 for each domain by a factor of 1.5. Similarly, Bu et al. found it necessary to increase the friction coefficients of Taq polymerase domains by a factor of ~2, which is possibly caused by the water displacement between domains due to the close proximity between domains. The shape of D eff (Q) calculated with three domains is similar to that obtained with two domains, although the increase in D eff with larger Q 92 values is more dramatic in the two domain case. The location of the calculated peaks at Q ~ 0.17 Å -1 concurs with the regions where large deviations from the rigid-body model are seen in the measured D eff values, which would seem to indicate that this deviation is a result of protein domain motion. However, the relatively larges values of D eff obtained with this model along with the dramatic decrease of D eff as Q approaches zero, violating the notion that the effective diffusion coefficient should approaching the center-of-mass translational diffusion coefficient measured with dynamic light scattering as Q approaches zero, indicates that the soft-linker-domain model without bond length constraints over predicts the magnitude of the domain motions in lysozyme. To take into account the effect of interdomain contacts, a freely-jointed-domain model 2 was also used to model lysozyme domain motions. The red dashed lines in Figures 4.3 (d) and (f) represent these calculations obtained by assuming a bond length of 25 Å and a friction constant taken as the average value calculated from Kirkwood formula of the three individual domains, again increased by a factor of 1.5 to account for an ~3 Å thick hydration layer around each domain. 58, 97 The results show a similar peak at Q ~ 0.17 Å -1 although in this case with a more realistic values of D eff throughout the entire Q relative to the experimental data, suggesting that the freely-jointed-domain model better models the domain motion within lysozyme in the presence of trans azoTAB. Although it should be pointed out that even this model only qualitatively resembles the trends in the measured D eff values of lysozyme. This is to be expected as the internal motions within lysozyme are much more complex than those represented by the freely-jointed-domain model, which assumes a polymer-like chain with the 93 identically-sized beads (domains) and bond lengths. Furthermore, the assumption of freely rotating bonds without a constraint on the bond angle likely oversimplifies the domain motions with lysozyme. One of the key aspects connected with protein dynamics is flexibility. Without flexibility a protein cannot perform small-scale fluctuations such as breathing or large- scale domain motions such as hinge bending or shearing. As such, a number of studies have indicated the importance of protein flexibility in enzyme activity. 56, 122, 131, 170 For instance, a lysozyme mutant was found to exhibit a decreased activity with increased stability as well as rigidity by replacing residues near the hydrophobic core with residues of higher hydrophobicity. 115 Study of another mutant with residues Arg14 and His 15 deleted indicated the importance of flexibility to enzyme function by correlating enhanced mobility with increased activity relative to wild-type lysozyme. 78, 109 In Chapter 3, the influence of azoTAB and light illumination on the activity of lysozyme was examined with different assays. In the presence of azoTAB an enhanced enzymatic activity was observed, which was possibly a combined result of a more open active site and enhanced dynamics or flexibility of the protein upon surfactant binding. To examine the effect of azoTAB on lysozyme flexibility, H/D exchange kinetics of lysozyme with different azoTAB concentrations were collected with FT-IR spectroscopy. H/D exchange has been used as an efficient technique to monitor proteins dynamics, flexibility and stability. 55, 71, 126, 157 Figure 4.4 (a) shows a representative H/D exchange spectra of lysozyme with 8.0 mM azoTAB. The replacement of amide protons (N-H) with deuterium (N-D) results in a shift of the amide II peak from 1550 cm -1 to 94 1450 cm -1 , while the amide I peak shifts by only about 5-10 cm -1 (e.g., see Figure 4.4(a)). This provides a convenient monitor of the exchange rate of hydrogen to deuterium by observing the disappearance of the amide II peak at 1550 cm -1 , thereby providing a measure of the global flexibility of a protein. 25, 90, 98 Figure 4.4(b) shows the time dependent change of the fraction of unexchanged hydrogen in lysozyme with different azoTAB concentration and light illumination (see: Experimental Details). The protein was observed to exhibit an exchange rate faster than the native state when in the presence of trans azoTAB, while cis azoTAB induced little influence in the exchange rate. This implies that trans azoTAB induces greater enhanced flexibility of the protein relative to the native state. 0 0.05 0.1 0.15 0.2 1300 1400 1500 1600 1700 1800 Wavenumber (cm -1 ) Amide I Amide II Absorbance 0 0.1 0.2 0.3 0.4 0.5 0.6 0 100 200 300 400 500 600 Pure lysozyme 5mM trans 8mM trans 12mM trans 5mM cis 8mM cis 12mM cis Fraction of Unexchanged Hydrogen time (min) Figure 4.4: (a) Representative lysozyme H/D exchange spectra in the time range of 5 min to 10 hr at 20 °C. Arrows in the spectra indicate direction of peak intensity changes. [azoTAB] = 8.0 mM. (b) H/D exchange kinetics of lysozyme as a function of azoTAB concentration and light illumination (closed symbols for trans and open symbols for cis azoTAB). Pure lysozyme (•), 5.0 mM azoTAB ( ¡ , ), 8.0 mM azoTAB ( S , U ), and 12.0 mM azoTAB ( T , V ). [Lysozyme] = 10 mg/mL in pH 7.2 buffer. (a) (b) 95 The hydrogen atoms within a protein are usually divided into three groups: ultra fast exchanging protons located on the surface of the protein, fast exchanging protons contained within flexible structural elements, and slow exchanging protons in the core of the protein formed by very rigid clusters. 52, 156 Since the ultra fast exchange is known to complete within a few seconds, 92 only the fast and slow exchange are observed in Figure 4.4. Kinetic rates were obtained for the protein at each surfactant concentration by fitting the H/D exchange data with a biexponential decay function, with the results shown in Table 4.1. The rate constants of the fast exchanging components within the protein are relatively constant at all azoTAB concentrations (both trans and cis). The exchange rate of the slow group, however, increases by about an order of magnitude when high concentrations of trans azoTAB are present (8.0 and 12.0 mM). In contrast, little change in the kinetic parameters relative to the native state was observed in the presence of cis azoTAB. The mean kinetics parameter, <k>, also showed an enhanced exchange rate in the presence of trans azoTAB and relatively constant exchange rate for pure lysozyme and cis azoTAB-lysozyme complex. H/D exchange parameters are usually viewed as an indication of protein dynamics, flexibility and stability. An enhancement in exchange rate is considered as a result of destabilization of secondary structure elements or unfolding of a compact core of a rigid cluster and, hence, a decrease in stability/rigidity and an increase in flexibility of proteins. 25, 26, 36, 71, 90, 134 For example, cytochrome c chemically modified at methionine- 80 exhibited an exchange rate constant of the slow exchanging group 2-3 times higher 96 Table 4.1: Analysis of H/D exchange kinetics of lysozyme in the presence of azoTAB surfactant Fast Slow [azoTAB] (mM) Percentage k×10 2 (min -1 ) Percentage k×10 4 (min -1 ) <k>×10 4 (min -1 ) 0 41 3.4 59 4.0 6.7 trans 5.0 32 3.0 68 10 14.3 8.0 40 8.7 60 56 89 12.0 38 4.8 62 52 76.5 cis 5.0 35 2.3 65 4.6 7.0 8.0 21 6.1 79 1.9 2.4 12.0 29 4.1 71 3.6 5.1 than the native protein as a result of a reduced tertiary stability at the core region of the protein. 36 Additives have also been found to change the H/D exchange properties of proteins. 25, 26, 134 The flexibility of bovine serum albumin either slightly decreased due to the formation of a more compact and stable structure of the protein, or increased due to the destabilized α-helices induced by additives of different anilinonaphthalene sulfonate derivatives (ANS). 25 The results presented in Figure 4.4 indicate that trans azoTAB induces an enhancement of lysozyme flexibility/dynamics, while cis azoTAB does not significantly change the protein flexibility. Moreover, the dramatically enhanced slow exchange rate constant and the largely unchanged fast exchange constant of the trans azoTAB-lysozyme complex may imply that the increased flexibility is due to the protein swelling and destabilizing in the rigid cluster region protected from solvent, while the fast exchanging components in mobile flexible region were not significantly influenced. The 97 H/D exchange results are in agreement with the SANS result of increased radius of gyration (R g ) of swollen lysozyme induced by trans azoTAB and the constant R g of native-like lysozyme with cis azoTAB (up to 12mM). To further examine effects of azoTAB on lysozyme flexibility, ensemble- averaged fluorescence spectra of T4 lysozyme labeled with donor-acceptor dye pairs are showed in Figure 4.5 (a). Lysozyme from bacteriophage T4 is slightly larger although similar in structure and function to hen egg white lysozyme. The dye labeled pure protein excited at 520 nm exhibits an emission peak of the donor dye at 554 nm accompanied by an additional emission peak at 617 nm typical of the acceptor dye, evidence of the intramolecular FRET within the protein. Addition of azoTAB into the protein solution decreases of the acceptor emission peak, with the trans azoTAB inducing a more rapid decrease than cis azoTAB. The relative ensemble-averaged FRET efficiency was evaluated as a function of azoTAB concentration, as shown in Figure 4.5(b). With increasing concentration of trans azoTAB, the relative FRET efficiency exhibits a dramatic drop even at very low surfactant concentrations (~0.005 mM, or a surfactant to protein ratio of S/P = 59/1) and reaches a constant value at a surfactant concentration of ~0.05 mM. In contrast, a slow and steady decrease of FRET efficiency is induced by cis azoTAB. Although these surfactant concentrations seems to be low relative to the previous data, the S/P values in this experiment are actually more than 3 times larger that those of the highest azoTAB concentration used in SANS and NSE experiments (12.0 mM azoTAB, S/P = 18/1), a result of the low protein concentration necessary in the FRET experiments. 98 0 5 10 4 1 10 5 1.5 10 5 2 10 5 2.5 10 5 550 600 650 700 750 0.02mM trans 0.04mM trans 0.11mM trans pure 0.02mM cis 0.04mM cis 0.12mM cis counts (s -1 ) λ (nm) 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.05 0.1 0.15 trans average cis averafe relative FRET efficiency [azoTAB] (mM) Figure 4.5: (a) Ensemble-averaged fluorescence spectra of Alexa 532/Alexa 594 labeled T4 lysozyme with varying concentration of azoTAB. The inserted picture represents the structure of T4 lysozyme (PDB code 6LZM) with the red and blue dots locating cys 54 and cys 97 where the dye pair binds, respectively. (b) Ensemble-averaged FRET efficiency as a function of azoTAB concentration. Considering the effect of azoTAB on lysozyme activity discussed in Chapter 3 and the SANS results presented here, decreases in the relative FRET efficiency induced by azoTAB can be a result of two factors: enhanced protein dynamics and/or increased swelling of the tertiary structure. Similar to hen egg white lysozyme, the active site of T4 lysozyme is located in the cleft between the α and β domains. The donor-acceptor dye pair is labeled with one dye located on each domain at positions sensitive to domain motions, as shown in the inserted picture in Figure 4.5(a). According to molecular dynamics simulation and single-molecule spectroscopy, native T4 lysozyme exhibit an opened active cleft when no substrate is present, thus, the dye pair is in close proximity in the substrate-free “open” mode. With substrates present in solution, the average distance (a) (b) 99 between the dye pair increases due to the protein performing not only the substrate-free “open” mode but the substrate-bound “close” mode (which increases the distance between the dye pair) resulting in lower ensemble FRET efficiency as compared to the FRET of native T4 lysozyme in absence of substrate. Thus, the distance between the dye pair is estimated to be increased by ~5.5 Å by the internal motions governing the enzymatic function of the protein. 31 Without substrate present, a possible factor for the decrease of FRET efficiency observed upon addition of azoTAB is an increase in protein flexibility and, hence, conformational fluctuation similar to a domain motion observed in the above NSE data, which would increase the time averaged distance of the dye pair. Interestingly, Figure 4.5 (b) measured with a 0.0016 mg/mL (85 nM) T4 lysozyme solution has a very similar shape compared to the lysozyme activity plot in Figure 3.5 (lysozyme concentration = 0.002 mg/mL, 139 nM). The effect of azoTAB on the ensemble-averaged FRET efficiency of T4 lysozyme may not be directly linked to the dynamics observed from NSE or the enzyme activity, however, the comparison of these graphs points to the possibility that enhanced flexibility as well as conformational fluctuations induced by azoTAB may play a role in the enhancement of protein function. The second factor that may also cause decreases of FRET efficiency is protein swelling. In the aforementioned discussion, lysozyme-azoTAB complexes were observed to exhibit an increased radius and maximum dimension as shown in Figure 4.1 (b). With 12.0 mM trans azoTAB the maximum dimension of lysozyme increased from 47.5 Å to 54.3 Å, a dimensional change larger than the change of the dye pair induced by 100 the internal motion of T4 lysozyme. At the level of ensemble measurement, it is not possible to determine the dominant effect causing the decrease of FRET efficiency with increasing azoTAB concentration. However, the ensemble-averaged FRET gives some insight into protein internal dynamics as well as protein folding. 4.5 Conclusions Lysozyme dynamics induced by photoresponsive azoTAB surfactant and light illumination have been examined using neutron spin-echo combined with in vitro protein conformations determined with small-angle neutron scattering to study the protein structure-dynamics-function relationship. Shape-reconstruction analysis applied to the SANS data indicates that the addition of trans azoTAB causes lysozyme to partially unfold relative to the native state, with swelling observed primarily away from the active site in the hinge region connecting the α and β domains. Upon illumination with UV light, trans azoTAB converts to the cis isomer causing lysozyme to refold into a native- like conformation. NSE experiments reveal enhanced dynamics of the partially unfolded lysozyme induced by trans azoTAB, while the native-like lysozyme in the presence of cis azoTAB exhibited dynamics similar to a rigid body. Domain models employed to calculate the effective diffusion coefficient of the unfolded lysozyme consisting of either two or three domains give a good approximation of the enhanced protein dynamics induced by trans azoTAB. H/D exchange measurements suggest that the protein exhibits higher flexibility induced by trans azoTAB, while the flexibility with cis azoTAB is similar to native lysozyme. Ensemble-averaged FRET observed a combined effect of a 101 swollen tertiary structure and enhanced dynamics of lysozyme induced by azoTAB under visible light. 4.6 Acknowledgements We acknowledge the generous advice and instruction of Dr. Boualem Hammouda on domain motion analysis. This work utilized facilities supported in part by the National Science Foundation under Agreement No. DMR-0454672. We acknowledge the support of the National Institute of Standards and Technology, U.S. Department of Commerce, in providing the neutron research facilities used in this work. 102 Chapter 5 Conclusion and Future Work 5.1 Conclusion The ability to control the structure-dynamics-function relationship of proteins with light illumination using photoresponsive surfactants has been demonstrated in the preceding Chapters. The visible-light (trans) form of the azobenzene surfactant (azoTAB) is more hydrophobic and undergoes more interactions with proteins than the UV-light (cis) form of the surfactant. As a consequence, exposing the surfactant-protein solution to visible light results in a greater degree of protein unfolding, while exposing to UV-light causes surfactants molecules to dissociate from the protein and the protein to refold. The ability to control protein structure with azoTAB and light illumination has been thoroughly examined with bovine serum albumin and lysozyme. With trans azoTAB binding the protein unfolds mainly in the α-helices regions, while β structures are largely unaffected. The structure-function and structure-dynamics relationships induced by azoTAB and light illumination have also been examined with lysozyme. In the presence of trans azoTAB, lysozyme exhibits a partially-unfolded structure with a more exposed active site, which leads to ~7 fold enhanced activity. The direct measurement of lysozyme dynamics indicated that this increased activity might be a result of enhanced internal motions induced by azoTAB. Overall, in this dissertation the 103 ability to provide insight into the complex structure-dynamics-function relationship of proteins with the light-tunable azoTAB surfactants has been successfully demonstrated. 5.2 Future Work Bovine pancreatic ribonuclease A (RNase A) is a small monomeric enzyme (124 amino acid residues, ~13.7 kDa) that catalyzes the hydrolysis of the phosphodiester linkage of single-strand RNA. Being one of the first proteins having the amino acid sequence and three dimensional structure determined, 1,2 the protein has been an important model system for the study of protein structure and function. Therefore, this protein was chosen to study the effect of azoTAB and light illumination on the structure-dynamic- function relationship. 5.2.1 Effect of azoTAB on Ribonuclease A structure RNase A consists of three helices at the N-terminal half of the protein packed against the central β-sheet of three β-strands, all of which fold into a kidney-shaped overall structure stabilized by four disulfide bonds. To investigate tertiary structural changes of the RNase A induced by azoTAB and light, two major techniques have been employed: dynamic light scattering (DLS) and small-angle neutron scattering (SANS). DLS is a convenient on-site technique that provides valuable structural information through the hydrodynamic radius of the protein-surfactant complexes, determined through the Stokes-Einstein equation D= k b T/6πηR H , where D represents the diffusion coefficient of protein in solution, k b is Boltzmann’s constant, T is temperature, η is 104 solvent viscosity and R H is the hydrodynamic radius of protein. The major technique for studying protein tertiary structure, however, will be SANS, which provides high- resolution tertiary structure information of the protein as stated in Chapter 3 and 4. Preliminary SANS data has been collected at the user facilities of NIST as shown in Figure 5.2. The ab initio shape reconstruction technique GA_STRUCT has also been applied to obtain the tertiary structure of the protein in solution (see below). 0.001 0.01 0.1 0.01 0.1 pure RNase A 2mM visible 2mM UV 4mM visible 4mM UV 8mM visible 8mM UV I(Q) / cm -1 Q / A -1 o Figure 5.1: SANS data of azoTAB-ribonuclease A solutions as a function of surfactant concentration and light illumination. [RNase A] = 10 mg/mL. The structure of RNase consists of relatively low amounts of α-helices (23%) and high amounts of β structures (46% β-sheet and 21% β turn) compared to proteins we have previously studied, such as BSA (65% α-helices, 25% β-strand, 6% β-turn) and lysozyme (45% α-helices, 19% β-sheet, 23 % β-turn). 20, 161 In the study of azoTAB-BSA 105 interactions, azoTAB was found to induce a significant decrease in α-helical content (65% to 47%), while the interaction of azoTAB with β structures was found to be relatively weak. Studying secondary structure changes of RNase A induced by azoTAB surfactants will allow better understanding of the unfolding process of a high β structure protein, and could also address unresolved issues for lysozyme such as the potential of azoTAB to interact and unfold the β domain. Preliminary secondary structure changes of RNase A induced by azoTAB have been examined through FT-IR spectroscopy. The FSD spectra of RNase A with 0, 1, 8 and 34 mM azoTAB are shown in Figure 5.2. The amide I peaks at all azoTAB concentration were normalized such that the areas were the same. The spectra of pure RNase A is almost identical to the spectra of RNase A in presence of azoTAB, even up to a relatively high concentration of 34 mM. Although the peak α-helices at 1652 cm -1 slightly decreases relative to the β-sheet peak at 1635 cm -1 , it is hard to conclusively determine the secondary structure changes induced by azoTAB. Thus, qualitatively the preliminary data indicate that azoTAB dose not induce significant changes of the protein secondary structure. The small changes of RNase A secondary structure are possibly due to the low content of α-helices (where azoTAB likes to bind) and high content of β structures (where azoTAB does not show much influence). 106 0 0.05 0.1 0.15 0.2 1600 1620 1640 1660 1680 1700 Pure RNase A 1mM 8mM 34mM FSD wavenumber (cm -1 ) Figure 5.2: FT-IR FSD spectra of RNase A without varying concentration of azoTAB. [RNase A] = 10 mg/mL, pH 7.2. 5.2.2 Effect of azoTAB on dynamics of Ribonuclease A To investigate the effect of azoTAB on the RNase A dynamics, preliminary neutron spin-echo (NSE) measurements were collected as shown in Figure 5.3. The Q- dependent effective diffusion coefficient (D eff ) of native RNase A is found to fluctuate around 10 Å 2 /ns, while in the presence of trans azoTAB RNase exhibits a trend of increasing D eff with higher Q values. Considering the similarity of tertiary structure between RNase A (the red feature in Figure 5.3) and lysozyme (in Chapter 4), the NSE data imply that azoTAB induces enhanced protein dynamics in RNase A, probably a domain motion similar to what was observed in lysozyme. To further determine the effect of azoTAB on RNase A dynamics, more data will be collected from NSE 107 measurements with dynamic model analysis including rigid-body calculation and the freely-jointed-domain model. 7 8 9 10 11 12 13 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Pure RNase A D eff (A 2 /ns) Q (A -1 ) o o 7 8 9 10 11 12 13 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 RNase A w/ 10mM S1 trans Q (A -1 ) o D eff (A 2 /ns) o 7 8 9 10 11 12 13 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 RNase A w/ 10mM S1 cis Q (A -1 ) o D eff (A 2 /ns) o Figure 5.3: NSE data of the effective diffusion coefficient of RNase A with (a) 0, (b) 10 (trans), (c) 10 (cis) mM azoTAB. The space filled structure next to each plot represents the protein conformation at the corresponding azoTAB concentration obtained from the SANS data in Figure 5.1 using GA_STRUCT. (a) (b) (c) 108 5.2.3 Effect of azoTAB on function of Ribonuclease A The preliminary SANS and NSE data have shown increased unfolding and enhanced dynamics of RNase A in the presence of trans azoTAB. To examine how these structure and dynamics changes may effect enzyme function, an activity assay is proposed below that can be applied to the complex azoTAB-enzyme system. RNase A serves as a digestive enzyme that catalyzes the hydrolysis of single- stranded RNA by first cleaving the P-O 5 ′ bond of the RNA strand followed by hydrolysis of the nucleoside 2 ′,3 ′-cyclic phosphodiester intermediate to form a 3 ′-phosphomonoester. The enzyme kinetics are usually examined by using the natural substrate “yeast RNA” or the intermediate product nucleoside 2 ′, 3 ′-cyclic phosphodiesters. Attempt to examine the kinetics of RNase A in the presence of azoTAB has been tried with both of the substrates. The relative activity of RNase A shown in Figure 5.4(a), measured with increasing azoTAB concentration, would seem to be higher than the pure protein using yeast RNA as the substrate. However, due to the insensitivity of the substrate and potentially strong surfactant-substrate interactions (positively-charged azoTAB was found to cause substrate aggregation, likely due to neutralization of the negatively-charged long chain RNA), attempts to obtain precise kinetic parameters of RNase A in the presence of azoTAB were not successful due to the small difference and the large error bars of the data collected at different azoTAB concentrations (as seen in Figures 5.4(a) and (b)). Other small molecule substrates such as nucleoside 2 ′, 3 ′-cyclic phosphodiesters require observation of the absorption at 260-290 nm, a region where the 109 (a) (b) 80 100 120 140 160 180 01234 trans cis Relative Activity (%) [azoTAB] (mM) 0.4 0.8 1.2 1.6 2 0 0.1 0.20.3 0.40.5 Pure RNase A 2.5mM trans 2.5mM cis 1/V 1/[Substrate] Figure 5.4: (a) Relative activity of RNase A against yeast RNA as a function of azoTAB concentration (•) and cis azoTAB ( ○). (b) Linweaver-Burk plot of native RNase A (•) and RNase A with 2.5 mM azoTAB (trans ( S ) and cis ( T )). signal would be strongly influenced by the absorption of azoTAB. Thus, attempts to use these small molecules as substrates also failed. Concluding from the trials on obtaining kinetic parameters of RNase A in the presence of azoTAB, a hypersensitive substrate with minimal substrate-surfactant interaction is needed. A fluorescence-based substrate is proposed to be used in enzyme activity measurements. The base of the substrate is a short ribonucleotide having the 5 ′ end labeled with a fluorophore and the 3 ′ end labeled with a quencher as seen in Figure 3(a). 87, 117 The fluorophore is quenched when the substrate is intact due to the close proximity to the quencher and fluorophore. When the substrate is cleaved, however, a dramatic increase in fluorescence would result, as seen in Figure 3(b). A set of commercially available substrates using 5 ′, 6-carboxyfluorescein (6-FAM) and 3 ′, 6- 110 Figure 5.5: (a) Chemical structure of the fluorescence based substrate, 6-FAM- dArU(dA) 2 -6TAMRA. (b) Illustration of the basic idea of the fluorescence based substrate. 117 carboxytetramethylrhodamine (6-TAMRA) as the fluorophore and the quencher may be used to examine the azoTAB influenced ribonuclease activity. Figures 5.6(a) and (b) show the emission spectra of the substrate before and after cleavage and the time course of the enzyme kinetic, respectively. The advantage of this substrate is that it is very sensitive and the enzyme kinetics can be determined by continuously monitoring the fluorophore (6-FAM) fluorescence emission at 515 nm with excitation at 490 nm. The UV-vis spectrum of azoTAB indicates that the surfactants may slightly absorb in the 450-500 nm region, especially when in the cis form. However, the kinetics time course can still be obtained if the surfactant concentration is kept low, which is usually the case for enzyme activity examination. In case that the azoTAB seriously alters the fluorescence of the product, an alternative fluorophore-quencher pair excited at higher wavelength may be considered using the same substrate base above. (a) (b) 111 Figure 5.6: (a) Emission spectra of 6-FAM~IArU(dA)2~5-TAMRA (solid line) and its cleavage product (dashed line) on excitation at 490 nm. (b) Exponential rise analysis of the cleavage of 6-FAM-dArU(dA)2-6-TAMRA by ribonuclease A. "CPS" refers to photon counts per second. 117 (a) (b) 112 Bibliography (1) Agarwal, P. K. Enzymes: An Integrated View of Structure, Dynamics and Function. Microbial Cell Factories 5, No pp given (2006). (2) Akcasu, A. Z., Hammouda, B., Stockmayer, W. H. and Tanaka, G. First Cumulant for Chains with Constraints. Journal of Chemical Physics 85, 4734-4743 (1986). (3) Ananthapadmanabhan, K. P. in Interactions of Surfactants with Polymers and Proteins (eds. Goddard, E. D. and Anathapadmanabhan, K. P.) 319-365 (CRC, Boca Raton, Fla, 1993). (4) Arnold, G. E. and Ornstein, R. L. Protein Hinge Bending as Seen in Molecular Dynamics Simulations of Native and M6i Mutant T4 Lysozymes. Biopolymers 41, 533- 544 (1997). (5) Banerjee, S. K., Kregar, I., Turk, V. and Rupley, J. A. Lysozyme-Catalyzed Reaction of the N-Acetylglucosamine Hexasaccharide. Dependence of Rate on Ph. The Journal of biological chemistry 248, 4786-4792 (1973). (6) Barth, A. and Zscherp, C. What Vibrations Tell About Proteins. Quarterly Reviews of Biophysics 35, 369-430 (2002). (7) Bendedouch, D. and Chen, S. H. Structure and Interparticle Interactions of Bovine Serum Albumin in Solution Studied by Small-Angle Neutron Scattering. Journal of Physical Chemistry 87, 1473-1477 (1983). (8) Benkovic, S. J. and Hammes-Schiffer, S. Enzyme Motions inside and Out. Science 312, 208-209 (2006). (9) Benkovic, S. J. and Hammes-Schiffer, S. A Perspective on Enzyme Catalysis. Science 301, 1196-1202 (2003). (10) Bernath, F. R. and Vieth, W. R. Lysozyme Activity in the Presence of Nonionic Detergent Micelles. Biotechnology and Bioengineering 14, 737-752 (1972). (11) Boehr, D. D., Dyson, H. J. and Wright, P. E. An Nmr Perspective on Enzyme Dynamics. Chemical Reviews (Washington, DC, United States) 106, 3055-3079 (2006). (12) Boulkanz, L., Balcar, N. and Baron, M.-H. Ft-Ir Analysis for Structural Characterization of Albumin Adsorbed on the Reversed-Phase Support Rp-C6. Applied Spectroscopy 49, 1737-1746 (1995). 113 (13) Boulkanz, L., Vidal-Madjar, C., Balcar, N. and Baron, M.-H. Adsorption Mechanism of Human Serum Albumin on a Reversed-Phase Support by Kinetic, Chromatographic, and Ftir Methods. Journal of Colloid and Interface Science 188, 58-67 (1997). (14) Bramanti, E. and Benedetti, E. Determination of the Secondary Structure of Isomeric Forms of Human Serum Albumin by a Particular Frequency Deconvolution Procedure Applied to Fourier Transform Ir Analysis. Biopolymers 38, 639-653 (1996). (15) Brown, J. R. (eds. Rosenoer, V. M., Oratz, M., Rothschild, M. A. and Editors) 397 pp (Pergamon, Oxford, Engl., 1977). (16) Bruccoleri, R. E., Karplus, M. and McCammon, J. A. The Hinge-Bending Mode of a Lysozyme-Inhibitor Complex. Biopolymers 25, 1767-1802 (1986). (17) Bruice, T. C. and Benkovic, S. J. Chemical Basis for Enzyme Catalysis. Biochemistry 39, 6267-6274 (2000). (18) Bu, Z., Biehl, R., Monkenbusch, M., Richter, D. and Callaway, D. J. E. Coupled Protein Domain Motion in Taq Polymerase Revealed by Neutron Spin-Echo Spectroscopy. Proceedings of the National Academy of Sciences of the United States of America 102, 17646-17651 (2005). (19) Byler, D. M. and Susi, H. Examination of the Secondary Structure of Proteins by Deconvolved Ftir Spectra. Biopolymers 25, 469-487 (1986). (20) Cai, S. and Singh, B. R. A Distinct Utility of the Amide Iii Infrared Band for Secondary Structure Estimation of Aqueous Protein Solutions Using Partial Least Squares Methods. Biochemistry 43, 2541-2549 (2004). (21) Cameron, C. E. and Benkovic, S. J. Evidence for a Functional Role of the Dynamics of Glycine-121 of Escherichia Coli Dihydrofolate Reductase Obtained from Kinetic Analysis of a Site-Directed Mutant. Biochemistry 36, 15792-15800 (1997). (22) Cannon, W. R., Singleton, S. F. and Benkovic, S. J. A Perspective on Biological Catalysis. Nature Structural Biology 3, 821-833 (1996). (23) Carrasco, B. and De la Torre, J. G. Hydrodynamic Properties of Rigid Particles: Comparison of Different Modeling and Computational Procedures. Biophysical journal 76, 3044-3057 (1999). (24) Carter, D. C. Structure of Serum Albumin. Advances in Protein Chemistry 45, 153-203 (1994). 114 (25) Celej Maria, S., Montich Guillermo, G. and Fidelio Gerardo, D. Protein Stability Induced by Ligand Binding Correlates with Changes in Protein Flexibility. Protein Science 12, 1496-1506 (2003). (26) Celej, M. S., Montich, G. G. and Fidelio, G. D. Conformational Flexibility of Avidin: The Influence of Biotin Binding. Biochemical and Biophysical Research Communications 325, 922-927 (2004). (27) Chen, G. Q. and Gouaux, E. Probing the Folding and Unfolding of Wild-Type and Mutant Forms of Bacteriorhodopsin in Micellar Solutions: Evaluation of Reversible Unfolding Conditions. Biochemistry 38, 15380-15387 (1999). (28) Chen, L., Hodgson, K. O. and Doniach, S. A Lysozyme Folding Intermediate Revealed by Solution X-Ray Scattering. Journal of Molecular Biology 261, 658-672 (1996). (29) Chen, M. C. and Lord, R. C. Laser-Excited Raman Spectroscopy of Biomolecules. Viii. Conformational Study of Bovine Serum Albumin. Journal of the American Chemical Society 98, 990-992 (1976). (30) Chen, S. H. and Teixeira, J. Structure and Fractal Dimension of Protein-Detergent Complexes. Physical Review Letters 57, 2583-2586 (1986). (31) Chen, Y., Hu, D., Vorpagel, E. R. and Lu, H. P. Probing Single-Molecule T4 Lysozyme Conformational Dynamics by Intramolecular Fluorescence Energy Transfer. Journal of Physical Chemistry B 107, 7947-7956 (2003). (32) Chen, Y.-H., Yang, J. T. and Chau, K. H. Determination of the Helix and B Form of Proteins in Aqueous Solution by Circular Dichroism. Biochemistry 13, 3350-3359 (1974). (33) Cole, R. and Loria, J. P. Evidence for Flexibility in the Function of Ribonuclease A. Biochemistry 41, 6072-6081 (2002). (34) Curry, S., Mandelkow, H., Brick, P. and Franks, N. Crystal Structure of Human Serum Albumin Complexed with Fatty Acid Reveals an Asymmetric Distribution of Binding Sites. Nature Structural Biology 5, 827-835 (1998). (35) Daniel, R. M., Dunn, R. V., Finney, J. L. and Smith, J. C. The Role of Dynamics in Enzyme Activity. Annual Review of Biophysics and Biomolecular Structure 32, 69-92 (2003). (36) de Jongh, H. H. J., Goormaghtigh, E. and Ruysschaert, J.-M. Tertiary Stability of Native and Methionine-80 Modified Cytochrome C Detected by Proton-Deuterium Exchange Using Online Fourier Transform Infrared Spectroscopy. Biochemistry 34, 172- 179 (1995). 115 (37) De la Torre, J. G., Huertas, M. L. and Carrasco, B. Calculation of Hydrodynamic Properties of Globular Proteins from Their Atomic-Level Structure. Biophysical journal 78, 719-730 (2000). (38) Di Nola, A., Gavuzzo, E., Mazza, F., Pochetti, G. and Roccatano, D. Internal B- Turn Hydration: Crystallographic Evidence and Molecular Dynamics Simulation. Journal of Physical Chemistry 99, 9625-9631 (1995). (39) Dobson, C. M. Protein Folding. Solid Evidence for Molten Globules. Current Biology 4, 636-640 (1994). (40) Durchschlag, H., et al. Effects of X- and Uv-Irradiation on Proteins. Radiation Physics and Chemistry 47, 501-505 (1996). (41) Durchschlag, H., Hefferle, T. and Zipper, P. Comparative Investigations of the Effects of X- and Uv-Irradiation on Lysozyme in the Absence or Presence of Additives. Radiation Physics and Chemistry 67, 479-486 (2003). (42) Eastoe, J., Dominguez, M. S., Wyatt, P., Beeby, A. and Heenan, R. K. Properties of a Stilbene-Containing Gemini Photosurfactant: Light-Triggered Changes in Surface Tension and Aggregation. Langmuir 18, 7837-7844 (2002). (43) Eastoe, J., et al. Photoresponsive Microemulsions. Langmuir 19, 6579-6581 (2003). (44) Eastoe, J. and Vesperinas, A. Self-Assembly of Light-Sensitive Surfactants. Soft Matter 1, 338-347 (2005). (45) Eisenthal, R., Danson, M. J. and Editors. Enzyme Assays: A Practical Approach (Oxford University Press, New York, 1992). (46) Faber, H. R. and Matthews, B. W. A Mutant T4 Lysozyme Displays Five Different Crystal Conformations. Nature 348, 263-266 (1990). (47) Fabian, H., et al. Secondary Structure and Temperature-Induced Unfolding and Refolding of Ribonuclease T1 in Aqueous Solution. A Fourier Transform Infrared Spectroscopic Study. Journal of Molecular Biology 232, 967-981 (1993). (48) Farago, B. Neutron Spin Echo Study of Well Organized Soft Matter Systems. Physica B: Condensed Matter (Amsterdam, Netherlands) 385-386, 688-691 (2006). (49) Fukamizo, T., Minematsu, T., Yanase, Y., Hayashi, K. and Goto, S. Substrate Size Dependence of Lysozyme-Catalyzed Reaction. Archives of Biochemistry and Biophysics 250, 312-321 (1986). 116 (50) Gabel, F., et al. Protein Dynamics Studied by Neutron Scattering. Quarterly Reviews of Biophysics 35, 327-367 (2002). (51) Gebe, J. A. and Schurr, J. M. Intramolecular Interference Effects in Dynamic Light Scattering from Rigid Rings. Biopolymers 33, 1757-1764 (1993). (52) Gekko, K., Obu, N., Li, J. and Lee, J. C. A Linear Correlation between the Energetics of Allosteric Communication and Protein Flexibility in the Escherichia Coli Cyclic Amp Receptor Protein Revealed by Mutation-Induced Changes in Compressibility and Amide Hydrogen-Deuterium Exchange. Biochemistry 43, 3844-3852 (2004). (53) Glinka, C. J., et al. The 30 M Small-Angle Neutron Scattering Instruments at the National Institute of Standards and Technology. Journal of Applied Crystallography 31, 430-445 (1998). (54) Goto, N. K., Skrynnikov, N. R., Dahlquist, F. W. and Kay, L. E. What Is the Average Conformation of Bacteriophage T4 Lysozyme in Solution? A Domain Orientation Study Using Dipolar Couplings Measured by Solution Nmr. Journal of Molecular Biology 308, 745-764 (2001). (55) Grimard, V., et al. Structure and Dynamics of the Membrane-Embedded Domain of Lmrainvestigated by Coupling Polarized Atr-Ftir Spectroscopy and (1)H/(2)H Exchange. Biochemistry 40, 11876-11886 (2001). (56) Gross, M., Auerbach, G. and Jaenicke, R. The Catalytic Activities of Monomeric Enzymes Show Complex Pressure Dependence. FEBS letters 321, 256-260 (1993). (57) Haliloglu, T. and Bahar, I. Structure-Based Analysis of Protein Dynamics: Comparison of Theoretical Results for Hen Lysozyme with X-Ray Diffraction and Nmr Relaxation Data. Proteins: Structure, Function, and Genetics 37, 654-667 (1999). (58) Hamill, A. C., Wang, S.-C. and Lee, C. T., Jr. Probing Lysozyme Conformation with Light Reveals a New Folding Intermediate. Biochemistry 44, 15139-15149 (2005). (59) Hamill, A. C., Wang, S.-C. and Lee, C. T., Jr. Solution Structure of an Amyloid- Forming Protein During Photoinitiated Hexamer-Dodecamer Transitions Revealed through Small-Angle Neutron Scattering. Biochemistry 46, 7694-7705 (2007). (60) Hammes, G. G. Multiple Conformational Changes in Enzyme Catalysis. Biochemistry 41, 8221-8228 (2002). (61) Hammes-Schiffer, S. Impact of Enzyme Motion on Activity. Biochemistry 41, 13335-13343 (2002). (62) Hammes-Schiffer, S. and Benkovic, S. J. Relating Protein Motion to Catalysis. Annual review of biochemistry 75, 519-541 (2006). 117 (63) Haris, P. I. and Chapman, D. The Conformational Analysis of Peptides Using Fourier Transform Ir Spectroscopy. Biopolymers 37, 251-263 (1995). (64) Hayashi, K., Kugimiya, M., Imoto, T., Funatsu, M. and Bigelow, C. C. The Inhibitory Interaction of Cationic Detergents with the Active Center of Lysozyme. I. Site of Interaction. Biochemistry 7, 1461-1466 (1968). (65) Hayashi, K., Kugimiya, M., Imoto, T., Funatsu, M. and Bigelow, C. C. The Inhibitory Interaction of Cationic Detergents with the Active Center of Lysozyme. Ii. The Ph Dependence of the Interaction. Biochemistry 7, 1467-1472 (1968). (66) Hayashita, T., Kurosawas, T., Miyata, T., Tanaka, K. and Igawa, M. Effect of Structural Variation within Cationic Azo-Surfactant Upon Photoresponsive Function in Aqueous Solution. Colloid and Polymer Science 272, 1611-1619 (1994). (67) Hayward, S., Kitao, A. and Berendsen, H. J. C. Model-Free Methods of Analyzing Domain Motions in Proteins from Simulation: A Comparison of Normal Mode Analysis and Molecular Dynamics Simulation of Lysozyme. Proteins: Structure, Function, and Genetics 27, 425-437 (1997). (68) He, X. M. and Carter, D. C. Atomic Structure and Chemistry of Human Serum Albumin. Nature 358, 209-215 (1992). (69) He, X. M. and Carter, D. C. Atomic Structure and Chemistry of Human Serum Albumin. Nature (London, United Kingdom) 358, 209-215 (1992). (70) Hedoux, A., et al. Evidence of a Two-Stage Thermal Denaturation Process in Lysozyme: A Raman Scattering and Differential Scanning Calorimetry Investigation. Journal of Chemical Physics 124, 014703/014701-014703/014707 (2006). (71) Heimburg, T. and Marsh, D. Investigation of Secondary and Tertiary Structural Changes of Cytochrome C in Complexes with Anionic Lipids Using Amide Hydrogen Exchange Measurements: An Ftir Study. Biophysical Journal 65, 2408-2417 (1993). (72) Heller, W. T., Krueger, J. K. and Trewhella, J. Further Insights into Calmodulin- Myosin Light Chain Kinase Interaction from Solution Scattering and Shape Restoration. Biochemistry 42, 10579-10588 (2003). (73) Hellweg, T., Gradzielski, M., Farago, B. and Langevin, D. Shape Fluctuations of Microemulsion Droplets: A Neutron Spin-Echo Study. Colloids and Surfaces, A: Physicochemical and Engineering Aspects 183-185, 159-169 (2001). (74) http://www.ncnr.nist.gov/dave. 118 (75) Huang, K. S., Bayley, H., Liao, M. J., London, E. and Khorana, H. G. Refolding of an Integral Membrane Protein. Denaturation, Renaturation, and Reconstitution of Intact Bacteriorhodopsin and Two Proteolytic Fragments. Journal of biological chemistry 256, 3802-3809 (1981). (76) Ibel, K., et al. Protein-Decorated Micelle Structure of Sodium-Dodecyl-Sulfate - Protein Complexes as Determined by Neutron Scattering. European Journal of Biochemistry 190, 311-318 (1990). (77) Imoto, T., Johnson, L. N., North, A. C. T., Phillips, D. C. and Rupley, J. A. Vertebrate Lysozymes. Enzymes, 3rd Ed. 7, 665-868 (1972). (78) Imoto, T., et al. Lysozyme Requires Fluctuation of the Active Site for the Manifestation of Activity. Protein Engineering 7, 743-748 (1994). (79) Inada, T., Terabayashi, T., Yamaguchi, Y., Kato, K. and Kikuchi, K. Modulation of the Catalytic Mechanism of Hen Egg White Lysozyme (Hewl) by Photochromism of Azobenzene. Journal of Photochemistry and Photobiology, A: Chemistry 175, 100-107 (2005). (80) Ito, Y., Yamada, H. and Imoto, T. Colorimetric Assay for Lysozyme Using Micrococcus Luteus Labeled with a Blue Dye, Remazol Brilliant Blue R, as a Substrate. Chemical & Pharmaceutical Bulletin 40, 1523-1526 (1992). (81) Jackson, M., Haris, P. I. and Chapman, D. Fourier Transform Infrared Spectroscopic Studies of Ca(2+)-Binding Proteins. Biochemistry 30, 9681-9686 (1991). (82) Johnson, L. N. and Phillips, D. C. Structure of Some Crystalline Lysozyme- Inhibitor Complexes Determined by X-Ray Analysis at 6 Angstrom Resolution. Nature 206, 761-763 (1965). (83) Jones, M. N. and Manley, P. Interaction between Lysozyme and N-Alkyl Sulphates in Aqueous Solution. Journal of the Chemical Society, Faraday Transactions 76, 654-664 (1980). (84) Jones, M. N., Prieto, G., Rio, J. M. d. and Sarmiento, F. Denaturation of Lysozyme by N-Alkyltrimethylammonium Bromides in Alkaline Solution. Journal of the Chemical Society, Faraday Transactions, 2805-2809 (1995). (85) Kamatari, Y. O., Konno, T., Kataoka, M. and Akasaka, K. The Methanol-Induced Transition and the Expanded Helical Conformation in Hen Lysozyme. Protein Science 7, 681-688 (1998). (86) Kang, J., et al. Interactions of Human Serum Albumin with Chlorogenic Acid and Ferulic Acid. Biochimica et biophysica acta 1674, 205-214 (2004). 119 (87) Kelemen, B. R., et al. Hypersensitive Substrate for Ribonucleases. Nucleic acids research 27, 3696-3701 (1999). (88) Khan, M. Y. Direct Evidence for the Involvement of Domain Iii in the N-F Transition of Bovine Serum Albumin. Biochemical journal 236, 307-310 (1986). (89) Kirkwood, J. G. The General Theory of Irreversible Processes in Solutions of Macromolecules. Journal of Polymer Science 12, 1-14 (1954). (90) Kohen, A. and Klinman, J. P. Protein Flexibility Correlates with Degree of Hydrogen Tunneling in Thermophilic and Mesophilic Alcohol Dehydrogenases. Journal of the American Chemical Society 122, 10738-10739 (2000). (91) Komura, S., Takeda, T., Seto, H. and Nagao, M. Dynamical Fluctuation of Cylindrical Micelles and Membranes in Binary and Ternary Amphiphilic Microemulsion Systems. Lecture Notes in Physics 601, 302-311 (2003). (92) Kossiakoff, A. A. and Spencer, S. A. Neutron Diffraction Identifies Histidine-57 as the Catalytic Base in Trypsin. Nature 288, 414-416 (1980). (93) Kumagai, I. and Miura, K. Enhanced Bacteriolytic Activity of Hen Egg-White Lysozyme Due to Conversion of Trp62 to Other Aromatic Amino Acid Residues. Journal of Biochemistry (Tokyo, Japan) 105, 946-948 (1989). (94) Kuroki, R., Weaver, L. H. and Matthews, B. W. A Covalent Enzyme-Substrate Intermediate with Saccharide Distortion in a Mutant T4 Lysozyme. Science 262, 2030- 2033 (1993). (95) Kuroki, R., Weaver, L. H. and Matthews, B. W. A Covalent Enzyme-Substrate Intermediate with Saccharide Distortion in a Mutant T4 Lysozyme. Science (New York, N.Y.) 262, 2030-2033 (1993). (96) Le Ny, A.-L. M. and Lee, C. T., Jr. Photoreversible DNA Condensation Using Light-Responsive Surfactants. Journal of the American Chemical Society 128, 6400-6408 (2006). (97) Lee, C. T., Jr., Smith, K. A. and Hatton, T. A. Photocontrol of Protein Folding: The Interaction of Photosensitive Surfactants with Bovine Serum Albumin. Biochemistry 44, 524-536 (2005). (98) Liang, Z.-X., et al. Evidence for Increased Local Flexibility in Psychrophilic Alcohol Dehydrogenase Relative to Its Thermophilic Homologue. Biochemistry 43, 14676-14683 (2004). (99) Lin, V. J. C. and Koenig, J. L. Raman Studies of Bovine Serum Albumin. Biopolymers 15, 203-218 (1976). 120 (100) Liu, H., Yang, W. and Chen, J. Effects of Surfactants on Emulsification and Secondary Structure of Lysozyme in Aqueous Solutions. Biochemical Engineering Journal 2, 187-196 (1998). (101) London, E. and Khorana, H. G. Denaturation and Renaturation of Bacteriorhodopsin in Detergents and Lipid-Detergent Mixtures. Journal of biological chemistry 257, 7003-7011 (1982). (102) Maenaka, K., et al. Dissection of Protein-Carbohydrate Interactions in Mutant Hen Egg-White Lysozyme Complexes and Their Hydrolytic Activity. Journal of Molecular Biology 247, 281-293 (1995). (103) Marti, T. Refolding of Bacteriorhodopsin from Expressed Polypeptide Fragments. Journal of biological chemistry 273, 9312-9322 (1998). (104) Matthews, B. W. Studies on Protein Stability with T4 Lysozyme. Advances in Protein Chemistry 46, 249-278 (1995). (105) Matthews, C. R. Pathways of Protein Folding. Annual review of biochemistry 62, 653-683 (1993). (106) McCammon, J. A., Gelin, B. R., Karplus, M. and Wolynes, P. G. The Hinge- Bending Mode in Lysozyme. Nature 262, 325-326 (1976). (107) McHaourab, H. S., Oh, K. J., Fang, C. J. and Hubbell, W. L. Conformation of T4 Lysozyme in Solution. Hinge-Bending Motion and the Substrate-Induced Conformational Transition Studied by Site-Directed Spin Labeling. Biochemistry 36, 307-316 (1997). (108) Meinhold, L. and Smith, J. C. Protein Dynamics from X-Ray Crystallography: Anisotropic, Global Motion in Diffuse Scattering Patterns. Proteins: Structure, Function, and Bioinformatics 66, 941-953 (2007). (109) Mine, S., Tate, S., Ueda, T., Kainosho, M. and Imoto, T. Analysis of the Relationship between Enzyme Activity and Its Internal Motion Using Nuclear Magnetic Resonance: 15n Relaxation Studies of Wild-Type and Mutant Lysozyme. Journal of Molecular Biology 286, 1547-1565 (1999). (110) Monkenbusch, M. Soft Matter and Biology. Lecture Notes in Physics 601, 246- 267 (2003). (111) Moriyama, Y., Kawasaka, Y. and Takeda, K. Protective Effect of Small Amounts of Sodium Dodecyl Sulfate on the Helical Structure of Bovine Serum Albumin in Thermal Denaturation. Journal of Colloid and Interface Science 257, 41-46 (2003). 121 (112) Murayama, K. and Tomida, M. Heat-Induced Secondary Structure and Conformation Change of Bovine Serum Albumin Investigated by Fourier Transform Infrared Spectroscopy. Biochemistry 43, 11526-11532 (2004). (113) Muzammil, S., Kumar, Y. and Tayyab, S. Molten Globule-Like State of Human Serum Albumin at Low Ph. European Journal of Biochemistry 266, 26-32 (1999). (114) Neault, J. F. and Tajmir-Riahi, H. A. Interaction of Cisplatin with Human Serum Albumin. Drug Binding Mode and Protein Secondary Structure. Biochimica et biophysica acta 1384, 153-159 (1998). (115) Ohmura, T., Ueda, T., Ootsuka, K., Saito, M. and Imoto, T. Stabilization of Hen Egg White Lysozyme by a Cavity-Filling Mutation. Protein Science 10, 313-320 (2001). (116) Olsson, M. H. M., Parson, W. W. and Warshel, A. Dynamical Contributions to Enzyme Catalysis: Critical Tests of a Popular Hypothesis. Chem. Rev. 106, 1737-1756 (2006). (117) Park, C., et al. Fast, Facile, Hypersensitive Assays for Ribonucleolytic Activity. Methods in enzymology 341, 81-94 (2001). (118) Parker, W. and Song, P. S. Protein Structures in Sds Micelle-Protein Complexes. Biophysical journal 61, 1435-1439 (1992). (119) Pauling, L. and Corey, R. B. Two Hydrogen-Bonded Spiral Configurations of the Polypeptide Chain. Journal of the American Chemical Society 72, 5349 (1950). (120) Pearson, W. R. Rapid and Sensitive Sequence Comparison with Fastp and Fasta. Methods in enzymology 183, 63-98 (1990). (121) Peters, J. t. All About Albumin: Biochemistry, Genetics, and Medical Applications (Academic Press, San Diego, California, 1995). (122) Poole, P. L. and Finney, J. L. Hydration-Induced Conformational and Flexibility Changes in Lysozyme at Low Water Content. International Journal of Biological Macromolecules 5, 308-310 (1983). (123) Pribic, R., van Stokkum, I. H., Chapman, D., Haris, P. I. and Bloemendal, M. Protein Secondary Structure from Fourier Transform Infrared and/or Circular Dichroism Spectra. Analytical biochemistry 214, 366-378 (1993). (124) Purcell, M., et al. Interactions of Atrazine and 2,4-D with Human Serum Albumin Studied by Gel and Capillary Electrophoresis, and Ftir Spectroscopy. Biochimica et biophysica acta 1548, 129-138 (2001). 122 (125) Purcell, M., Neault, J. F. and Tajmir-Riahi, H. A. Interaction of Taxol with Human Serum Albumin. Biochimica et biophysica acta 1478, 61-68 (2000). (126) Raussens, V., Narayanaswami, V., Goormaghtigh, E., Ryan, R. O. and Ruysschaert, J. M. Hydrogen/Deuterium Exchange Kinetics of Apolipophorin-Iii in Lipid-Free and Phospholipid-Bound States. An Analysis by Fourier Transform Infrared Spectroscopy. The Journal of biological chemistry 271, 23089-23095 (1996). (127) Reed, R. G., Feldhoff, R. C., Clute, O. L. and Peters, T., Jr. Fragments of Bovine Serum Albumin Produced by Limited Proteolysis. Conformation and Ligand Binding. Biochemistry 14, 4578-4583 (1975). (128) Reynolds, J. A., Herbert, S., Polet, H. and Steinhardt, J. The Binding of Divers Detergent Anions to Bovine Serum Albumin. Biochemistry 6, 937-947 (1967). (129) Richards, P. G., Walton, D. J. and Heptinstall, J. The Effects of Tyrosine Nitration on the Structure and Function of Hen Egg-White Lysozyme. Biochemical journal 315, 473-479 (1996). (130) Richter, D., Monkenbusch, M., Arbe, A. and Colmenero, J. Neutron Scattering and the Glass Transition in Polymers - Present Status and Future Opportunities. Journal of Non-Crystalline Solids 287, 286-296 (2001). (131) Roh, J. H., et al. Influence of Hydration on the Dynamics of Lysozyme. Biophysical journal 91, 2573-2588 (2006). (132) Rosov, N., Rathgeber, S. and Monkenbusch, M. in Scattering from Polymers: Characterization by X-Rays, Neutrons, and Light (Acs Symp. Series 739) (eds. Cebe, P., Hsaio, B. S. and Lohse, D. J.) 103-116 (American Chemical Society: Washington D. C., 2000). (133) Sadana, A. Biocatalysis: Fundamentals of Enzyme Deactivation Kinetics (Prentice-Hall, Inc, Englewood Cliffs, New Jersey, 1991). (134) Scheirlinckx, F., Buchet, R., Ruysschaert, J.-M. and Goormaghtigh, E. Monitoring of Secondary and Tertiary Structure Changes in the Gastric H+/K+-Atpase by Infrared Spectroscopy. European Journal of Biochemistry 268, 3644-3653 (2001). (135) Schramm, V. L. Enzymatic Transition States and Transition State Analogues. Current Opinion in Structural Biology 15, 604-613 (2005). (136) Shang, T., Smith, K. A. and Hatton, T. A. Photoresponsive Surfactants Exhibiting Unusually Large, Reversible Surface Tension Changes under Varying Illumination Conditions. Langmuir 19, 10764-10773 (2003). 123 (137) Shin, J. Y. and Abbott, N. L. Using Light to Control Dynamic Surface Tensions of Aqueous Solutions of Water Soluble Surfactants. Langmuir 15, 4404-4410 (1999). (138) Shugar, D. Measurement of Lysozyme Activity and the Ultraviolet Inactivation of Lysozyme. Biochimica et biophysica acta 8, 302-309 (1952). (139) Smeller, L., Meersman, F. and Heremans, K. Refolding Studies Using Pressure: The Folding Landscape of Lysozyme in the Pressure-Temperature Plane. Biochimica et Biophysica Acta, Proteins and Proteomics 1764, 497-505 (2006). (140) Stenstam, A., Montalvo, G., Grillo, I. and Gradzielski, M. Small Angle Neutron Scattering Study of Lysozyme-Sodium Dodecyl Sulfate Aggregates. Journal of Physical Chemistry B 107, 12331-12338 (2003). (141) Stuhrmann, H. B. and Fuess, H. Neutron Small-Angle Scattering Study of Hen Egg-White Lysozyme. Acta Crystallographica, Section A: Crystal Physics, Diffraction, Theoretical and General Crystallography A32, Pt. 1, 67-74 (1976). (142) Susi, H. and Byler, D. M. Resolution-Enhanced Fourier Transform Infrared Spectroscopy of Enzymes. Methods in enzymology 130, 290-311 (1986). (143) Svergun, D. I. and Koch, M. H. J. Small-Angle Scattering Studies of Biological Macromolecules in Solution. Reports on Progress in Physics 66, 1735-1782 (2003). (144) Svergun, D. I., Petoukhov, M. V. and Koch, M. H. Determination of Domain Structure of Proteins from X-Ray Solution Scattering. Biophysical journal 80, 2946-2953 (2001). (145) Takeda, K., Harada, K., Yamaguchi, K. and Moriyama, Y. Conformational Changes of Bovine Serum Albumin in an Aqueous Solution of Sodium Bis(2-Ethylhexyl) Sulfosuccinate and in the Reverse Micelle of the Same Surfactant. Journal of Colloid and Interface Science 164, 382-386 (1994). (146) Takeda, K. and Moriyama, Y. The Interaction of Bovine Serum Albumin with Sodium Dodecyl Sulfate: Binding of the Surfactant and Conformational Change of the Protein Induced by the Binding. Current Topics in Colloid & Interface Science 1, 109- 135 (1997). (147) Takeda, K., Moriyama, Y. and Hachiya, K. in Encyclopedia of Surface and Colloid Science (eds. Hubbard, A. and Somasundaran, P.) 2558-2574 (Marcel Dekker, New York, 2002). (148) Takeda, K., Sasa, K., Kawamoto, K., Wada, A. and Aoki, K. Secondary Structure Changes of Disulfide Bridge-Cleaved Bovine Serum Albumin in Solutions of Urea, Guanidine Hydrochloride, and Sodium Dodecyl Sulfate. Journal of Colloid and Interface Science 124, 284-289 (1988). 124 (149) Takeda, K., Shigeta, M. and Aoki, K. Secondary Structures of Bovine Serum Albumin in Anionic and Cationic Surfactant Solutions. Journal of Colloid and Interface Science 117, 120-126 (1987). (150) Tanford, C. and Wager, M. L. Hydrogen Ion Equilibrium of Lysozyme. Journal of the American Chemical Society 76, 3331-3336 (1954). (151) Tousignant, A. and Pelletier, J. N. Protein Motions Promote Catalysis. Chem. Biol. 11, 1037-1042 (2004). (152) Turro, N. J., Lei, X.-G., Ananthapadmanabhan, K. P. and Aronson, M. Spectroscopic Probe Analysis of Protein-Surfactant Interactions: The Bsa/Sds System. Langmuir 11, 2525-2533 (1995). (153) Vendruscolo, M. and Dobson, C. M. Dynamic Visions of Enzymatic Reactions. Science 313, 1586-1587 (2006). (154) Verhamme, I. M. A., Van Dedem, G. W. K. and Lauwers, A. R. Ionic-Strength- Dependent Substrate Inhibition of the Lysis of Micrococcus Luteus by Hen Egg-White Lysozyme. European Journal of Biochemistry 172, 615-620 (1988). (155) Vermeer, A. W. P. and Norde, W. The Influence of the Binding of Low Molecular Weight Surfactants on the Thermal Stability and Secondary Structure of Igg. Colloids and Surfaces, A: Physicochemical and Engineering Aspects 161, 139-150 (2000). (156) Vigano, C., Goormaghtigh, E. and Ruysschaert, J. M. Detection of Structural and Functional Asymmetries in P-Glycoprotein by Combining Mutagenesis and H/D Exchange Measurements. Chemistry and Physics of Lipids 122, 121-135 (2003). (157) Vigano, C., et al. Hydrogen-Deuterium Exchange in Membrane Proteins Monitored by Ir Spectroscopy: A New Tool to Resolve Protein Structure and Dynamics. Biopolymers 74, 19-26 (2004). (158) Wagner, G., Hyberts, S. G. and Havel, T. F. Nmr Structure Determination in Solution: A Critique and Comparison with X-Ray Crystallography. Annual Review of Biophysics and Biomolecular Structure 21, 167-198 (1992). (159) Wang, D. and Geva, E. Protein Structure and Dynamics from Single-Molecule Fluorescence Resonance Energy Transfer. Journal of Physical Chemistry B 109, 1626- 1634 (2005). (160) Wang, S.-C. and Lee, C. T., Jr. Enhanced Enzymatic Activity through Photoreversible Conformational Changes. Biochemistry 46, 14557-14566 (2007). 125 (161) Wang, S.-C. and Lee, C. T., Jr. Protein Secondary Structure Controlled with Light and Photoresponsive Surfactants. Journal of Physical Chemistry B 110, 16117-16123 (2006). (162) Warshel, A. and Levitt, M. Theoretical Studies of Enzymic Reactions: Dielectric, Electrostatic and Steric Stabilization of the Carbonium Ion in the Reaction of Lysozyme. Journal of Molecular Biology 103, 227-249 (1976). (163) Watanabe, M., Aizawa, T., Demura, M. and Nitta, K. Effect of Hydrostatic Pressure on Conformational Changes of Canine Milk Lysozyme between the Native, Molten Globule, and Unfolded States. Biochimica et Biophysica Acta, Proteins and Proteomics 1702, 129-136 (2004). (164) Willner, I. and Rubin, S. Reversible Photoregulation of the Activities of Proteins. Reactive Polymers 21, 177-186 (1993). (165) Willner, I., Rubin, S. and Riklin, A. Photoregulation of Papain Activity through Anchoring Photochromic Azo Groups to the Enzyme Backbone. Journal of the American Chemical Society 113, 3321-3325 (1991). (166) Wirth, S. J. and Wolf, G. A. Dye-Labeled Substrates for the Assay and Detection of Chitinase and Lysozyme Activity. Journal of Microbiological Methods 12, 197-205 (1990). (167) Xie, M.-X., Xu, X.-Y. and Wang, Y.-D. Interaction between Hesperetin and Human Serum Albumin Revealed by Spectroscopic Methods. Biochimica et Biophysica Acta, General Subjects 1724, 215-224 (2005). (168) Yamasaki, N., Tsujita, T. and Takakuwa, M. Simple Colorimetric Method for the Determination of Lysozyme Activity. Agricultural and Biological Chemistry 37, 1507- 1508 (1973). (169) Yoshida, Y., Ohkuri, T., Kino, S., Ueda, T. and Imoto, T. Elucidation of the Relationship between Enzyme Activity and Internal Motion Using a Lysozyme Stabilized by Cavity-Filling Mutations. Cellular and Molecular Life Sciences 62, 1047-1055 (2005). (170) Zavodszky, P., Kardos, J., Svingor, A. and Petsko, G. A. Adjustment of Conformational Flexibility Is a Key Event in the Thermal Adaptation of Proteins. Proceedings of the National Academy of Sciences of the United States of America 95, 7406-7411 (1998). (171) Zhang, X. and Houk, K. N. Why Enzymes Are Proficient Catalysts: Beyond the Pauling Paradigm. Accounts of Chemical Research 38, 379-385 (2005). 126 (172) Zhang, X.-j., Wozniak, J. A. and Matthews, B. W. Protein Flexibility and Adaptability Seen in 25 Crystal Forms of T4 Lysozyme. Journal of Molecular Biology 250, 527-552 (1995).

Abstract (if available)

Abstract The interaction of a light-responsive azobenzene-based surfactant ( azoTAB ) with proteins has been investigated as a means to photoreversibly control protein structure, dynamics, and function. AzoTAB undergoes a reversible photoisomeriztion upon exposure to appropriate wavelength of light, with the visible-light, trans isomer being more hydrophobic and, thus, inducing a greater degree of protein unfolding than the UV-light, cis form. AzoTAB is found to induce triggered and localized protein unfolding, measured directly in solution with small-angle neutron scattering (SANS) experiments, and to further influence the biological function and dynamics of proteins. For example, the relationship betweenphotoreversible changes in secondary and tertiary structure of BSA, found to exist as one of three discrete forms depending on the azoTAB concentration, has been examined. Similarly, photo-control of the form-function relationship of lysozyme has been studied. With azoTAB in the trans form under visible light a partially unfolded intermediate conformation of lysozyme with an exposed active site was found, while lysozyme was observed to refolded to a native-like structure upon UV illumination. In addition, enhanced dynamics within the partially-unfolded form of lysozyme were observed with neutron spin echo (NSE) measurements and thought to contribute to a nearly 8-fold enhancement in the enzyme activity compared to the native state. Combined, these results provide insight into a unique light-based method of controlling the complete structure-dynamics-function relationship of proteins.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Control the function, dynamics, aggregation of proteins with light illumination

PDF

Photocontrol of protein conformation through the use of photoresponsive surfactants, investigated by small angle neutron scattering

PDF

Enhancement of biofuel enzyme activity and kinetics with azoTAB surfactants

PDF

Photocontrol of protein conformation and enzyme activity in the presence of light-responsive surfactants

PDF

A molecular dynamics study of interactions between the enzyme lysozyme and the photo-responsive surfactant azobenzene trimethylammonium bromide (azoTAB)

PDF

Controlling membrane protein folding with light illumination and catanionic surfactant systems

PDF

Molecular dynamics simulation study of initial protein unfolding induced by the photo-responsive surfactants, azoTAB

PDF

The interaction of photo-responsive surfactants with biological macromolecules

PDF

Controlling membrane protein folding using photoresponsive surfactant

PDF

Molecular dynamics studies of protein aggregation in unbounded and confined media

PDF

Photocontrol of DNA compaction using light-responsive surfactants

Asset Metadata

Creator Wang, Shao-Chun (author)

Core Title Controlling the form-dynamics-function relationship of proteins with light illumination

School Andrew and Erna Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Chemical Engineering

Publication Date 04/18/2008

Defense Date 03/26/2008

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag azobenzene,FTIR,light control,neutron scattering,OAI-PMH Harvest,photoresponsive surfactant,protein conformation

Language English

Advisor Lee, C. Ted, Jr. (committee chair), Shing, Katherine S. (committee member), Stephens, Philip J. (committee member)

Creator Email shaochuw@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m1132

Unique identifier UC1445123

Identifier etd-Wang-20080418 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-57832 (legacy record id),usctheses-m1132 (legacy record id)

Legacy Identifier etd-Wang-20080418.pdf

Dmrecord 57832

Document Type Dissertation

Rights Wang, Shao-Chun

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu