Ligand-Binding Cooperativity Effects in Polymer-Protein Conjugation

Subscriber access provided by UNIV OF NEW ENGLAND ARMIDALE

Article

Ligand-Binding Cooperativity Effects in Polymer-Protein Conjugation Jörg Reichenwallner, Anja Thomas, Tobias Steinbach, Jana Eisermann, Christian E. H. Schmelzer, Frederik Wurm, and Dariush Hinderberger Biomacromolecules, Just Accepted Manuscript • DOI: 10.1021/acs.biomac.9b00016 • Publication Date (Web): 10 Jan 2019 Downloaded from http://pubs.acs.org on January 13, 2019

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

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

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

Biomacromolecules

Ligand-Binding Cooperativity Effects in PolymerProtein Conjugation Jörg Reichenwallner,€ Anja Thomas,§ Tobias Steinbach,$,§ Jana Eisermann,€ Christian E. H. Schmelzer,&,# Frederik Wurm,$ and Dariush Hinderberger€,* €Institute

of Chemistry, Martin Luther University Halle-Wittenberg,

Von-Danckelmann-Platz 4, 06120 Halle (Saale), Germany. $Max

Planck Institute for Polymer Research, Ackermannweg 10, 55128 Mainz, Germany.

§Institute

of Organic Chemistry, Johannes Gutenberg-University, Duesbergweg 10-14, 55128 Mainz, Germany.

&Fraunhofer

Institute for Microstructure of Materials and Systems (IMWS),

Walter-Hülse-Strasse 1, 06120 Halle (Saale), Germany. #Institute

of Pharmacy, Martin Luther University Halle-Wittenberg,

Wolfgang-Langenbeck-Strasse 4, 06120 Halle (Saale), Germany.

KEYWORDS: albumin nanoparticles; protein-polymer conjugation; ligand binding; cooperativity; posttranslational modification; ESR/EPR spectroscopy

ACS Paragon Plus Environment

1

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

Page 2 of 52

ABSTRACT:

We present an electron paramagnetic resonance (EPR) spectroscopic characterization of structural and dynamic effects that stem from posttranslational modifications of bovine serum albumin (BSA), an established model system for polymer-protein conjugation. Beyond the typical drug delivery and biocompatibility aspect of such systems, we illustrate the causes that alter internal dynamics and therefore functionality in terms of ligand-binding to the BSA protein core. Uptake of the paramagnetic fatty acid derivative 16-doxyl stearic acid by several BSAbased squaric acid macroinitiators and polymer-protein conjugates was studied by EPR spectroscopy, aided by dynamic light scattering (DLS) and Zeta potential measurements. The conjugates were grafted from oligo(ethyleneglycol) methyl ether methacrylate (OEGMA), forming an overall core-shell-like structure. It is found that ligand-binding and associated parameters such as binding affinity, cooperativity and the number of binding sites of BSA change drastically with the extent of surface modification. In the course of processing BSA, the ligands also change their preference for individual binding sites, as observed from a comparative view of their spatial alignments in double electron electron resonance (DEER) experiments. The protein-attached polymers constitute a diffusion barrier that significantly hamper ligand uptake. Moreover, Zeta potentials (ζ) decrease linearly with the degree of surface modification in protein macroinitiators and an effective dielectric constant can be estimated for the polymer layer in the conjugates. All this suggests that ligand uptake characteristics in BSA can be fine-tuned by the extent and nature of such posttranslational modifications (PTMs). We show that EPR spectroscopy is suitable for quantifying these subtle PTM-based functional effects from selfassembly of substrate and ligand.


2

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

Biomacromolecules

INTRODUCTION Posttranslational modifications (PTMs) in proteins are natural, physiological and important biochemical events that include a vast variety of chemical and physical processes.1 After successful translation of a protein, these processes mainly comprise cleavage of a signal peptide, formation of disulfide bridges2 and covalent and non-covalent attachment of small organic and inorganic compounds.3,4 Furthermore, PTMs may define the onset of the duty cycle of a protein, e.g. by formation of quaternary complexes,2 or its imminent termination by ubiquitination.5 Thus, each induced change in the protein’s structural integrity and purity may have fundamental influence on its inherent functionality. This is best exemplified by the histone code hypothesis in epigenetics where combined or sequential modification processes on proteins are thought to lead to distinct downstream events.6 On a more general level, PTMs are often more abundant in disordered proteins (IDPs) or disordered regions in proteins such as albumins.4 Artificial PTMs like the covalent attachment of inert synthetic polymers in bioconjugation procedures constitute efficient methods to protect proteins from metabolic degradation, epitope accessibility, plasma protein binding and may furthermore lead to an improvement of pharmacokinetic properties compared to unmodified proteins.7–9 In this context, it is generally expected that e.g. polyethyleneglycol (PEG)-based polymer-protein conjugates10,11 and dendritic polyglycerol (PG)-based polymer conjugates12,13 exhibit high biocompatibilities. A recent publication of Panganiban et al.14 highlights progress in polymer-protein conjugation with surface-bound random heteropolymers aiming at protein solubilization even in non-native environments As major transport proteins in blood plasma, albumins are subjected to a broad variety of subtle physiological PTMs as e.g. glycosylation,15,16 acetylation,17,18 non-covalent uptake of


3


Page 4 of 52

cofactors, ligands and ions,3,4 modification of the redox-active Cys34 residue19–22 and the highly reactive lysine residues.23,24 While the primary structure of albumin contains as much as 59 lysine residues,25 altogether 25 – 35 of them can be targeted for PTM-related reactions in vitro.10,26,27 Through the versatility of its PTMs, albumin has become a model protein for many applications ranging from the construction of such polymer-protein conjugates in materials science, nano-medicine and polymer chemistry8 to e.g. initial spin-labeling studies in EPR spectroscopy.28–30 Recent progress in controlled radical polymerization techniques including atom transfer radical polymerization (ATRP),31 has triggered increased interest in the development of protein macroinitiators that provide access to polymer-protein conjugates in a “grafting-from” approach.32–39 Therefore, we chose BSA as a widely used model substrate in such polymerprotein conjugations. To realize this grafting from approach, BSA-macroinitiators with lysinebound, squaric acid (SA)-based ATRP initiating sites are here used for grafting oligo(ethyleneglycol) methyl ether methacrylate (OEGMA) monomers from the functionalized protein surface. These macroinitiators and graft copolymers were prepared as described previously40 by selective squaric acid coupling to surface lysines27 to build up suitable coreshell-like structures with a versatile protein core and a hydrophilic and biocompatible shell.11 In the course of introducing such artificial PTMs, a critical issue is the maintenance of the functional structure of the protein, while its function is generally assumed to be positively or negatively manipulated in its native bioactivity.41,42 The causes for functional alteration of proteins by PTMs are of an exceptionally intricate nature and may be, if at all, only detected with a combination of a broad range of methods from biophysical chemistry. At this strategic point, we pitch in and aim to corroborate the intrinsic potential of EPR spectroscopy to unravel and


4

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

Biomacromolecules

quantify such complex issues in a purposeful way. This is done to detect how functionality of the BSA core may change with the degree of modification beyond the aspect of general stability and solubility of the resulting polymer-protein conjugates.43 Since the early 1940s it is well known that albumins bind fatty acids (FAs) in their crystalline,44 as well as in their native solution form.45 As BSA is intrinsically diamagnetic and thus EPR-silent, the commercial availability of appropriately spin-labeled fatty acids (SLFAs),46–48 led to a vast amount of EPR spectroscopic studies about this paramagnetic, selfassembled, i.e. spin probed albumin system.49–60 Here, we proceed with our established albuminbased EPR spectroscopic research platform,61 that utilizes continuous wave (CW) EPR experiments in combination with DEER spectroscopy.62,63 This enables us to further approach and understand the functional solution structure of albumins on a functional nanoscopic scale. The rather abstract EPR-based picture of structural plasticity of albumin surfaces64,65 will be considered to be of major interest in this study. Thus, the 16-DSA spin probe was chosen as an appropriate EPR-active reporter due to its intrinsic propensity for entering albumins’ hydrophobic ligand binding pockets.51,56,66 Furthermore, 16-DSA bears its dynamically sensitive nitroxide group (doxyl) close to its hydrophobic tail end (C16-position). The protein’s dynamic surface structure is thus monitored,56,67 where the applied chemical alterations (PTMs) take place. With this EPR spectroscopic approach, posttranslationally modified albumins have been recently investigated, revealing that e.g. cationized albumin surfaces enhance68 and dendronized albumin surfaces hamper FA uptake.69 Based on these studies, we refine our analytic strategies towards a more quantitative level, regarding effects on ligand binding affinity (KA) and capacity (NT) and correlate these effects with the dynamic solution nanostructure as obtained from dipolar interactions of bound paramagnetic ligands in DEER. In principle, this can be understood as a


5


Page 6 of 52

type of motion capture snapshot from the dynamic ensemble of PTM constructs in terms of protein-bound paramagnetic centers, i.e. a superposition of individual nitroxide positions. To complement our EPR spectroscopic findings, the hydrodynamic and electrophoretic properties of the protein-based core-shell structures are characterized as well by combining results from DLS and LASER Doppler velocimetry (LDV) with an electrophoretic light scattering (ELS) setup.


6

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

Biomacromolecules

EXPERIMENTAL SECTION Materials. Lyophilized powders of native BSA (BSA-N, Sigma-Aldrich), 16-DSA (SigmaAldrich) and glycerol (87 wt% in water, ACROS Organics) were used without further purification. DPBS buffer at pH 7.4 was thoroughly prepared according to the original protocol.70 The preparations of BSA macroinitiators (BSA-Ix, x = 9, 10 and 14), BSA-gP(OEGMA)n polymer-protein conjugates (BSA-Pn) are schematically described below. Synthesis of BSA-Ix and BSA-Pn. The BSA-ATRP macroinitiators (BSA-Ix) were prepared as described previously.40,71 As an example, the preparation of BSA-I9 was conducted by dropwise addition of 10 equivalents of 2-(2-ethoxy-3,4-dioxocyclobut-1-enylamino) ethyl 2bromo-2-methyl propanoate (a squaric acid (SA) derivative, 5.2 mg, 15.4 μmol) dissolved in 0.708 ml DMSO to a stirred solution of BSA (102.4 mg, 1.54 μmol, 1 eq) in borate buffer (2.5 ml, 10 mM, pH 9.1). The suspension was vigorously stirred at room temperature for 48 hours. Afterwards, the homogeneous suspension was centrifuged (5 min, 3500 rpm) to remove insoluble residuals and the clear solution was dialyzed against deionized water (MWCO = 15.000, or 25.000 kDa). The solution was centrifuged again for 30 min at 3500 rpm and lyophilized to obtain the pure macroinitiator powder. The preparation of BSA-I14 was conducted by addition of 20 eq of squaric acid and was treated in the same way. Therefore, the yield can be considered as almost quantitative. Afterwards, the BSA-I10 macroinitiator was used for atom transfer radical polymerization (ATRP) experiments with 1000 initiator sites (x = 10), or 100 BSA equivalents of OEGMA (MWOEG = 475 Da). The catalyst solution was prepared separately from CuCl,2,2´-bipyridine and ascorbic acid in a degassed 0.1 M phosphate buffer (pH 6.5) – methanol (6 : 4) mixture.40 After addition of the dark brown catalyst solution (40 BSA eq) to the reaction mixture, the polymerization was allowed to proceed for 30 minutes. The reaction was


7


Page 8 of 52

quenched by addition of water and consequent air exposure. The reaction mixture was then dialyzed against water (MWCO = 50.000 kDa dialysis tubing) for 24 hours and was lyophilized to obtain the pure BSA-g-P(OEGMA)n graft copolymer powders (BSA-Pn). MALDI-TOF Mass Spectrometry. The molecular weights of the corresponding macroinitiators (BSA-Ix, BSA-N for x = 0) were determined with MALDI-TOF mass spectrometry (see Table 1 and Figure S1). Prior to the experiments all stock solutions were diluted with ultrapure water (MilliQ) to final concentrations of cB,Ix = 2 mg/ml. MALDI-MS experiments were carried out using a delayed extraction TOF mass spectrometer Voyager-DE PRO (Sciex) equipped with a pulsed nitrogen laser (λ = 337 nm). Sinapinic acid was used as matrix solution and mixed with the sample solution 10:1 (v/v) and then dried in a stream of air. Measurements were performed operating in the positive ion linear mode at a total acceleration voltage of 25 kV, grid voltage set to 92%, 0.15% guide wire voltage and an extraction delay of 700 ns. A low mass gate was set to m/z = 10.000 to prevent detector saturation from low mass compounds. The instrument was externally calibrated using BSA and calibration mixture 3 of the Sequazyme Peptide Standards Kit (Sciex). All mass spectra were smoothed using a 15-point Gaussian smoothing routine prior to subsequent peak picking. The calculation of the number of squaric acid residues was conceived by the relation: x = (MWIx – MWN )/MWSA assuming that each ATRP initiator (x) increases the molecular weight by MWSA = 288.11 Da. SDS PAGE. The successful “grafting from” reaction of OEGMA475 from BSA-I10macroinitiators was observed with a SDS PAGE. The gels were run on a Biorad system containing 8% bisacrylamide as a crosslinker and silver staining (silver nitrate) was used afterwards according to the scheme given by Chevallet et al.72 for visualizing the bands (see Figure S2). As graphical references, BSA-N and BSA-I10 have been added to highlight the


8

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

Biomacromolecules

weight increase after conjugation. The approximate degree of polymerization (DP = n) was calculated according to the relation: n = (MWPn – MWIx ) / x·MWOEG. For convenience, x = 10.4, as all polymer-protein conjugates were grafted from the BSA-I10 macroinitiators. Therefore, the number of added OEGMA molecules to a single BSA is x∙n. All such derived molecular weights (BSA-P11 and BSA-P18) and the respective yield is summarized in Table 1. The signal of unreacted BSA, or BSA macroinitiator completely disappeared in the BSA-g-P(OEGMA)n samples, thereby revealing high initiation efficiency. In both cases (lane 5 and 6 in Figure S2), the bands of the biohybrids are very broad. The molecular weights given in Table 1 are thus only seen as averaged values. Higher molecular weight bands in lane 3 and lane 4 can be attributed to dimerization of the BSA samples. Sample Preparation for EPR Spectroscopy. Aqueous stock solutions of 1 mM BSA-N, 0.5 mM BSA-Ix macroinitiators, 0.929 mM BSA-P11 and 0.426 mM BSA-P18 polymer-protein conjugates were prepared in 0.137 M DPBS buffer pH 7.4, depending on the accessible amount of lyophilized powders (all powders: mk  2.0 mg). All final BSA-N, modified BSA-Ix and BSA-Pn protein concentrations were set in the range of 0.06 – 0.32 mM for CW EPR and 0.15 – 0.40 mM for DEER experiments with sample volumes of V = 0.06 – 0.20 ml. Upon addition of appropriate amounts of 8 – 26 mM stock solutions of 16-DSA spin probe dissolved in 0.1 M KOH, nominal equivalent concentrations of 1:1 to a maximum of 1:7 (BSA:16-DSA) were adjusted for CW EPR samples (c16-DSA,CW = 0.10 – 0.74 mM). For DEER samples the loading ratio was set to 1:2 (c16-DSA,DEER = 0.30 – 0.40 mM) apart from the 1:1 reference sample of 0.4 mM BSA that was prepared for gaining sufficient SNR. The nominal equivalent concentrations of 16-DSA were used for Scatchard plot construction and were optimized according to the added volumes from the sample preparation protocols (see Figure S6). At these


9


Page 10 of 52

low 16-DSA concentrations micelle formation that hampers spectral evaluation can be excluded and overall more than 80% of the ligands are bound to the BSA cores of all samples (ϕb > 0.8). Each sample was prepared individually with a titration volume of 10 – 20% for adjusting pH values with a predefined set of acidic and basic 0.12 M DPBS buffers in the range of 0.5 < pH < 13.5. All samples were adjusted to pH 7.39  0.03 with a well-calibrated pH microelectrode (Mettler Toledo InLab®Micro pH 0 – 14, in combination with the EL20 Mettler Toledo pH meter). The final 0.1 M DPBS protein solutions were equipped with 16-DSA and 20% v/v glycerol to prevent crystallization upon freezing for potential DEER experiments. For CW EPR measurements, about 15 µl of sample were filled into a quarz capillary (BLAUBRAND® IntraMARK) with ca. 1 mm outer diameter. About 80 μl of the final solutions were filled into 3 mm (outer diameter) quartz tubes (Heraeus Quarzglas) and shock-frozen in liquid-nitrogen-cooled 2-methylbutane for subsequent DEER measurements. EPR Spectroscopy. CW EPR: The Miniscope benchtop spectrometers MS200 and MS400 (Magnettech GmbH) were used for X-band CW EPR measurements at microwave frequencies of about 9.4 GHz that were recorded with appropriate frequency counters (RACAL DANA, model 2101, or Magnettech FC400). All experiments were performed at 25°C using microwave powers of PMW = 3.16 – 10.00 mW, modulation amplitudes of 0.1 mT and sweep widths of 12 – 15 mT. DEER:

In

order

to

obtain

nanoscale

distance

information

from

16-DSA-probed,

posttranslationally modified BSA molecules the 4-pulse DEER sequence:62,63 (/2)obs–τ1–()obs,1–(td+t0+Nt∙Δt)–()pump–(t´–Nt∙Δt+td)–()obs,2–τ2–echo was used to obtain dipolar time evolution data. All 4-pulse DEER experiments were also performed at X-band frequencies of 9.1 – 9.4 GHz using a BRUKER Elexsys E580 spectrometer


10

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

Biomacromolecules

equipped with a BRUKER Flexline splitring resonator ER4118X–MS3. The temperature was set to 50 K for all experiments by cooling with a closed cycle cryostat (ARS AF204, customized for pulse EPR) and the resonator was overcoupled to Q < 300. The pump frequency pump was set to the maximum of each field swept electron spin echo (ESE)-detected spectrum. The observer frequency obs was set to pump + Δ with Δ being in the range of 65 MHz and therefore coinciding with the low field local maximum of the nitroxide ESE spectrum. The observer pulse lengths for each DEER experiment were set to 32 ns for both /2– and –pulses and the pump pulse length was 12 ns. Additionally, a 2-step phase cycle () was applied to the first /2 pulse of the observer frequency for cancelling out receiver offsets and unwanted echoes. Proton modulation was averaged by addition of eight time traces of variable τ1 starting with τ1,1 = 200 ns, incrementing by Δτ1 = 8 ns and ending up at τ1,8 = 256 ns. For all samples the pump pulse position td + t0 after the first observer -pulse deadtime td was typically incremented for Nt timesteps of Δt = 8 ns in the range t0 + t´ = τ1 + τ2 – 2td, whereas τ1 and τ2 were kept constant. DEER time traces were recorded for 12 – 48 hours giving rise to reliable distance information in between about 1.6 and 5.0 nm as the maximum accessible dipolar evolution time was about tmax  2 µs throughout.73 Evaluation of Data from EPR Spectroscopy. Explicit spectral simulations of CW EPR data from 16-DSA-probed BSA-N samples (see Figure S4A) have been exclusively evaluated in MATLAB 2008b (v.7.7) utilizing the MATLAB-based Easyspin v5.2.11 software package.74 This program comprises a toolkit that implements the application of slow tumbling nitroxide theory75–77 for specific sample requirements in feasible and straightforward procedures. All MATLAB codes were optimized for micelle-free 3-component nitroxide spectra comprising one (b) or two immobilized components b1 and b2 and a free component f as it is nowadays routinely


11

Page 12 of 52


conducted.59,78 Best fits were obtained with aiso values of about 15.3 G for components bi and 15.8 G for component f, as it was suggested earlier for 16-DSA spin probes interacting with BSA.54 Individual subspectra were double integrated to extract the spectral fractions i of corresponding dynamic regimes i as described in ESI S4. The errors for individual spectral fractions bi are estimated as Δi = 3% and the accuracy for the other parameters is Δaiso,bi = 1.00 MHz and Δτc = 8%. For an exemplary set of simulation parameters and the general simulation approach79 the reader is additionally referred to Table S2. These rather laborious spectral simulations can be circumvented by a corresponding readout scheme that is realized by introducing a correlation constant kL = 0.02486. The strategy to extract kL is presented in Figure S4C and Figure S5 (see ESI S5). The Scatchard plots were constructed as described in the main text with all fit parameters for linear regressions given in Table S3. Furthermore, an explanation and short description of the applied cooperativity test80,81 is given in the main text. Raw time domain DEER data were processed with the MATLAB-based program package DeerAnalysis2013 utilizing the Tikhonov regularization procedure.82 The background dimensionalities were set to D = 3.71  0.01 that emerge due to excluded volume effects of the albumin (BSA) molecules.56,67,83 No differences in background dimensionalities in between native (BSA-N) and posttranslationally modified BSA samples (BSA-Ix and BSA-Pn) have been noticed. Additionally, the resulting distance distributions P(r) were validated to assert the convincing significance of their individual shapes (Figure S7). The theoretical distance distribution (CS-1:7 in Figure 6) was constructed as described in Junk et al.56 Dynamic Light Scattering (DLS). All DLS data from BSA, BSA-Ix and BSA-Pn samples were obtained with an ALV-NIBS high performance particle sizer (HPPS) equipped with an ALV-5000/EPP Multiple Tau Digital Correlator (ALV-Laser Vertriebsgesellschaft m. b. H.).


12

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

Biomacromolecules

This device facilitates HeNe-LASER irradiation with a typical wavelength of λ = 632.8 nm and 3 mW output power with an automatic attenuator for optimum count rates recorded in a backscattering detection angle of 173° relative to the incident monochromatic light. The sample cell temperature was kept constant at 25°C with a Peltier temperature control unit. All samples have been prepared from the aforementioned stock solutions giving final BSA concentrations of 4.2 – 11.4 µM at pH 7.39  0.03. The final sample volumes of V = 300 – 430 µl were measured in 1.5 ml PMMA semi-micro cuvettes (BRAND). Data were extracted from the intensity correlation functions by a g2(t)-DLS exponential and a mass weighted regularized fit in the ALVNIBS software v.3.0 utilizing the CONTIN algorithm.84 Each sample was measured at least four times at constant temperature for 60 s and was averaged at least over four individual values as suggested by Bhattacharjee.85 The mean values ak = RH,k of the most prominent particle size peaks and their statistical fluctuations are given as the standard deviation (Table 1). The corresponding diffusion coefficients were calculated according to the Stokes-Einstein relation Dk = kBT∙(6∙RH,k)–1 from individual values of ak.86 Zeta Potential Measurements. Zeta potentials (ζk) were obtained from electrophoretic mobility measurements at 25°C utilizing the LASER Doppler velocimetry (LDV) technique in an electrophoretic light scattering (ELS) setup (see Table 2 and Table S1). Therefore, a constant voltage (U = 87 V) was applied between the electrodes of an Omega cuvette in a LitesizerTM 500 device (Anton Paar GmbH). This Omega cuvette prevents electric field gradients at the measurement position and allows for precise and reproducible experiments. All data were analyzed with the novel continuously monitored phase-analysis light scattering technology (cmPALS),87 as implemented in the provided software package KalliopeTM 1.8.0. All samples were prepared at identical concentrations (cB = 1.5 µM, i.e. ranging from 0.10 – 0.25 g/l) in 0.05


13


Page 14 of 52

M DPBS buffer at pH 7.51  0.04 with an ionic strength of I = 0.062 M. Consistently, the electrical double layer (DL) thickness is λD = κ–1 = 1.222 nm and κa > 2.6 for all samples (see Table S1). All given Zeta potential values were averaged over 5 – 7 individual experimental values. The measured conductivities of all solutions were in the range of κel = 6.7  0.3 mS cm–1. Furthermore, the dielectric constant εr = 78.4,88 refractive index nH20 = 1.332 (for λ = 632.8 nm)89 and viscosity  = 0.89 mPa∙s90 were assumed to be constant for all solutions (also in DLS experiments). Calculations of corresponding Henry functions FSF(κa) were conducted according to Swan and Furst,91 however, in all experiments reproducible Zeta potentials were only obtained for F(κa) = 1.09 as calculated for BSA-N. All experimental parameters, as µe,k, κel,k and calculated values of κak, FSF(κak), Nc,k, Zk and εeff,Pn are explained, derived and summarized in the ESI S3.

RESULTS AND DISCUSSION Characterization of Modified BSA Samples. Three different squaric acid-based BSA-Ix macroinitiators (x is the number of modifications per BSA) were prepared according to a procedure described in Wurm et al.27 and serve as reference standards to monitor influences on protein functionality depending on the degree of surface modification. These ATRP initiating sites were covalently attached to the lysine residues of the protein. Thereupon, two BSA-gP(OEGMA)n graft copolymers (BSA-Pn) of different sizes were synthesized by ATRP initiation of the BSA-I10 macroinitiator.40 The chemical composition and schematic representations of all investigated structures are presented in Figure 1. Principally, the numbers (x = 9, 10 and 14) of ATRP sites in BSA-Ix macroinitiators are determined from the respective weight increase measured in MALDI-TOF spectrometry


14

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

Biomacromolecules

compared to the native precursor BSA-N (MWN = 66.431 kDa, Table 1 and Figure S1). Each attached squaric acid (SA) residue increases the molar mass of the macroinitiator by MWSA = 288.11 Da. Molecular weights of the BSA conjugates (BSA-Pn) were just determined by SDS PAGE. A significant increase in molecular weight towards about 125 – 160 kDa is observed for both polymer-protein conjugates as compared to pure BSA, confirming the successful grafting reaction with high initiation efficiency (Figure S2). This allows for an approximate estimation of the degrees of polymerization (DP = n) from the molecular weight of OEGMA475 monomers (MWOEG = 475 Da) yielding about n = 11 (BSA-P11) and n = 18 (BSA-P18).

Figure 1. Chemical structures of macroinitiators BSA-Ix and polymer-protein conjugates BSA-Pn. (A) The native protein BSA-N (blue sphere, PDB ID: 3v03)92 is depicted with the chemical structures of a SA-based ATRP initiator sites (red squares) from where OEGMA475 monomers (blue) can be grafted from to yield the polymer-protein conjugates BSA-Pn. The EPR-active 16-DSA spin probe is depicted in orange with a schematic affinity (KA) towards the protein core. (B) The native precursor apoprotein BSA-N (x = 0) and available macroinitiators BSA-Ix are schematically presented with the colored symbols shown in (A).


15


Page 16 of 52

The exact topological location of the PTMs was not investigated and is assumed to occur in a quantitative, statistical manner. DLS measurements suggest a growth in hydrodynamic radii ak with increased surface modification x of BSA. While measurements of the native apoprotein yield values that are in good agreement with those of previously published studies,93,94 the slight increase in molecular weights for all BSA-Ix macroinitiators already leads to significant changes in ak. Note, that the analogous rate of diffusion Dk is strongly reduced in the polymer-protein conjugates. Therefore, it can be assumed that surface modifications on BSA lead to strong alterations of solvent shells and entanglements, especially in polymer-protein conjugates.10 This is further underlined by the strong decrease in molecular densities k upon surface modification, exhibiting a significant fraction of water-pervaded volume, especially in case of the BSA-Pn nanostructures as shown in Table 1.

Table 1. Characterization of modified BSA samples k k

MWk [kDa]

x

n

yield [wt%]

aka [nm]

Dka [107 cm2 s–1]

kb [kg m–3]

N

66.431c,d

0

–

–

3.24  0.24

7.56  0.56

774  172

I9

68.976d

8.8

–

–

3.85  0.30

6.34  0.49

479  112

I10

69.415d

10.4

–

–

3.95  0.28

6.21  0.44

447  95

I14

70.357d

13.6

–

–

4.38  0.33

5.60  0.42

332  75

P11

125.000e

10.4

11.3

55

7.78  0.37

3.15  0.15

105  15

P18

160.000e

10.4

18.4

95

7.75  0.44

3.17  0.18

136  23

aHydrodynamic

radii ak and diffusion coefficients Dk were obtained from DLS experiments with the relation Dk = kBT∙(6ak)–1 that interconverts both parameters.86 bk = the molecular density was calculated from the relation k = 3∙MWk∙(4πak3)–1. cA reference value MWBSA = 66.463 kDa can be found in Chafer-Pericas et al.95 dMWIx of BSA-Ix from MALDI-TOF mass spectra in Figure S1. eMWPn of BSA-Pn from the SDS PAGE in Figure S2.


16

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

Biomacromolecules

The Zeta potential of BSA-N is found to be negative at ζN = –18.6 mV (Table 2) as it is commonly observed at physiological pH and ionic strengths about I ≈ 0.1 M.96,97 The associated electrophoretic mobility µe,k (with k denoting the respective species) is defined as:97 µe ,k 

2 0 r k F ( ak ) 3

,

(1)

where ε0 represents the vacuum permittivity, F(κak) = 1.09 is Henry’s function, εr = 78.4 is the relative permittivity88 and  = 0.89 mPa∙s is the dynamic viscosity for water at 25°C.90 For BSAN, µe,N = –1.05∙10–8 m2 V–1 s–1 (Table S1) and corresponds well with previously reported values.94,98

Table 2. Results from Zeta potential measurements k

ζka [mV]

Zkb

Q,kc [10–7 C m–3]

εeff,Pnd

N

–18.6  1.5

–10.5  2.2

–1.18  0.51

–

I9

–27.4  1.5

–18.6  3.9

–1.25  0.55

–

I10

–28.7  1.1

–20.1  3.7

–1.25  0.50

–

I14

–30.9  0.9

–24.2  4.6

–1.10  0.46

–

P11

–12.4  1.6

–18.3  5.8

–0.149  0.069

69.7  26.8

P18

–13.2  1.3

–19.4  5.9

–0.159  0.075

73.0  28.2

aζ k

= Zeta potential. bZk = Qk∙e–1 = Nc,k∙kQ–1 number of net charges calculated from eqn 5 (see also Table S1). cQ,k is the charge density as calculated from the relation Q,k = Qk∙Vk–1 = 3Zke∙(4πak3)–1. dεeff,Pn = effective dielectric constant of the polymer layer as estimated from eqn 6 (for details, see also ESI S3).

For BSA-Ix, ζk decreases with the degree of modification x, while BSA-Pn has values similar to those of BSA-N (Table 2). Generally, BSA-Ix suspensions gain stability (ca. ± 30 mV) by increasing their electrostatic repulsion upon further modification.99–101 This behavior primarily


17


Page 18 of 52

stems from covalent attachment of squaric acids to positively charged lysines resulting in a more negative net surface charge Zk. The number of uncompensated charges Nc,k can be determined from Zeta potentials ζk with a strategy that was suggested by Adamczyk et al.102 giving the relation: N c ,k 

6 ak  µe ,k e

,

(2)

where e is the elementary charge. As it was also seen by Jachimska et al.,94 for BSA-N we find a rather low uncompensated charge number Nc,N = –3.6  0.6 (Table S1). Unlike the results from titration experiments on BSA,103 the number of uncompensated charges is here quite low due to counter ion condensation104–106 of buffer ions to BSA.97,98 The stepwise lysine-blocking across the BSA-Ix samples reveals (Figure S3) that there are linear relationships between the degree of surface modification x and both the Zeta potential ζk and the number of uncompensated charges Nc,k:

 k (x) = k x   N

(3)

N c ,k (x)  kQ x  N c ,N

,

(4)

where kζ = –(0.938  0.052) mV SA–1, ζN = –(18.76 ± 0.46) mV, kQ = –(0.342  0.021) SA–1 and Nc,N = –(3.46 ± 0.21). Therefore, each SA-modified lysine residue lowers ζ by about 1 mV, while Nc,k decreases by a full charge ΔQ = –1∙e for each three consecutively blocked lysines in the observed range from 0 < x < 14. It can be estimated from this kQ value that about two-thirds ( = 1 – kQ = 0.658) of the BSA surface charges are screened by buffer ions (assuming one charge/lysine). Thus, the reduced net charge Zk = Qk∙e–1 can be directly obtained with the expression:


18

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

Biomacromolecules

Zk 

N c ,k kQ



4 ak  0 r k F ( ak ) ekQ

,

(5)

yielding ZN = –10.5  2.3, which is virtually identical with values for BSA at physiological pH seen from other experimental methods (–18 < ZN,exp < –8).93,103,107 However, the charge values calculated from amino acid sequences alone seem to be slightly higher (ZN,calc = –17)3 but serve as a good estimate. An explicit treatment of the charge calculations is given in the ESI S3 and the calculated values from eqn 5 are shown in Table 2. Moreover, ZPn corresponds quite well with the value for BSA-I10, so the “grafting from” approach obviously does not change the charge state of the protein core as observed from the macroinitiators. In direct comparison to BSA-I10 the Zeta potential of the polymer-protein conjugates ζPn amounts to only about 45% of this value. This is due to the presence of the P(OEGMA)n-shell that increases the separation of the charge-bearing macroinitiator surface (aI10 = 4.0 nm) from the hydrodynamic slipping plane (aPn = 7.8 nm) by dPn = 3.8 nm (Figure 2). The slipping plane alone separates the mobile bulk solvent from the colloid,108 or in this case the polymer-protein conjugate structure. The charge density is almost constant at Q,k = –(1.19  0.07)∙107 C m–3 for BSA-N and BSA-Ix, but experiences a significant drop of almost one order of magnitude for BSA-Pn (see Table 2, Q,Pn = –(1.54  0.08)∙106 C m–3). Generally, the potential V generated by a (BSAbased) charge density decreases with this spatial separation as d–1.109 A consistency check (see ESI S3) proves that the separation d = aPn – aI10 between BSA-I10 macroinitiator surface and slipping plane of BSA-Pn leads to a decrease in Zeta potential that can be approximated via the relation aI10ζI10 ≈ aPnζPn for ak > aN (see also eqn S18).


19


Page 20 of 52

Figure 2. Schematic representation of the Zeta potential from modified BSA samples in DPBS buffer. The BSA core (dark blue) with radius aN bears the charge Qk and is surrounded by an electrical double layer (DL, pale red) composed of different DPBS buffer ions Yz of varying valency (–z charge red, +z charge green spheres). For simplicity only the negative net charge (Zk) is shown here on the protein surface. The ATRP initiators (SA, red squares) of BSA-Ix interrupt the electrical DL and the Zeta potential ζIx is measured at aIx. The water-pervaded polymer layer (grey) of BSA-Pn is assumed to bear no charge at all but is permeable for solvent and ions with an average effective dielectric constant of εeff,Pn = 71.3. The value of ζPn at the slipping plane (SP) between polymer and bulk solvent (pale blue) is therefore much lower, as indicated by the red curve at the bottom.

Furthermore, the distance-dependent decrease in ζPn allows for an estimation of the dielectric constant εeff,Pn of the water-pervaded polymer layer, assuming an average equivalent number of uncompensated charges Z∙kQ = Nc = –6.58 for BSA-I10, BSA-P11 and BSA-P18 with the


20

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

Biomacromolecules

relation:  eff ,Pn 

Nc  e 4 0Δ n

  1 1      aPn F ( aPn ) aI10 F ( aI10 ) 

(6)

with values of εeff,P18 = 73.0 and εeff,P11 = 69.7 (Table 2), being only slightly lower than that of pure water (εr = 78.4). Here, the parameter Δζn = ζPn – ζI10 is the Zeta potential difference at individual slipping planes of samples k. However, the intrinsic errors Δεeff,Pn appear to be quite large at about 38%. A derivation of eqn 6 is also given in ESI S3 (see eqn S19) from an approach emphasizing several electrostatic considerations.109 The basis of this calculation is an uncompensated point charge that is located at the coordinate system origin of a spherical BSAI10 macroinitiator being surrounded by a dielectric shell of thickness d = aPn – aI10 (Figure 2). Analyses of the involved corresponding electrostatic free energies110 are yet considered to be beyond the scope of this work. This result shows that the polymer shell does not have a large effect on the dielectric properties of the medium that surrounds the BSA protein core. This again reveals that water widely pervades the polymer shell, which in turn serves as an attenuator for the electrostatic potential at the core. The next section summarizes how ligand-binding dynamics are affected by these PTM-induced physicochemical properties.

CW EPR Spectroscopy on Modified BSA. A further specific characterization of intrinsic dynamic changes upon BSA modification was obtained by adding 16-DSA to the solutions of the PTM BSA samples to study their fatty acid binding properties. CW EPR spectra of all 16-DSAloaded samples are shown in Figure 3 (at a 1:1 ratio of 16-DSA:BSA with an equivalent concentration of 0.13 mM). The emergence of clearly detectable bound (b) and freely tumbling (f) 16-DSA spin probes already allows for a qualitative estimate of the ligand-binding affinity KA,k. Primarily, the free component f increases with further protein modification, already


21


Page 22 of 52

indicating a decrease in KA,k. This decrease is most pronounced for the polymer-protein conjugates BSA-Pn.

Figure 3. CW EPR spectra of modified BSA samples loaded with the 16-DSA spin probe. Recorded CW EPR spectra are shown for a nominal 1:1 loading ratio of 16-DSA comprising all PTM BSA samples BSA-Ix and BSA-Pn together with the unmodified precursor BSA-N. Significant spectral features as the bound (b, grey) and freely tumbling species (f, green) are highlighted. The low-field peak height of the free spectral fraction (h+1,f) and the center-field peak height of the bound spectral fraction (h0,b) are shown to visualize the peak-picking strategy shown in eqs 7 and 8.

Furthermore, it is known that the rate of ligand association ka is proportional to the diffusion constant111,112 , i.e. ka  Dk in the respective receptor(R)-ligand(L) association reaction:113

ˆ ˆˆkaˆ† R+L ‡ ˆ ˆ RL k

,

d

(7)

whereas KA = ka/kd. This circumstance alone should lead to a decrease in KA,k of a factor of about DN/DP18 = KA,N/KA,P18 = 2.38 (see also Table 1).


22

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

Biomacromolecules

As these individual CW EPR spectra do not provide a complete view on the ligand-binding properties of the substrates we measured a concentration-dependent spectral series for each sample and constructed conventional Scatchard plots.114 Individual samples were successively loaded in the range from about one to a maximum of seven equivalents of 16-DSA to prevent micelle formation of the ligand. As it is sufficient to only extract the free (f) and bound (b) spectral fractions for Scatchard plot construction in EPR spectroscopy,52,53,115 explicit spectral simulations were only conducted on the reference sample BSA-N (see ESI S4 and Figure S4A) according to a strategy presented in previous work.79 However, due to the large data sets that were recorded, a simple strategy was devised from reference simulations that allows for quick extraction of the required spectral fractions by directly correlating peak height ratios to spectral fractions (see also Figure 3). It is found, that the ligand concentrations [L]i of free (i = f) and bound (i = b) 16-DSA can be directly obtained by the relations: [L] f  kL  [L]t ,k 

h1, f h0 ,b

  f  [L]t ,k

 h1, f [L]b  [L]t ,k 1  kL   h0 ,b 

(8)

   b  [L]t ,k  

,

(9)

where [L]t,k is the total ligand concentration in a sample, h+1,f is the peak height of the low-field line of the free fraction (f) and h0,b is the peak height of the center-field line of the bound spectral fraction (b) as depicted in Figure 3. Here, kL = 0.02486 is a numerical constant correlating the values of h+1,f and h0,b with corresponding peak heights from “dummy” spectral simulations. This simple method helps to avoid cumbersome and time-expensive analyses. An explicit description of this approach is given in ESI S5 that further substantiates this strategy.


23


Page 24 of 52

The relative deviations of values from explicit simulations on experimental spectra and from the peak-picking routine are between 2 and 9%. However, this strategy is for now restricted to only 16-DSA-probed BSA molecules in DPBS buffer at pH 7.4 equipped with 20% v/v glycerol in the absence of ligand aggregates or micelles. After extraction of the relative concentrations of the spectral fractions ([L]i = [L]t,k∙i), a Scatchard analysis is then conducted in a plot of v = [L]b/([L]f ∙cB) versus the number of bound ligand equivalents NL where cB is the protein (BSA) concentration in the sample. The resulting Scatchard plots from all six samples are shown in Figure 4 and all corresponding CW EPR spectra are given in Figure S4A and Figure S6. Primarily, the linear phases p in the Scatchard plots of samples k can be evaluated with following generalized expression:113  

[L]b,k [L] f ,k  cB ,k





[L]t ,k  [L] f ,k [L] f ,k  cB ,k



  K A ,k ,p  N L  N E ,k ,p   k ,p,N L0

,

(10)

where KA,k,p is the association constant, NE,k,p the number of equivalent binding sites and vk,p,NL→0 are the y-axis intercepts (of sample k and Scatchard phase p). The individual phases shown in Figure 4 are either denoted as a single phase spanning the whole data range (I), individual linear phases that span at least three data points (Ia, Ib, or Ic), or a linear phase that is usually identified as the weak binding part in an apparent biphasic Scatchard plot (II).116 A typical biphasic Scatchard plot is generally expected to have an exponentially decreasing curve shape due to the overlap of two classes of a specific number of independent strong (NS, KS) and weak (NW, KW) binding sites.53,80,116,117 Several strategies have been developed to extract Ni and Ki,116,118 but in terms of comparability their use is precluded in these BSA-based systems as the number of non-interacting binding


24

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

Biomacromolecules

systems is unknown, especially in BSA-Ix. Therefore, emphasis is placed on the qualitative change in the number of equivalent binding sites NE,k,p. From the pure phases I or II a total number of binding sites (NT,k,p) can be extracted deliberately, as the x-axis intercept of these phases in principle correspond to NT = NE,k,I = NE,k,II = NS + NW.117

Figure 4. Scatchard plots of native BSA, BSA-Ix and BSA-Pn interacting with 16-DSA. All Scatchard plots were constructed from CW EPR spectra shown in Figure S4A and Figure S6. Individual samples were loaded with 16-DSA in different molar ratios for testing the ligand

concentration-dependent

response

of

the

substrates

(A)

native

BSA-N

(cB,N = 0.15 mM), polymer-protein conjugates (B) BSA-P11, (C) BSA-P18 (cB,Pn = 0.06 – 0.32 mM) and the macroinitiators (D) BSA-I9, (E) BSA-I10 and (F) BSA-I14, each at cB,Ix = 0.1 mM. Each CW EPR spectrum is here reduced to a single data point (black) with appropriate error bars (grey). The dashed black lines are linear extrapolations of Scatchard phases I, Ia, Ib, Ic, or II for data point ranges q  3 (see Table 3 and Table S3).


25


Page 26 of 52

Furthermore, a classical Tanford-type cooperativity test was conducted.81 Cooperativity, as observed in proteins with multiple ligand-binding sites, is believed to originate from binding site heterogeneity and conformational adaptability,64 but may also be induced by modifications in local environments due the mere presence of ligands119 that mutually affect their binding properties. This effect is encountered when the number of occupied binding sites of a macromolecule is a non-linear function of the relative ligand-to-protein concentration.120 Cooperativity is therefore understood as an effective interaction between ligands in (often remote) binding sites that mutually affect their individual binding affinities by allosteric modulation of the substrate. Additionally, following the strategy of De Meyts and Roth80 one may differentiate between negative cooperative (–), positive cooperative (+), as well as locked, non-cooperative (N.C.) phases p. The physical basis of cooperativity is defined by an intrinsic affinity constant KA,int that is the limiting value of KA for NL → 0. The intrinsic change in standard Gibbs energy ΔG°A,int is then used to introduce the effect of binding site interaction with an arbitrary function (NL) for NL > 0: o ΔGAo  ΔGA,int  RT  Φ(N L )

.

(11)

The variable association constant is then given as:

K A  K A,int  e Φ(N L )

(12)

and (NL) = 0 for NL = 0. Here, ΔG°A and (NL) are average properties across all binding sites and protein configurations. Combining eqn 12 with the equilibrium condition that defines the degree of association:81 K A [L] f 

[L]b N T cB  [L]b

,

(13)


26

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

Biomacromolecules

leads to following expression: lnK A,int*  lnK A,int  Φ(N L )  [L]b  ln   N T cB  [L]b

   ln[L] f 

.

(14)

When the right-hand side of eqn 14 is then plotted versus NL, a so-called cooperativity plot is obtained. The result from this cooperativity test is shown for all samples (Figure 5). In case the observed slope is negative, cooperativity is also negative (–) and (NL) is an increasing function that lowers lnKA,int*. Positive cooperativity (+) coincides with (NL) being a decreasing function leading to a positive slope. Non-cooperative ligand binding (N.C.) is observed, when (NL) = 0 for all NL. Therefore, lnKA = lnKA,int and the slope in the cooperativity plot is locked at a constant value of lnKA,int*.80 For convenience, the cooperativity parameter Ck,p is simply defined as the sign of the slope of lnKA,int* (i.e. C = 0 (N.C.), C < 0 (–) and C > 0 (+)). Unfortunately, not all phases of the samples can be appropriately resolved, but an extensive set of results from these analyses are summarized in Table 3 together with corresponding Gibbs energy changes of association (ΔG°A,k,p). The fit parameters for calculations of KA,k,p and NE,k,p are given in Table S3. Note, that the association reaction always remains exergonic in nature (ΔG°A < 0) throughout all phases, although individual values may decisively decrease (e.g. in BSA-I10 and BSA-I14). Native BSA-N (Figure 4A) is the only sample that exhibits a linear Scatchard plot (I), while all PTM BSA derivatives show multi-modal curve shapes indicating the chemically (PTM) induced heterogeneity in individual binding site affinity values (KA,k,p). Evaluation of the linear Scatchard plot for BSA-N and 16-DSA is straightforward and yields NE,N = 6.0 equivalent, non-cooperative


27


Page 28 of 52

binding sites with a locked macroscopic association constant of KA,N = 8.4∙105 M–1. This reference number of binding sites was also used for the cooperativity test on all BSA, or BSAderivatives (NE,N = NT = 6).

Figure 5. Cooperativity test of 16-DSA-binding to BSA and modified BSA. (A) Cooperativity of 16-DSA binding to BSA-N is compared to macroinitiators BSA-I9 (grey), BSA-I10 (orange) and BSA-I14 (green). (B) Cooperativity of 16-DSA binding to BSA-N is compared to polymerprotein conjugates BSA-P11 (blue) and BSA-P18 (red). Individual regions of negative (–, C < 0) and positive (+, C > 0) cooperativities are separated by a colored dotted line for each sample indicating the number NL of bound ligands that lead to an inversion in cooperativity from (–) to (+). Specifically, BSA-N exhibits non-cooperative binding with a locked value of lnKA,int* (N.C.). A comparison with data obtained from explicit spectral simulations yields almost identical values (NE,N = 6.4, see Table 3) and is in reasonable correspondence with values for stearic acids that were reported in earlier studies.52,53,121,122 Interestingly, a biphasic Scatchard plot was reported in a study for unmodified BSA interacting with 16-DSA.52 This deviant behavior may


28

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

Biomacromolecules

originate from experimental differences like protein purity, treatment, temperature, buffer composition, sample viscosity, as well as in the applied experimental method. The polymerprotein conjugates BSA-Pn exhibit non-linear Scatchard plots (Figure 4B and Figure 4C), each with a steep initial decrease (Ia) indicating about 2.9 equivalent strong binding sites.

Table 3. Parameters from phase-specific (p) linear regressions in Scatchard plots of samples k k

p

NT,k,p

NE,k,p

KA,k,p [M–1]

ΔG°A,k,pa [kJ mol–1]

[L]t,kb [eq]

qc

Ck,pd

BSA-N

Ie

6.37  0.33

–

(7.50  0.23)∙105

–33.54  0.08

1.59 – 5.29

10

N.C.

I

6.05  0.33

–

(8.38  0.28)∙105

–33.81  0.08

1.59 – 5.29

10

N.C.

Ia

–

2.97  0.82

(7.58  1.43)∙105

–33.56  0.47

1.00 – 2.00

3

–

Ib

–

4.77  0.31

(2.52  0.10)∙105

–30.83  0.10

2.00 – 4.50

6

–

Ia

–

6.34  0.72

(1.51  0.13)∙105

–29.56  0.21

1.45 – 2.34

3

N.C.

Ib

–

4.45  0.82

(2.87  0.32)∙105

–31.15  0.28

2.34 – 4.15

5

(+)

II

9.02  1.31

–

(3.20  0.30)∙104

–25.71  0.23

4.15 – 6.38

6

+

Ia

–

3.80  0.47

(2.56  0.22)∙105

–30.87  0.21

1.00 – 2.50

6

(–)

Ib

–

3.48  0.01

(5.15  0.01)∙105

–32.60  0.01

2.80 – 3.41

3

(+)

Ic

–

7.00  1.29

(5.89  0.65)∙104

–27.23  0.27

4.61 – 5.80

5

(N.C.)

II

12.87  0.94

–

(1.51  0.08)∙104

–23.85  0.13

3.41 – 7.00

8

+

BSA-P11

Ia

–

2.92  0.25

(4.20  0.23)∙105

–32.10  0.14

1.08 – 2.69

4

–

BSA-P18

Ia

–

2.86  0.25

(2.90  0.14)∙105

–31.18  0.12

1.00 – 2.35

3

–

Ib

–

4.99  0.73

(6.34  0.58)∙104

–27.41  0.23

2.35 – 4.70

3

(+)

BSA-I9

BSA-I10

BSA-I14

aGibbs

energy changes of ligand association are calculated from association constants KA,k,p with ΔG°A,k,p = –RT∙lnKA,k,p (see also Table S3 for all obtained fit parameters). b[L]t,k [eq] = the range of total ligand concentration is given in added molar equivalents to BSA. cq = number of fitemployed data points (q  3). dCk,p = type of cooperativity for observed ligand-binding phases p from Figure 5. Cooperativities in parentheses only give tendencies, whereas N.C., – and + are clearly observed. eResult from explicitly simulated CW EPR spectra (see Figure S4).


29


Page 30 of 52

It can be anticipated that due to the higher degree n of OEGMA side-chain polymerization that may effectively shield the protein core from ligand penetration, the 16-DSA affinity is a bit lower for BSA-P18. The differences in strong KA values in phase Ia of BSA-Pn compared to phase I in BSA-N correspond astonishingly well with the observed differences in diffusion constants (KA,N,I/KA,Pn,Ia ≈ 2.4  0.5, see also Table 1). Nevertheless, a second linear region can be found for BSA-P18 (Ib), indicating the formation of almost exactly 5.0 equivalent binding sites with slight positive cooperativity, but with a much lower affinity. The slightly higher molecular density (Table 1) of BSA-P18 compared to BSA-P11 may be a significant feature that prevents 16-DSA entering the binding sites in the bulk protein. A more complicated behavior is found for the polymer-free macroinitiators BSA-Ix. Overall, each investigated macroinitiator shows a different Scatchard plot appearance with varying multiphasic characteristics. Phase Ia in BSA-I9 reveals 3.0 equivalent binding sites with a dissociation constant that is comparable to native BSA (Figure 4D, KA,I9,Ia = 7.6∙105 M–1). In phase Ib the number of equivalent binding sites is increased to 4.8 but only one third of the native 16-DSA affinity (KA,I9,Ib = 2.5∙105 M–1). The cooperativity test shows that the gain in equivalent binding sites seems to over-compensate this decrease in binding affinity, still leading to an overall negative cooperativity. Here, the region above NL > 4.2 is coined by positive cooperativity, yet, a clearly detectable slope that would be needed for an exact determination of NT is not observed in this high-loading region. Therefore, a total number of binding sites is unfortunately inaccessible for BSA-I9. Phases similar to Ib in BSA-I9 are observed in the Scatchard plots of BSA-I10 and BSA-I14 as a slightly positive cooperative upward bend for medium loading ratios (Figure 4E and Figure 4F). In principle, all macroinitiators BSA-Ix exhibit a transition from negative cooperativity at


30

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

Biomacromolecules

low NL to positive cooperativity at higher NL. However, upon increasing the degree of modification x, this ()-transition region in cooperativity is shifted to lower NL (Figure 5). While the curve for BSA-I9 changes to positive cooperativity at NL = 4.2, BSA-I14 passes this threshold already at NL = 3.2. This largely coincides with the decreasing number of equivalent binding sites NE in phase Ib with increasing x (Table 3). Simultaneously, the individual affinities of the remaining binding sites notably increase, so the loss in the number of equivalent binding sites may be compensated to some extent. Phase II in BSA-I10 and BSA-I14 gives the maximum number of 16-DSA binding sites of a macroinitiator, i.e. NT,I10,II = 9.0  1.3 and NT,I14,II = 12.9  0.9. This can be clearly stated without any knowledge about the individual numbers of weak and strong binding sites.117 A closer look at phase II in BSA-I14 additionally reveals an intriguing feature that is for now termed as phase Ic, interrupting the linearity in phase II in the range from 4.3 < NL < 5.3 bound molar equivalents of 16-DSA (or 4.6 – 5.8 total ligand equivalents added). Phase Ic is also slightly non-cooperative (Figure 5A, green), and is intermittent to the positive cooperative phase II, which shows only NE,I14,Ic = 7.0 equivalent binding sites. These findings imply that positive cooperativity in BSA-Ix leads to a hierarchized and staggered formation of new binding sites. The total number of binding sites NT,x may be a function of the loading status (NL) and the degree of modification x. However, these additional binding sites are only formed at the expense of their equivalent binding affinity at high NL. From the characterization results above it can be concluded that 16-DSA affinity decreases as the values of the Zeta potentials ζIx become more negative with increasing x. As 16-DSA is partially negatively charged (COO–) at physiological pH as it is used here, the Zeta potential at the slipping plane, or rather the surface charge Zk, has decisive influence on ligand-binding properties. This would also explain why BSA-Pn exhibits differently


31


Page 32 of 52

shaped Scatchard plots compared to BSA-Ix, as it was shown that the water-pervaded polymer shell decisively weakens the electric field emerging from the charged BSA sphere in the applied physical model (see ESI S3). Overall, ligand-binding remains strong with at least 80% of ligands bound for all investigated BSA samples in the observed loading range from 1 to 7 equivalents of 16-DSA. It can be concluded that the chemical modifications applied to BSA here allow for subtle functional adjustments of the fatty acid, or general ligand-binding affinities and capacities. This effect is supposed to be mainly regulated by sterical hindrance (d in BSA-Pn) and electrostatic repulsion (ζIx in BSA-Ix). DEER Experiments on Modified BSA. In a next step, the modified BSA samples are screened for potential differences in ligand distributions and alignments in the bulk protein matrix. This can be realized with DEER experiments as it was shown earlier for native HSA and BSA,56,57,67 as well as for cationized HSA constituting another example of a posttranslational functional modification of the protein surface.68 In Figure 6, results from DEER experiments are shown for BSA-N, BSA-Ix macroinitiators and BSA-Pn protein-polymer conjugates. Despite the similarities in time traces and dipolar evolution functions (Figure 6A and Figure 6B), the resulting distance distributions in Figure 6C reveal some interesting features and differences. Compared to BSA-N, all macroinitiators exhibit similar, broad distance distributions (P(r)) with a strong relative increase of a distance peak at 2.34  0.11 nm. For BSA-N, this peak is only similarly pronounced for fatty acid loadings above four equivalents as it was shown in an earlier study.67 This observation may already indicate a change in the order of binding site occupation in modified BSA-Ix molecules, although the surface has just a few modifications that increase its molecular weight by only about 2.5 – 3.9 kDa (see Table 1). Extending the comparison to the BSA-Pn conjugates, again a mutual similarity in P(r) is observable, however, with an even more


32

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

Biomacromolecules

prominent short distance peak that emerges at about 2.0 nm. Two general remarks have to be made in terms of comparing distance distributions of all compounds.

Figure 6. DEER experiments on modified BSA samples. These experiments were conducted on samples that contain two nominal equivalents of 16-DSA (1:2, for BSA-N also 1:1) Experimental results as (A) raw time domain data V(t)/V(0), (B) dipolar evolution functions F(t)/F(0) (black) with regularized fits (red) and (C) distance distributions P(r) are shown for all samples. Additionally, the red trace in (C) is the theoretical distribution (CS-1:7) obtained from the crystal structure of HSA co-crystallized with seven stearic acids (PDB ID: 1e7i)66 as given in Junk et al.56 As an aid to the eye, regions with corresponding peaks are color shaded with green (a), blue (b), orange (c) and purple (d).

Firstly, all distance distributions with 1:2 loading equivalents (k-1:2) exhibit the features at 2.22  0.18 nm (a), 3.24  0.06 nm (b), 3.74  0.06 nm (c) and 4.76  0.02 nm (d) that were also validated (Figure S7). It was already shown by Akdogan et al.,67 that the 16-DSA-derived distance distribution in BSA represents the distance distribution of HSA’s crystal structure quite nicely as obtained with seven co-crystallized stearic acids (PDB ID: 1e7i).56,66 Therefore, this


33


Page 34 of 52

HSA crystal structure is here again used as a distribution shape reference. Here, it is worth mentioning that P(r) from HSA’s crystal structure (CS-1:7, red, Figure 6C) is constructed and discussed under the assumption that all individual binding sites possess equal ligand-binding affinities (KA) and that all fatty acid binding sites are fully occupied (1:7).56,67 Unfortunately, a crystal structure of BSA co-crystallized with stearic acids is not available to date. Secondly, the obvious comparability of all P(r) shapes strongly suggests, that the macroinitiators BSA-Ix and the BSA-Pn polymer-protein conjugates force the ligand distribution and therefore the BSA core into a more crystal-like configuration (compare to CS-1:7). The same effect was observed in DEER experiments on cationized human serum albumin (HSA) selfassembled with 16-DSA.68 It is probable that differences in fatty acid distance distributions also emerge due to modified binding site selectivities in terms of local electrostatic repulsion (BSA-Ix) and sterical hindrance (BSA-Pn), or a combination of both effects as mentioned above. Finally, each set of fatty acid distance distributions (BSA-N, -Ix and -Pn) owns a characteristic shape and can be seen as a fingerprint for all three modification groups with specific features, mainly at short distances (Figure 6C, a, green). Nevertheless, all fatty acid binding sites and interspin distances of 16-DSA seem to remain largely conserved throughout the different stages of modification with all of them indicating a compact, native-like structure of the BSA-core as indicated by the regions (a–d). A crystal-like, or dynamically frozen behavior of BSA is further substantiated by the multiple staggered changes in ligand binding affinities as e.g. observed in the Scatchard plots of BSA-Ix (Figure 4D–F). From these available data sets it is therefore concluded that the applied PTMs freeze the structural plasticity of the BSA core in solution.


34

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

Biomacromolecules

CONCLUSIONS The successful utilization of EPR spectroscopic strategies are presented together with further analytic methods to unravel the dualism in structure-function relationships of the effect that posttranslational modifications exert on protein (BSA) molecules regarding polymer chemistrybased applications. It is found that changes in the ligand-binding behavior of BSA can be mainly traced back to electrostatic and sterical effects. This suggests that ligand affinity (KA), ligand capacity (NT) and ligand-binding cooperativity (C) in albumins can be fine-tuned by chemical modifications that alter net charge and binding site accessibility. Independent of the extent of the investigated modifications, linearity in Scatchard plots compared to native BSA appears to be perturbed by generally inducing heterogeneity to the individual binding site affinities. Multiphasic ligand-binding curves are obtained that reveal negative cooperativity at lower levels of bound ligand equivalents and positive cooperativity at higher loadings above 3 – 4 bound ligands (16-DSA). It is also shown that the rate of diffusion plays a significant role in protein-ligand interactions. In the course of processing BSA towards its proposed role as a polymer-coated drug delivery device, the protein-core itself is forced into a native-like and compact crystalline state, as observed from a comparative view of spatial ligand alignments in DEER experiments. The binding site selection and preference appears to get shuffled with the degree of modification. Together with the results from CW EPR, a picture can be drawn that features an alteration of BSA’s functional adaptability and therefore its structural plasticity. The functionalization of only a few surface lysines as the main target for PTMs leads to marginal increases in molecular weight (e.g. BSA-Ix) but may already induce strong functional effects. This analytic strategy is thought to be not only restricted to the investigation of albumins but may be adopted for any self-assembling system that relies on non-covalent interactions between


35


Page 36 of 52

substrate and ligand. The single prerequisite would be that a paramagnetic drug or natural ligandderivative is accessible,115 so that the system of interest may be investigated with EPR spectroscopy. Likewise, the physicochemical effects resulting from the aforementioned processes with e.g. physiological relevance as glycosylation or acetylation could be strategically investigated and rationalized in a similar way by just varying the degree of chemical modification. EPR spectroscopy reveals itself as a powerful quantitative tool that may provide information on a nanoscopic analytic level. More specifically, this study illustrates that PTMs as applied in polymer chemistry do not only alter bulk physicochemical properties of the constructs but also modify subtle ligand-substrate interactions on various functional levels. The concept of protein shape plasticity and ligand binding cooperativity is shown to exert decisive effects on protein functionality. A present tentative notion arises that these intricate PTM-induced functional aspects can be fine-tuned as well as the bulk properties for various purposes and should be rigorously investigated in future conceivable studies of this kind.

ASSOCIATED CONTENT Supporting Information Following Supporting Information is available: MALDI-TOF mass spectra of BSA and BSAbased macroinitiators (BSA-Ix), SDS PAGE of different stages of BSA polymer-protein conjugation (BSA-Pn), calculational details about Zeta potential characterization of modified BSA samples and technical details about EPR data evaluation, as referred to in the main text.

AUTHOR INFORMATION Corresponding Author *E-mail [email protected] (D.H.)


36

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

Biomacromolecules

ORCID ID Dariush Hinderberger: 0000-0002-6066-7099 Frederik Wurm: 0000-0002-6955-8489 Jörg Reichenwallner: 0000-0002-6862-1802 Jana Eisermann: 0000-0001-7784-5720

Author Contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

Notes The authors declare no competing financial interest.

ACKNOWLEDGEMENTS The authors thank Heike Schimm, Stefanie Weber and Franziska Zeuner for technical support. We gratefully acknowledge financial support from the Fonds der Chemischen Industrie (FCI), from the Fraunhofer Internal Program under Grant No. Attract 069-608203 (C. E. H. Schmelzer), and from the Deutsche Forschungsgemeinschaft (DFG) under grant number HI 1094/5–1 and SFB TRR 102.


37


Page 38 of 52

REFERENCES (1)

Pejaver, V.; Hsu, W. L.; Xin, F.; Dunker, A. K.; Uversky, V. N.; Radivojac, P. The structural and functional signatures of proteins that undergo multiple events of posttranslational modification. Protein Sci. 2014, 23, 1077–1093.

(2)

Berg, J. M.; Tymoczko, J. L.; Stryer, L. Biochemie; Spektrum Akademischer Verlag: Heidelberg, 2013; pp 80–85.

(3)

Peters Jr., T. All about Albumin: Biochemistry, Genetics and Medical Applications; Academic Press, San Diego, 1995.

(4)

Litus, E. A.; Permyakov, S. E.; Uversky, V. N.; Permyakov, E. A. Intrinsically Disordered Regions in Serum Albumin: What Are They For? Cell Biochem. Biophys. 2018, 76, 39–57.

(5)

Hochstrasser, M. Origin and function of ubiquitin-like proteins. Nature 2009, 458, 422– 429.

(6)

Strahl, B. D.; Allis, C. D. The language of covalent histone modifications. Nature, 2000, 403, 41–45.

(7)

Stayton, P. S.; Shimoboji, T.; Long, C.; Chilkoti, A.; Chen, G.; Harris, J. M.; Hoffman, A. S. Control of protein-ligand recognition using a stimuli-responsive polymer. Nature 1995, 378, 472–474.

(8)

Broyer, R. M.; Grover, G. N.; Maynard, H. D. Emerging synthetic approaches for protein-polymer conjugations. Chem. Commun. 2011, 47, 2212–2226.

(9)

Roberts, M. J.; Bentley, M. D.; Harris, J. M. Chemistry for peptide and protein PEGylation. Adv. Drug Delivery Rev. 2012, 64, 116–127.


38

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

Biomacromolecules

(10)

Abuchowski, A.; van Es, T.; Palczuk, N. C.; Davis, F. F. Alteration of Immunological Properties of Bovine Serum Albumin by Covalent Attachment of Polyethylene Glycol. J. Biol. Chem. 1977, 252, 3578–3581.

(11)

Lutz, J. F. Polymerization of Oligo(Ethylene Glycol) (Meth)Acrylates: Toward New Generations of Smart Biocompatible Materials. J. Polym. Sci., Part A: Polym. Chem. 2008, 46, 3459–3470.

(12)

Sisson, A. L.; Steinhilber, D.; Rossow, T.; Welker, P.; Licha, K.; Haag, R. Biocompatible Functionalized Polyglycerol Microgels with Cell Penetrating Properties. Angew. Chem. Int. Ed. 2009, 48, 7540–7545.

(13)

Khandare, J.; Mohr, A.; Calderon, M.; Welker, P.; Licha, K.; Haag, R. Structurebiocompatibility relationship of dendritic polyglycerol derivatives. Biomaterials 2010, 31, 4268–4277.

(14)

Panganiban, B.; Qiao, B.; Jiang, T.; DelRe, C.; Obadia, M. M.; Nguyen, T. D.; Smith, A. A. A.; Hall, A.; Sit, I.; Crosby, M. G.; Dennis, P. B.; Drockenmuller, E.; Olvera de la Cruz, M.; Xu, T. Random heteropolymers preserve protein function in foreign environments. Science 2018, 359, 1239–1243.

(15)

Day, J. F.; Thorpe, S. R.; Baynes, J. W. Nonenzymatically Glucosylated Albumin. J. Biol. Chem. 1979, 254, 595–597.

(16)

Dolhofer, R.; Wieland, O. H. Glycosylation of Serum Albumin: Elevated GlycosylAlbumin in Diabetic Patients. FEBS Lett. 1979, 103, 282–286.

(17)

Hawkins, D.; Pinckard, R. N.; Farr, R. S. Acetylation of Human Serum Albumin by Acetylsalicylic Acid. Science 1968, 160, 780–781.


39


(18)

Page 40 of 52

Means, G. E.; Bender, M. L. Acetylation of Human Serum Albumin by p-Nitrophenyl Acetate. Biochemistry 1975, 14, 4989–4994.

(19)

Oettl, K.; Stauber, R. E. Physiological and pathological changes in the redox state of human serum albumin critically influence its binding properties. Br. J. Pharmacol. 2007, 151, 580–590.

(20)

Oettl, K.; Marsche, G. Redox State of Human Serum Albumin in Terms of Cysteine-34 in Health and Disease. Methods Enzymol. 2010, 474, 181–195.

(21)

Anraku, M.; Chuang, V. T. G.; Maruyama T.; Otagiri, M. Redox properties of serum albumin. Biochim. Biophys. Acta, Gen. Subj. 2013, 1830, 5465–5472.

(22)

Bonanata, J.; Turell, L.; Antmann, L.; Ferrer-Sueta, G.; Botasini, S.; Mendez, E.; Alvarez, B.; Coitino, E. L. The thiol of human serum albumin: Acidity, microenvironment and mechanistic insights on its oxidation to sulfenic acid. Free Radical Biol. Med. 2017, 108, 952–962.

(23)

Thordarson, P.; Le Droumaguet, B.; Velonia, K. Well-defined protein-polymer conjugates – synthesis and potential applications. Appl. Microbiol. Biotechnol. 2006, 73, 243–254.

(24)

Gauthier, M. A.; Klok, H. A. Peptide/protein-polymer conjugates: synthetic strategies and design concepts. Chem. Commun. 2008, 2591–2611.

(25)

Ko, S.; Gunasekaran, S. Preparation of sub-100-nm β-lactoglobulin (BLG) nanoparticles. J. Microencapsulation 2006, 23, 887–898.

(26)

D’Urso, E. M.; Fortier, G. New Hydrogel based on polyethylene glycol cross-linked with bovine serum albumin. Biotechnol. Tech. 1994, 8, 71–76.


40

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

Biomacromolecules

(27)

Wurm, F.; Steinbach, T.; Klok, H. A. One-pot squaric acid diester mediated aqueous protein conjugation. Chem. Commun. 2013, 49, 7815–7817.

(28)

Stone, T. J.; Buckman, T.; Nordio, P. L.; McConnell, H. M. Spin-labeled Biomolecules. Proc. Natl. Acad. Sci. USA 1965, 54, 1010–1017.

(29)

Griffith, O. H.; McConnell, H. M. A Nitroxide–Maleimide Spin Label. Proc. Natl. Acad. Sci. USA 1966, 55, 8–11.

(30)

Wallach, D. Effect of Internal Rotation on Angular Correlation Functions. J. Chem. Phys. 1967, 47, 5258–5268.

(31)

Matyjaszewski, K. Atom Transfer Radical Polymerization (ATRP): Current Status and Future Perspectives. Macromolecules 2012, 45, 4015–4039.

(32)

Heredia, K. L.; Bontempo, D.; Ly, T.; Byers, J. T.; Halstenberg, S.; Maynard, H. D. In Situ Preparation of Protein–“Smart” Polymer Conjugates with Retention of Bioactivity. J. Am. Chem. Soc. 2005, 127, 16955–16960.

(33)

Nicolas, J.; San Miguel, V.; Mantovani, G.; Haddleton, D. M. Fluorescently tagged polymer bioconjugates from protein derived macroinitiators. Chem. Commun. 2006, 4697–4699.

(34)

Le Droumaguet, B.; Nicolas, J. Recent advances in the design of bioconjugates from controlled/living radical polymerization. Polym. Chem. 2010, 1, 563–598.

(35)

Sumerlin, B. S. Proteins as Initiators of Controlled Radical Polymerization: Graftingfrom via ATRP and RAFT. ACS Macro Lett. 2012, 1, 141–145.

(36)

Averick, S.; Simakova, A.; Park, S.; Konkolewicz, D.; Magenau, A. J. D.; Mehl, R. A.; Matyjaszewski, K. ATRP under Biologically Relevant Conditions: Grafting from a Protein. ACS Macro Lett. 2012, 1, 6–10.


41


(37)

Page 42 of 52

Konkolewicz, D.; Magenau, A. J. D.; Averick, S. E.; Simakova, A.; He, Hongkun; Matyjaszewski, K. ICAR ATRP with ppm Cu Catalyst in Water. Macromolecules 2012, 45, 4461–4468.

(38)

Strozyk, M. S.; Chanana, M.; Pastoriza-Santos, I.; Perez-Juste, J.; Liz-Marzan, L. M. Protein/Polymer-Based Dual-Responsive Gold Nanoparticles with pH-Dependent Thermal Sensitivity. Adv. Funct. Mater. 2012, 22, 1436–1444.

(39)

Cummings, C. S.; Fein, K.; Murata, H.; Ball, R. L.; Russell, A. J.; Whitehead, K. A. ATRP-grown protein-polymer conjugates containing phenylpiperazine selectively enhance transepithelial protein transport. J. Controlled Release 2017, 255, 270–278.

(40)

Steinbach, T.; Wurm, F.; Klok, H. A. Squaric acid mediated bioconjugation expanded to polymers prepared by ATRP. Polym. Chem. 2014, 5, 4039–4047.

(41)

Gauthier, M. A.; Klok, H. A. Polymer-protein conjugates: an enzymatic activity perspective. Polym. Chem. 2010, 1, 1352–1373.

(42)

Liu, M.; Johansen, P.; Zabel, F.; Leroux, J. C.; Gauthier, M. A. Semi-permeable coatings fabricated from comb-polymers efficiently protect proteins in vivo. Nat. Commun. 2014, 5, 5526.

(43)

Morgenstern, J.; Alvaradejo, G. G.; Bluthardt, N.; Beloqui, A.; Delaittre, G.; Hubbuch, J. Impact of Polymer Bioconjugation on Protein Stability and Activity Investigated with Discrete Conjugates: Alternatives to PEGylation. Biomacromolecules 2018, 19, 4250– 4262.

(44)

Kendall, F. E. Studies on Human Serum Proteins. II. Crystallization of Human Serum Albumin. J. Biol. Chem. 1941, 138, 97–109.


42

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

Biomacromolecules

(45)

Goodman, D. S. The Interaction of Human Serum Albumin with Long-chain Fatty Acid Anions. J. Am. Chem. Soc. 1958, 80, 3892–3898.

(46)

Keith, A. D.; Waggoner, A. S.; Griffith, O. H. Spin-labeled Mitochondrial Lipids in Neurospora Crassa. Proc. Natl. Acad. Sci. USA 1968, 61, 819–826.

(47)

Griffith, O. H.; Waggoner, A. S. Nitroxide Free Radicals: Spin Labels for Probing Biomolecular Structure. Acc. Chem. Res. 1969, 2, 17–24.

(48)

Waggoner, A. S.; Kingzett, T. J.; Rottschaefer, S.; Griffith, O. H. A Spin-labeled Lipid for Probing Biological Membranes. Chem. Phys. Lipids 1969, 3, 245–253.

(49)

Gaffney, B. J.; McConnell, H. M. The Paramagnetic Resonance Spectra of Spin Labels in Phospholipid Membranes. J. Magn. Reson. 1974, 16, 1–28.

(50)

Morrisett, J. D.; Pownall, H. J.; Gotto, Jr., A. M. Bovine Serum Albumin. Study of the Fatty Acid and Steroid Binding Sites using Spin-labeled Lipids. J. Biol. Chem. 1975, 250, 2487–2494.

(51)

Rehfeld, S. J.; Eatough D. J.; Plachy, W. Z. The binding isotherms for the interaction of 5-doxyl stearic acid with bovine and human albumin. J. Lipid Res. 1978, 19, 841–849.

(52)

Perkins, Jr., R. C.; Abumrad, N.; Balasubramanian, K.; Dalton, L. R.; Beth, A. H.; Park, J. H.; Park, C. R. Equilibrium Binding of Spin-Labeled Fatty Acids to Bovine Serum Albumin: Suitability as Surrogate Ligands for Natural Fatty Acids. Biochemistry 1982, 21, 4059–4064.

(53)

Gantchev, T. G.; Shopova, M. B. Characterization of spin-labelled fatty acids and hematoporphyrin binding sites interactions in serum albumin. Biochim. Biophys. Acta, Protein Struct. Mol. Enzymol. 1990, 1037, 422–434.


43


(54)

Page 44 of 52

Ge, M.; Rananavare, S. B.; Freed, J. H. ESR studies of stearic acid binding to bovine serum albumin. Biochim. Biophys. Acta, Gen. Subj. 1990, 1036, 228–236.

(55)

Kazmierczak, S. C.; Gurachevsky, A.; Matthes, G.; Muravsky, V. Electron Spin Resonance Spectroscopy of Serum Albumin: A Novel New Test for Cancer Diagnosis and Monitoring. Clin. Chem. 2006, 52, 2129–2134.

(56)

Junk, M. J. N.; Spiess, H. W.; Hinderberger, D. The Distribution of Fatty Acids Reveals the Functional Structure of Human Serum Albumin. Angew. Chem. Int. Ed. 2010, 49, 8755–8759.

(57)

Junk, M. J. N.; Spiess, H. W.; Hinderberger, D. DEER in biological multispin-systems: A case study on the fatty acid binding to human serum albumin. J. Magn. Reson. 2011, 210, 210–217.

(58)

Akdogan, Y.; Junk, M. J. N.; Hinderberger, D. Effect of Ionic Liquids on the Solution Structure of Human Serum Albumin. Biomacromolecules 2011, 12, 1072–1079.

(59)

Pavicevic, A. A.; Popovic-Bijelic, A. D.; Mojovic, M. D.; Susnjar, S. V.; Bacic, G. G. Binding of Doxyl Stearic Spin Labels to Human Serum Albumin: An EPR Study. J. Phys. Chem. B 2014, 118, 10898–10905.

(60)

Reichenwallner, J.; Oehmichen, M. T.; Schmelzer, C. E. H.; Hauenschild, T.; Kerth, A.; Hinderberger, D. Exploring the pH-induced Functional Phase Space of Human Serum Albumin by EPR Spectroscopy. Magnetochemistry 2018, 4, 47.

(61)

Reichenwallner, J.; Hinderberger, D. Using bound fatty acids to disclose the functional structure of serum albumin. Biochim. Biophys. Acta, Gen. Subj. 2013, 1830, 5382–5393.


44

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

Biomacromolecules

(62)

Martin, R. E.; Pannier, M.; Diederich, F.; Gramlich, V.; Hubrich, M.; Spiess, H. W. Determination of End-to-End Distances in a Series of TEMPO Diradicals of up to 2.8 nm Length with a New Four-Pulse Double Electron Electron Resonance Experiment. Angew. Chem. Int. Ed. 1998, 37, 2833–2837.

(63)

Pannier, M.; Veit, S.; Godt, A.; Jeschke, G.; Spiess, H. W. Dead-Time Free Measurement of Dipole-Dipole Interactions between Electron Spins. J. Magn. Reson. 2000, 142, 331– 340.

(64)

Karush, F. Heterogeneity of the Binding Sites of Bovine Serum Albumin. J. Am. Chem. Soc. 1950, 72, 2705–2713.

(65)

Karush, F. The Interaction of Optically Isomeric Dyes with Human Serum Albumin. J. Am. Chem. Soc. 1954, 76, 5536–5542.

(66)

Bhattacharya, A. A.; Grüne, T.; Curry, S. Crystallographic Analysis Reveals Common Modes of Binding of Medium and Long-chain Fatty Acids to Human Serum Albumin. J. Mol. Biol. 2000, 303, 721–732.

(67)

Akdogan, Y.; Reichenwallner, J.; Hinderberger, D. Evidence for Water-Tuned Structural Differences in Proteins: An Approach Emphasizing Variations in Local Hydrophilicity. PLoS One 2012, 7, e45681.

(68)

Akdogan, Y.; Wu, Y.; Eisele, K.; Schaz, M.; Weil, T.; Hinderberger, D. Host-guest interactions in polycationic human serum albumin bioconjugates. Soft Matter 2012, 8, 11106–11114.

(69)

Kuan, S. L.; Stöckle, B.; Reichenwallner, J.; Ng, D. Y. W.; Wu, Y.; Doroshenko, M.; Koynov, K.; Hinderberger, D.; Müllen, K.; Weil, T. Dendronized Albumin Core–Shell Transporters with High Drug Loading Capacity. Biomacromolecules 2013, 14, 367–376.


45


(70)

Page 46 of 52

Dulbecco, R.; Vogt, M. Plaque Formation and Isolation of Pure Lines with Poliomyelitis Viruses. J. Exp. Med. 1954, 99, 167–182.

(71)

Steinbach, T. New Methods for Bioconjugation via Squaric Acid Derivatives. Diploma Thesis, Johannes Gutenberg-Universität Mainz and École Polytechnique Fédérale de Lausanne, August 2011.

(72)

Chevallet, M.; Luche, S.; Rabilloud, T. Silver staining of proteins in polyacrylamide gels. Nat. Protoc. 2006, 1, 1852–1858.

(73)

Jeschke, G. DEER Distance Measurements on Proteins. Annu. Rev. Phys. Chem. 2012, 63, 419–446.

(74)

Stoll, S.; Schweiger, A. EasySpin, a comprehensive software package for spectral simulation and analysis in EPR. J. Magn. Reson. 2006, 178, 42–55.

(75)

Freed, J. H. Anisotropic Rotational Diffusion and Electron Spin Resonance Linewidths. J. Chem. Phys. 1964, 41, 2077–2083.

(76)

Freed, J. H. Theory of Slow Tumbling ESR Spectra for Nitroxides. In Spin Labeling: Theory and Applications; Berliner, L. J., Ed.; Academic Press: New York, 1976; Vol. 1, pp 53–132.

(77)

Schneider, D. J.; Freed, J. H. Calculating Slow Motional Magnetic Resonance Spectra: A User’s Guide. In Spin Labeling: Theory and Applications; Berliner, L. J., Reuben, J., Eds.; Plenum Press: New York, London, 1989; Vol. 8, pp 1–76.

(78)

Muravsky, V.; Gurachevskaya, T.; Berezenko, S.; Schnurr, K.; Gurachevsky, A. Fatty acid binding sites of human and bovine albumins: Differences observed by spin probe ESR. Spectrochim. Acta, Part A 2009, 74, 42–47.


46

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

Biomacromolecules

(79)

Reichenwallner, J.; Thomas, A.; Nuhn, L.; Johann, T.; Meister, A.; Frey, H.; Hinderberger, D. Tunable dynamic hydrophobic attachment of guest molecules in amphiphilic core-shell polymers. Polym. Chem. 2016, 7, 5783–5798.

(80)

De Meyts, P.; Roth, J. Cooperativity in Ligand Binding: A New Graphic Analysis. Biochem. Biophys. Res. Commun. 1975, 66, 1118–1126.

(81)

Tanford, C. Physical Chemistry of Macromolecules; John Wiley & Sons, Inc.: New York, 1961; pp 526–586.

(82)

Jeschke, G.; Chechik, V.; Ionita, P.; Godt, A.; Zimmermann, H.; Banham, J.; Timmel, C. R.; Hilger, D.; Jung, H. DeerAnalysis2006 – a Comprehensive Software Package for Analyzing Pulsed ELDOR Data. Appl. Magn. Reson. 2006, 30, 473–498.

(83)

Kattnig, D. R.; Reichenwallner, J.; Hinderberger, D. Modeling Excluded Volume Effects for the Faithful Description of the Background Signal in Double Electron-Electron Resonance. J. Phys. Chem. B 2013, 117, 16542–16557.

(84)

Provencher, S. W. A Constrained Regularization Method for Inverting Data Represented by Linear Algebraic or Integral Equations. Comput. Phys. Commun. 1982, 27, 213–227.

(85)

Bhattacharjee, S. DLS and zeta potential – What they are and what they are not? J. Controlled Release 2016, 235, 337–351.

(86)

Berne, B. J.; Pecora, R. Dynamic light scattering – with applications to chemistry, biology, and physics; Dover Publications, Inc.: New York, 2000.

(87)

Noack, H.; Moitzi, C. Modulator Monitoring During Measuring Electromobility. U.S. Patent, US20140144780A1, May 29, 2014.


47


(88)

Page 48 of 52

Lemmon, E. W. Thermophysical Properties of Water and Steam. In CRC Handbook of Chemistry and Physics; Haynes, W. M., Lide, D. R., Bruno, T. J., Eds.; CRC Press: Boca Raton, 2014; Vol. 95, p 6–1.

(89)

Schiebener, P.; Straub, J.; Levelt Sengers, J. M. H.; Gallagher, J. S. Refractive Index of Water and Steam as Function of Wavelength, Temperature and Density. J. Phys. Chem. Ref. Data 1990, 19, 677–717.

(90)

Sheely, M. L. Glycerol Viscosity Tables. Ind. Eng. Chem. 1932, 24, 1060–1064.

(91)

Swan, J. W.; Furst, E. M. A simpler expression for Henry’s function describing the electrophoretic mobility of spherical colloids. J. Colloid Interface Sci. 2012, 388, 92–94.

(92)

Majorek, K. A.; Porebski, P. J.; Dayal, A.; Zimmerman, M. D.; Jablonska, K.; Stewart, A. J.; Chruszcz, M.; Minor, W. Structural and immunologic characterization of bovine, horse, and rabbit serum albumins. Mol. Immunol. 2012, 52, 174–182.

(93)

Tanford, C.; Buzzell, J. G. The Viscosity of Aqueous Solutions of Bovine Serum Albumin Between pH 4.3 and 10.5. J. Phys. Chem. 1956, 60, 225–231.

(94)

Jachimska, B.; Wasilewska, M.; Adamczyk, Z. Characterization of Globular Protein Solutions by Dynamic Light Scattering, Electrophoretic Mobility, and Viscosity Measurements. Langmuir 2008, 24, 6866–6872.

(95)

Chafer-Pericas, C.; Balaguer, A.; Maquieira, A.; Puchades, R. Dispersive solid-phase extraction and immunoassay with internal reference calibration using fatty acid-coated inorganic fluorescent nanoparticles. Anal. Biochem. 2013, 432, 31–37.


48

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

Biomacromolecules

(96)

Salis, A.; Boström, M.; Medda, L.; Cugia, F.; Barse, B.; Parsons, D. F.; Ninham, B. W.; Monduzzi, M. Measurements and Theoretical Interpretation of Points of Zero Charge/Potential of BSA Protein. Langmuir 2011, 27, 11597–11604.

(97)

Jachimska, B.; Pajor, A. Physico-chemical characterization of bovine serum albumin in solution and as deposited on surfaces. Bioelectrochemistry 2012, 87, 138–146.

(98)

Atmeh, R. F.; Arafa, I. M.; Al-Khateeb, M. Albumin Aggregates: Hydrodynamic Shape and Physico-Chemical Properties. Jordan J. Chem. 2007, 2, 169–182.

(99)

Derjaguin, B.; Landau, L. Theory of the Stability of Strongly Charged Lyophobic Sols and of the Adhesion of Strongly Charged Particles in Solutions of Electrolytes. Acta Physicochim. URSS 1941, 14, 633–662.

(100) Honary, S.; Jahanshahi, M.; Golbayani, P.; Ebrahimi, P.; Ghajar, K. Doxorubicin-Loaded Albumin Nanoparticles: Formulation and Characterization. J. Nanosci. Nanotechnol. 2010, 10, 7752–7757. (101) Larsson, M.; Hill, A.; Duffy, J. Suspension Stability; Why Particle Size, Zeta Potential and Rheology are Important. Ann T. Nord. Rheol. Soc. 2012, 20, 209–214. (102) Adamczyk, Z.; Bratek, A.; Jachimska, B.; Jasinski, T.; Warszynski, P. Structure of Poly(acrylic acid) in Electrolyte Solutions Determined from Simulations and Viscosity Measurements. J. Phys. Chem. B 2006, 110, 22426–22435. (103) Tanford, C.; Swanson, S. A.; Shore, W. S. Hydrogen Ion Equilibria of Bovine Serum Albumin. J. Am. Chem. Soc. 1955, 77, 6414–6421. (104) Manning, G. S. Counterion Binding in Polyelectrolyte Theory. Acc. Chem. Res. 1979, 12, 443–449.


49


Page 50 of 52

(105) Manning, G. S. Counterion Condensation on Charged Spheres, Cylinders, and Planes. J. Phys. Chem. B 2007, 111, 8554–8559. (106) Lamm, G.; Pack, G. R. Counterion Condensation and Shape Within Poisson–Boltzmann Theory. Biopolymers 2010, 93, 619–639. (107) Böhme, U.; Scheler, U. Effective charge of bovine serum albumin determined by electrophoresis NMR. Chem. Phys. Lett. 2007, 435, 342–345. (108) Bergethon, P. R. The Physical Basis of Biochemistry: The Foundations of Molecular Biophysics; Springer Science + Business Media, LLC: New York, 2010; pp 713–737. (109) Demtröder, W. Experimentalphysik 2: Elektrizität und Optik; Springer-Verlag: Berlin, Heidelberg, 2013; pp 1–42. (110) Tanford, C. The Electrostatic Free Energy of Globular Protein Ions in Aqueous Salt Solution. J. Phys. Chem. 1955, 59, 788–793. (111) Shoup, D.; Szabo, A. Role of Diffusion in Ligand Binding to Macromolecules and CellBound Receptors. Biophys. J. 1982, 40, 33–39. (112) Berg, O. G.; von Hippel, P. H. Diffusion-Controlled Macromolecular Interactions. Ann. Rev. Biophys. Biophys. Chem. 1985, 14, 131–160. (113) Copeland, R. A. Enzymes: A Practical Introduction to Structure, Mechanism, and Data Analysis; Wiley-VCH: New York, 2000; pp 76–108. (114) Scatchard, G. The Attractions of Proteins for Small Molecules and Ions. Ann. N.Y. Acad. Sci. 1949, 51, 660–672. (115) Hauenschild, T.; Reichenwallner, J.; Enkelmann, V.; Hinderberger, D. Characterizing Active Pharmaceutical Ingredient Binding to Human Serum Albumin by Spin-Labeling and EPR Spectroscopy. Chem. Eur. J. 2016, 22, 12825–12838.


50

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

Biomacromolecules

(116) Schreier, A. A.; Schimmel, P. R. Interaction of Manganese with Fragments, Complementary

Fragment

Recombinations,

and

Whole

Molecules

of

Yeast

Phenylalanine Specific Transfer RNA. J. Mol. Biol. 1974, 86, 601–620. (117) Norby, J. G.; Ottolenghi, P.; Jensen, J. Scatchard Plot: Common Misinterpretation of Binding Experiments. Anal. Biochem. 1980, 102, 318–320. (118) Rosenthal, H. E. A Graphic Method for the Determination and Presentation of Binding Parameters in a Complex System. Anal. Biochem. 1967, 20, 525–532. (119) Klotz, I. M.; Royer, G. P.; Sloniewsky, A. R. Macromolecule–Small Molecule Interactions. Strong Binding and Cooperativity in a Model Synthetic Polymer. Biochemistry 1969, 8, 4752–4756. (120) Stefan, M. I.; Le Novere, N. Cooperative Binding. PLoS Comput. Biol. 2013, 9, e1003106. (121) Spector, A. A.; John, K.; Fletcher, J. E. Binding of long-chain fatty acids to bovine serum albumin. J. Lipid Res. 1969, 10, 56–67. (122) Cistola, D. P.; Small, D. M.; Hamilton, J. A. Carbon 13 NMR Studies of Saturated Fatty Acids Bound to Bovine Serum Albumin. I. The Filling of Individual Fatty Acid Binding Sites. J. Biol. Chem. 1987, 262, 10971–10979.


51


Page 52 of 52

ToC graphic:


52

Ligand-Binding Cooperativity Effects in Polymer-Protein Conjugation

Recommend Documents