Embed Size (px)
Citation preview
Protein Structure Dynamics and Function at Micelle and Particle Interfaces
Michael Chin
Submitted in partial fulfillment of the requirements
for the degree of Doctor of Philosophy
in the Graduate School of Arts and Sciences
COLUMBIA UNIVERSITY
2015
© 2015
Michael Chin
All rights reserved
ABSTRACT
Protein Structure Dynamics and Function at Micelle and Particle Interfaces
Michael Chin
This work is an investigation into how protein function and structural dynamics are
affected by their interactions with detergents and particles. Recent studies have reported
the addition of certain non-ionic surfactants increase the catalytic rates of enzymes. This
is a departure from the established understanding of surfactant/protein interactions
wherein charged-surfactants tend to denature the macro-molecular structure, while other
detergents can be used to isolate and preserve protein structures. The mechanism of
denaturation by ionic surfactants is well understood, as is the preservation and
crystallization of protein structures with detergents. But why should non-ionic surfactants
increase observed activity? This question remains unanswered. Furthermore, the
explanations provided with previous studies where enzyme activity enhancement via
surfactant interactions are inconsistent with each other, and are highly specific to the
system in which they are studied. These contradicting conclusions highlight the need to
propose a general theory of protein-surfactant interactions which can explain these
reported changes in enzyme activity that can be applied to any of these reported systems.
The theory proposed in this thesis is as follows: Unlike ionic surfactants, non-ionic
surfactants minimally bind to the protein structure. Instead, bound water regions
surrounding non-ionic surfactant aggregates (micelles) interact with the surface hydration
of adjacent proteins, thus affecting the free energy of solvent forces at the protein surface.
This, in turn, affects the small scale fluctuations of the proteins structures which
influenced properties important to enzyme function, such as flexibility and substrate
binding energy. Aside from enzyme interactions with surfactant, the implications of this
proposed theory extend to any system in which functional proteins readily interact with a
surface. The two parallel examples also reported in this dissertation are the structural
dynamics of carrier proteins adsorbed onto metal phosphate adjuvant particles and a
mathematical model describing changes to FRET efficiency due to the fluctuating
motions of a light active protein
Three major objectives were established in order to test this hypothesis: 1) Establish
model system of enzyme and surfactants, and a means to test enzyme efficacy. 2) Study
the differences in surfactants and surfactant aggregates that would attribute to changes in
enzyme activity and 3) Test the flexibility or the enzyme structures as a function of
surfactant structure and concentration to see if there is, indeed, a connection between the
surfactant and the enzyme activity through a modification of the protein’s structural
dynamics.
Subtilisin proteinase and horseradish peroxidase were ultimately chosen as model
enzymes to study. Initial activity assays confirmed previously reported results in which
charged surfactants such as sodium dodecyl sulfonates had little effect or negatively
impacted enzyme activity. Longer exposures to these anionic surfactants result in a loss
of activity which can be attributed to enzyme denaturation. Enzyme activity in non-ionic
surfactants such as Polyoxyethylene (4) lauryl ether (Brij-30) and alkyl polyglucosides
(APG), increased. Curiously, dodecyl-β-D-maltoside (DM), a sugar-based non-ionic
surfactant of similar head group structure to APG, did not induce enhanced enzyme
activity, giving us two similar surfactants with which we can compare to determine what
factors, if not surfactant molecular structure, affect enzyme activity.
Surface tension, dynamic light scattering, electron spin resonance and NMR were used to
elucidate the structural differences between alkyl polyglucoside (APG) and dodecyl
maltoside micelles (DM), and which of these characteristics are attributing to the
differences in activity. It was found that APG transitions between two micelle structure
conformations as a function of concentration. After surpassing an initial critical micelle
concentration, APG forms loosely packed aggregates that are attracted to each other to
form even larger super-aggregates. Passing a second transitional concentration, the
micelles form dense, rod-like structures. It should be noted that enzyme activity peaks in
the presence of the loosely-packed super-aggregates and drops back to normal values
once the micelles transition into their rod-like configuration. Conversely, dodecyl
maltoside, which did not induce greater enzyme activity, does not exhibit this dual phase
transition. Instead, the surfactants in DM solutions exceeding the critical micelle
immediately form tightly packed, single layered spherical micelles. From these
experiments we conclude that the densely packed micelles like those formed by DM and
APG at higher concentrations do not affect enzyme structure the same way as those
loosely packed ones at lower APG concentrations.
Life-time fluorescence techniques were employed to fulfill the third objective and
measure the structural dynamics of the model proteins. In both subtilisin and horseradish
peroxidase, we observed a similar pattern of enzyme activity changes as a function of
APG and Brij-30 concentration. Anionic SDS surfactant did not affect this flexibility
parameter. Furthermore, we demonstrated that the flexibility of HRP is not affected by
dodecyl maltoside much in the same way dodecyl maltoside did not affect the enzyme
activity in any significant way.
The connection between micelle structure and enzyme flexibility can be attributed to the
structure and dynamics of the solvent between the micelle and enzyme. The final extent
of influence of solvent dynamics on enzyme function is still contentious; however the
experimental results from this work lend support this theory. Building off models
presented by Viparelli on the influence of micelle bound water on enzyme activity, and a
model describing the rate of “solvent-slaved” protein functions as a function of solvent
relaxation times, a new model which relates changes in enzyme activity as a function of
surfactant concentration is developed and presented here in this body of work.
i
Contents
List of Figures ................................................................................................................................... iii
1 Introduction ............................................................................................................................. 1
1.1 Initial findings and what is in the literature .................................................................... 2
1.2 Gaps in knowledge .......................................................................................................... 7
1.3 Central goal of dissertation: .......................................................................................... 10
2 Hypothesis and rationale ....................................................................................................... 11
2.1 Rationale for Hypotheses: ............................................................................................. 12
3 Research plan and objectives ................................................................................................ 18
3.1 Sub-objectives: .............................................................................................................. 18
4 Methods and Materials ......................................................................................................... 21
4.1 Objective 1 - Find suitable model system to test enzyme hyperactivity via interaction
with surfactants ......................................................................................................................... 21
4.1.1 Enzyme selection ................................................................................................... 22
4.1.2 Surfactants ............................................................................................................. 24
4.1.3 Subtilisin assay: ...................................................................................................... 26
4.1.4 Subtilisin initial rate activities and kinetics ........................................................... 27
4.1.5 HRP Kinetics: .......................................................................................................... 28
4.1.6 Subtilisin assay side note: Behavior of Fluorescein Probe (optimizing subtilisin
assay method) ....................................................................................................................... 29
4.2 Objective 2: Characterizing surfactant / particle properties in protein solutions ........ 40
4.2.1 DLS ......................................................................................................................... 40
4.2.2 Surface tension ...................................................................................................... 42
4.2.3 Electron paramagnetic resonance ......................................................................... 43
4.3 Objective 3: Measuring structural dynamics of model proteins ................................... 46
4.3.1 Life time fluorescence anisotropy ......................................................................... 46
4.3.2 Life time fluorescence decay ................................................................................. 49
4.4 Objective 3a: Dynamics and function changes at inorganic particle surfaces ............. 54
4.4.1 UV-Vis: ................................................................................................................... 56
4.4.2 Zeta potential measurement ................................................................................. 57
ii
4.4.3 FTIR ........................................................................................................................ 57
5 Results and discussion ........................................................................................................... 59
5.1 Objective 1: Establishing a model system ..................................................................... 59
5.1.1 5.1.1 Enzyme product yield assays. ....................................................................... 60
5.1.2 Initial rate assays – Subtilisin protease ................................................................. 63
5.1.3 HRP activity in Brij-30, APG and DM ...................................................................... 73
5.2 Discussion of subtilisin and HRP assay and activity results ........................................... 78
5.2.1 Sub section: kinetics parameters Km and Vmax ................................................... 82
5.3 Objective 2: Aggregation / micelle properties .............................................................. 95
5.3.1 Surface tension: ..................................................................................................... 95
5.3.2 Micelle size: ........................................................................................................... 97
5.3.3 ESR investigation of micelle structure: APG vs DM ............................................... 99
5.3.4 NMR: (Cooperative work with C. Totland of University of Bergen) .................... 101
5.3.5 IR investigation of Bio-Surfactant degradation via enzyme hydrolysis: .............. 106
5.4 Objective 3: Protein structure dynamics: .................................................................... 114
5.4.1 Life-time fluorescence: Anisotropy ..................................................................... 114
5.4.2 Time resolved fluorescence to measure flexibility of horseradish peroxidase: .. 119
5.5 Objective 3a: Protein flexibility and orientation at particle surfaces:......................... 127
5.6 Objective 3b: Modeling enzyme activity as a function of surfactant interaction and
concentration. ......................................................................................................................... 132
5.7 Results and discussion conclusions ............................................................................. 138
6 Conclusions and Next Steps ................................................................................................ 149
6.1 Next steps: ................................................................................................................... 152
7 References ........................................................................................................................... 154
iii
List of Figures
Figure 1-1 - Initial subtilisin activities in anionic and non-ionic surfactants ................................... 2
Figure 1-2 - Increased hydrolytic rate seen in Bacillus amyloliquefaciens α-amylase (BAA) as a
function of Brij-35 concentration as reported by Hosino et al. ...................................................... 3
Figure 1-3 - Three phases that enzyme hydrolysis can occur. Bulk (Sf), Bound water interface (Sb)
and within the micelle as described by Viparelli et al. .................................................................... 6
Figure 1-4 - Top down view of three protein chains which compose of a bacteriorhodopsin unit.
One is shaded blue with a white outline indicating the location of the retinol chromophore. ...... 9
Figure 2-1 - Range of time scales associated with protein structure dynamics. This study with
focus on elastic fluctuations occurring in the ps range. ................................................................ 13
Figure 2-2 - General schematic and description of induced fit model. Active site conformation
changes play an important role in minimizing energy barrier to product formation [33] ............ 14
Figure 4-1 - Dodecyl maltoside ...................................................................................................... 25
Figure 4-2 - Example of colormetric activity assay ........................................................................ 27
Figure 4-3 - Schematic illustrating liberation of free fluoresein probe from labled casein .......... 28
Figure 4-4 - Fluoresence spectroscopy of fluorescein fluorophore dilution series ....................... 30
Figure 4-5 - Fluorescein peak wavelength vs. concentration ........................................................ 31
Figure 4-6 - Fluorescein peak intensity vs. Concentration ............................................................ 32
Figure 4-7 - Divergence from intensity / concentration linarity starting at 0.1 mM of flourescein
....................................................................................................................................................... 33
Figure 4-8 - Reaction profiles of subtilisin catalyzed proteolysis (0.001 to 0.1 mM substrate
concentrations) ............................................................................................................................. 34
Figure 4-9 - End of reaction fluoresence intensity vs. substrate concentration (.0001 to 0.1 mM)
....................................................................................................................................................... 35
Figure 4-10 - Casein hydrolysis reaction profile (0.0005 to 0.025 mM substrate concentration) 36
Figure 4-11 - End of reaction fluoresence intensity (low substrate concentration 0.0005 to 0.025
mM) ............................................................................................................................................... 37
Figure 4-12 - Set of initial rates plotted against substrate concentration drawn with error
distribution. ................................................................................................................................... 38
Figure 4-13 - Schematic of Zeeman Effect energy splitting ........................................................... 43
Figure 4-14 - Typical nitroxyl group ESR spectrum and peak parameters used to calculate
rotational correlation times .......................................................................................................... 44
Figure 4-15 - Dansyl fluoride structure.......................................................................................... 46
Figure 4-16 - example of parallel (VV) and perpendicular(VH) fluorescence decays with
instrument response function (IRF)............................................................................................... 47
Figure 4-17 - anisotropy (r(t)) derived from VV(t) and VH(t). [62] ............................................... 48
Figure 4-18 - Structure of horseradish peroxidase. Trp residue and heme group are highlighted.
....................................................................................................................................................... 50
Figure 4-19 - Structure of CRM 197 carrier protein used in vaccine formulations ....................... 55
Figure 4-20 - Concentration standards of BSA measured by UV-Vis at 280nm ............................ 56
iv
Figure 5-1 - Subtilisin activity with and without SDS anionic surfactant as a function of exposure
time ............................................................................................................................................... 60
Figure 5-2 - Subtilisin activity with and without SDS anionic surfactant as a function of detergent
concentration ................................................................................................................................ 62
Figure 5-3 - Enzyme activity assays of subtilisin in various surfactants (relative to enzyme in
buffer) ............................................................................................................................................ 63
Figure 5-4 - Fluorescence spectrum of casein substrate proteolysis as a function of time. Peak
intensity is found at 519 nm. ......................................................................................................... 65
Figure 5-5 - Fluorescence spectrum of casein substrate proteolysis as a function of time. Peak
intensity is found at 519 nm. ......................................................................................................... 65
Figure 5-6 - Initial subtilisin rate profile experiments confirm faster initial proteolysis rates in
alkyl polyglucoside solutions. ........................................................................................................ 67
Figure 5-7 - Subtilisin catalyzed proteolysis reaction profile in APG solutions ............................. 67
Figure 5-8 - Normalized values of the initial rates derived from reaction profiles shown in Fig. 5.7
....................................................................................................................................................... 68
Figure 5-9 - Subtilisin catalyzed proteolysis reaction profile in Brij-30 surfactant ....................... 69
Figure 5-10 - Initial reaction rates of subtilisin mediated proteolysis in Brij-30 non-ionic
surfactants ..................................................................................................................................... 69
Figure 5-11 - Proteolysis reaction profile in Brij-35 surfactant ..................................................... 70
Figure 5-12 - Initial reaction rates of subtilisin catalyzed proteolysis in Brij-35 surfactant .......... 71
Figure 5-13 - Normalized reaction rates of Subtilisin in low concentrations. APG CMC is
approximately 0.2 mM. ................................................................................................................. 72
Figure 5-14 - Subtilisin activity in Brij-30 solutions ....................................................................... 72
Figure 5-15 - HRP initial reaction rates in APG surfactant ............................................................ 74
Figure 5-16 - - Initial reaction rates of HRP in APG (columns) and surface tension measurements
of APG (diamonds) ......................................................................................................................... 75
Figure 5-17 - Normalized initial rate values of HRP activity in dodecyl maltoside concentrations
....................................................................................................................................................... 76
Figure 5-18 - Normalized initial reaction rates of subtilisin (red) compared to those of HRP
enzyme (black) ............................................................................................................................... 77
Figure 5-19 - Figure 5.17. Comparison of normalized initial rates of HRP enzyme in APG (black
squares) vs DM (red triangles) ...................................................................................................... 77
Figure 5-20 - Subtilisin mediated proteolysis reaction profiles as a function of substrate
concentration ................................................................................................................................ 84
Figure 5-21 - Non-linear regression fitting of reaction rates of subtilisin in buffer vs. casein
substrate concentration done in Origins 8.0 software ................................................................. 84
Figure 5-22 - Initial rates of subtilisin catalyzed proteolysis vs substrate concentration in 10mM
of APG ............................................................................................................................................ 85
Figure 5-23 - Initial rates of subtilisin catalyzed proteolysis vs substrate concentration in 20mM
of APG ............................................................................................................................................ 86
Figure 5-24 - Subtilisin Vmax as a function of APG concentration ................................................ 87
Figure 5-25 - Subtilisin Vmax as a function of DM concentration ................................................. 88
v
Figure 5-26 - Subtilisin Vmax as a function of Brij-30 concentration ............................................ 88
Figure 5-27 - Subtilisin Vmax as a function of Brij-35 concentration ............................................ 89
Figure 5-28 - Subtilisin Km as a function of APG concentration .................................................... 90
Figure 5-29 - Subtilisin Km as a function of DM concentration ..................................................... 91
Figure 5-30 - Subtilisin Km as a function of Brij-35 concentration ................................................ 91
Figure 5-31 - Subtilisin Km as a function of Brij-30 concentration ................................................ 92
Figure 5-32 - Surface tension vs APG concentration ..................................................................... 96
Figure 5-33 - Surface tension comparisons of dodecyl maltoside (single surfactant) to alkyl
polyglucosides, a surfactant mixture of similar composition. ....................................................... 97
Figure 5-34 - APG micellesize vs concentration ............................................................................ 98
Figure 5-35 - DM micelle size vs concentration ............................................................................ 98
Figure 5-36 - Rotational correlation times (measure of local viscosity) as a function of surfactant
(APG and DM) concentration in buffer only. ................................................................................. 99
Figure 5-37 - Rotational correlation times (measure of local viscosity) as a function of surfactant
(APG and DM) concentration in buffer with 0.01 mg/ml HRP. ................................................... 100
Figure 5-38 - Comparison of the line widths of the main APG CH2 peak (2-12) for the 0.15 and 12
mM APG sample in 0.01 mg/ml HRP in 100mM potassium phosphate buffer ........................... 102
Figure 5-39 - The line width here is the line width of the main CH2 peak (2-12) in the NMR
spectrum. The line widths are measured by using the deconvolution tool in the Topspin software
(version 1.3). ................................................................................................................................ 103
Figure 5-40 - Overlay of the aggregate mobility data presented in figure 5.31 with the activity of
HRP in APG (triangles) ................................................................................................................. 104
Figure 5-41 - NOESY spectra with water signal suppression using excitation sculpting with
gradients, at different concentrations of APG at constant HRP concentration. A possible
interpretation in terms of APG aggregate structure is presented to the left. ............................ 105
Figure 5-42 - Sodium Dodecyl Sulfate ......................................................................................... 107
Figure 5-43 - n-Dodecyl β,D Maltoside ........................................................................................ 107
Figure 5-44 - IR spectrum of SDS and Dodecyl Maltoside between 1350 and 850 cm-1
wavenumber range. .................................................................................................................... 108
Figure 5-45 - IR spectrum of SDS and Dodecyl Maltoside between 1160 and 960 cm-1
wavenumber range. .................................................................................................................... 109
Figure 5-46 - Schematic of cellulase hydrolysis mechanism ....................................................... 110
Figure 5-47 - FTIR of Dodecyl maltoside with a-Cellulase indicating possible loss of C-O-C bond
..................................................................................................................................................... 111
Figure 5-48 - Proposed reaction of surfactant catalysis .............................................................. 111
Figure 5-49 - FTIR from 1800 to 1500 cm-1 possibly showing the carboxyl group located at the
cellulase active site. ..................................................................................................................... 112
Figure 5-50- Rotational correlation time (blue diamonds) and subtilisin activity (orange squares)
as a function of Brij-30 concentration. ........................................................................................ 117
Figure 5-51 - Decay correlation times of dansyl probe attached to subtilisin as a function of APG
detergent. .................................................................................................................................... 118
vi
Figure 5-52 -Top. Raw data (Blue) and fitting of sum of 3 exponential decay functions (orange) of
hrp in buffer. Bottom: residual of fitted line vs raw decay data. ................................................ 119
Figure 5-53 - Decay data and multiple exponential fitting (including residuals below) of HRP
enzyme in 0.05 mM APG. ............................................................................................................ 122
Figure 5-54 - Decay data and multiple exponential fitting (including residuals below) of HRP
enzyme in 2.0 mM APG. .............................................................................................................. 123
Figure 5-55 - Decay data and multiple exponential fitting (including residuals below) of HRP
enzyme in 10.0 mM APG. ............................................................................................................ 124
Figure 5-56 - Fluorescence decay times as a function of surfactant (APG) concentration ......... 125
Figure 5-57 - Fluorescence decay times as a function of surfactant (dodecyl maltoside)
concentration .............................................................................................................................. 126
Figure 5-58 - Amount of CRM-197 adsorbed onto AlPO4 particles at pH 4 and 5 as a function of
initial protein concentration. ...................................................................................................... 127
Figure 5-59 - CRM-197 structure with Trp residues highlighted ................................................. 128
Figure 5-60 - Fluorescence decay times resolved from CRM in pH 7 and 5.8 buffer and adsorbed
onto AlPO4 at pH 5.8 ................................................................................................................... 129
Figure 5-61 - The decay contribution of each decay time. .......................................................... 130
Figure 5-62 - a) (left) hydration sphere of densely packed micellese such as that of dodecyl
maltoside and high conc. APG consists of tightly bound water. b) (right) loosely packed micelles
result in more mobile hydration sphere ..................................................................................... 137
Figure 5-63 - Minimization of loosely bound hydration layers leading to enzyme/micelle
interaction ................................................................................................................................... 137
Figure 5-64 - Schematic differentiating micell formation in DM (top) vs APG (bottom) ............ 144
vii
Acknowledgements
I would like to extend my deepest thanks to all those who have supported this
undertaking and made this work possible: Firstly, to my advisor, Professor P.
Somasundaran who has guided my efforts throughout the years with wisdom, knowledge,
and patience.
I would also like to acknowledge my committee members, Professor’s Arthur Palmer III,
John Hunt, Xi Chen, and Raymond Farinato. Their advice and insight throughout the
course of this research have been invaluable. Similarly, the input from Dr. Nagaraj,
Professor Moudgil, my fellow co-workers, and all of the participating members of the
Columbia Center for Particulate and Surfactant Systems (CPASS), has helped to refine
this work into what it is today.
Finally I wish to thank my family and friends: My parents and sister for their ever-present
emotional and financial support, and my best-friend, who convinced me to pursue this
unique opportunity during a time when I might have decided otherwise.
1
1 Introduction
Enzymes are transitioning from niche technologies to a prominent tool in industrial
processing, energy production, and pharmaceuticals and medicine. One common use for
enzymes is in commercial laundry applications, where they are incorporated into
detergents to break down sugars, lipids, and proteins, so that stains ingrained into fabrics
are removed with less thermal and physical agitation [1], thereby conserving water and
energy. Improving upon the function, stability and reliability of enzymes and other
functional proteins is an increasingly important goal to fields of alternative energy as
well. In the case of bio-fuels, enzymes are a prominent strategy towards efficiently
liberating free sugars from cellulosic bio-feed stocks under low-energy conditions with
minimal impact to the environment. Other emerging concepts include bio-fuel cells [2,
3], in which enzymes catalyze sugar into free electrons [4], and bio-solar[5], in which
light-harvesting proteins are used to capture light and transfer energy to reaction centers
in photovoltaic cells. Control of macromolecules behavior also plays a major role in
medical technology - such as the interactions between anti-gen and adjuvant particles in
vaccine formulation. Understanding the interactions of enzymes and other proteins with
detergents and at particle is becoming an increasingly important goal. This particular
work will focus on elucidating the interactions between enzymes and detergents,
particularly, the mechanism by which certain surfactants have been reported to increase
enzyme activity.
2
1.11.11.11.1 Initial Initial Initial Initial findingsfindingsfindingsfindings and what is in the literatureand what is in the literatureand what is in the literatureand what is in the literature
Surfactant dependent enzyme activity was observed by our own initial experiments when
testing the activity of a subtilisin protease in solutions of detergents typically used in
detergent applications, including: anionic sodium dodecyl sulfonates and bio-based, non-
ionic, alkyl polyglucosides. As expected, ionic surfactants slightly diminished the activity
of the enzyme relative to a buffer only solution, presumably from denaturation of the
enzyme structure[6] (Figure 1.1) the two non-ionic surfactants tested, including the
sugar-based APG, increased enzyme activity, however.
SDS LDS Brij-30 APG
-10
0
10
20
30
40
50
60
70
Activity r
ela
tive
to
su
btilis
in in
bu
ffe
r
Sufactant
Figure 1-1 - Initial subtilisin activities in anionic and non-ionic surfactants
Our finding is not the first. In the past 20 years there have been several reports of
elevated enzymatic activity in the presence of non-ionic surfactants. This phenomenon
has been observed in wide range of enzyme types and target applications, ranging from
3
amylases and proteases for digesting sugars and proteins suspended in bulk solutions and
from fabric surfaces[7-10] to cellulases used to liberate free sugars from crystalline
polysaccharides (figure 1.2) for use in bio-fuel production [11-13]. In our own
preliminary experiments, we have observed enhanced proteolysis in the presence of sugar
based non-ionic alkypolyglucoside (APG) as well as a polyoxyethylene lauryl ethers (Brij
series surfactants). This enzymatic hyperactivity has also been observed in reverse
micelle systems[14, 15]. Regardless of the enzyme, substrate, or solvent environment, it
is particular non-ionic surfactants that are consistently observed promoting this
enhancement. Ionic surfactants, on the other hand, denature protein structures [6, 16-18]
reducing the proteins function.
Figure 1-2 - Increased hydrolytic rate seen in Bacillus amyloliquefaciens α-amylase (BAA) as a function of Brij-35
concentration as reported by Hosino et al.
While there is an apparent and consistent connection between non-ionic surfactants and
increased enzyme activity in the literature, the underlying mechanism that explains how
these surfactants are inducing this phenomenon is not clear. Shome et al [14] conclude
4
that additions of non-ionic surfactants to the cationic surfactant mixtures reduce the
surface charge at the head-group / protein interface, limiting the chance of electro-static
inhibition of an active site. Kaar et al [13] suggests that the non-ionic Tween surfactants
helped to maintain cellulase structure as it digested cellulosic bio feed stock into free
sugars; Erikson and colleges disagree with this assessment [11], and show results
supporting their own hypothesis that the non-ionic surfactants prevent non-specific and
permanent cellulase adsorption onto the cellulose surface, rendering the enzyme inactive.
Alternatively, Hoshino et al [7] attribute their observations of enhanced amylase activity
in Brij-35 surfactants to both the enzyme and substrate having an affinity for the micelle
surface, concentrating both elements at micelle interface thereby increasing maximum
reaction rate. While these empirically derived conclusions explain the enzyme activity
enhancement in each system unique to their respective study, a universal theory that
might explain why non-ionic surfactants should increase activity in *ALL* of these
studies has not yet been produced.
Theoretical models have also been formulated to mathematically describe, patterns of
enzyme hyperactivity and activity loss as a function of non-ionic surfactant
concentrations, starting with theoretical modeling of enzymes in reverse phase systems.
In 1988, Kabanov and colleagues [10] pointed out that enzyme activity in reverse micelle
systems are maximized at specific water to surfactant ratios. They postulated that reverse-
micelles in specific solvent concentrations are just large enough to house the enzyme,
thereby maximizing the interactions between micelle headgroups and the enzyme surface.
5
Their proposed model predicts enzyme activity as a function of total surface area of
enzyme and micelle interaction.
In parallel to Kabanov’s work, Bru, Sanchez-Ferrer and Garcia-Carmona also outlined a
theoretical model explaining enzyme superactivity in reverse micelle systems [19],
emphasizing the role of three possible pseudo-phases: hydrophobic regions of surfactant
tails, bound water phase at the surfactant heads, and free water which occurs inside the
inverse micelles as they grow larger. By taking into account the kinetics of micelle
aggregation and enzyme / substrate partitioning throughout the three pseudo-phases Bru
and colleagues developed a model to express the concentration of enzyme and substrate
in any of the three phases. This knowledge was coupled with an assumption that catalytic
rates differ in each phase, their final simulation accurately match previous experimental
observations of super-activity and inhibition of enzymes in reverse micelle systems.
Viparelli et al built upon Bru’s work [20], applying the “pseudo-phase approach” to
standard micelles in aqueous solutions. They modeled the steady state concentrations of
enzymes within bound water boundaries at the micelle interface versus the free water
phases in the bulk over a range of surfactant concentrations, and enzyme and substrate
affinities. Also like Kabanov and Bru etal, Viparelli concluded that maximizing enzyme
concentration in the bound water pseudo-phase of the micelle surface plays a key role in
enzyme super-activity.
6
Figure 1-3 - Three phases that enzyme hydrolysis can occur. Bulk (Sf), Bound water interface (Sb) and within the
micelle as described by Viparelli et al.
These theoretical studies attempt to predict enzyme activity in the presence of non-ionic
surfactants and offer a general model that can be applied universally to different enzyme,
substrates and surfactants. However, all the studies mentioned here only provide cursory
physical mechanisms to explain the faster specificity and catalytic constants of the
enzymes within these bound water regions. One of the papers highlights this:
“Of course, there may be other "molecular mechanisms" that can explain the enzyme
catalytic activity as a bell-shaped function of micelle dimensions. Nevertheless, despite
the possible variety and complexity of these mechanisms, for a formal kinetic description
of observed regularities it is sufficient to consider only three types of micelles: "small",
"large" and "optimal". The catalytic activities are different for enzymes solubilized in
micelles of each type: the kcat value is maximal in optimal micelles”
– Kabanov et al. 1988
7
1.21.21.21.2 Gaps in knowledgeGaps in knowledgeGaps in knowledgeGaps in knowledge
In section 1.1 we mentioned a considerable body of work extending into the experimental
and theoretical aspects of this enhancement phenomenon. However, when comparing
them all side-by-side, a considerable gap in knowledge becomes evident. As previously
stated, there is still no consistent explanation for how a seemingly inert colloid surface,
such as at a non-ionic micelle surface, can influence the activity of an enzyme. And while
the theoretical studies offer models that predict how enzyme activity may respond to
surfactants as a function of surfactant concentration, the physical explanation as to why
such an enhancement phenomenon should occur are lacking. This work attempts to fill in
this gap in understanding by studying one aspect of enzyme function that has not been
studied in the aforementioned works: structural dynamics, and how they might be
influenced by interactions with particular surfactants.
The protein structures and structural dynamics play an integral role in the function of
many proteins. This can be clearly seen in enzymes, where conformational changes in the
protein structure are a perquisite to the substrate binding, transition state formation and
energy minimization steps required for enzymatic catalysis. Other proteins that rely on
structure dynamics include: proton pumps, ion transport channels and photosystems.
These functional proteins are becoming increasingly prominent in emerging alternative
energy technologies such as bio-fuel production, bio-solar and enzymatic fuel cells; as
such, they will be exposed to surfaces and systems outside the native cellular
8
environment. Understanding the influence of these various ex vivo conditions on the
structure dynamics and functionality of proteins also becomes an increasingly prominent
priority.
Aside from enzymes, many other proteins rely on conformational changes at various
scales to function. This project will aim to extend the mechanisms elucidated from these
enzyme/surfactant systems, to other protein/colloid interfaces. To do this, we will
construct a model to describe how changes to the structural dynamics of a light
harvesting protein such as bacteriorhodopsin can affect its overall energy transfer
mechanism. Bacteriorhodopsin is a membrane protein found in halophilic bacteria. They
use sunlight to pump protons across the cell membranes to maintain proper chemical
gradient, which is in turn used to create ATP through ATP synthase[21, 22]. The
structure is composed of three protein chains that are embedded in the bacteria
membranes. In each of these chains is the light sensitive chromophore retinol which
undergoes a Cis-Trans conformational change when excited by a photon. This
conformation change induces further structural shifts along the protein chain which
ultimately move protons in a particular direction. This retinal switch occurs in the time
scale of 200-500 femto-seconds while the subsequent structure changes related to proton
pumping occur within picoseconds[23, 24]. X-ray diffraction and NMR demonstrate that
the collective motions during these pumping cycles have an average deviation of 3Å.
Clearly, these mechanisms are fast and precise. It is not inconceivable that attaching these
proteins to inorganic surfaces for use in bio-solar applications, may alter with these
dynamics and hence their electron transfer properties.
9
Figure 1-4 - Top down view of three protein chains which compose of a bacteriorhodopsin unit. One is shaded blue
with a white outline indicating the location of the retinol chromophore.
In addition, changes in protein structure, structural dynamics, and inter-particular forces
are paramount to the stability of many drug formulations, such as vaccine carrier proteins
adsorbed onto a solid, metal phosphate, particle surfaces[25]. There is limited knowledge
on the protein structure, and dynamics of protein-particle, and protein-protein interactions
when absorbed onto a particle surface such as aluminum. Gradual changes in protein
structure as well as its structural dynamics on particle surface may result in shifts in
protein electrostatics, hydrodynamics and steric interactions may lead to protein-protein
attraction. The structural dynamics of these vaccine carrier proteins as well as those of
bacteriorhodopsin at particle surfaces are not well understood, study parallel to studying
protein interactions at micelle interfaces may give us greater insights into protein
behavior at a solid-liquid interface and how forces that arise at these interties affect the
structure, structural dynamics of macro-molecules.
10
1.31.31.31.3 Central gCentral gCentral gCentral goal of oal of oal of oal of dissertationdissertationdissertationdissertation::::
Our goal is to propose one such “molecular mechanism” that would explain this broadly
observed enzyme hyper-activity phenomenon in the presence of particular surfaces
observed in previous literature as well as through our own experiments. In addition, we
aim to study proteins such as CRM-197 diphtheria toxin mutant on the surfaces of hard
colloids such as metal oxide particles. If successful, we believe the lessons learned from
the comparisons of proteins at soft micelle surfaces vs. hard particles can be applied to
any system where functional proteins are utilized in close packed environments as it is in
the fields of renewable energy, waste management, home-personal care and medicine.
Explanations in the works cited above are varied and inconsistent, proposing non-specific
binding, structure preservation and enzyme substrate concentration at the bound water
interface, to explain enzyme hyperactivity while theoretical modeling infers faster Kcat
of the enzyme itself. What has not yet been considered in the realm of surfactant /
enzyme interactions is the role of structural dynamics in enzyme function and how they
can be affected by the bound water at these micelle interfaces. Furthermore, small scale
structure dynamics play a major role in the charge transfer mechanisms of proton
pumping proteins. How these motions are affected by colloid surfaces will also be
investigated.
11
2 Hypothesis and rationale
We have proposed and tested the following hypotheses:
1) The surfactant enhances enzyme activity by improving flexibility of the enzyme
without denaturing or distorting the native structure
2) These flexibility improvements are influenced by micelles that extend the range
of bound water regions around the protein structure which impart less kinetic
energy into the protein structure via thermal fluctuations and shifts in electro-
static bonding constants, minimizing the amplitude and frequency of atomic
collisions within the structure, thereby increasing rate and range of induced
conformation changes.
3) The implications of these outcomes extend to other applications where functional
proteins are used in dense and restricted environments, often coupled directly to a
surface. I.E bio-solar, cellulosic bio-fuels and vaccine applications
12
2.12.12.12.1 Rationale Rationale Rationale Rationale forforforfor HHHHypotheseypotheseypotheseypotheses:s:s:s:
It is well known that proteins are not static structures [26-29] but rather dynamic bodies
that exhibit motions over a wide range of time and length scales (figure 2.1). These
motions fall under three major categories [30]:
1) Fluctuations (i.e. atomic vibrations): These motions are random, very fast and
cover distances less than 0.5Å. The energy for these motions comes from kinetic
energy inherent in the system as a function of temperature. Spatial displacement
range between 0.01 – 1 Å, temporal scale 10-15 to 10-11s
2) Collective motions: This category includes the motions of groups of atoms that
are covalently linked in such a way the group moves as a unit. These collective
motions are further divided into two subcategories: those that occur quickly, but
infrequently, like the flipping of tyrosine ring, and those that occur slowly. The
energy for these transitions is also derived from the thermal energy of the protein.
Spatial displacement between 0.01 to 5.0 Å, temporal scale 10-12 to 10-3 s
3) Induced motions: The third category includes motions that are triggered
conformational changes. These are the motions of individual side-chains or whole
protein sections, which occur as a response to a specific stimulus. The energy for
triggered conformational change comes from specific interactions. 0.5 to >10Å,
temporal scale 10-9 to 103s
13
Figure 2-1 - Range of time scales associated with protein structure dynamics. This study with focus on elastic
fluctuations occurring in the ps range.
The roles of these dynamics are a subject of great interest for elucidating the many
aspects of protein function [31]. Enzyme function is closely linked to the dynamic nature
of its structure. In the late 19th century, it was thought that the highly specific nature of
enzyme / substrate binding predicated that the active site be shaped exactly to the
cleavage site, as a key hole is perfectly shaped for one specific key, otherwise known as
the “lock-and-key” model. However, as characterization techniques improved, this static
view of enzyme mechanic had to be reconsidered. And so in the 50 some years since the
introduction of the lock-and-key model, conceptualization of enzymes and proteins had
shifted away from those of static structures; to instead, as soft assemblies that are
constantly sampling random conformations about an mean structure position [32, 33].
In the 1950’s, the static “lock-and-key” view of enzyme/substrate interactions was
modified into the “induced fit” model, wherein enzyme structures undergo a
conformational change as the active site is reshaped during its interactions with the
14
substrate [27, 34]. This dynamic model of enzyme/substrate interaction opened up
pathways to the elucidation of the various mechanisms employed by enzymes to lower
the barrier energy of chemical reaction: Inducing bond strain, bringing reactive groups
into close proximity or by creating a covalently bonded intermediate complex are several
examples of energy barrier reduction pathways induced by conformation changes [33].
Because of this prominent role that structure dynamics has in substrate binding and
energy barrier minimization, investigation of the relationship between the enzymes’
structural dynamic properties, and their catalytic effectiveness has become an intensely
studied field of research [28, 29, 35-38].
Figure 2-2 - General schematic and description of induced fit model. Active site conformation changes play an
important role in minimizing energy barrier to product formation [33]
Since it has been established that flexibility plays a major role in many protein functions,
this project will focus on studying factors that affect flexibility, particularly those that
stem from interactions with interfaces within the proteins’ local environment. Studying
the link between a protein’s local environment and flexibility has been investigated in the
15
past. Using methods that measure local flexibility about the enzyme active site, coupled
with comparative enzyme kinetics studies, several groups have demonstrated that
solvents play a substantial role in dictating enzyme flexibility [39-42]. These studies
consistently cite solvent interaction with structurally bound water as the root cause of
flexibility change. In parallel, many other studies point to hydration as an integral part of
protein function [43-45] [46], affecting bond distances [47], stabilizing structure,
facilitating protein folding [48] and acting as a lubricant, facilitating conformation
transitions [49], whether or not these same effects can be induced or influenced by
detergent micelles or particles has yet to be considered.
However, many studies have demonstrated that inorganic particles and surfactant
aggregates manipulate bound water dynamics at their interfaces. Recent work done by
Fayer et al utilizes ultra-fast infrared spectroscopy to observe orientation dynamics of
water at the interface of reverse AOT micelles [50, 51]. They found two distinct
reorientation correlation times; slow restructuring is attributed to bound water near the
micelle interface while a significantly faster transitioning phase is assigned to bulk-like
water in the micelle core. By decreasing the size of these micelles from 10 to 1 nm
diameter, Fayer et al. observed a transition between the amounts of this slow interfacial
water relative to fast moving bulk. In micelles less than 3nm in diameter, the interfacial
water reorientation characteristics are dominant. Comparisons were also made between
ionic (AOT) and non-ionic (Igepal) and it was found that the restructuring times are more
influenced by total surface area in the system rather than the chemical composition of the
hydrophilic groups. Given the connection between protein structure dynamics and
16
hydrating water, it is reasonable to hypothesize that those colloid interfaces which do not
adsorb directly to protein structure, can still influence the dynamics of the structure
through bound water interactions at the micelle surface.
While solvent conditions may be considered an “external” effecter on protein structure
dynamics, the protein itself provides its own “internal” mechanism that regulates
flexibility and conformational change. For some time the effects of smaller, pico-second
scale fluctuations of local atomic groups on the range of movement of global structures
has been studied [52, 53]. These thermodynamically driven fluctuations have been shown
to dictate the range and rate of conformation changes [53-55] resulting in a hierarchy of
dynamics. In some cases, these fluctuations are the driving force facilitating the moment
of larger and slower movements, such as the hinge bending seen in enzymes like
adenylate kinase [30, 54]. Conversely, frequent internal collisions of adjacent groups are
responsible for limiting net displacement [56] Hindering overall flexibility of the overall
structure [57, 58]. On the importance of dynamics on function [59]:
“Understanding the role of enzyme dynamics in the chemical step is both important and
complex. (Garcia-Viloca etal) describes possible equilibrium and nonequilibrium
contributions and discusses motions of the enzyme that can lower the free energy of
activation of enzyme reactions. Crossing of the transition state region is usually very fast
(in the fs to ps range) in enzyme reactions, and the reaction seems slow only because the
activation barrier makes it improbable to reach the transition state. Thus, fs to ps
motions must play a role, although whether they contribute to the faster rate of the
reaction in the enzyme versus that in solution has yet to be determined.”
- Karplus (2010)
17
Our initial experiments with subtilisin protease have demonstrated greater enzyme
activity in solution with akyl polyglucoside surfactant, a phenomenon that has been seen
in multiple studies in the past decade. Surfactants already play a major role in protein
stability, crystallization and solubilization[60, 61], but the possibility of surfactants
improving enzymatic activity through modification of structural dynamics has not yet
been considered. This study intends to elucidate this connection. However it will not be
enough to only demonstrate a correlation between flexibility and activity in the presence
of surfactants when the underlying mechanism of surfactant / enzyme structure remains
unknown. Various studies provide clues to the nature of these interactions. One likely
mechanism is through the protein structure hydration, which affects the overall free
energy of the structure[62], lubricating internal motions and directing folding pathways.
Measurements of protein structural dynamics in different solvent environments also
indicate changes occur because of bound water being displaced from adsorption sites -
this denotes the importance of bound water to structure dynamics [39, 47]. As mentioned
above, the surfaces of micelles play a large role in influencing the volume of bound water
in the system[42, 50], as well as the rate at which hydrogen bonds are exchanged[47].
This project intends to study these interfacial bound water dynamics as the primary
mechanism by which non-ionic surfactant interact with protein structures, thereby
influencing structural dynamics and dependent functions, such as enzyme activity. Since
ceramic and oxide nanoparticles are also known to attract bound water or affect water
structure at interfaces [63, 64], the structure dynamics of various functional proteins,
including bacteriorhodopsin, subtilisin enzyme and BSA have also been studied at hard
particle surfaces in parallel with the surfactant studies.
18
3 Research plan and objectives
Overall Goal: To demonstrate that proteins at a non-ionic interface are not completely
free of influence from its local environment, and that these interactions play can still play
a major role in affecting the steady-state dynamics of the proteins structure and thereby
it’s function. In addition this project aims to propose a mechanism by which non-ionic
surfaces may affect protein structural dynamics.
3.13.13.13.1 SubSubSubSub----objectives: objectives: objectives: objectives:
1. Select a system where enzyme activity / kinetics parameters are enhanced or retarded
by surfactant interactions
2. Determine the link between changes in local and/or global structure flexibility and
enzyme activity relative to surfactant concentration
2.1. Study the possible link of micelle structure vs. surfactant concentration
2.2. Correlate changes in protein structural flexibility to changes to bound water
dynamics with increasing surfactant concentration
3. Find parallel changes in function and structure dynamics of proteins
3.1. Demonstrate that changes in steady state conformation fluctuations and overall
structural mobility can be influenced by solid particle surfaces as well
4. Propose mechanism by which colloid surfaces can affect conformation fluctuations
without denaturing protein structure
19
4.1. Propose model which describes FRET efficacy as a function of protein flexibility
and fluctuation for use in bio-solar applications
Our initial goal was to elucidate the mechanism dictating the “hyperactivity”
phenomenon seen in enzymes in non-ionic surfactant solutions. We hypothesized that
changes in the enzyme structural flexibility are induced by bound water interactions
found at the micelle/water interface. From there, we extend the findings from this
particular surfactant / enzyme system to develop a holistic understanding of the
interactions between colloid surfaces and protein structural dynamics. To test our
hypotheses the project was divided into a series of four objectives. The first was to
determine a model system with which to test enzyme/surfactant interactions. This was
done primarily through enzyme activity and kinetics assays. The second objective was to
measure the structural mobility of the enzyme and observe how these changes with
respect to surfactant structure and concentration. A correlation was found between these
enzyme flexibilities and enzyme kinetics properties observed from the activity
measurements found while fulfilling our first objective. Since this correlation between
flexibility and surfactant concentration was found, we proceeded to our final objective
which was to describe a general mechanism by which colloid surfaces influence protein
structural dynamics. Our hypothesis, based on theoretical work of Bru, Viparelli and
Kabanov and their respective groups cited that the bound water region at the micelle
interface is instrumental in this hyperactivity phenomenon. Work to further study the
effects of bound water structure on enzyme activity and flexibility are now on going,
however, preliminary tests of enzyme activity in Hofmeister series of ions confirm the
20
connection. Additionally, cooperative work with the University of Bergen is underway to
use NMR to measure which specific amino acid residues are affected most.
21
4 Methods and Materials
The methods described in this section are divided into the sub-objective for which they
were designed to achieve. The first objective is to find suitable model enzyme and
surfactant systems which induce faster enzyme activities and compare them to those that
do not. The primary technical challenge of this objective will be to devise or adapt a
suitable method for quantifying enzyme activity. The second objective is to elucidate the
interactions between surfactant and enzymes, or the physical characteristics of the
surfactant-enzyme aggregates that form upon mixing the two components. The third
objective aims to investigate the role of the enzyme’s structural mobility in determining
activity and how surfactants maybe affecting this structural dynamic.
4.14.14.14.1 Objective 1 Objective 1 Objective 1 Objective 1 ---- Find suitable model system to test enzyme hyperactivity via Find suitable model system to test enzyme hyperactivity via Find suitable model system to test enzyme hyperactivity via Find suitable model system to test enzyme hyperactivity via
interaction with surfactants interaction with surfactants interaction with surfactants interaction with surfactants
The first objective of this study was to select enzyme and surfactant pairings that will
result in measureable changes in catalytic activity. For this purpose two enzymes,
subtilisin protease and horse radish peroxidase (HRP) were studied in the presence of 6
surfactants: two commonly used anionic surfactants, sodium dodecyl sulfate and linear
alkylbenze sulfonate, and four non-ionic surfactants, alkyl polyglucoside, dodecyl
maltoside, Polyoxyethylene (4) lauryl ether (Brij-30) and its larger headgroup
22
counterpart, Polyoxyethylene (23) lauryl ether. The rationale for selecting these enzymes
and surfactants is explained in the following sub-sections.
4.1.14.1.14.1.14.1.1 Enzyme selectionEnzyme selectionEnzyme selectionEnzyme selection
The enzymes subtilisin and HRP were both chosen because their structural and functional
characteristics have been well studied and documented. Subtilisin proteases are
commonly used in commercial and industrial applications. Likewise, horseradish
peroxidase is commonly used for bio-chemical research applications. Furthermore, both
of these enzymes are readily available in large, relatively pure quantities. The subtilisin
families of enzymes are a non-specific protease initially obtained from the Bacillus
Subtilis bacteria strains and are commonly used for a large variety of application [65],
including as a components in detergent formulation to aid in protein based stain removal.
Over the last decades, several variations of the subtilisin protein have been isolated.
Subitilisin Carlsberg refers to enzyme derived from Bacillus Subtilis bacteria. Subtilisin
BPN’ originates from the B. Subtilis strain N’ and subtilisin Novo refers to the enzyme
from the Novo pharmaceuticals company. For this study, subtilisin Carlsberg is
predominately used and will be referred to from here on in as subtilisin unless specified
otherwise.
23
Subtilisin Carlsberg has a total molecular weight of 27,287 in the form of a single peptide
chain devoid of cysteine or cysteine, phosphorous and metals and an isoelectric point of
9.4 kpH. The tertiary structure is approximately 30% alpha-helical resulting in a protein
that is close to spherical and approximately 42Å in diameter[66]. The catalytic center is a
Asp-His-Ser catalytic triad with the nucleophilic serine-OH group located in a shallow
groove in the enzyme surface[40]. The subtilisin used for this study was all products
ordered from Sigma Aldrich in different forms. Initially P4860 was used, this is a
subtilisin product produced by Novozymes under the trademarked name of Alcalase.
However, since the exact composition of the proteins and suspension additives were not
known, P4860 was later replaced with P5459, another subtilisin product suspended in a
polyethylene glycol solution. For the time resolved fluorescence experiments P5459 was
initially used, but the polyethylene was thought to interfere with the probe attachment
process, so P5380 (Sigma-Aldrich), a lypholized subtilisin, was used predominately for
those experiments.
Horseradish peroxidase is a glycoprotein isolated from plants and is commonly used in
biochemistry applications. The majority of reactions catalyzed by HRP can be
summarized as such:
H2O2 + 2AH2 � 2H2O + 2AH●
24
Where AH2 is a substrate, typical aromatic phenols, phenolic acid, amines or sulfonates
and AH● is the radical product.
There are several isoforms of HRP[60], but the most commonly studied is the ‘C’ variant
derived from the horseradish plant. The primary structure of HRP consists of 308 amino
acid residues forming a single polypeptide chain. Unlike subtilisin, HRP incorporates
carbohydrates and metals into its structure, as well as multiple cysteine disulfide bridges
to maintain structure. Four disulfide bridges are found in HRP between eight cysteine
residues. Three metal centers are also present in the HRP structure, an iron (III)
protoporphyrin (heme) group located near the center between two clearly defined
domains of the structure. On either side of the heme group are two calcium ion binding
sites, one distal and the other proximal to the heme group. Both of these metal groups are
essential to the peroxidase structural integrity while the heme group is also central to the
catalytic mechanism. The weight of the polypeptide chain is 33,890 Da, the heme group
is 572 Da, and the calcium ions 80 Da. The total weight of HRP C is approximately 42
kDa assuming 20% contribution from the carbohydrate moieties[60].
4.1.24.1.24.1.24.1.2 SurfactantsSurfactantsSurfactantsSurfactants
Several surfactants have been reported in literature to improve enzyme activity. The large
majority of these are non-ionic structures including: Tween-20, Tween-80, and Brij-30.
In our own preliminary experiments, we have identified improved subtilisin activity in
solutions of alkyl polyglucosides (APG). For this study the synergistic properties of
APG’s as well as Brij-30 on enzymes were further studied. Dodecyl maltosides and Brij-
25
35 were also added to compare the effects of long chain ethoxylate chains of varying
lengths vs. the sugar groups of APG and DM on the enzyme activity.
Figure 4-1 - Dodecyl maltoside
The alkyl polyglucosides are a class of non-ionic surfactants derived from sugar sources
and usually in the form of a mixture consisting of a range of linear tail chain lengths and
sugar head group oligomer size. The APG used in this study was a commercial product
supplied by the Cognis specialty chemicals company with a tail chain length between 12
and 14 carbons and head group sizes between one and three sugar groups. However the
exact composition of tail and head group size distributions are not known, for this reason,
beta-D-maltoside a pure, single sugar, is also used in this study to better control the
surfactant composition in solution.
The polyethylene glycol dodecyl ethers known by their trade name, Brij, has been used in
several of the previous papers that have reported increased enzyme activity in non-ionic
surfactants. Brij-30 (polyoxyethylene (4) lauryl ether was most commonly cited of the
Brij surfactants and is used in this study as well. The effects of larger head group sizes
26
were also studied with Brij-35, which also features a linear polyoxyethylene headgroup,
but several times longer with 23 oxyethylene groups. In parallel to these four non-ionic
surfactants, two ionic surfactants were tested to compare the effects of charged
surfactants on protein structure. Sodium dodecyl sulfate (SDS) and linear alkylbenzne
sulfonate (LAS) were selected for this reason.
4.1.34.1.34.1.34.1.3 Subtilisin assay:Subtilisin assay:Subtilisin assay:Subtilisin assay:
Subtilisin activity was measured primarily in two ways, product formation assays and
initial rate activities. Product formation assays were measured by the amount of tyrosine
liberated from casein substrate by subtilisin in buffer and various surfactants. The
liberated tyrosine reacts with Folin-Ciocalteu reagent, turning it from from yellow to
blue. The color change is measured with UV-Vis and presented below normalized versus
the enzyme in buffer only.
27
Figure 4-2 - Example of colormetric activity assay
4.1.44.1.44.1.44.1.4 Subtilisin Subtilisin Subtilisin Subtilisin initial rate activities and kineticinitial rate activities and kineticinitial rate activities and kineticinitial rate activities and kineticssss
The second method of quantifying enzyme catalyzed proteolysis rates was done via a
method modified from Broos etal [67], which utilizes fluorescein labeled casein as a
substrate of proteolysis. In its native state the fluorophores are in very close proximity
and their fluorescence is quenched by adjacent fluorescein. Changes in FITC
concentration, which is released from the casein substrate during the course of the
reaction, is measured as a function of time during the relation. Fluorescence intensity is
monitored with Horriba Yvoin fluorolog -3 fluorometer. The reaction takes place in a 3ml
quartz cuvette (1cm path length) where an aliquot of the enzyme (1.0ml) is added to the
FTIC casein suspension (2.0ml) under constant stirring. Alternately, the reaction is
28
started in 1.5 ml plastic vials by adding 250µl of enzyme to 500µl of substrate, shaking
and quickly transferring to a 0.56 ml reduced volume quartz cuvette with a 4mm path
length (Starna cell). A series of emission intensity measurements at 519nm is taken once
every 2 seconds for 10 minutes, starting shortly after the enzyme addition. This reaction
is measured at 22 degrees C.
Figure 4-3 - Schematic illustrating liberation of free fluoresein probe from labled casein
4.1.54.1.54.1.54.1.5 HRP Kinetics: HRP Kinetics: HRP Kinetics: HRP Kinetics:
Kinetics rates were obtained from the aminoantipyrine method [68], spectroscopically
measuring the increase of absorbance at 510nm resulting from the formation of the
colored complex AMPH. Two solutions are prepared: one consisting of 2.5 mM
aminoantipyrine and 170 mM phenol in DI water and 1.7 mM Hydrogen peroxide also
diluted in DI water. 0.01 mg/ml. A 1:1 mixture of the two reagents were prepared shortly
before a small aliquot of HRP in buffer is added to the substrate mixture. The final
concentration of HRP was 0.01 mg/ml. The resulting colorimetric evolution was
monitored by UV-Vis spectrometer at 510nm.
Fluorescein isothiocyanate
29
4.1.64.1.64.1.64.1.6 Subtilisin assay side note: Behavior of Fluorescein Probe (optimizing subtilisin Behavior of Fluorescein Probe (optimizing subtilisin Behavior of Fluorescein Probe (optimizing subtilisin Behavior of Fluorescein Probe (optimizing subtilisin
assay method)assay method)assay method)assay method)
Fluorescein is a common fluorophore dye used in a wide variety of applications, ranging
from biology and geology to medicine and oil recovery. It has also been critical to our
enzyme kinetics experiments done up until this point, as activity is actually the measure
the release of fluorescein as an indicator of enzyme activity. Previously, we have reported
the results as changes in intensity over time, but this is only relative information. Using a
concentration standard these relative intensity results can be correlated to quantitative
“freed” flourescein concentrations.
Furthermore, it is important to find the concentration parameters in which fluorescence
intensity and probe concentration deviates from Beer-Lamberts law. An additional
concern when using the fluorescien probe is that they can interact with one another,
quenching emission and giving rise to another limitation to the experimentally observable
concentration range. These properties need to be carefully considered when selecting
experimental concentration ranges and interpreting the intensity results.
MethodsMethodsMethodsMethods:
0.33 g of fluorescein was added to 50mM potassium phosphate buffer with pH adjusted
to 7.5. Mixture stirred for 60 minutes allowing probe to dissolve and saturate solvent, any
after an hour any undissolved fluorescein was centrifuged, filtered out, dried and weighed
30
to determine actual probe solution concentration. 0.18 grams of fluorescein probe
dissolved into the buffer, making a max solution concentration of 25mM.
This concentrated solution was measured with steady state fluorescence. Excited by 490
nm light and emission spectrum between 500 and 600 nm was measured. Lamp and
detector slit widths where 0.5 and 1 mm respectively. Each successive measurement was
diluted to half the concentration of the previous and measured again. The resulting
spectra are shown below.
Figure 4-4 - Fluoresence spectroscopy of fluorescein fluorophore dilution series
0
100000
200000
300000
400000
500000
600000
500 520 540 560 580 600
Inte
nsi
ty
wavelength
25
12.5
6.25
3.125
1.5625
0.78125
0.390625
0.1953125
mMMax intensity at
approximately 100x dilution
31
These over lapping plots are hard to visually deconstruct two crucial elements are
immediately observable: 1) The highest concentration does NOT result in greatest
emission intensity and the 2) fluorescence peak is shifting with concentration.
First we plot changes in plot intensity with concentration. Below 0.1 mM, the peak
emission wavelength is approximately at 512 nm, the value reported in literature. At
higher concentrations, however, we see an exponential shift in the peak positions which
strongly indicate probe excimer formation, resulting in energy transfer between adjacent
probes producing emitted light of lower energy.
Figure 4-5 - Fluorescein peak wavelength vs. concentration
Next, we look at the peak intensities with concentration: As mentioned before, the
greatest emission intensity occurs at 0.2 mM and quickly declines afterwards. This
32
concentration threshold also corresponds to the point where the peak intensities begin to
drop off. These results suggest that, not only are excimers formed after 0.2 mM, but they
also self-quench.
Figure 4-6 - Fluorescein peak intensity vs. Concentration
The figure below indicates that the linear concentration/intensity relationship breaks
down above 0.075mM, hence any measurements above 300k cps should be reevaluated
as it may provide a misleading measurement of actual probe concentration.
33
Figure 4-7 - Divergence from intensity / concentration linarity starting at 0.1 mM of flourescein
In conclusion: These properties are important to know for using fluorescein probe.
Concentrations should be kept below the 0.075 mM range to avoid quenching effects and
peak position shifts.
Obtaining reliable enzyme kinetics parameters such as binding constants and max
reaction rates requires repeated acquisition of data points over a range of substrate
concentrations. In addition, we are studying the effect of surfactants on these kinetics
parameters, which adds an additional dimension to our work on enzyme kinetics.
Previously, a Novozymes product provided from Sigma Aldrich (P5860) was used. This
is a serine based subtilisin listed under the trade name Esperase. However, because it is a
commercial product the suspension medium and total enzyme concentration not
disclosed.
34
To eliminate possible interactions with unknown stabilizing agents, enzyme kinetic
experiments are repeated with lyophilized Subtilisin Carslberg protease (p5380) of
known wt% and activity. Since the product is received as a dry powder, they must be
freshly dissolved in buffer solution before each experiment. Steady state fluorometery is
used to monitor enzyme catalyzed proteolysis of fluorescein labeled casein insitu. While
bound to the casein substrate, the probe’s fluorescence is quenched due to the close
proximity of adjacently bound probe molecues. As the protease acts on the substrate,
probe is released into solution and can freely fluoresce. The resulting increase in emission
intensity can be repeatedly measured over the course of the reaction to measure the
progress of protein hydrolysis.
Initially, the reaction profiles of subtilisin catalyzed casesin hydrolysis are measured for
casein concentrations from 0.001 to 0.1 wt%
Figure 4-8 - Reaction profiles of subtilisin catalyzed proteolysis (0.001 to 0.1 mM substrate concentrations)
0
100000
200000
300000
400000
500000
600000
0 50 100 150 200 250 300 350
Inte
nsi
ty (
CP
S)
Time (s)
.001
.0025
.005
.01
.025
.05
.1.001
.0025
.005
.01
.025
.05
.1
35
The reaction profiles indicate more products being formed up until 0.025 wt%. Above
this concentration the amount of product rapidly drops off. A comparison of these final
intensities can be seen in the figure 4.11. This apparent loss of enzyme function might be
mistaken for an inhibition mechanism that would require additional study. However, our
previous experiments with the fluorescein probe indicate that we have far exceeded the
ideal concentration parameters (not to exceed 300k CPS intensity), and that this loss in
intensity is trivial, due to an over saturation of liberated fluorescein probe and not
indicative of enzyme reaction kinetics.
Figure 4-9 - End of reaction fluoresence intensity vs. substrate concentration (.0001 to 0.1 mM)
0
100000
200000
300000
400000
500000
600000
0 0.02 0.04 0.06 0.08 0.1 0.12
Inte
nsi
ty @
10
min
ute
s
Substrate wt %
End reaction intensity vs Substrate conc.
36
To correct this, the substrate concentration window was adjusted to less than 0.025 wt%
or approximately 320nM of casein substrate. This would ensure that the amount of
liberated fluorescein will not exceed a concentration where probe aggregates can form.
The amount of enzyme used is also reduced by 10 times, to ensure that substrate
saturation, or Vmax, can be achieved.
Figure 4-10 - Casein hydrolysis reaction profile (0.0005 to 0.025 mM substrate concentration)
0
50000
100000
150000
200000
250000
300000
350000
400000
0 100 200 300 400
Inte
nsi
ty (
CP
S)
Time (s)
.0005
0.001
0.0025
0.005
0.01
.025
A
37
Figure 4-11 - End of reaction fluoresence intensity (low substrate concentration 0.0005 to 0.025 mM)
Above are reaction rate profiles of the reduced substrate concentration range. To the left
is a plot of the final intensities after the 10 minute reaction.
In this optimized concentration window, we see a linear correlation between substrate
concentration and total amount of probe freed after the 10minute reaction for
concentrations below 0.01 wt%. The highest concentration deviates from this linearity;
however, since the intensity remains below 300k cps, this deviation can be attributed to
enzyme saturation phenomenon and not an aberration of probe interactions. This
kinetics experiment was repeated three times to establish an error range. Enzyme kinetics
0
50000
100000
150000
200000
250000
300000
350000
400000
0 0.005 0.01 0.015 0.02 0.025 0.03
Inte
nsi
ty (
CP
S)
Substrate wt %
B
38
parameters, Vmax and Km are needed to comparatively study enzyme efficiency. Kcat,
or the turn over number, is derived from Vmax, while Kcat/Km is a quantification of the
enzyme’s overall efficiency. These parameters are determined by plotting the initial
steady state portion of the reaction profile (change in intensity over the first 20 seconds)
as a function of substrate concentration. Below is such a plot derived from the reaction
profiles of subtilisin carlsburg illustrated by figure 4.13 on page 44.
Figure 4-12 - Set of initial rates plotted against substrate concentration drawn with error distribution.
39
Fitting the Michaelis-Menten equation to the data gives us the kinetics parameters Vmax
and Km, (Vmax = .00015 mM/sec and Km= 45 um.) for subtilisin Carlsberg in buffer
solution. Moving forward we repeated the same experiment in different surfactant
solution and inorganic particle suspensions of various concentrations. In the meantime
the reaction rates of subtilisin in surfactant have been measured at just one substrate
concentration in the presence of varying concentrations of Brij-30. The substrate
concentration tested here is relatively high, (160 um) so the reaction rates can be
indicative of changes in Vmax but a range of substrate concentrations need to be tested to
ascertain Km.
40
4.24.24.24.2 OOOObjective 2: bjective 2: bjective 2: bjective 2: Characterizing surfactant / particle properties in protein Characterizing surfactant / particle properties in protein Characterizing surfactant / particle properties in protein Characterizing surfactant / particle properties in protein
solutionssolutionssolutionssolutions
In addition to studying how surfactants affect protein conformation, conformation
dynamics and function, we are also investigating how additions of proteins ultimately
affect surfactant solutions properties. Because proteins are essentially bio-polymers, with
domains of varying hydrophobicity and polarity, they are expected to interact with
surfactants much in the way any polymer would. Some examples may include, shifting
CMC, affecting micelle shape and size or affecting foaming properties.
To achieve this particular objective this study will incorporate dynamic light scattering
(DLS) for particle size measurements and surface tension to measure surface activity of
surfactant/protein mixtures.
4.2.14.2.14.2.14.2.1 DLSDLSDLSDLS
Dynamic light scattering measurements were taken of surfactant solutions in 200mM
phosphate buffer with a Malvern Zetasizer Nano ZS90 and Brookhaven instrument
consisting of a 324 nm laser coupled with a BI-9000AT autocorrelator. DLS
measurements are often done to characterize the size and distribution of sub-micrometer
sized particles. The technique involves shining a monochromatic and coherent source of
light onto a sample suspension. The resulting Reyleigh scattering of the photons are
dependent on the size and potion of the particles off which the light is scattered. The
scattered photons then undergo either constructive or destructive interference. The
41
fluctuation of this pattern contains information of the Brownian motions of the particles.
Extrapolation of the particle size and distribution is commonly done with either the
Malvern or Brooklyn haven software utilizing the non-negative least squares (NNLS) or
CONTIN methods. Dynamic light scattering techniques have several limitations,
however; high particle concentrations will result in indirect scattering that cannot be
distinguished from the initial source, requiring DLS be done on relatively dilute
suspensions. Also our DLS instruments are they are can only assume particles of
spherical shapes.
Rayleigh scattering techniques are useful for particles smaller than the wavelength of
visible light, approximately 50 nm and smaller. The resulting scattered light is isotropic,
i.e. uniform in intensity regardless of the angle from the incident beam. This is feasible
for this system of micelles and proteins if there is no interaction between each other and
the colloids remain dispersed. However, if there is any further aggregation between the
surfactants and/or protein, then the final particle size may easily cause the scattering
behavior to transition out of the Rayleigh scattering size range into particles sizes that
may inelastically scatter radiation. In this case, Mie light scattering theory is considered.
Mie theory is a more general treatment of Rayleigh theory and covers larger particles
which absorb light and would maybe of more practical interests. For this application, the
detector position can be rotated about the sample to gather anisotropic scattering
intensities.
42
4.2.24.2.24.2.24.2.2 Surface tensionSurface tensionSurface tensionSurface tension
Surface tension is a phenomenon that occurs at the interphase between two continuous
bulk phases. The anisotropic forces that occur as a result of this transition is commonly
expressed in units of area (dynes / cm2) or as a linear form (mN/m). Surface tension
applies to all phases of matter, but most commonly studied at liquid-gas interface or the
interfaces of non-miscible liquids. Surface tension measurements are especially relevant
in studying detergent applications because of an amphiphillic nature of detergents. These
molecules will find their way to the water-air or water-oil interface of the solution they
are in, lowering the surface tension. This trend of decreasing surface tension vs.
increasing concentration continues until the interface is saturated with surfactant
molecules. At this point, the surface tension ceases to decrease, as the excess detergent
molecules can no longer fit at the interface. Instead, they begin to form aggregates. These
aggregates are referred to as “micelles”, and form in such a way to minimize exposure of
the hydrophobic tails with the aqueous solvent. The most commonly represented form of
such aggregates is in monolayer spheres, consisting of a hydrophobic tail core, and
hydrophilic headgroups in the solvent. This is not always the case, and structures such as
elongated rods, bilayer sheets, or hollow vesicles are often formed.
The Whilhelmy plate technique is used to measure the surface tension of surfactant
solutions as a function of concentration to determine the point at which surface tension no
longer decreases with concentration, indicating the critical micelle concentration (CMC).
This is particularly useful for this project in determining if the non-ionic surfactants are
affecting enzyme activity through free monomers or in their aggregated micelle form.
43
4.2.34.2.34.2.34.2.3 Electron paramagnetic resonance Electron paramagnetic resonance Electron paramagnetic resonance Electron paramagnetic resonance
Electron paramagnetic resonance (EPR), also known as Electron spin resonance (ESR) is
a probe technique that takes advantage of the Zeeman Effect wherein an unpaired
electron is oriented by a magnetic field. In measuring surfactant aggregation behavior,
ESR probes can provide information on the degree of molecular crowding about its
immediate environment.
Figure 4-13 - Schematic of Zeeman Effect energy splitting
The ESR absorption characteristic is heavily influenced by the unpaired electrons
environment and is employed as a probe to measure complex molecular structures and
aggregates.
The interactions of the probe influence the shape and intensities of an ESR spectral line
which can yield information about local viscosity adjacent to the probe and quantified in
44
the form of a rotational correlation time. Common classes of probes used in surfactant
and protein research are stabilized nitroxyl radical compounds functionalized to form
specific covalent bonds, or attached along the carbon chains of stearic acid. Common
examples of the latter are 16 and 5 Doxyl stearic acids (DSA), where a 4,4-dimethyl-3-
oxazolidinyloxy group is attached to either to 16th or 5th carbon of the hydrophobic chain.
A typical ESR spectrum of 16-DSA is shown in figure 3a
The probe mobility can be determined from this spectrum by comparing the ratios of
hyperfine splitting peaks. This can be calculated manually from the equations:
Figure 4-14 - Typical nitroxyl group ESR spectrum and peak parameters used to calculate rotational correlation
times
45
These functionalized surfactant probes are extensively used in many areas of colloid
research as in the study of self-assembled co-polymers [69] and SDS aggregates on
alumina surfaces [70]. Probe mobility is a common indicator of surfactant packing or
protein flexibility in these studies. This is quantified by comparing the hyperfine splitting
peaks of the nitrogen radical. Intensities of these peaks vary as the head group is hindered
by adjacent compounds, the degree of interaction can quantified as a rotation correlation
time by the relationship given in figure 3b and further defined by equation 3b. This
technique is also applied to measuring protein structure flexibility[35].
EPR spectrum of 0.1, 2.5, 5.0, and 20 mM surfactant (APG or DM) was measured in 200
mM sodium phosphate buffer as well as with the addition of 0.01 mg/ml HRP in also in
buffer solution. 16-DSA probe was introduced to the system at a final concentration of
1.0x10-5 M. For this investigation, a Bruker EMX-EPR spectrometer that operated at a
microwave field of 9.5 GHz (X-band). Dynamic light scattering measurements were
taken of surfactant solutions in 200mM phosphate buffer with a Malvern Zetasizer Nano
ZS90. Spectrum were analyzed with “Topspin” software installed onto Matlab. Initial
parameters: a, g, tcorr. Line fitting was done using Nelder/Mead simplex until at least .05
RMSD was achieved.
46
4.34.34.34.3 Objective 3Objective 3Objective 3Objective 3: : : : Measuring structural dynamics of model proteinsMeasuring structural dynamics of model proteinsMeasuring structural dynamics of model proteinsMeasuring structural dynamics of model proteins
An important aspect of this project focuses on measuring the “flexibility”, or rate and
range of internal motion, if it is affected by the selected surfactants, and whether these
changes in dynamics correlate with changes in enzyme kinetics explored in the first
objective. However, these must be measured in various surfactant solutions, so an
analytical method capable of measuring these dynamic structure properties in an aqueous
solution.
4.3.14.3.14.3.14.3.1 Life time Life time Life time Life time fluorescencefluorescencefluorescencefluorescence anisotropyanisotropyanisotropyanisotropy
Previous works have used electron spin resonance or time-resolved fluorescence to
measure subtilisin flexibility in the presence of different organic solvents [17, 43, 45, 67]
This study focuses on time resolved florescence decay, measuring protein flexibility
through probe anisotropy and intrinsic fluorescence decay.
Figure 4-15 - Dansyl fluoride structure
Anisotropic fluorescence decay is a probe mediated technique wherein a fluorophore
probe is attached to a polymer structure; in this case the enzyme. Irradiated by vertically
polarized light the directional (anisotropy) aspect of the fluorescent emission will decay
over time as the probe randomly spins, reorienting in space. This is typically done by
47
comparing the fluorescence decay of vertically polarized component of this emission
versus the horizontal[71].
Equation 4-1 anisotropy as a function of vertical and horizontally polarized fluorescence decay
Where VV(t) and VH(t) denote fluorescence decay kinetics of the sample measured using
vertically polarized light from the excitation source and emission component is vertically
(V) or horizontally (H) polarized. VV(t) and VH(t) are often referred to as parallel and
perpendicular polarized decays. The measure of anisotropy is the ratio between the
difference between parallel and perpendicular decays over the sum of all emission.
Figure 4-16 - example of parallel (VV) and perpendicular(VH) fluorescence decays with instrument response
function (IRF)
48
Figure 4-17 - anisotropy (r(t)) derived from VV(t) and VH(t). [62]
The anisotropic decay kinetics can be attributed to rotation diffusion of a spherical
particle and described as :
Equation 4-2 - anisotropy as a function of initial anisotropy, time and decay rate
Ro is the initial anisotropy value, t is elapsed time and τr is the decay correlation time. In
the case of labeled proteins, this decay function is primarily the sum of two rotation
mechanisms: A fast decay, originating from the spin of the probe about its covalent bond
to the protein structure. A much slower rotation occurs as the whole protein tumbles in
solution. The fast rotation is indicative of the flexibility about the probes immediate area
and therefore of most interest in this study.
49
This study emulates a procedure described by Broos and coworkers[72] where the
flexibility of subtilisin carlsburg is measured in different solvent environments via
fluorescence anisotropy decay. A major advantage of this procedure is in the dansyl
fluoride probe used; it specifically binds to the serine residue at the enzyme active site
allowing observation of changes to the active site’s flexibility. This specific interaction is
useful as active-site conformational changes can play a major role in enzyme activity.
Some additional factors that need to be considered include the treatment of anisotropy
decay time model [71, 73] and geometric considerations, such as the probe rotation in a
confined angle [74, 75].
From this portion of the work we explored the possibility that non-ionic surfactants also
affects enzyme structure dynamics, which opens the possibility of a new mechanism by
which enzymes are regulated by surrounding structures. Secondly, we investigated a
correlation between enzyme activity and flexibility as a function of colloid concentration.
4.3.24.3.24.3.24.3.2 Life time fluorescence decayLife time fluorescence decayLife time fluorescence decayLife time fluorescence decay
Fluorescence decay time of the single tryptophan residue in the HRP structure can also be
used as a qualitative measure of this protein’s overall flexibility. This decay is
characterized by the summation of three distinct decay parameters, dominated by a
picosecond-scale lifetime component, with small contributions from two other lifetime
components in the nanosecond range. The fastest decay time is the result of energy
transfer via Förster theory (FRET) from the excited Trp to the heme moiety located at the
50
enzyme active center[76, 77]. FRET transfer efficiencies are highly dependent on the
distances from acceptor and donor with transfers typically occurring between 10-70Å.
Figure 4-18 - Structure of horseradish peroxidase. Trp residue and heme group are highlighted.
� = 11 + ��|� 6
Equation 4-3 - relationship between energy transfer efficency and acceptor donor distance
Within this spatial range, energy transfer efficiency, E, is dictated by proportional to r,
the distance between the excited donor and acceptor, and by the ‘Forster distance’, Ro,
the point at which the transfer efficiency is 50%. Ro itself is defined by
51
Equation 4-4 Determination of Forster distance
Where Qo is the fluorescence quantum yield of the donor in the absence of an acceptor.
k2 is the dipole orientation factor, n is the refractive index of the medium and Na is
Avogadros’ number. The energy transfer efficiency can be related to fluorescence
lifetimes via the following relationship:
Equation 4-5 - Forster resonance energy transfer efficiency (E) as a function of fluorescence decay times
Where τ’D and τD are the donor decay life times with and without acceptors, respectively.
Many studies have taken advantage of the relationship between FRET efficiency and
decay times, by using time resolved fluorescence spectroscopy as a method of monitoring
protein unfolding. However, in this work, we are interested in changes in steady state
protein dynamics such as flexibility and fluctuations, on which life time fluorescence can
also provide information.
General thermodynamic expressions for energy and volume fluctuations about the mean
have been proposed over the past 50 years since revealing the dynamic nature of proteins,
52
such as the following models for the distribution functions of internal energy and
volumes of a protein reported by A. Cooper in 1976:
� ������� = ������
Equation 4-6 Discrete changes in protein's free energy
�������� = �����
Equation 4-7 Thermodynamically driven changes in protein volume
Where m and V are the mass and volume of the system, respectively, Cv is the heat
capacity at constant volume, �� the isothermal bulk compressibility, T the absolute
temperature and k is the Boltzmann’s constant. The change in protein radius resulting
from global protein structure fluctuations may be large enough to fall into the range that
effects FRET efficiency, and therefore decay time resulting from FRET mechanism, as is
the case of intrinsic fluorescence decay of HRP, which is predominately driven by the
energy transfer from Trp to the heme group of the enzyme active site. The range and rate
of atomic fluctuations will drive the distribution function of distances between Trp and
Heme at any given time. A more flexible structure with greater degrees of atomic
mobility will result in a spread in this distribution function. While the mean distance
between groups may not change from more a more “flexible” protein structure, the
53
overall time two residue will spend in closer proximity will increase, decreasing the
overall Trp fluorescence decay time with increasing protein local structure mobility.
While donor/acceptor distance, r, is a commonly studied parameter using time resolved
fluorescence. Ro is also an important, and at times controversial, parameter in measuring
FRET efficiency between donors and acceptors within proteins. A degree of uncertainty
arises when considering the dipole orientation factor, κ2. If the donor and acceptor are
both freely and isotropically rotating, then κ2 is averaged to 2/3. However, this is unlikely
to be the case in horseradish peroxidase, where both the donor (tryptophan) and acceptor
(heme group) are intrinsically bound within the protein structure. Molecular dynamic
simulation studies of intrinsic Trp in tetracycline repressor indicated the possible rotation
modes of the fluorescence donor and calculated a κ2 value of 0.36. A lower κ2 value of
hindered acceptors and donor would ultimately lower the transfer efficiency of the
Forster radiative energy transfer. Alternatively, should the flexibility of the protein
increase so will the mobility of the fluorophore and quencher, thus the transfer efficiency
will increase as a function of protein flexibility. Through measuring changes in the
intrinsic fluorescence decay rates of horseradish peroxides, we can determine changes in
the proteins relative structural flexibility.
54
4.44.44.44.4 OOOObjective 3bjective 3bjective 3bjective 3aaaa: Dynamics and function changes at inorganic : Dynamics and function changes at inorganic : Dynamics and function changes at inorganic : Dynamics and function changes at inorganic particle surfaces particle surfaces particle surfaces particle surfaces
A parallel objective is to measure protein flexibility and structure changes at TiO2 and
AlPO4 surfaces; TiO2 because it is a common electron scavenger employed in dye
sensitized solar cell (DSSC) and proposed bio sensitized solar cell designs; AlPO4
because it is prominently used as an adjuvant compound in vaccine formulations [78, 79].
A common problem in vaccine formulations featuring these particles is their aggregation
during long term storage; currently we hypothesize that gradual changes in protein and
polysaccharide structure and structure dynamics on particle surface create shifts in
electrostatics, hydrodynamics and steric interactions, these gradual shifts can be
influenced by the particle surfaces and lead to protein-protein attraction [80, 81].
Similarly, in an effort to utilize bacteriorhodopsin as a light sensitizer in greener
photovoltaic cells [82, 83], the protein needs to be attached to an electrode. In DSSC, this
is usually in the form of a sintered network of TiO2 nanoparticles. While it has been
established through circular dichromism that the secondary structures remain intact, the
internal dynamics that drive the proton pumping have not yet been studied at this particle
interface. For this work we will also employ time-resolved fluorescence and ESR to
measure flexibility changes in protein structures.
55
Figure 4-19 - Structure of CRM 197 carrier protein used in vaccine formulations
56
4.4.14.4.14.4.14.4.1 UVUVUVUV----Vis: Vis: Vis: Vis:
UV-Vis detection of proteins via tyrosine and tryptophan is a well-known method [84,
85]. For this project we will use it to quantify the concentration of proteins in solutions
and how much is adsorbed onto the AlPO4 particle surfaces. Using BSA as a model
protein, standard curve was prepared to correlate concentration with intensity. With that
curve we were able to determine the concentration of BSA adsorbed into AlPO4 by
measuring the concentration left in solution after mixing protein and particles and
separating the particles out. By subtracting from the known concentration of the original
solution, we determined that approximately 48 mg of BSA is absorbed onto ever 1 gram
of AlPO4
Figure 4-20 - Concentration standards of BSA measured by UV-Vis at 280nm
57
4.4.24.4.24.4.24.4.2 ZetaZetaZetaZeta potential measurementpotential measurementpotential measurementpotential measurement
A Zeta-Meter electrophoresis instrument and a Malvern Zetasizer have been used to
characterize the zetapotential of AlPO4 suspension as a function of pH and ionic strength
to determine the IEP of adjuvants alone. The change in surface charge of AlPO4 by
adsorbed protein have been studied as a function of concentration of protein, pH and
aging. The adsorption of protein and modification of surface charge can have an effect on
the agglomeration behavior of adjuvant-protein complex.
In electrophoresis measurement the electrodes are positioned at the opposite ends of a
glass capillary tube and electric current is applied to the dilute solution. The mobility of
particles under the influence of the electric field is measured manually using a
microscope. This technique is limited to approximately 1-micron size particles due to the
limitations of simple microscopy. A light scattering based Zetasizer instrument has been
used to measure the average zeta potential of micron to submicron AlPO4 particles. It can
yield an electrophoretic mobility distribution.
4.4.34.4.34.4.34.4.3 FTIR FTIR FTIR FTIR
A second method used to examine secondary structures in proteins (with and without
adjuvants) is infrared spectroscopy. FTIR analyzes the normal modes of vibrations of
covalent bonds, changes in dipole moment are necessary for IR absorption intensities to
be observed and most commonly involve stretching and bending modes of
nonsymmetrically bonded atoms. In the case of proteins, the amide I, II, and III bands
58
near 1700–1600 (C=O str), 1575–1475 (N-H str), and 1300–1220 cm−1 (C-N and N-H)
are the most commonly employed signals to monitor secondary structure in solutions. By
employing different sample geometry, Attenuated Total Reflectance (ATR), the changes
in protein structure in conjugation with adjuvant particles has been monitored as a
function of pH and degree of adsorption onto alumina phosphate particles.
59
5 Results and discussion
The results and discussions described in this section follow the four main objectives
outlined in previous sections. 1) Establishing a model system and methods for
quantifying enzyme activity in surfactant solutions. 2) Elucidating the nature of
interaction between surfactant, and surfactant aggregates with enzyme structures. 3)
Investigating changes in enzyme structure flexibility as a function of surfactant structure
and concentration. 4) Comparing how these interactions differ in situations where
proteins interact with a solid surface as opposed to surfactant micelle solutions.
5.15.15.15.1 Objective 1: Establishing a model systemObjective 1: Establishing a model systemObjective 1: Establishing a model systemObjective 1: Establishing a model system
In this first objective we present the data from enzyme “assays” and wherein the enzyme
catalyzed reaction is allowed to continue for a set number of minutes and the amount of
product formed during this time period is quantified and compared. This is described and
discussed in section 5.1.1. In contrast, the bulk of our enzyme activity measurements
referred to in this project involve real time measurement of product formation as the
reaction progresses, and the initial rates of these reaction profiles are what are compared
to each other. The term enzyme “activity” refers to this latter form of enzyme function
quantification.
60
5.1.15.1.15.1.15.1.1 5.1.1 Enzyme 5.1.1 Enzyme 5.1.1 Enzyme 5.1.1 Enzyme product yieldproduct yieldproduct yieldproduct yield assays. assays. assays. assays.
Activity is measured by the extent of polyphenolic residues such as tryptophan and
tyrosine liberated from the casein substrate through subtilisin catalyzed proteolysis in
buffer and various surfactants. The liberated tyrosine changes the color of Folin-
Ciocalteu reagent from yellow to blue. The color change is measured with UV-Vis and
presented below normalized versus the enzyme in buffer only.
When the color treated solutions were analyzied with UV-Vis, significant drops in
tyronise concentration were observed as protease solutions were exposed to 20mM
solutions of SDS. In the figure 5.1, the amount of tyrosine liberated by enzyme in SDS solution
is compared to that in pure water. It can been seen that the activity rate decreases significantly
faster in the presence of the anionic surfactant than in buffer alone.
Figure 5-1 - Subtilisin activity with and without SDS anionic surfactant as a function of exposure time
61
The effect of concentration was also studied. Enzyme was introduced to a 40 mM
solution of SDS, the activity tested before another 40 mM of SDS was added to the
solution. This was repeated until the enzyme was in a 120 mM solution of SDS.
Surprisingly, the enzyme activity *increased* with surfactant concentration. The error of
these experiments are determined by two additional repeats of the experiment.
Since it was additions of surfactant to a solution as opposed to three separate solutions, it
is implied that the deactivation seen in the 40 mM is actually *reversed* upon addition of
SDS. This will need to be confirmed and further studied.
62
20 40 60 80 100 1200
10
20
30
40
50
60
70
80
90
100
Tyro
sin
e L
ibera
ted
(u
M)
SDS Conc (mM)
Protease activity vs. SDS concentration
Figure 5-2 - Subtilisin activity with and without SDS anionic surfactant as a function of detergent concentration
This subtilisin assay experiment were carried out in solutions of linear alkyl benzene
(LAS), Brij-30 (Poly(oxyethylene)(4) lauryl ether), and APG (alkyl polyglucosides).
63
SDS LDS Brij-30 APG
-10
0
10
20
30
40
50
60
70
Activity r
ela
tive
to
su
btilis
in in
bu
ffe
r
Sufactant
Figure 5-3 - Enzyme activity assays of subtilisin in various surfactants (relative to enzyme in buffer)
Both the anionic surfactants, sodium dodecyl sulfonate (SDS) and linear alkylbenze
sulfonate (LAS) reduced the amount of product formed while the two non-ionic
surfactants Brij-30 and alkyl polyglucoside (APG) improved the yield. However, these
activity assays provide only a measure of product formed after a given time. Moving
forward, this dissertation will focus on activity quantifications that can provide rates of
product formation that is more applicable to extrapolating kinetics data.
5.1.25.1.25.1.25.1.2 Initial rate assaysInitial rate assaysInitial rate assaysInitial rate assays –––– Subtilisin proteaseSubtilisin proteaseSubtilisin proteaseSubtilisin protease
As previously described by section 4.1.3 in methods and materials section, the majority
of enzyme assays are carried out with casein substrate labeled with fluorescein
fluorphores. It is hydrolyzed by the subtilisin enzyme. The concentration of liberated
64
fluorophores is measured with a steady state fluorescence spectrometer at regular time
intervals over the course of the reaction. This procedure is repeated over a range of
substrate concentrations. The initial reaction rates are plotted as a function of substrate
concentration. Initial experiments are conducted with P4860, a commercial product of
subtilisin A from Novozymes in 50 mM potassium phosphate buffer solution. A <1 ml>
aliquot of <150ng/ml> enzyme solution is introduced to 2 ml of the substrate solution of
<0.005 wt%> FITC casein. The resulting changes in fluorescence are recorded by a
Horiba-Yvon Fluorolog 3, fluoro-spectrometer at 490nm excitation. Initial fluorescence
spectrum were measured between 500 and 580 nm to determine the correct emission
range as well as changing intensity with time. The experiment was then repeated and the
intensity at 520nm was measured every 30 seconds for the duration of the experiment,
1000 seconds in all. This experimental method will now be the standard measurement of
subtilisin activity for the remainder of this project, either as an absolute value of
▲intensity/second, Δ moles of fluorophore released/second or a normalized value
relative to the activity rate of the enzyme in buffer without surfactant.
65
Figure 5-4 - Fluorescence spectrum of casein substrate proteolysis as a function of time. Peak intensity is found at
519 nm.
Figure 5-5 - Fluorescence spectrum of casein substrate proteolysis as a function of time. Peak intensity is found at
519 nm.
0.00E+00
5.00E+05
1.00E+06
1.50E+06
2.00E+06
2.50E+06
3.00E+06
3.50E+06
4.00E+06
480 500 520 540 560 580 600
Inte
nsi
ty (
cou
nts
)
Emission wavelenth (nm)
Spectra of FTC-Casein w/ subtilisin protease
20 minutes
5 mintues
1 minute
1.30E+07
1.50E+07
1.70E+07
1.90E+07
2.10E+07
2.30E+07
2.50E+07
0 200 400 600 800 1000
Inte
nsi
ty (
cou
nts
)
Time (s)
Absorption @ 519 nm
Initial rate
measured
66
With an experimental method of measuring enzyme activity established, the next portion
of the first objective will be to confirm activity changes in the chosen surfactant systems.
Of our initial assay tests, APG and Brij-30 exhibited increased enzyme activity over
subtilisin in water alone. Thus, these surfactants were added to the system and the
changes in initial rate were measured. This initial experiment was done with P5459
subtilisin enzyme in 5mM of surfactant solution versus in water suspension. The reaction
profiles can be seen in figure 5.6, which demonstrates the difference between initial rates
of subtilisin as influenced by APG. To further investigate the dependence of subtilisin
mediated proteolysis rates on surfactant concentration, changes in initial activity are
measured over a concentration range of APG from 0.2 to 20mM. The reaction profiles
and corresponding initial rate measurements, normalized to the activity in buffer, is
plotted in figures fig 5.7 and 5.8 below. The differences of intensity at time zero can be
attributed to the delay between addition of enzyme to substrate and the start of
measurement. This interval is maintained to 10 seconds.
67
Figure 5-6 - Initial subtilisin rate profile experiments confirm faster initial proteolysis rates in alkyl polyglucoside
solutions.
Figure 5-7 - Subtilisin catalyzed proteolysis reaction profile in APG solutions
0
10000
20000
30000
40000
50000
60000
70000
80000
0 100 200 300 400 500 600 700
Inte
nsi
ty (
CP
S)
Time (seconds)
In APG
0.02
0.025
0.03
0.035
0.04
0.045
0.05
0.055
0.06
0 100 200 300 400 500 600 700
Pro
du
ct f
orm
ed
(m
M)
Time (seconds)
Subtilisin in APG
buffer
0.2mM
2.0 mM
5.0 mM
10 mM
20 mM
In Buffer only
68
Figure 5-8 - Normalized values of the initial rates derived from reaction profiles shown in Fig. 5.7
This experiment was also run in concentrations of Brij-30 at 0.007, 0.067, 0.67 and 6.7
mM of surfactant. At the lowest concentrations of this surfactant, there appears to be a
slight loss in relative activity at 0.95 relative units. The activity increase with surfactant
concentration, although not as much as APG, only reaching 1.24 relative activity units at
6.7mM. The reaction profiles and corresponding initial rate measurements, normalized to
the activity in buffer, are plotted in figures 5.9 and 5.10 below. These activity
experiments in APG were repeated to obtain experimental error which is reported in table
5.1
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
0.2 2 5 10 20
rela
tiv
e a
ctiv
ity
un
it
APG concentration (mM)
APG
69
Figure 5-9 - Subtilisin catalyzed proteolysis reaction profile in Brij-30 surfactant
Figure 5-10 - Initial reaction rates of subtilisin mediated proteolysis in Brij-30 non-ionic surfactants
0.00E+00
5.00E+06
1.00E+07
1.50E+07
2.00E+07
2.50E+07
0 100 200 300 400 500 600 700
Flu
ore
sce
in m
ark
er
inte
nsi
ty (
CP
S)
Reaction time (s)
Water
0.007 mM
0.067 mM
0.66 mM
6.7mM
0.951.03 1.07
1.24
0
0.2
0.4
0.6
0.8
1
1.2
1.4
0.007 0.067 0.67 6.7
No
rma
lize
d i
nit
ial
rea
ctio
n r
ate
Initial rate relative to enzyme in water
70
To compare the activity in a surfactant with a similar structure that is extensively used for
protein extraction and preservation This experiment was also run in concentrations of
Brij-35 at 0.026, 0.26, 2.62 and 26.0 mM of surfactant. At the lowest concentrations of
this surfactant, there appears to be a slight loss in relative activity at 0.95 relative units.
The activity increase with surfactant is even lower, only reaching 1.15 relative activity
units at 0.26mM, not significantly greater than the activity in buffer alone. At 2.6 and
26mM of Brij-35 surfactant the activity dramatically drops to 0.58 and 0.53 relative
activity units respectively. The reaction profiles and corresponding initial rate
measurements, normalized to the activity in buffer, is plotted in figures 5.8 and 5.9
below.
Figure 5-11 - Proteolysis reaction profile in Brij-35 surfactant
0
10000
20000
30000
40000
50000
60000
70000
0 100 200 300 400 500 600 700
Reaction time (seconds)
Digestion of 0.01 wt% casein via subtilisin in conc. of
Brij 35
Water
0.026 mM
0.26 mM
2.60 mM
26.0 mM
71
Figure 5-12 - Initial reaction rates of subtilisin catalyzed proteolysis in Brij-35 surfactant
To investigate the effects of free surfactant monomers on subtilisin function, the activities
were recorded in APG concentrations below CMC, at 0.015, 0.03, 0.06, 0.12, 0.24, 0.5
and 1.0 mM (figures 5.13). A rapid loss in activity was observed, reaching a minimum of
0.092 relative activity units at 0.1 mM of surfactant. This trend reverses upon increasing
surfactant concentration. It is important to note that the CMC of APG is measured to be
between 0.2 and 0.3 mM. This preliminary was the first indication that the activity is
negatively influenced by interactions with surfactant monomers, but also suggest that the
hyperactivity observed is linked to the formation of micelles. The reaction rate is
compromised at low Brij-30 (fig. 5.14) concentrations below CMC (0.3 mM) but can be
seen increasing at higher surfactant concentrations suggesting that enzymatic
hyperactivity is linked to micelle concentration and not surfactant monomer. However,
0
0.2
0.4
0.6
0.8
1
1.2
1.4
0 0.026 0.26 2.6 26
Init
ial
rate
re
lati
ve
to
en
zym
e i
n b
uff
er
Brij - 35 (mM)
Enzyme activity in Brij-35
72
the presence of micelles does not exclude the presence of surfactant monomers still in
solution and interactions between monomers and enzyme should not be ignored.
Figure 5-13 - Normalized reaction rates of Subtilisin in low concentrations. APG CMC is approximately 0.2 mM.
Figure 5-14 - Subtilisin activity in Brij-30 solutions
0
0.1
0.2
0.3
0.4
0.5
0.6
0.01 0.1 1
Act
ivit
y (
rela
tiv
e t
o b
uff
er)
APG concentration (mM)
0
0.00005
0.0001
0.00015
0.0002
0.00025
0.0003
0.00035
201020.020
Re
act
ion
ra
te (
mM
/ S
ec)
Brij-30 concentration (mM)
73
5.1.35.1.35.1.35.1.3 HRP activity in BrijHRP activity in BrijHRP activity in BrijHRP activity in Brij----30, APG and DM 30, APG and DM 30, APG and DM 30, APG and DM
To compare the effects of surfactant on another model protein of different size, catalytic
mechanism and substrate, horseradish peroxidase was selected and activities tested in the
presence of surfactants APG, dodecyl maltoside and Brij-30. Figures 5.12, and 5.13 show
the product formation profile and extrapolated initial reaction rates for HRP in 2.0mM
and 0.5mM of APG. Further trials were measured for APG concentrations 0.17, 0.33,
0.66, 3.3, 6.7 and 20mM and their values are listed in Table 5.1. The relative activity
units of those reactions are plotted in figure 5.14.
Figure 5.15. Reaction profile of horseradish peroxidase mediated phenol oxidation.
0
0.2
0.4
0.6
0.8
1
1.2
1.4
0 1 2 3 4 5 6
Pro
du
ct f
orm
ati
on
Time (minutes)
Buffer
2.0 mM APG
.5 mM APG
74
Figure 5-15 - HRP initial reaction rates in APG surfactant
APG
conc
trial 1 2 3 Avg StD
Measured Buffer 0.21 0.23854 0.244 0.23 0.02
04/16,
04/17
.17mM 0.23 0.23988 0.22 0.23 0.01
.33 mM - 0.21165 0.19 0.20 0.01
.67 mM 0.24 0.25051 0.24 0.24 0.01
APG
conc.
trial 1 Trial 2 Trial 3 Avg StD
Measured Buff 2 0.18 0.17 - 0.18 0.006
18-Apr 3.3mM 0.26 0.25 - 0.25 0.003
6.7 mM 0.19 0.20 - 0.19 0.008
20 mM 0.19 0.20 - 0.19 0.001
Table 5.1 - Results of multiple HRP activity tests.
0.2
0.205
0.21
0.215
0.22
0.225
0.23
0.235
0.24
0.245
Buffer 2.0 mM APG .5 mM APG
Init
ial
rea
ctio
n r
ate
75
Figure 5-16 - - Initial reaction rates of HRP in APG (columns) and surface tension measurements of APG (diamonds)
As seen with the subtilisin protease model enzyme, initial introduction of APG surfactant
decreased HRP enzyme activity up until the critical micelle concentration. Upon
formation of detergent aggregates, this shift in activity so closely linked to the
surfactant’s CMC strongly implicates surfactant aggregates, or micelles, as the primary
driver in the observed enzyme hyper-activity, whereas interactions with surfactant
monomers appear to suppress it. Surface tension measurements were also conducted to
investigate the effect of added enzyme has on micelle formation. Preliminary tests did not
indicate an appreciable shift in CMC, however additional experiments will have to be
carried out to evaluate the validity of those initial findings.
0.6
0.7
0.8
0.9
1
1.1
1.2
1.3
1.4
1.5
27
29
31
33
35
37
39
41
43
0.01 0.1 1 10 100
HR
P a
ctiv
ity
(n
orm
ali
zed
)
Su
rfa
ce t
en
sio
n n
N/m
APG concentration mM
.3 mM
76
0.1 1
0.85
0.90
0.95
1.00
No
rma
lize
d H
RP
in
itia
l re
actio
n r
ate
Dodecyl Maltoside Conc. (mM)
D
Figure 5-17 - Normalized initial rate values of HRP activity in dodecyl maltoside concentrations
77
0.1 1 10
0.6
0.8
1.0
1.2
1.4
1.6
Enyzm
e A
ctivity (N
orm
aliz
ed r/r
o)
APG conc. (mM)
Figure 5-18 - Normalized initial reaction rates of subtilisin (red) compared to those of HRP enzyme (black)
0.1 1 10
0.7
0.8
0.9
1.0
1.1
1.2
1.3
1.4
1.5
En
zym
e a
ctivity (
no
rma
lize
d r
ate
)
Surfactant Concentration (mM)
Figure 5-19 - Figure 5.17. Comparison of normalized initial rates of HRP enzyme in APG (black squares) vs DM (red
triangles)
78
5.25.25.25.2 Discussion of subtilisin and HRP assay and activity resultsDiscussion of subtilisin and HRP assay and activity resultsDiscussion of subtilisin and HRP assay and activity resultsDiscussion of subtilisin and HRP assay and activity results
The primary objective of this portion of work was to devise methods for measuring
enzyme activity and changes of enzymatic catalytic rates in the presence of the selected
surfactants: APG, DM, Brij-30 and Brij-35. Initially, activity of subtilisin was measured
by a colorimetric method of reduction of Foilin-Ciocalteu reacgent via liberated phenolic
residues. However, this method required that the reaction be manually halted at discrete
intervals and could not provide reasonably accurate reaction rates required for the fast
subtilisin mediated proteolysis. To overcome these limitations, a real time assay method
was adapted, which utilized the fluorescein-labled casein as a proteolysis substrate. This
allowed us to observe the reaction profile as it was occurring in real-time. Furthermore, it
was a much more streamlined process that was easier to measure reaction rates over
substrate concentrations to extract kinetics parameters. These kinetics experiments are
discussed in the next section. However, the parameters for using fluorescein needed to be
optimized. This process is described in the methods and materials section in the methods
and materials section
Using this assay method, we were able to reliably measure the initial proteolysis rates
mediated by subtilisin and perform a systematic study of enzymatic activity in the
presence of APG, Brij-30 and Brij-35 at various concentrations. As was demonstrated by
the initial assay tests, all three surfactants were shown to increase enzyme activity to
some degree, 48% in APG, 24% in Brij-30 and 15% in Brij-35. This already illustrates a
fundamental difference between how the various surfactant structures interact with either
the substrate, enzyme or combination of the two. What is also important to note is that the
79
overall increase in reaction rate occurs at specific surfactant concentrations. APG is the
most intense example, where the 48% increase in relative reaction rate occurs at 2mM
and quickly drops after this concentration range, returning to normal or sub-normal
activities at 5, 10 and 20 mM. Subtilisin activity in Brij-30 also appears to reach a
possible maximum of an additional 24% at 6.7 mM surfactant concentration. However
the Brij-30 cloudpoint, a temperature above which non-ionic surfactant micelles self-
aggregate, is below room temperature. As a result, higher concentrations of Brij-30 result
in increased turbidity, causing complications in the spectroscopic methods that we have
adapted. Alternatively, Brij-35, with its larger polyexthoxylate head group has a lower
cloud point. Like APG it appears to increase enzyme activity the most at a particular
concentration, only by 15% at 0.26 mM. However, further repeats of this experiment will
be needed to confirm if this value is significantly greater than the normal value. It is also
worthy to note that there is some activity loss at the very low concentration of 0.026 mM
of Brij-35 surfactant. While this also needs to be retested to determine if the difference is
significant, it prompted us to return to the APG system and measure enzyme activity
below the critical micelle concentration.
Since APG is a mixed system of surfactants, ranging in headgroup and tail group sizes,
the CMC can vary depending on the APG mixture. Using surface tension measurements
and pyrene probe fluorescence techniques to be discussed in the next section, it was
determined that micelle formation begins at 0.2 mM of the APG sample we were using.
This is confirmed in literature for APG mixtures ranging from 12-14 carbon units of
length. Measuring the activity below this concentration at 0.015, 0.03, 0.06, 0.13, 0.25,
80
0.5 and 1.0 mM indicates rapidly diminishing activity with surfactant concentration, up to
0.13 mM. This is near the CMC range and indeed, the trend in activity loss reverses at
0.25 mM, continuing to rise with each increase of surfactant concentration. This loss of
activity in pre-micelle concentrations, as well as the apparent dependence of subtilisin
activity on surfactant concentration strongly suggests that it is the micelles that play a
major role in the observed enzyme hyper activity phenomena. Moreover, free surfactant
monomers may be responsible for negatively affecting enzyme activity. The concept of
micelles being responsible for enzyme hyperactivity has been proposed by theoretical
models reported by Viparelli, et al. However, direct observation through experimentation
has been lacking, and, as previously mentioned, a precise physical reason explaining why
or how micelle surfaces should influence the enzyme’s catalytic properties is also
missing. Finally, the activity experiments presented in this thesis demonstrate an activity
increase maxima, most notably in the APG mixed surfactant system. Theoretical models
do not take into account possible complex interactions in surfactant aggregate structures,
such as to rod-like or liquid crystal formations. This will be explored in subsequent
sections of this dissertation.
Horseradish peroxidase (HRP) was chosen as an alternative model protein to be studied
in parallel with subtilisin. Aside from being well characterized, both structurally and
functionally, HRP differs from subtilisin in several key factors, such as having disulfide
bonds, metal complexation sites, a heme group, and most importantly, requires hydrogen
peroxide as a oxidizing reagent. For an HRP mediated oxidation, both substrate and
oxidant are water soluble molecules, unlike the large casein protein used in our subtilisin
81
activity tests, where the surfactant may have a significant impact on the reaction rate
through interaction with the substrate, by either improved dispersion and/or partially
denaturing the substrate before proteolysis. The differences of activity between these two
enzyme/substrate systems in various surfactants have provides valuable comparison
points for final elucidation of the surfactant / protein interactions.
A comparison of subtilisin and HRP activities in the presence of a range of APG
concentrations yield two very similar activity vs. concentration profiles. As was shown
previously, subtilisin activity suffers a loss at lower concentration ranges before the
CMC. Between 0.7mM and 1.5 mM of APG, this decrease is reversed and reaches a
maximum in activity centered around 2.0 mM. This activity drops back to less than
normal rates at 5 and 10 mM. The relative activities of HRP in APG solutions follow a
very similar pattern, with activity rates that decrease with additions of surfactant up until
the critical micelle concentration. Upon the formation of micelles, as was in the case of
subitilisin, the activity begins to increase reaching a maximum of 1.43 relative activity
units at 2mM and, also like the protease system, decreases after this point to 1.1 and 1.1
relative activity unites for 5.0 and 10.0 mM of APG respectively. The results can be seen
in figures 5.16 and 5.17.
The implications of these similar activity profiles again strongly indicate that micelle
structure plays an essential role in this mechanism of enzyme activity enhancement.
However, APG is a mixed surfactant system, known to undergo micelle structure
82
transitions and form worm-like structures at higher concentrations. Furthermore, the APG
product we started experiments with was sourced from an industrial product with an
unknown mixture composition, only an estimate provided by the manufacturer. For this
reason, dodecyl maltoside was also added to HRP reaction solution to observe induced
changes in the enzyme activity and how it compares to the APG mixed surfactant system.
Contrary to our initial assumption that the enzyme in dodecyl maltoside would react
similarly as in APG, the activities of HRP steadily decreased with additions of the
surfactant, from 0.015 to 5.0mM. This is a stark contrast from enzyme activity in APG,
despite having similar headgroup structures. In subsequent sections 5.3 and 5.4, this
thesis will explore the differences in micelle structure between the two sugar based
surfactants and how these differences correlate to these diverging activities. This next
section 5.2.1, will describe our investigation into how enzyme activity kinetics are
affected by the introduction of surfactants.
5.2.15.2.15.2.15.2.1 Sub sectionSub sectionSub sectionSub section:::: kinetics parameters Km and Vmaxkinetics parameters Km and Vmaxkinetics parameters Km and Vmaxkinetics parameters Km and Vmax
Vmax and Km are important kinetic parameters which reflect the binding efficiency and
forward rate of reaction in catalytic reactions. These two variables fall under the
Michaelis-Menten model of enzyme kinetics :
Equation 5-1 – Michaelis-Menten relationship between
Where υ is the rate of the reaction, P is the concentration of product, S is the
concentration of substrate, Vmax is the rate of reaction at substrate saturation and Km is
83
the substrate concentration at half Vmax. The overall scheme of an enzyme mediated
reaction is as follows:
K cat is the rate of forward reaction, also known as the turnover number, and is associated
with Vmax by:
Where [E]0 is the enzyme concentration. The Michaelis constant is the substrate
concentration at half Vmax and is a measure of the affinity of the enzyme for the substrate
molecule. The only controlled variable in this model is the substrate concentration. In the
subsequent experiments we used the fluorescence activity measurement techniques to
monitor changes in initial reaction rates of subtilisin on a casein substrate in substrate
concentrations ranging from 20, 40, 80 and 160 and sometimes 320 ng/ml. The data from
these experiments were then fitted with the Micealis-Menten equation in Origins 8.5
software, giving us the values of Km and Vmax for subtilisin acting on casein in buffer.
84
Figure 5-20 - Subtilisin mediated proteolysis reaction profiles as a function of substrate concentration
0 20 40 60 80 100 120 140 160 180
0.00004
0.00005
0.00006
0.00007
0.00008
0.00009
0.00010
0.00011
0.00012
Data: Data1_B
Model: Dhyperbl
Equation:
y = P1*x/(P2 + x) + P3*x/(P4 + x) + P5*x
Weighting:
y No weighting
Chi^2/DoF = 1.9857E-11
R^2 = 0.98737
P1 0.00015 ±0.00001
P2 45.4249 ±7.88149
P3 0 ±0
P4 0 ±0
P5 0 ±0
Initia
l re
action r
ate
(m
M/s
ec)
Casesin Substrate (ng/ml)
B
Figure 5-21 - Non-linear regression fitting of reaction rates of subtilisin in buffer vs. casein substrate concentration
done in Origins 8.0 software
0
0.005
0.01
0.015
0.02
0.025
0.03
0.035
0.04
0.045
0 100 200 300 400 500 600 700
Est
pro
du
ct f
orm
ati
on
(m
M)
Time (s)
20 ng/ml
40
80
160
85
To measure the effects of various surfactants on Vmax and Km, the same series of
experiments were conducted in solutions of APG (2, 10 and 20 mM), DM (10 and 20
mM), Brij-30 (2,5, 10, and 20mM) and Brij-35 (5, 10 and 20mM). The initial rates and
the Michaelis-Menten model fittings can be seen for the APG 10 and 20 experiments in
figures 5.20 and 5.21. In figures 5.22 a-d, the Vmax values are plotted as a function of
APG, DM, Brij-30 and Brij-35 surfactant concentration, respectively. The Km values are
plotted in Figures 5.23 a-d.
0 20 40 60 80 100 120 140 160 180
0.00002
0.00004
0.00006
0.00008
0.00010
0.00012
0.00014
0.00016
Data: Apg10mM_B
Model: Hyperbl
Equation:
y = P1*x/(P2 + x)
Weighting:
y No weighting
Chi^2/DoF = 1.0838E-10
R^2 = 0.96282
P1 0.0002 ±0.00003
P2 58.71577 ±20.21745
Substrate conc. nM
B
Figure 5-22 - Initial rates of subtilisin catalyzed proteolysis vs substrate concentration in 10mM of APG
86
Figure 5-23 - Initial rates of subtilisin catalyzed proteolysis vs substrate concentration in 20mM of APG
87
5.2.1.15.2.1.15.2.1.15.2.1.1 VVVVmaxmaxmaxmax
The changes in Vmax reflect the overall activity changes that have been demonstrated in
the previous sections. Activity tests of subtilisin in APG solutions results in higher Vmax
values, and therefore faster overall Kcat, for intermediate APG concentrations, 5.0 and
10.0 mM, and returns to nominal levels at higher concentrations. As seen in the activity
tests with HRP, dodecyl maltosides do not do much to increase the Vmax of subtilisins,
however, there is not strong evidence that it significantly hinders it either. Brij-30 also
increases subtilisin Vmax but only after an initial loss in before 5mM. Brij-35, as with the
activity tests, demonstrated little effects on the Vmax.
Figure 5-24 - Subtilisin Vmax as a function of APG concentration
1E-08
6E-08
1.1E-07
1.6E-07
2.1E-07
2.6E-07
0 5 10 15 20 25
Max r
eaction r
ate
(M
/sec)
Surf. Conc (mM)
APG
88
Figure 5-25 - Subtilisin Vmax as a function of DM concentration
Figure 5-26 - Subtilisin Vmax as a function of Brij-30 concentration
1E-08
6E-08
1.1E-07
1.6E-07
2.1E-07
2.6E-07
0 5 10 15 20 25
Ma
x R
ea
ctio
n r
ate
(M
/se
c)
Surf Conc. (mM)
Dodecyl Maltoside
1E-08
6E-08
1.1E-07
1.6E-07
2.1E-07
2.6E-07
0 5 10 15 20 25
Re
ac
tio
n R
ate
(M
/se
c)
Surf. Conc. (mM)
Brij-30
89
Figure 5-27 - Subtilisin Vmax as a function of Brij-35 concentration
1E-08
6E-08
1.1E-07
1.6E-07
2.1E-07
2.6E-07
0 5 10 15 20 25
Vm
ax R
ea
ctio
n r
ea
te (M
/ s
ec)
Surf. Conc (mM)
Brij-35
90
5.2.1.25.2.1.25.2.1.25.2.1.2 KmKmKmKm
For the most part, Km, which is an inverse measure of the binding efficiency between
substrate and enzyme, appears to be proportional with Vmax. That is, there is an inverse
relationship between the binding efficiency of subtilisin and Vmax in the presence of the
surfactants. In the case of dodecyl maltosides and Brij-35, which did not significantly
improve Vmax at any concentration, their Km values are increased significantly,
particularly at 5.0 and 10 mM concentrations. This is also true for APG, although this
decrease in binding efficiency is matched with an increase in Vmax. Brij-30 is the only
surfactant that appears to improve binding efficiency, lowering Km between 2.5 and 10
mM. This trend reverses and by 20mM Km is again higher than subtilisin in buffer alone.
Figure 5-28 - Subtilisin Km as a function of APG concentration
0.00E+00
2.00E-08
4.00E-08
6.00E-08
8.00E-08
1.00E-07
0 5 10 15 20 25
Km
(M
)
Surf. Conc
Alkyl Polyglucoside
91
Figure 5-29 - Subtilisin Km as a function of DM concentration
Figure 5-30 - Subtilisin Km as a function of Brij-35 concentration
0.0E+00
2.0E-08
4.0E-08
6.0E-08
8.0E-08
1.0E-07
0 5 10 15 20 25
Km
(M
)
Surf. Conc (mM)
Dodecyl Maltoside
0.0E+00
1.0E-08
2.0E-08
3.0E-08
4.0E-08
5.0E-08
6.0E-08
7.0E-08
8.0E-08
9.0E-08
1.0E-07
0 5 10 15 20 25
Km
(M
)
Surf Conc. (mM)
Brij-35
92
Figure 5-31 - Subtilisin Km as a function of Brij-30 concentration
0.0E+00
2.0E-08
4.0E-08
6.0E-08
8.0E-08
1.0E-07
0 5 10 15 20 25
Km
(M
)
Surf. Conc. (mM)
Brij-30
93
5.2.1.35.2.1.35.2.1.35.2.1.3 Discussion of kinetics in surfactants: Discussion of kinetics in surfactants: Discussion of kinetics in surfactants: Discussion of kinetics in surfactants:
The changes in Vmax in the four surfactants tested reflect what was demonstrated in the
activity tests done at a single substrate concentration (section 5.2): APG induces higher
maximum reaction rates at 2 and 5 mM and this increased rate drops back to normal
values at higher concentrations. Brij-30 increases the forward catalytic rate the most at
higher surfactant concentrations (20mM). These increases suggest that the surfactant
maybe interacting with the enzyme and/or substrate. Since Vmax is the product of forward
catalysis rate, Kcat, and enzyme concentration, one likely explanation is that the
introduction of surfactants helps disperse the enzyme, increasing the overall availability
per unit of enzymes introduced to the system. This hypothesis is explored in later sections
with particle sizing experiments. However, this is possibly an incorrect explaination, as
one might expect to see binding efficiency improve if more active site were exposed
through surfactant interactions. Alternately, the surfactant may be dispersing the
substrate, as opposed to the enzyme. As with dispersion of enzymes, dispersion of
substrates will increase the number of binding sites per unit of substrate added, which
would theoretically lower Km. Furthermore, at higher substrate concentrations, the
relationship between initial rate and substrate reaches a zero-order relationship and thus
Vmax should not be affected by the availability of more substrate attachment sites. Both of
these are not reflected the changes of kinetics in APG, which demonstrates greater Vmax
AND Km at 5 and 10mM. Both of these phenomena, especially the increase in Km suggest
that the surfactant is directly affecting the enzyme to increase Kcat while also lowering
binding efficiency. One physical explanation for these two kinetics parameters is that the
surfactant is in some way changing the flexibility of the subtilisin structure. Specifically,
94
a stiffer enzyme structure may account for further reduction of activation energy,
inducing a faster Kcat, while making binding with the substrate more difficult, thus
increasing the Km parameter.
Brij-30 is the only surfactant seen to lower Km. As mentioned before, this can be the
result of dispersion, either of the enzyme or substrate. Alternatively, the binding
efficiency is directly affected at those particular surfactant concentrations. This is another
distinct possibility, as Vmax is seen to decrease from a buffer only solution at those
surfactant concentrations. This maybe the opposite of the hypothetical “structure-
stiffening” scenario proposed earlier to explain greater Vmax and greater Km. A “softer”,
more “flexible” enzyme structure can conceivably have a greater substrate binding
affinity while slowing the forward rate of catalysis.
Finally, in these kinetics experiments, it is apparent that Vmax and Km are affected only at
particular concentrations of APG and Brij-30, while DM and Brij-35 with their similar
head group chemical compositions did not. This, combined with the drop in activities at
pre-micelle concentrations demonstrated in activity assays, strongly indicates that the
changes in enzyme activities are driven by micelle formations and presumably, the size,
shape and structure of these surfactant aggregates. The next section will describe our
efforts in elucidating the differences in the micelle properties of the four surfactants and
if and how they might correlate to the changes in activities that we have observed.
95
5.35.35.35.3 Objective 2: AggregatObjective 2: AggregatObjective 2: AggregatObjective 2: Aggregation /ion /ion /ion / micelle properties micelle properties micelle properties micelle properties
The second objective of this project is to elucidate the nature of surfactant – enzyme
interactions. Enzyme activity vs. surfactant concentration measurements discussed in
section 5.2 indicate a strong correlation between changes in enzyme activity and the
formation of APG micelles. The differences between APG and DM aggregates will be
measured by DLC, ESR, NMR and surface tension, and the results discussed in this
following section.
5.3.15.3.15.3.15.3.1 Surface tension: Surface tension: Surface tension: Surface tension:
The results of APG surface tension vs concentration can be seen in figure 5.32. The
break-point which indicates the concentration of micelle formation can be identified at
0.3 mM. However, surface tension continues to decrease up to 3mM of surfactant
concentration. This phenomena has been observed in mixed surfactant APG systems and
is indicative of micelle structure transitions which result from the varying solubility of
the heterogeneous surfactants[86].
96
In comparison, dodecyl maltoside do not exhibit this transition range. The surface tension
break point occurs at 0.1 mM and remains constant at all higher surfactant
concentrations. This is consistent with our previous assertion that the transition phase
seen in APG is the result of APG being a mixture. Furthermore, the surface activity of
DM is lower when compared to APG figure 5.25.
CMC I CMC II
Figure 5-32 - Surface tension vs APG concentration
0.1 1 10
30
40
Surf
ace T
ensio
n (
mN
/m)
APG Conc. (mM)
97
Figure 5-33 - Surface tension comparisons of dodecyl maltoside (single surfactant) to alkyl polyglucosides, a
surfactant mixture of similar composition.
5.3.25.3.25.3.25.3.2 Micelle sizeMicelle sizeMicelle sizeMicelle size::::
In order to better understand the interactions between the surfactants and enzyme
structure DLS was used to measure the differences between APG and DM micelles. 1.5
and 3.0 mM APG micelles appear to agglomerate into large particles, 1.4 and 1.7 in
diameter respectively. This concentration range is within the ‘transition’ phase between
0.3 and 3.0 mM as identified by surface tension. At 5 mM, these large aggregates
disperse and the micelle sizes rapidly reduce to 50-60nm in size. This dimension is still
large for a monolayer, spherical micelle which should only range between 5-10nm for
APG. The formation of rod-like micelles is a common for APGs and 60nm is a likely size
for the length of such conformations. The rod width was not picked up by this method,
however.
25
30
35
40
45
50
0.01 0.1 1 10 100
Su
rfa
ce t
en
sio
n (
mN
/m)
Surfactant Conc. (mM)
APG DM
98
Figure 5-34 - APG micellesize vs concentration
In comparison, the micelle size of DM at all concentrations above CMC were between 6
and 8 nm, which would be consistent with monolayer, spherical micelles.
Figure 5-35 - DM micelle size vs concentration
1443
1770
56 52 56
0
200
400
600
800
1000
1200
1400
1600
1800
2000
1.5 3 5 10 20
Pa
rtic
le d
iam
ete
r (n
m)
APG conc. (mM)
8.59
8.29
0
2
4
6
8
10
12
2 5 10 20
Mic
ell
e S
ize
(n
m)
DM conc. (mM)
99
5.3.35.3.35.3.35.3.3 ESR investigation of micelle structure: APG vs DM ESR investigation of micelle structure: APG vs DM ESR investigation of micelle structure: APG vs DM ESR investigation of micelle structure: APG vs DM
To further investigate the nature of these micelles and how the differences in these
structures ultimately influence the enzyme activity we performed electron spin resonance
experiments to measure the rigidity and packing density of the two surfactants. In figure
5.28 the correlation time parameter of the nixtroxile stearic acid probe, which is a
measure of the probe’s mobility, is plotted as a function of both APG and DM surfactant
concentration.
Figure 5-36 - Rotational correlation times (measure of local viscosity) as a function of surfactant (APG and DM)
concentration in buffer only.
0
5E-10
1E-09
1.5E-09
2E-09
0.01 0.1 1 10
Ro
t. C
orr
ela
tio
n t
ime
(s)
Surfactant Conc. (mM)
APG only
DM only
100
Figure 5-37 - Rotational correlation times (measure of local viscosity) as a function of surfactant (APG and DM)
concentration in buffer with 0.01 mg/ml HRP.
Figure 5.28 further highlights a significant difference in the evolution of APG and DM
micelle aggregates. In both surfactants, the transition from free monomers to aggregates
is apparent starting at 1mM. This is higher than the expected CMC values of APG and
DM (0.2 and 0.3 mM respectively) and this may be due to the significant concentration of
16-DSA probe at these lower concentrations.
Above 1.0 mM, the local viscosity around the probe in dodecyl maltoside solution
increases dramatically plateauing after a quick transition between 1.0 and 2.5 mM. In
contrast, the packing density of APG makes a gradual transition starting at 1.0 mM and
appears to begin to plateau by 5mM. Currently, it is hypothesized that Dodecyl maltoside
makes an immediate transition to spherical, monolayer micelles at CMC, while, APG
undergoes an extended transition phase before the bulk concentration of the mixed
0.00E+00
5.00E-10
1.00E-09
1.50E-09
2.00E-09
0.01 0.1 1 10
Ro
t. c
orr
ela
tio
n t
ime
(s)
Surfactant concentration (mM) + 0.01mg/ml HRP
APG +
HRP
101
surfactant is sufficient to form rod like micelles. The reasoning and implications of this
structure shift are further discussed in Section 5.6, result discussions.
5.3.45.3.45.3.45.3.4 NMR:NMR:NMR:NMR: (Cooperative work with C. Totland of University of Bergen)(Cooperative work with C. Totland of University of Bergen)(Cooperative work with C. Totland of University of Bergen)(Cooperative work with C. Totland of University of Bergen)
NMR was used as a tool to understand the precise mechanisms of interaction between
surfactant and micelles. Its major advantage over probe techniques such as electron spin
resonance, is that it can give detailed data on all molecular interactions in the system, not
only about the probe molecule. However, the resulting spectrum of complex systems,
such as in surfactant-enzyme solutions, can be very dense and difficult to analyze. With
the help of Dr. C. Totland[87, 88] from the university of Bergen, NMR was successfully
used to investigate APG and HRP interactions. Furthermore, NMR revealed unique head-
tail group interactions which implies a unique anti-parallel packing that has not yet been
observed in APG systems.
5.3.4.15.3.4.15.3.4.15.3.4.1 Line WidthsLine WidthsLine WidthsLine Widths
The line width data strongly suggest the formation of rod-shaped aggregates. Line widths
are inversely proportional to the T2 (spin-spin) relaxation time of those protons. In
contrast to T1 (spin-lattice) relaxation, which requires fluctuations near the Larmor
frequency (rotational speed of the nucleus) to occur, T2 is affected by slower motions as
well. Hence, while changes in T1 upon micellization reflect changes in the local
environment of that proton, changes in T2 (or line width) primarily reflect changes in
102
mobility of the aggregate as a whole. An increase in line-width therefore means that the
micelle rotation is slowing down due to increased size.
Furthermore, a recent study [89] showed using both gemini and conventional surfactants
that large increases in line width at concentrations above CMC only occur when a sphere-
to-rod transition takes place. Surfactants without the ability to form rods does not display
such line width increases. A comparison with more sophisticated techniques such as
NMR diffusometry showed that line width data qualitatively gives the same information.
Figure 5-38 - Comparison of the line widths of the main APG CH2 peak (2-12) for the 0.15 and 12 mM APG sample in
0.01 mg/ml HRP in 100mM potassium phosphate buffer
10080604020020406080100 Hz
0.15 mM
12 mM
103
Figure 5.30 shows the line widths of APG just below CMC (0.15 mM) and well above
the hypothesized sphere-to-rod transition and illustrates the differences in line width
between a mobile surfactant tail and those slowly spinning in a confined micelle. Figures
5.31 and 5.32 plot the T2 relaxation times derived from these peak widths as a function of
surfactant concentration.
Figure 5-39 - The line width here is the line width of the main CH2 peak (2-12) in the NMR spectrum. The line
widths are measured by using the deconvolution tool in the Topspin software (version 1.3).
10
15
20
25
30
35
40
45
50
0.01 0.1 1 10 100
APG + HRP
APG
[APG] mM
Lin
e w
idth
(H
z)
104
Figure 5-40 - Overlay of the aggregate mobility data presented in figure 5.31 with the activity of HRP in APG
(triangles)
Figure 5.31 shows that the sphere-to-rod transition of APG (white diamonds). Starting at
0.4 mM the average aggregate mobility begins to increase steadily until ~4.0 mM
wherein the rate of aggregate size increase vs. surfactant concentration decreases. This
can be attributed to the shift from one micelle transition to a final rod like conformation.
The presence of HRP may delay this transition with APG concentration (0.6 mM without
HRP, about 1 mM with HRP). This may be due to adsorption of APG monomers to HRP
prior to the transition, thus reducing the APG concentration in the bulk. Furthermore, the
overall line width is higher when HRP is present below the sphere-to-rod transition. This
may be caused both by an increase in viscosity and by interaction between APG and
HRP.
10
15
20
25
30
35
40
45
50
0.01 0.1 1 10 100
APG + HRP
APG
[APG] mM
Lin
e w
idth
(H
z)
105
5.3.4.25.3.4.25.3.4.25.3.4.2 NOESY (Nuclear Overhouser Effect Spectroscopy)NOESY (Nuclear Overhouser Effect Spectroscopy)NOESY (Nuclear Overhouser Effect Spectroscopy)NOESY (Nuclear Overhouser Effect Spectroscopy)
Figure 5-41 - NOESY spectra with water signal suppression using excitation sculpting with gradients, at different
concentrations of APG at constant HRP concentration. A possible interpretation in terms of APG aggregate
structure is presented to the left.
Head and tail closer in space
ppm
0.51.01.52.02.53.03.54.04.55.0 ppm
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
NOESY - 3 mM APG + HRP
NOESY - 1 mM APG + HRP
ppm
0.51.01.52.02.53.03.54.04.55.0 ppm
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
ppm
0.51.01.52.02.53.03.54.04.55.0 ppm
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
NOESY - 9 mM APG + HRP
106
As shown in Figure 5.41 the NOESY spectra show that the sphere-to-rod transition
results in a cross-peak in the 3 and 9 mM APG samples which suggests that the tail
methyl group is fairly close to the head group. This further supports the formation of a
lamellar bilayer structure. Put together with the rod formation indicated by the line width
data, the above combined results point towards the formation of tubes. These line
broadening effects observed in phospholipid bilayer systems by NOSEY NMR have also
been reported by Chen and Stark [90]. They have also reported a degree of interaction
between the CH2N of the head group and CH3 from the alkyl chain that is unexpected for
the traditional bilayer model. They have also proposed an intercalated, anti-parallel,
spacing of the lipids so that the tail end of one layer extends through the bilayer to
interact with the head group of the other surface, much in the same way as our proposed
APG tube like structure. In addition to this formation, Chen and Stark also suggest the
possibility of tail chains bending backwards to interact with its own head group.
5.3.55.3.55.3.55.3.5 IR investigation of BioIR investigation of BioIR investigation of BioIR investigation of Bio----Surfactant Surfactant Surfactant Surfactant degradationdegradationdegradationdegradation via enzyme hydrolysisvia enzyme hydrolysisvia enzyme hydrolysisvia enzyme hydrolysis: : : :
In a parallel study, the possibility of enzymes digesting bio-surfactants that have been
proposed for use in detergents was also investigated. IR spectroscopy of surfactant-
enzyme combinations were carried out to determine if there is any bond breaking or
formation when enzyme and surfactants are introduced to each other. Two different
surfactants were studied, n-Dodecyl β,D Maltoside and SDS. IR spectrum were taken of
107
both individually and then each with approximately 5.0 g/L concentration of cellulase
enzyme. Enzyme alone in water was also studied.
Figure 5-42 - Sodium Dodecyl Sulfate
Figure 5-43 - n-Dodecyl β,D Maltoside
108
Figure 5-44 - IR spectrum of SDS and Dodecyl Maltoside between 1350 and 850 cm-1 wavenumber range.
SDS peaks can be identified by S=O, and S-O stretching in the 1250-1212 and 972 – 955
Cm-1 range respectively. There are a series of C-O-C stretching modes that can attribute
to the series of peaks between 1200 and 950 cm-1.
Comparing the IR spectrum of dodecyl maltoside before and after being introduced to the
enzyme:
1370.0 1340 1320 1300 1280 1260 1240 1220 1200 1180 1160 1140 1120 1100 1080 1060 1040 1020 1000 980 960 940 920 900 870.0cm-1
A
SDSS=O Stretch
S-O Stretch
Dodecyl MaltosidePossible C-O-C stretching
109
Figure 5-45 - IR spectrum of SDS and Dodecyl Maltoside between 1160 and 960 cm-1 wavenumber range.
In this plot most of the major peaks associated with the DM head group are present when
in the presence of enzyme. Some notable differences worth further investigation are shifts
in peaks at 1038 and 1027, as well as the total disappearance of a peak at 1054. cm-1
Thorough investigation into the cellulase cleavage mechanisms will be necessary to give
any relevance to these peaks shifts. Initial searches indicate that the enzyme cleaves the
o-glycosidic bonds via a donation of a proton by a carboxyl group of glutamic acid to the
one to be cleaved. There is a stabilization of the carbonium ion intermediate by aspartic
acid. The reaction sequence is finished by an atack of nucleophilic water.
1161.7 1150 1140 1130 1120 1110 1100 1090 1080 1070 1060 1050 1040 1030 1020 1010 1000 990 980 970 959.1cm-1
A
DM Before Enzyme
DM W/ enzyme 1075.701075.70
1040.75
992.85
1024.18
1107.78
1148.30
1075.70
1038.041027.05
1055.21
1107.78
1148.30
995.26
110
From this understanding, it might be possible to speculate that the peak at 1054 that
disappears may belong to the intermediate o-glycosidic bond shown in figure 5.39 and
disappears after introduction to the enzyme and it is cleaved. This is speculative at this
point and will require more study to confirm.
Figure 5-46 - Schematic of cellulase hydrolysis mechanism
HATR Spectroscopy of Dodecyl Maltoside and Enzyme mixtures further suggest
enzymatic activity and cleavage of the head group. Comparing the IR spectrum of
dodecyl maltoside before and after being introduced to the enzyme:
111
1200 1100 1000 900
0.000
0.005
0.010
0.015
0.020
0.025
-0.010
-0.005
0.000
0.005
0.010
0.015
0.020
0.025
DM
DM
cm-1
DM + Enzyme
DM
+ E
nzym
e
Dissapearing Peak After Enzyme Exposure
Figure 5-47 - FTIR of Dodecyl maltoside with a-Cellulase indicating possible loss of C-O-C bond
This disappearing peak may be attributed to the C-O-C stretching connecting the sugar
groups in the DM head structure.
Figure 5-48 - Proposed reaction of surfactant catalysis
112
Evidence of enzymatic activity can also be seen in the IR spectrum of the cellulase
enzyme before and after introduction to dodecyl maltoside surfactant. C-O Stretching
peaks are indicative of the carboxyl group responsible for glycosidic bond catalysis. In
the plot below (Figure 5.37), we can see a shift in this peak after introduction of DM
surfactant strongly suggesting there is a deprotonation of the carboxyl group, which is
step in the catalysis process.
1800 1750 1700 1650 1600 1550 1500
-0.004
-0.003
-0.002
-0.001
0.000
0.001
0.000
Care
nzym
e C
ellu
lase
Wavenumber (cm-1)
Shift after exposure
to DM surfactant
Figure 5-49 - FTIR from 1800 to 1500 cm-1 possibly showing the carboxyl group located at the cellulase active site.
113
Evidence of enzymatic activity can also be seen in the IR spectrum of the cellulase
enzyme before and after introduction to dodecyl maltoside surfactant. C-O Stretching
peaks are indicative of the carboxyl group responsible for glycosidic bond catalysis. In
the above plot, we can see a shift in this peak after introduction of DM surfactant strongly
suggesting there is a deprotonation of the carboxyl group, which is step in the catalysis
process.
114
5.45.45.45.4 Objective 3: Objective 3: Objective 3: Objective 3: Protein structure dynamics:Protein structure dynamics:Protein structure dynamics:Protein structure dynamics:
The third objective of this study is to investigate the “flexibility” of an enzyme, as a
physical characteristic that is being affected by interactions with surfactants. This next
section describes two time resolved fluorescence techniques used to study the structural
mobility of subtilisin and horseradish peroxidase enzymes used for this work.
5.4.1 Life-time fluorescence: Anisotropy
The anisotropic decay of dansyl probe attached to subtilisin active site has been measured
in buffer solutions and compared to a solution of dansyl probe in ethanol. Fitting a decay
profile to the anisotropy curve is done through Datamax software. The number of decay
correlation times between 1 and 3 were tested and the parameters with the chi-squared
values closest to 1 are displayed in the chart below.
Column1 Probe protein Protein plus 8mM brij 30
chi sq 1.005 1.020 1.090
θ 1(s) 4.90E-10 6.80E-10 5.19E-10
θ 2(s) 3.60E-08 2.30E-06
θ 3(s) 6.99E-06
Table 5.2- Significant anisotropy decays from labled subsitlin protease
115
Table 5.2 Shows the decay correlation times for the models which best fit the anisotropy
loss of dansyl labeled subtilisin. The first column denotes decay times for probe
suspended in ethanol indicating a singular decay rate. Center: probe labeled subtilisin
yield two decay rates. Presumably one for the probe spinning about a covalent bond and
the second a result of the whole protein rotation. Right: Labeled protein in 8mM
surfactant solution. Here, the fast motion of the probe is less hindered relative to the
probe without surfactant.
A singular decay time is the best fit for the probe in solution, shown above as θ1. This is
expected as the unbound probe should spin randomly at a uniform rate. Moreover, the
decay time is rapid, about half a nanosecond; this indicates that the probe in solution is
relatively unhindered by its local environment. Alternatively, a function that is the sum of
two decay times’ best fit the dansyl labeled enzyme. A decay function with multiple
decay correlation times is typically an indication of non-spherical particles. However,
considering that subtilisin is a globular protein, and that the differences between the two
decay times are vastly different (two orders of magnitiude), it is more likely that the
faster correlation time corresponds to the probe rotating about the covalent bond while
the slow rotation can be attributed to the whole protein tumbling.
It is important to note the decay rate increase in θ1 from the free probe, indicating a
slower rotation of the probe. This is consistent with the probe being attached to the
enzyme active site, hindered by adjacent peptide residues. Repeating this experiment at a
relatively high concentration of Brij-30 surfactant resulted in a reduction in rotational
116
correlation time, from .68ns to .52ns, indicating that the probe is more freely spinning.
This is change is regarded as the active site becoming more flexible, possible influenced
by the non-ionic surfactant. It might be argued that the surfactant is unfolding the protein
structure, hence the increased probe mobility. But, enzyme activity if found to be
improved in this surfactant range, strongly suggesting that the structure is preserved.
Further experiments include expanding the selection and concentration range of
surfactants as discussed in objective 1. It should also be noted that, in the presence of the
Brij-30 surfactant, a three exponential decay was the best fitted solution with a chi-sq
value of 1.09. The global rotation of protein is slowed by two orders of magnitude, to the
order of microseconds, presumably from interaction with surfactant micelles.
This initial experiment demonstrates subtilisin interaction with the surfactant resulting in
both slower global rotation of the protein and faster probe mobility about the active site.
The next set of these experiments were to monitor changes in the active site mobility as a
function of surfactant concentration.
117
Figure 5-50- Rotational correlation time (blue diamonds) and subtilisin activity (orange squares) as a function of
Brij-30 concentration.
Figure 5.50 plots the fast rotational correlation time as a function of surfactant
concentration (blue) for reference the initial activities of subtilisin are also plotted here
(orange). The initial correlation time as previously measured was .68 ns and .52 at the
highest Brij-30 concentration measured. The correlation times measured at the
concentrations in between 0.0 and 8.0 mM indicate an exponential increase in probe
mobility with increasing surfactant concentration. In parallel, the reaction rates of
subtilisin increase at a similar rate with additional surfactant. This correlation strongly
indicates a connection between active site flexibility and enzyme activity as a function of
surfactant.
0.9
0.95
1
1.05
1.1
1.15
1.2
1.25
1.3
0.5
0.54
0.58
0.62
0.66
0.7
0.001 0.01 0.1 1 10 100
Reacti
on
Rate
(r/
r 0)
Co
rrela
tio
n t
ime (
ns)
Brij - 30 Conc. (mM)
118
The same preparation and measurements were carried out for APG, the sugar based
surfactant also shown to increase enzyme activity. Like Brij-30, additions of APG
reduced the local viscosity of the protease active site with the addition of surfactant.
Figure 5-51 - Decay correlation times of dansyl probe attached to subtilisin as a function of APG detergent.
Unlike Brij-30, APG activity increases in a specific concentration at 1.0 – 2.0 mM and
drops back to normal activity values above 3 mM (see figure 5.16). We attribute this drop
in activity to the formation of rod-like micelles that increase the solution viscosity,
slowing down the mass transfer kinetics, limiting enzyme – substrate interactions.
5.70E-10
5.80E-10
5.90E-10
6.00E-10
6.10E-10
6.20E-10
6.30E-10
0.1 1 10
Anis
otr
opy D
ecay C
orr
ela
tion (
s)
APG Conc (mM)
119
5.4.25.4.25.4.25.4.2 Time resolved fluorescence to measure flexibility of horseradish peroxidase: Time resolved fluorescence to measure flexibility of horseradish peroxidase: Time resolved fluorescence to measure flexibility of horseradish peroxidase: Time resolved fluorescence to measure flexibility of horseradish peroxidase:
Horse radish peroxidase is another model enzyme that was used in this study. It has been
shown in the enzyme activity portion of this thesis that HRP is also positively influenced
by APG and Brij-30 surfactant, (figure 5.16). However, there is no serine residue located
at the active site for convenient dansyl probe attachment, limiting the application of
fluorescence anisotropy decay analysis. However, as discussed in the methods section,
the intrinsic decay of HRP’s naturally occurring fluorescence can be measured to
compare overall mobility of the structure. Figure 5.40a is the raw data of this decay of
HRP in buffer. Also in this graph is the fitted model of three exponential decays used to
extrapolate decay correlation times τ.
Figure 5-52 -Top. Raw data (Blue) and fitting of sum of 3 exponential decay functions (orange) of hrp in buffer.
Bottom: residual of fitted line vs raw decay data.
0.1
1
10
100
1000
10000
0 1000 2000 3000 4000 5000
Co
un
ts
channels
Decay
Fit
-10
0
10
0 1000 2000 3000 4000 5000
Residuals
120
Exponential decay of fluorescence intensity as a function of time, I(t), can be expressed
as a sum of exponential decays
Equation 5-2
Where k is the number of distinct decay rates, t is the time in seconds, τ is the decay
parameter in seconds, and ak is the contribution factor for that particular decay function.
From the data above the decay parameters and percent contributions are listed below.
T1 (s) a1 T2 (s) a2 T3 (s) a3
2.60E-09 12% 9.10E-09 8% 5.50E-11 80%
Table 5.3 - Three parameters of intrinsic HRP fluorescence decay. The majority of the decay is attributed to the τ3
parameter.
The majority of the decay is attributed to the τ3 parameter occurring at 5.5e-11 seconds,
which is two orders of magnitude faster than the second fastest, happening at 2.6
nanoseconds. This fast decay is attributed to resonant energy transfer of from the excited
tryptophan to the heme group located at the enzyme’s active site.
121
Using this technique we will measure changes in the decay parameter of HRP in dodecyl
maltoside and alkyl polyglucosides. Referring back to <section> it was shown that APG
will induce greater enzyme activity while DM, with a very similar head group structure,
does not.
The fast decay parameter was measured and plotted over the range of these surfactant
concentrations to further investigate the correlation between activity and flexibility in
HRP. The results are plotted below in figures 5.47 and 5.48. To reiterate, I hypothesize
that a faster decay rate can be attributed to a more flexible enzyme structure. The
opposite is also postulated to be true: that a slower decay rate is indicative of a more rigid
structure. At the lowest concentration measured (0.05) mM the decay correlation times of
HRP in both DM and APG have decreased (i.e. faster decay) to ~4.2e-11s. The decay
data and the multiple exponential fitting, as well as the residual data can be seen below
for APG concentrations: 0.05, 2.0 and 10.0 mM.
122
Figure 5-53 - Decay data and multiple exponential fitting (including residuals below) of HRP enzyme in 0.05 mM
APG.
-2000
0
2000
4000
6000
8000
10000
12000
2E-08 2.5E-08 3E-08 3.5E-08 4E-08
Fit Decay
-6
-4
-2
0
2
4
2E-08 2.5E-08 3E-08 3.5E-08 4E-08
123
Figure 5-54 - Decay data and multiple exponential fitting (including residuals below) of HRP enzyme in 2.0 mM APG.
-2.00E+03
0.00E+00
2.00E+03
4.00E+03
6.00E+03
8.00E+03
1.00E+04
1.20E+04
2E-08 2.5E-08 3E-08 3.5E-08 4E-08
Fit Decay
-8
-6
-4
-2
0
2
4
2E-08 2.5E-08 3E-08 3.5E-08 4E-08
124
Figure 5-55 - Decay data and multiple exponential fitting (including residuals below) of HRP
enzyme in 10.0 mM APG.
-2.00E+03
0.00E+00
2.00E+03
4.00E+03
6.00E+03
8.00E+03
1.00E+04
1.20E+04
2E-08 2.5E-08 3E-08 3.5E-08 4E-08
Fit Decay
-10
-5
0
5
2E-08 2.5E-08 3E-08 3.5E-08 4E-08
125
0.1 1
0.6
0.8
1.0
1.2
1.4
Avg (
AP
G)
Concentration (mM)
Figure 5-56 - Fluorescence decay times as a function of surfactant (APG) concentration
126
0.1 1
0.6
0.8
1.0
1.2
1.4avg (
DM
)
Concentration (mM)
Figure 5-57 - Fluorescence decay times as a function of surfactant (dodecyl maltoside) concentration
127
5.55.55.55.5 Objective 3aObjective 3aObjective 3aObjective 3a: : : : Protein flexibility and orientation at particle surfacesProtein flexibility and orientation at particle surfacesProtein flexibility and orientation at particle surfacesProtein flexibility and orientation at particle surfaces: : : :
Model protein CRM 197 was introduced to AlPO4 in pH controlled buffer solution. UV-
Vis was used to quantify the amount of protein adsorbed onto the particle surfaces.
Figure 5-58 - Amount of CRM-197 adsorbed onto AlPO4 particles at pH 4 and 5 as a function of initial protein
concentration.
These results indicate that pH 5 is the optimal condition for binding of the two. CRM
adsorbs onto the particles at a linear rate up to 150 ppm of initial protein concentration.
Knowing the optimal conditions for adsorption, time resolved fluorescence and FTIR
techniques were employed to probe the orientation of the adsorbed protein, and perhaps
change in protein structure dynamics on the particle surface.
0
0.05
0.1
0.15
0.2
0.25
0
20
40
60
80
100
120
140
0 50 100 150 200 250 300 350
Ad
sorb
ed
CR
M (
g/g
AlP
O)
Ad
sorb
ed
CR
M (
pp
m)
Initial CRM 197 (ppm)
pH 5
Series2pH 4
128
Intrinsic time resolved fluorescence is a suitable technique to measure conformation
changes of CRM-197 or the protein’s interactions with particle surfaces because of the
few Trp residues located throughout the structure. Trp 206 and 281 are located within an
alpha-helix dense region, Trp 398 amongst a series of beta sheet conformations, while
Trp 150 and 53 are attached to a random coil at different region of the protein. Changes
in the fluorescence of these residues can give information such as the polarity of its
immediate environment. Lifetime-fluorescence will provide the ability to differentiate
between the decays of the 5 Trp throughout the structure, allowing us to better define
which regions are interacting with solvent, or particle surfaces.
Figure 5-59 - CRM-197 structure with Trp residues highlighted
129
Time resolved fluoresence data for CRM in buffer at pH 7.0 and 5.8 as well as CRM on
AlPO4 particles at pH 5.8.
Figure 5-60 - Fluorescence decay times resolved from CRM in pH 7 and 5.8 buffer and adsorbed onto AlPO4 at pH
5.8
1.6E-091.9E-09
6.1E-095.8E-09
4.5E-09
6.1E-11 9.9E-110
1E-09
2E-09
3E-09
4E-09
5E-09
6E-09
7E-09
CRM pH 7 CRM pH 5.8 CRM pH 5.8 ALPO
De
cay
lif
e t
ime
(s)
t1
t2
t3
130
Figure 5-61 - The decay contribution of each decay time.
Figure 5.44a illustrates the time decays that are resolved from the time resolved
fluoresence data of CRM-197 at pH 7, 5.8 and on a particle surface. In buffer, at a neutral
pH, there are two distinct decay times, one at 1.6 ns and the other 6.1 ns contributing
30% and 70% of the decay respectively. Both of these decay rates are similar to each
other within the nanosecond range, which is typical for free tryptophan in solution. The
slower decay time can be attributed to tryptophans buried further in the protein’s
structure, away from the polar solvent.
Adjusting the pH to 5.8 did not drastically affect these decay times. However a new,
much faster decay time (6.1e-11s), arises. This new decay pathway contributes
30.0% 30.0%
70.0%
55.0%
20.0%15.0%
80.0%
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
CRM pH 7 CRM pH 5.8 CRM pH 5.8 ALPO
De
cay
Pa
ram
ete
r C
on
trib
uti
on
(%
)
t1
t2
t3
131
significantly to the over decay profile (15%), taking away from the slower of the two
original decay contributors (T2) which is now reduced from 70% to 55% from lowering
the pH.
The appearance of such a short decay time can be attributed to a tryptophan species
becoming more mobile as the enzyme structure as a whole gains greater flexibility.
132
5.65.65.65.6 Objective 3b:Objective 3b:Objective 3b:Objective 3b: Modeling enzyme activity as a function of surfactant Modeling enzyme activity as a function of surfactant Modeling enzyme activity as a function of surfactant Modeling enzyme activity as a function of surfactant
interaction and concentration.interaction and concentration.interaction and concentration.interaction and concentration.
Viparelli and coworkers created mathematical models to describe changes in enzyme
activity as a function of surfactant concentration. These models were heavily based on the
kinetics of protein/micelle interactions as well as the partition of enzymes between bound
water / free water / and micelle core. Ultimately, changes in enzyme activity would be a
function of the degree of interaction between the various phases of water found about a
micelle or bulk solution. In this work, the volume of bound water about the micelle is
given as:
�� = ����� ��� !"#$
Equation 5-3 - Determination of bound water volume in a surfactant system
Where Dn is the concentration of surfactant, ψ is the surfactant number per micelle, ϴ,
the number of bound water molecules, and �� !"#$ the molecular volume of the bound
water. Furthermore the partition between enzyme at a micelle bound water layer and
those in bulk water can be expressed as:
%�&' + �()� *+⇔ ����
Equation 5-4 - Equilibrium reaction between micelles and free enzyme.
Where [Ef] is the concentration of free micelles, [Mn] is the concentration of micelles,
[Eb] is the concentration of bound enzyme, and KE is the reaction constant of binding.
Viparelli derived equations for the concentrations of free and bound enzymes as follows:
133
%�&' = ��"�1 + -.�()�
Equation 5-5 - Concentration of free enzyme for a given, Micelle concentration
���� = �-.�()���"�1 + -.�()�
Equation 5-6 - Concentration of enzyme in bound water for a given concentration of micelles
What is missing from this theoretical work is the mechanism that would explain why
enzyme activity should change in these different water environments is not explored, nor
is there an explanation as to the driving force that would compel the enzyme to seek out
the micelle core, surface layer or solution bulk.
This dissertation investigated and identified those mechanisms, allowing for the
expression of a more comprehensive model of interaction that would take into account
the physical factors that drive enzyme partition throughout the water layers as well as
identifying changes in flexibility as the mechanism that drives the actual change in
enzyme activity. Furthermore, we now know that micelles that induce more flexible
protein structures are those that have lower cloud point temperatures, that is, their bound
water layers are less stable than those micelles that do not increase enzyme flexibility and
activity. These findings further strengthen past assertions that the bound water layers
about the micelles are responsible for changes in enzyme activity in surfactant/enzyme
systems. This work has demonstrated that changes in protein structural flexibility is a key
factor that links bound water partitions to changes in enzyme activity. However, this
leads to an additional question not yet explored by this work: Why does the hydration
sphere about a micelle affect the flexibility?
134
The relationship between solvent dynamics and protein function is still a contentious one.
With many proponents arguing that small and fast motions of the protein structure add up
to quantifiable effects on enzyme catalysis [54], while others argue that they are driven
primarily through changes in electro-static interactions and bond-formation energies [91,
92]. Since our system does not involve a change in ionic strength, rather a change in
water structure induced by non-ionic surfactants, I have further investigated changes in
dynamics as the link between micelle structure and enzyme activity.
A recently proposed theory solvent slaving links solvent mobility with protein processes
and is explored by Fenimore, Fraudenfelder, and coworkers in 2002 [93, 94]. That study
proposed that there are two classes of protein motions, those that follow the classic
Arrhenius relationship between heat and molecular mobility and those that are linked to
the rotational frequency of solvent motions. Myoglobin was cited as an example of this
phenomenon, where the release of CO is proportional to the dielectric relaxation times of
the surrounding solvent. The rate of the release and of the respective protein motions
where related to solvent motions by the following equation:
��� = �01#2 3��⁄
Equation 5-7 - Frudenfelder relationship between solvent slaved processes and
solvent relaxation rates
Where ���is the slaved protein function, kdiel is the relaxation rate of the solvent’s
dielectric constant, and n(T) is a correlation constant translating the faster water rotation
rates to the slower motions of protein dynamics.
135
Solvent-slaving theory
Putting together elements from Viparelli’s and Fraudenfelder’s work, a model that
describes surfactant modified enzyme catalytic rate (����) as a function of micelle
number and how it relates to the volume of pseudo phase water around the micelle
surface. By combining equations 5.3 and 5.7, we now have an expression for the enzyme
reaction rate we a function of enzyme concentration in micelle bound water phase and in
the bulk:
Equation 5-8 - Enzyme reaction rate as a function of enzyme in micelle bound and water regions
Where kdb and kdf are the dielectric relaxation rates of the solvent in bound water and bulk
regions, respectively. NT is the relaxation correlation factor that translates the rate of
solvent motion to observed protein function rate. By plugging in equations 5-5 and 5-6
into bound and free enzyme concentrations, giving a new relationship between enzyme
activity as a function of micelle number:
Equation 5-9 - enzyme reaction rate as a function of micelle number
What has not yet been considered is the actual very small percentage of bound water
volume in the APG surfactant concentrations in which enhanced activity is observed.
���� = 5�0�3� 6 ���� + 7�0&3� 8 %�&'
���� = 5�0�3� 6 9�-��(:���;�1 + -��(:� < + 7�0&3� 8 9 ��;�
1 + -��(:�<
136
Also, what factors that arise between densely packed and loosely packed DM and APG
surfactants have not yet been differentiated. Some mechanism of interaction between
micelles that drives enzyme and micelles together so that their water structures would
affect one another.
To reconcile this inconsistency we further investigate water structure and its role in
determining aggregation. K. Collins proposed a theory on how water structure is a
stronger driving force that influences ion attraction than previously assumed. Collins
proproses ‘law of matching water affinity’ where the similarity of water affinity of
oppositely charged ions play just a critical role in their ability to form pairs as do their
electrostatic interactions[95, 96]. Ultimately, ‘like sees like’, small ions with greater
charge densities result in highly ordered water about them, but their electro-static
attractions overcome this boundary and push out the intermediate water. Alternatively,
large ions of low charge density poorly adhere to water, thus big ions of opposite charges
also tend to form pairs. It is small, well hydrated ions, and large loosely hydrated ions
that do not form pairs.
It is not unreasonable that the same phenomenon is happening in this system of micelles
and enzymes. Non-ionic surfactants do not have strong electro static forces that induce
attraction or repulsion, so sphere of hydration is the only factor that determines
dispersion. In this, it is possible that micelles carrying strongly bound water will remain
dispersed. Those with weakly bound hydration shells will succumb to Van Der Waals
forces, and aggregate to minimize the surface area of weakly bound water. This would
not be unlike the forces that result in the surfactant “clouding” phenomenon commonly
observed in non-ionic surfactant systems. In the case of APG’s, the greater degree of
137
head group exthoxylation resulting in higher cloud point temperatures, i.e. greater
temperature needed to disrupt hydration layer and induce clouding. Also longer chain
lengths lower cloud point. This is consistent with what has been observed in
heterogeneous APG where micelles initially form from the longest chain lengths with the
shortest head groups. This aggregation between micelle and enzyme would be reflected
in the ‘interaction rate’ parameter Ke
Figure 5-62 - a) (left) hydration sphere of densely packed micelles such as that of dodecyl maltoside and high conc.
APG consists of tightly bound water. b) (right) loosely packed micelles result in more mobile hydration sphere
Figure 5-63 - Minimization of loosely bound hydration layers leading to enzyme/micelle interaction
138
Expanding on this concept, the water structuring not only around the micelle but relative
to that in the bulk needs also be considered. Because of the high ionic strength of
potassium phosphate buffer, the bulk water is rather ordered. This results in further
increasing the driving force that draws hydration water away from the softly formed APG
micelles, resulting in diminished hydration spheres.
At this point it is speculated that the hydration about the enzyme is also relatively
disordered and weakly bound, and thus seeks out the spaces near the micelle. The waters
near the micelles maybe fewer and very likely more mobile. It’s likely the one or both of
these factors interact with the enzyme structure to globally effect fluctuation modes of the
overall structure.
5.75.75.75.7 Results and discussion conclusionsResults and discussion conclusionsResults and discussion conclusionsResults and discussion conclusions
In this dissertation the effects of various surfactants on enzymes have been measured via:
• Enzyme activity & kinetics
• Surface tension
• Time resolved fluorescence
• Electron spin resonance
• FTIR
• Dynamic light scattering
139
The main goals of this project was to 1) Identify which surfactants induced increased
enzyme activity 2) Elucidate the structures or surfactant properties that are associated
with this enhancement and 3) Explore protein flexibility as a mechanism by which
surfactants are affecting enzyme activities.
From the activity experiments, several non-ionic surfactants were shown to improve the
product yield within the experiment time, long chain exthoxylate surfactants Brij-30 and
Brij-35, as well as an alkyl polyglucoside, sugar-based surfactant mixture, provided to us
by an industrial partner. Of the three surfactants, interaction with APG resulted in the
greatest degree of enzyme activity increases in the two model enzymes tested, subtilisin
protease and horseradish peroxidase. In addition, APG induced enzyme hyper-activity in
both enzymes in a narrow surfactant concentration, peaking at 2-3mM. In comparison,
the activities horseradish peroxidase was also tested in the lab grade, single surfactant,
dodecyl maltoside, also a sugarbased surfactant with a sized head and tail group similar
to the average of what would be found in the APG mixture. This control surfactant did
not induce significant changes to the activity of the HRP enzyme.
To better understand the differences between these two surfactants that would cause one
to effect greater enzyme activity while the other not, surface tension, dynamic light
scattering, NMR and ESR were employed to characterize aggregation behavior of the two
surfactants. APG surface tension results reveal two breaks in the surface tension vs
concentration plot, suggesting that there are two significant aggregation events that occur
140
as bulk concentration increases. We will refer to these transition points as critical micelle
concentrations (CMC) I and II.
The first CMC at 0.3 mM indicates an initial formation of surfactant aggregates. After
this point, the surface tension is expected to remain constant with surfactant
concentration. In the case of this APG sample, the surface tension continues to decrease
with concentration from 0.3 mM to 5.0 mM. This phenomenon is common in mixed
surfactant systems (ref mixed systems APG) where the Gibbs free energy of micelle
formation is not uniform for each component of the mixture. This results in the less
soluble moieties in the mixture to precipitate into micelles first, while detergents with
higher CMC continue to stay in monomer form, collecting at the surface and lowering
surface tension. This trend continues until the solubility limit for the most soluble
surfactant moiety is reached, signified by the “second” CMC depicted in this APG
system. It is also worthy to note that it is during this micelle transition phase between 0.3
and 5.0 mM where HRP activity is shown to increase, peaking at 3mM of surfactant
concentration.
Alternatively, DM demonstrates a singular inflection point suggesting that the
homogeneous distribution of surfactants immediately forms one structure of micelle
above the CMC and there is no further change in that structure with increasing surfactant
concentration, only an increase in the number of aggregates.
141
Particle size measurements were done of these surfactant systems to further elucidate the
differences between APG and DM. APG at higher concentrations, 5, 10 and 20 mM form
micelles that range from 60 to 70 nm. This is considered a large micelle, approximately
10 times larger than expected if the surfactant were to aggregate into a spherical,
monolayer configuration. Since rod-like morphologies of APG has been well observed,
(ref on APG) these large particle sizes observed can be attributed to the formation of such
elongated structures.
However, these well-defined, rod-like structures, can only be seen at concentrations
greater than 5 mM, that is, past the second “CMC” identified with surface tension. At
3mM and below, DLS can only pick up extremely large particles averaging in 1.5
microns in size, approximately 2 orders of magnitude larger than the already large rod-
like configuration formed at higher concentrations. A continuous structure with a core of
hydrophobic tails and surface of hydrophilic head is unlikely at this scale. This
phenomenon suggests a “clouding” phenomenon [97] wherein micelles of non-ionic
surfactants further aggregate with each other because they lack the surface charge nor a
stable hydration cage to overcome the forces attracting them to neighboring aggregates.
The result is “super-aggregates” large enough to render the solution cloudy. This is would
explain the hazy appearance of the solution at these middle concentration ranges.
Dodecyl maltoside did not exhibit any of the multi-phase size shifts as seen in the APG
mixtures. DLS indicated that the single surfactant formed small (~8-10 nm) micelles
upon increasing the concentrations past CMC. Further additions of surfactant did not
change micelle size, but increased DLS signal counts per second, indicating and increase
in micelle count.
142
Thus far, surface tension and particle sizing experiments have demonstrated significant
physical differences between the aggregation characteristics of APG and DM. Moreover,
it appears that the transition phase between the two CMC’s observed in APG has greatest
effect on the enzyme activities. Alternately, while dodecyl maltoside has similar head-
group structure and tail lengths, the micelle structures that this homogeneous surfactant
formed are significantly different than the mixed APG system, which correlates well to
the differences in enzyme activity that the two surfactants induce. ESR and NMR were
used to further investigate the differences in physical properties between the aggregates
of these two surfactants, and how they might ultimately affect enzyme activity.
ESR was used to probe the formation of micelles and their packing density with and
without the presence of HRP enzyme. Surface tension measurements indicate micelle
formation starts around 0.2 – 0.3mM for both APG and DM. However, ESR indicated the
formation of micelles occurring at a higher concentration, 1mM, for both surfactants.
This discrepancy may be attributed to the stearic acid probe, which maybe suppressing
micelle formation. Above 1mM the differences in micelle evolution can be observed. In
DM, the probe mobility rapidly decreases (inverse of rotation correlation time) and then
plateaus, supporting surface tension and DLS observations of formation of spherical,
monolayer micelles as the final conformation of the micelle aggregates. For APG, it was
inferred from surface tension and particle sizing that an intermediate micelle structure
phase is formed before the surfactant concentration is sufficiently high to support the
final, rod-like, micelle structures. ESR further supports this hypothesis, demonstrating
that the micelle packing increases gradually after 1mM appearing to stabilize as the
143
concentration approaches 10mM. This slow increase in packing also supports the
proposed hypothesis of gradual incorporation of surfactants from the APG mixture, with
less soluble portions aggregating in the bulk first. As surfactant concentration increases
the more soluble portions start to integrate into the micelles, increasing packing density
of the aggregate as reflected in ESR. Moreover, the mobility of the probe in the DM
micelles is more restricted that those in APG. This agrees with the initial assertion that
DM forms mono-layered, spherical micelles, a configuration that most densely packs the
hydrophobic tail portions. Local viscosity of the probe in APG does not reach the same
degree of immobility and the 2D NMR spectrum indicates significant proximity between
the C13-C14 and hydrophobic head groups. Combined with the ~70 micelle size diameter
measured by DLS, we can infer an anti-parallel, bilayer morphology associated with the
final aggregate structure. This can come in the form of a hollow spherical vesicle or tube
like structure. It is important to note that it is not this bilayer formation that induces
greater enzyme activity. When the APG concentration is sufficiently high for these large
structures to form, the activity is seen to return to initial values. Putting the information
from surface tension, DLS and ESR together we propose the following schematic to
illustrate the differences in micelle structure between APG and DM.
144
Figure 5-64 - Schematic differentiating micelle formation in DM (top) vs APG (bottom)
NMR also provides greater insights into the formation of APG micelles and how this is
affected by the presence of proteins such as the horseradish peroxidase enzyme. Peak
line-widths indicate changes in slow rotation of whole aggregates and were used to
characterize the evolutions of APG structure as a function of concentration. Mirroring the
ESR results discussed above, NMR revealed a dual transition phase. The first occurs
shortly after the first CMC at approximately 0.4 mM resulting in aggregates that
gradually, but steadily, decrease in mobility with surfactant concentration. This agrees
with the current hypothesis that this concentration range results in super aggregates,
loosely packed micelles that further aggregate with adjacent micelles. Further increasing
the concentration at this range only adds more mass and volume to the super-aggregate,
resulting in an increasingly slower rotation of the particle. The rate of decreasing particle
145
motion with respect to concentration slows at 5mM, approximately where the surfactant
is expect to transition to the rod-like conformation.
While surface tension, DLS, ESR and NMR has yielded valuable information on the
distinct differences between the surfactant structures that do initiate enhanced enzyme
activity (APG) versus those that do not (DM), they have not been able to explore the
precise molecular mechanism by which these particular micelle conformations should
enact on the system to increase the rate of enzyme catalyzed reaction. As to date, several
theories have been provided to explain the mechanism that governs the interactions
between these non-ionic surfactants and the enzyme activity, ranging from influencing
local substrate concentrations about the micelles to preventing unproductive adsorption of
enzymes onto a substrate surface. These factors discussed in previous literature will all
play a role in determining enzyme activity, but what has yet to be suggested or
investigated in significant detail is how surfactant interactions with the enzyme may
affect the structural dynamics, or “flexibility” of the protein structure, thus affecting the
activity. This study closely examines the relationship between enzyme internal structure
mobility and function as a possible explanation for the changes in activity as they interact
with the surfactants and seeks to find a generic correlation between surfactant interaction
and enzyme activity that will be applicable to all systems involving a functional protein
and surface active agents.
146
Time resolved fluorescence was used to measure the changes in the internal structural
mobility of the two model enzymes, subtilisin protease and horseradish peroxidase. For
the subtilisin, a dansyl fluoride probe was selectively attached to the serine residue of the
active site, and its rotational mobility was measured with time resolved anisotropy
fluorescence. Exposing the labeled proteinase to two surfactants known to initiate an
increased activity response, APG and Brij-30, resulted in faster rotational correlation
times (i.e. increased mobility) of the dansyl probe.
HRP does not have a convenient attachment site for a fluorescent probe to take advantage
of. However, it does have a heme-group that quenches the intrinsic fluorescence of local
tryptophan residues. The rate of fluorescence decay is attributed to the physical distance
between between the trp and heme group and has been used by other groups as a means
of measuring conformation changes. This work theorizes that changes to this decay time
can also be indicative to changes in the rate and range of internal fluctuations. The greater
the degree of these motions, indicate a more flexible structure and would result in faster
energy transfers reflected by time resolved fluorescence measurements. Additions of
APG to the system increased the fluorescence decay rate, indicating a more flexible
structure, much in the same way that APG affected the subtilisin enzyme. As a
comparison, dodecyl maltoside, which did not induce faster enzyme activities, was
introduced to the fluorescence experiment and did not increase the HRP flexibility. This
is the first indication that these surfactants control the internal mobility of protein
structures.
147
Currently, we have rationalized that the interactions which induce greater enzyme activity
are primarily between the enzyme and micelles, as activity is closely linked to surfactant
concentration in APG, brij-30 and Brij-35 post CMC, this is especially evident in APG
systems, where the activities changes significantly according to the conformation of the
surfactant micelle. This opens a significant question as to how does a non-charged
micelle surface, or any surface for that matter, interact with a protein structure in such a
way as to manipulate the dynamics of its internal structures. Enzyme activity and
flexibility measurements indicate that this enhancement phenomena occurs at a very
specific concentration range for APG which is also associated with an increase in the
enzyme flexible mobility. While the factors that drive the inter-micellar aggregation
between non-ionic surfactant micelles, or “clouding”, are still not entirely clear, water
structure is commonly cited as the primary motivator in determining the stability of non-
ionic surfactant suspensions. More specifically, hydration spheres of ordered water about
a non-ionic surfactant provided the necessary repulsion to prevent non-ionic micelles
from aggregating. If this hydration sphere were disrupted by the addition of salts or an
increase in temperature above that surfactants “cloud-point”, this hydration sphere is
disrupted and micelle aggregation occurs, typically resulting in cloudy solutions.
This work as demonstrated that the APG surfactant mixture transitions through a micelle
morphology that results in a low cloud point, and it is during this surfactant concentration
range that increases enzyme flexibility, increasing activity. A wealth of studies highlight
148
the importance that hydration and solvent dynamics have on the structure and function of
proteins[39, 43, 46, 49]. In this work I hypothesize that the hydration spheres determined
by micelle interfaces dictate the solvent structure and dynamics which in turn interact
with the protein. The APG surfactants between 0.2 and 2.0 mM result in a depleted
hydration sphere, resulting in a localized volume of solvent which is more mobile than in
the bulk. The model that this work proposes describes protein function rates as a function
of two aspects of micelle – bound water – protein interactions: volume of bound water as
a function of surfactant concentration, and interaction kinetics between micelle and
protein. This interaction kinetics is an indication of the attractive forces between micelle
and protein which has been proposed by this work to arise from the ‘likenes’ of hydration
layers about protein and micelles. This concept of attraction of like hydration was
discussed by Collins and coworkers and creates a physical frame work which predicts ion
pairs form based not only on charge, but the similarities of their hydration
149
6 Conclusions and Next Steps
This project was initially focused on the possible degradation of bio-based surfactants by
enzymes often included in detergent formulations. Preliminary observations with FTIR
indicate the loss of the β-acetyl linkage of n-dodecyl-β-d-maltoside upon introduction of
a cellulase enzyme. In parallel, we also measured the activity of subtilisin protease in the
presence of various detergents. We anticipated that anionic detergents such as SDS would
eventually bind to, and denature the enzyme, reducing catalytic activity, while non-ionic
surfactants such as Brij-30 (Polyethylene glycol dodecyl ether), and APG (alkyl
polyglucoside) will preserve the enzymes native structure and thus, activity, as these
findings are often described in literature. Indeed, prolonged exposure (>24 hrs) of
subtilisin to the charged surfactants resulted in diminished activity relative to the enzyme
in buffer alone. While the precise mechanisms for this activity drop in anionic surfactant
has not been investigated experimentally in this work, literature consistently cites
denaturation of proteins in ionic surfactants. Unexpectedly, enzyme in solutions of the
non-ionic surfactants did not just maintain their activity, but exhibited increased activity
relative to that of the buffer only normal. This result could not be as easily explained as
with the loss of activity induced by the charged surfactants. Thus it is this “enhancement”
phenomenon and mechanism behind it that has become the central point of investigation
of this work.
150
Our group is not the first to observe this enzymatic “hyperactivity” induced by non-ionic
surfactants. Within the last decade several groups have reported similar results with
different enzymes including amylases and celluases, in the presence of a range of non-
ionic surfactants, such as: several from the Brij series of poly ethylene ethers, and Triton
X-100. While the phenomenon of enhanced enzyme activity in non-ionic surfactant
solutions were consistently observed by all of these studies, the explanations provided by
each one rarely agreed with one another. The goal of this work was then to propose a
mechanism that would universally explain this enzyme hyperactivity, and to test this
novel hypothesis. One mechanism that the previous studies have not considered was the
effect of the enzymes’ conformational mobility in determining overall activity. As such,
this was the concept central to this thesis: the observed increase in enzyme activity is the
result of changes to the enzyme’s structural dynamics induced by interactions with the
surfactants.
The pathway to testing this hypothesis was divided into three distinct objectives. 1) To
determine appropriate test systems of enzyme, substrate, surfactant and buffer, as well as
to develop method to measure enzyme activity of each system. 2) Elucidate the nature of
surfactants and enzyme aggregates if there are any. 3) Measure enzyme structure
flexibility and determine if there is a correlation between surfactant structure, flexibility
and enzyme activity.
151
By working towards these three objectives, we were able to: 1) define two model
enzymes that exhibited varied enzyme activity in certain surfactants, and a variety of
surfactants that generated different responses in this enzyme activity. 2) Identify the
differences between these surfactants, specifically those of their micelle structures, and
how these differences may be responsible for their interactions with enzymes affecting
enzyme activity. 3) Identify changes in enzyme structure flexibility as a prominent
mechanism by which certain surfactants are manipulating enzyme activity. In parallel to
these main objectives, we have also explored the effects that a hard- particle surface has
on the structural flexibility of adsorbed proteins and proposed a model on how Forster
resonant energy transfer can be affected by changes in protein flexibility.
Ultimately, completion of these objectives revealed a positive connection between
surfactant interaction with a protein structure flexibility and the impact that these changes
have on enzyme activity. They also illustrated the differences between the interaction of
enzymes with non-ionic surfactant micelles and protein’s adsorbed onto hard particle
surfaces via electro-static interactions. While the former increased the protein’s structural
flexibility, the latter is seen to decrease it. It is hypothesized in this work that bound
water interactions are a primary path way of interaction that allows non-ionic micelles to
increase flexibility of the enzyme structure and we have built model linking changes in
rates of enzyme function to the volume of water pseudo-phases generated by micelle
surfaces. The claim that solvent dynamics is the primary conduit between micelle
structure and enzyme flexibility is further substantiated by our initial enzyme activity
tests in Hoffmeister series ion solutions. Experiments to conclusively explore the role of
152
bound water structure between surfactant micelles and enzyme structures are planned for
the next steps for this study, but with the work described in this study, we have
determined that surfactants can affect the activity enzyme activity by manipulating
structural flexibility, a mechanism that should be considered in any system where a
surfactant is seen to have an impact on the activity of an enzyme or a functional protein.
Other studies have also observed differences in activity and flexibility as a function of the
Hoffmeister series [98].
6.16.16.16.1 NNNNext steps: ext steps: ext steps: ext steps:
In this body of work we have investigated the mechanisms that drive changes enzyme
activity in non-ionic surfactants. This behavior might easily be attributed to unfolding
and conformation changes of the protein structure in the presence of charged surfactants,
except that non-ionic surfactants typically do not disrupt the steady state structure of
proteins. On the contrary, many are used to preserve proteins Ex-Vivo. And only
recently, has increased enzyme activity in the presence of non-ionic surfactants have been
observed. This phenomenon hints that a proteins immediate environment has more
influence on its functions than previously considered. The majority of the studies that
have reported this enzyme enhancement via surfactant interaction phenomena have
attributed this to factors external to the protein itself, i.e. prevention of non-specific
interactions, concentrating substrate through volume exclusion, or dispersion of substrate.
This study is the first to consider the effects that the surfactants might have on the
protein’s structure directly, that is through manipulation of the structures overall
153
flexibility without disrupting secondary or tertiary native structures. In the future, a
simpler, simple-cleavage peptide substrate should be considered for future activity
measurements, as casein substrate used in the experiments described in this dissertation
are large complex globular proteins, which may result in large deviations in activity
results. While it was demonstrated in this work that flexibility is being affected by
interaction with particles and surfactant micelles, what remains to be conclusively proven
is that these changes in the protein’s structural dynamics are motivated by water
structural dynamics and a disruption of hydration boundaries extending outwards from
non-ionic surfaces. To conclusively test for such a mechanism, a superfast spectroscopy
technique will be needed to measure the dynamics of water in different surfactant
environments. Fortunately such techniques exist, such as the super-fast, time resolved IR
techniques pioneered by Professor Fayer [51, 99, 100] In parallel, molecular dynamics
simulations [37, 101] may provide precise data on how the motions of solvent molecules
impact the motions of proteins.
154
7 References
1. Antrim, R.L., et al., Industrial Enzymes: Enzymes in Nonfood Applications, in Enzymes in
Industry. 2004, Wiley-VCH Verlag GmbH & Co. KGaA. p. 155-244.
2. Sivakumar, G., et al., Bioethanol and biodiesel: Alternative liquid fuels for future
generations. Engineering in Life Sciences, 2010. 10(1): p. 8-18.
3. Wilson, D.B., Cellulases and biofuels. Current Opinion in Biotechnology, 2009. 20(3): p.
295-299.
4. Leech, D., P. Kavanagh, and W. Schuhmann, Enzymatic fuel cells: Recent progress.
Electrochimica Acta, 2012. 84(0): p. 223-234.
5. Somasundaran, P., Chin, M., Latosiewicz, U. T., Tuller, H. L., Barbiellini, B. and
Renugopalakrishnan, V, Nanoscience and engineering for Robust Biosolar Cells. Bio-
nanotechnology: Global Perspectives II, ed. R. D.E. Vol. 2. 2012, Boca Raton Fl: CRC
Press.
6. Moore, P.N., S. Puvvada, and D. Blankschtein, Role of the Surfactant Polar Head
Structure in Protein−Surfactant Complexa&on: Zein Protein Solubilization by SDS and by
SDS/C12En Surfactant Solutions. Langmuir, 2003. 19(4): p. 1009-1016.
7. Hoshino, E. and A. Tanaka, Enhancement of enzymatic catalysis of <i>Bacillus
amyloliquefaciens α-amylase by nonionic surfactant micelles. Journal of Surfactants and
Detergents, 2003. 6(4): p. 299-303.
8. Martinek, K., et al., The principles of enzymes stabilization: VI. Catalysis by water-soluble
enzymes entrapped into reversed micelles of surfactants in organic solvents. Biochimica
et Biophysica Acta (BBA) - Enzymology, 1981. 657(1): p. 277-294.
9. Rodríguez, V.B., et al., Modification of the activity of an α-amylase from Bacillus
licheniformis by several surfactants. 2006. 2006.
10. Kabanov, A.V., et al., Enzymes entrapped in reversed micelles of surfactants in organic
solvents: A theoretical treatment of the catalytic activity regulation. Journal of
Theoretical Biology, 1988. 133(3): p. 327-343.
11. Eriksson, T., J. Börjesson, and F. Tjerneld, Mechanism of surfactant effect in enzymatic
hydrolysis of lignocellulose. Enzyme and Microbial Technology, 2002. 31(3): p. 353-364.
12. Mizutani, C., et al., Effect of a nonionic surfactant on Trichoderma cellulase treatments
of regenerated cellulose and cotton yarns. Cellulose, 2002. 9(1): p. 83-89.
13. Kaar, W.E. and M.T. Holtzapple, Benefits from Tween during enzymic hydrolysis of corn
stover. Biotechnology and Bioengineering, 1998. 59(4): p. 419-427.
14. Shome, A., S. Roy, and P.K. Das, Nonionic Surfactants: A Key to Enhance the Enzyme
Activity at Cationic Reverse Micellar Interface. Langmuir, 2007. 23(8): p. 4130-4136.
15. Shi, Y.-F., X.-J. Yu, and Q. Yuan, Enhancement Effect of Reverse Micelles on Enzymatic
Synthesis of Cephalexinaa. Annals of the New York Academy of Sciences, 1998. 864(1):
p. 497-501.
16. Deo, N., et al., Surfactant Interactions with Zein Protein. Langmuir, 2003. 19(12): p.
5083-5088.
17. De, S., A. Girigoswami, and S. Das, Fluorescence probing of albumin–surfactant
interaction. Journal of Colloid and Interface Science, 2005. 285(2): p. 562-573.
155
18. Ruso, J.M., N. Deo, and P. Somasundaran, Complexation between Dodecyl Sulfate
Surfactant and Zein Protein in Solution. Langmuir, 2004. 20(21): p. 8988-8991.
19. Bru R., S.-F.A., Garcia-Carmona F., A theoretical study on the expression of enzymic
activity in reverse micelles. Biochem. J., 1989. 259: p. 355-61.
20. Viparelli P., A.F., Cantarella M., , Models for Enzyme Superactivity in Aqueous Solutions
of Surfactants. Biochem. J., 1999. 344: p. 765-773.
21. Lanyi, J.K., Bacteriorhodopsin. Annual Review of Physiology, 2004. 66(1): p. 665-688.
22. Hayashi, S., E. Tajkhorshid, and K. Schulten, Molecular Dynamics Simulation of
Bacteriorhodopsin's Photoisomerization Using Ab Initio Forces for the Excited
Chromophore. Biophysical Journal, 2003. 85(3): p. 1440-1449.
23. Edholm, O., O. Berger, and F. Jähnig, Structure and Fluctuations of Bacteriorhodopsin in
the Purple Membrane: A Molecular Dynamics Study. Journal of Molecular Biology, 1995.
250(1): p. 94-111.
24. Hampp, N., Bacteriorhodopsin as a Photochromic Retinal Protein for Optical Memories.
Chem Rev. , 2000. 100(5).
25. Audette, G., Lombardo S., Dudzik, J., Arruda, T.M., Kolinski M., Filipek, S., Mukerjee, S.,
Kannan, A.M., Thavasi, V., Ramakrishna, S., Chin, M., Somasundaran, P., Viswanathan,
S., Keles R.S., and Renugopalakrishnan V., Protein Hot Spots at Bio-nano Interfaces.
Materials Today, 2011. 14.
26. Koshland, D.E., Conformational changes: How small is big enough? Nature Medicine
1998. 4: p. 1112-1114.
27. Koshland, D.E., Application of a Theory of Enzyme Specificity to Protein Synthesis.
Proceedings of the National Academy of Sciences, 1958. 44(2): p. 98-104.
28. Daniel, R.M., et al., THE ROLE OF DYNAMICS IN ENZYME ACTIVITY. Annual Review of
Biophysics and Biomolecular Structure, 2003. 32(1): p. 69-92.
29. Agarwal, P., Enzymes: An integrated view of structure, dynamics and function. Microbial
Cell Factories, 2006. 5(1): p. 2.
30. Tang, K.E. and K.A. Dill, Native protein fluctuations: the conformational-motion
temperature and the inverse correlation of protein flexibility with protein stability.
Journal of biomolecular structure & dynamics, 1998. 16(2): p. 397-411.
31. Allen, T.W., Andersen O.S., Roux B., On the Importance of Atomic Fluctations, Protein
Flexibility, and Solvent in Ion Permeation. Journal of General Physiology, 2004. 124(6): p.
679-690.
32. Cooper, A., Conformational change, fluctuation and drift in biological macromolecules:
An empirical langevin approach. Journal of Molecular Liquids, 1988. 39(0): p. 195-206.
33. Lee, A.L. and A.J. Wand, Microscopic origins of entropy, heat capacity and the glass
transition in proteins. Nature, 2001. 411(6836): p. 501-504.
34. Koshland, D.E., The Key–Lock Theory and the Induced Fit Theory. Angewandte Chemie
International Edition in English, 1995. 33(23-24): p. 2375-2378.
35. Watanabe, K., T. Yoshida, and S.-i. Ueji, The role of conformational flexibility of enzymes
in the discrimination between amino acid and ester substrates for the subtilisin-
catalyzed reaction in organic solvents. Bioorganic Chemistry, 2004. 32(6): p. 504-515.
36. Fischer, S. and C.S. Verma, Binding of buried structural water increases the flexibility of
proteins. Proceedings of the National Academy of Sciences, 1999. 96(17): p. 9613-9615.
37. Tuckerman, M.E. and G.J. Martyna, Understanding Modern Molecular Dynamics:
Techniques and Applications. The Journal of Physical Chemistry B, 1999. 104(2): p. 159-
178.
38. Vickers, T. Diagram illustrating the induced fit model. 2006.
156
39. Yang, L., J.S. Dordick, and S. Garde, Hydration of Enzyme in Nonaqueous Media Is
Consistent with Solvent Dependence of Its Activity. Biophysical Journal, 2004. 87(2): p.
812-821.
40. Hudson, E.P., et al., Active-Site Motions and Polarity Enhance Catalytic Turnover of
Hydrated Subtilisin Dissolved in Organic Solvents. Journal of the American Chemical
Society, 2009. 131(12): p. 4294-4300.
41. Griebenow, K. and A.M. Klibanov, Can conformational changes be responsible for solvent
and excipient effects on the catalytic behavior of subtilisin Carlsberg in organic solvents?
Biotechnology and Bioengineering, 1997. 53(4): p. 351-362.
42. Marzola, P. and E. Gratton, Hydration and protein dynamics: frequency domain
fluorescence spectroscopy of proteins in reverse micelles. The Journal of Physical
Chemistry, 1991. 95(23): p. 9488-9495.
43. Affleck, R., et al., Enzymatic catalysis and dynamics in low-water environments.
Proceedings of the National Academy of Sciences, 1992. 89(3): p. 1100-1104.
44. Fitzpatrick, P.A., et al., Enzyme crystal structure in a neat organic solvent. Proceedings of
the National Academy of Sciences, 1993. 90(18): p. 8653-8657.
45. Broos, J., et al., Flexibility of Enzymes Suspended in Organic Solvents Probed by Time-
Resolved Fluorescence Anisotropy. Evidence That Enzyme Activity and Enantioselectivity
Are Directly Related to Enzyme Flexibility. Journal of the American Chemical Society,
1995. 117(51): p. 12657-12663.
46. Bone, S., Electrostatic interactions in protein structures: The influence of water on
enzyme fluctuations. Journal of Electrostatics, 1988. 21(2–3): p. 245-256.
47. Fernández, A. and H.A. Scheraga, Insufficiently dehydrated hydrogen bonds as
determinants of protein interactions. Proceedings of the National Academy of Sciences,
2003. 100(1): p. 113-118.
48. Baldwin, R.L., Temperature dependence of the hydrophobic interaction in protein folding.
Proceedings of the National Academy of Sciences, 1986. 83(21): p. 8069-8072.
49. Kurkal, V., et al., Low frequency enzyme dynamics as a function of temperature and
hydration: A neutron scattering study. Chemical Physics, 2005. 317(2–3): p. 267-273.
50. Fenn, E.E., et al., Dynamics of Water at the Interface in Reverse Micelles: Measurements
of Spectral Diffusion with Two-Dimensional Infrared Vibrational Echoes. The Journal of
Physical Chemistry B, 2011. 115(40): p. 11658-11670.
51. Thielges, M.C. and M.D. Fayer, Protein Dynamics Studied with Ultrafast Two-
Dimensional Infrared Vibrational Echo Spectroscopy. Accounts of Chemical Research,
2012.
52. Bey Temsamani, M., et al., Fourier Transform Infrared Investigation of Water States in
Aerosol-OT Reverse Micelles as a Function of Counterionic Nature. The Journal of
Physical Chemistry B, 1998. 102(18): p. 3335-3340.
53. Orecchini, A., et al., Low-frequency Vibrational Anomalies in β-Lactoglobulin:
Contribution of Different Hydrogen Classes Revealed by Inelastic Neutron Scattering. The
Journal of Physical Chemistry B, 2001. 105(48): p. 12150-12156.
54. Henzler-Wildman, K.A., et al., A hierarchy of timescales in protein dynamics is linked to
enzyme catalysis. Nature, 2007. 450(7171): p. 913-916.
55. Havsteen, B.H., A new principle of enzyme catalysis: Coupled vibrations facilitate
conformational changes. Journal of Theoretical Biology, 1989. 140(1): p. 101-127.
56. McCammon, J.A., B.R. Gelin, and M. Karplus, Dynamics of folded proteins. Nature, 1977.
267(5612): p. 585-590.
157
57. Doster, W., Viscosity scaling and protein dynamics. Biophysical Chemistry, 1983. 17(2):
p. 97-103.
58. McCammon, J.A. and M. Karplus, The dynamic picture of protein structure. Accounts of
Chemical Research, 1983. 16(6): p. 187-193.
59. Karplus, M., Role of conformation transitions in adenylate kinase. Proceedings of the
National Academy of Sciences, 2010. 107(17): p. E71.
60. Textor, L.C., et al., Purification, crystallization and preliminary crystallographic analysis
of peroxidase from the palm tree Chamaerops excelsa. Acta Crystallographica Section F,
2011. 67(12): p. 1641-1644.
61. Becker, T., et al., General Production Methods, in Enzymes in Industry. 2004, Wiley-VCH
Verlag GmbH & Co. KGaA. p. 37-82.
62. Alexov, E.G. and M.R. Gunner, Incorporating protein conformational flexibility into the
calculation of pH-dependent protein properties. Biophysical Journal, 1997. 72(5): p.
2075-2093.
63. Bickley, R.I., et al., A structural investigation of titanium dioxide photocatalysts. Journal
of Solid State Chemistry, 1991. 92(1): p. 178-190.
64. Kipp, J.E., The role of solid nanoparticle technology in the parenteral delivery of poorly
water-soluble drugs. International Journal of Pharmaceutics, 2004. 284(1–2): p. 109-122.
65. Smith, E.L., et al., Subtilisin Carlsberg. Journal of Biological Chemistry, 1968. 243(9): p.
2184-2191.
66. Bott, R., et al., The three-dimensional structure of Bacillus amyloliquefaciens subtilisin at
1.8 A and an analysis of the structural consequences of peroxide inactivation. Journal of
Biological Chemistry, 1988. 263(16): p. 7895-7906.
67. Twining, S.S., Fluorescein isothiocyanate-labeled casein assay for proteolytic enzymes.
Analytical Biochemistry, 1984. 143(1): p. 30-34.
68. Saito Y., M.M., Nakashima S., Odo J., Tanaka Y., Chikuma M., Tanaka H., Determination
of Hydrojen Peroxide with Phenol and 4-Aminoantipyrine by the Use of a Resin Modified
with Manganese-Tetrakis(sulfophenyl)porphine. Analytical Sciences 1987. 3.
69. Beghein, N., et al., Characterization of self-assembling copolymers in aqueous solutions
using Electron Paramagnetic Resonance and Fluorescence spectroscopy. Journal of
Controlled Release, 2007. 117(2): p. 196-203.
70. Waterman, K.C., et al., Use of a nitroxide spin probe to study the structure of the
adsorbed layer of dodecyl sulfate at the alumina-water interface. The Journal of Physical
Chemistry, 1986. 90(26): p. 6828-6830.
71. Cross, A.J. and G.R. Fleming, Analysis of time-resolved fluorescence anisotropy decays.
Biophysical Journal, 1984. 46(1): p. 45-56.
72. Broos, J., et al., Flexibility of enzymes suspended in organic solvents probed by time-
resolved fluorescence anisotropy. Evidence that enzyme activity and enantioselectivity
are directly related to enzyme flexibility. Journal of the American Chemical Society, 1995.
117(51): p. 12657-12663.
73. Wahl, P., Analysis of fluorescence anisotropy decays by a least square method.
Biophysical Chemistry, 1979. 10(1): p. 91-104.
74. Kinosita Jr, K., A. Ikegami, and S. Kawato, On the wobbling-in-cone analysis of
fluorescence anisotropy decay. Biophysical Journal, 1982. 37(2): p. 461-464.
75. Beechem, J.M., The Global Analysis of Fluorescence Intensity and Anisotropy Decay
Data: Second-Generation Theory and Programs. 2002, Springer. p. 241-305.
76. Das, T.K. and S. Mazumdar, pH-Induced Conformational Perturbation in Horseradish
Peroxidase. European Journal of Biochemistry, 1995. 227(3): p. 823-828.
158
77. Lasagna, M., et al., Apohorseradish peroxidase unfolding and refolding: intrinsic
tryptophan fluorescence studies. Biophysical Journal, 1999. 76(1 Pt 1): p. 443-450.
78. Bai, S. and A. Dong, Effects of immobilization onto aluminum hydroxide particles on the
thermally induced conformational behavior of three model proteins. International
Journal of Biological Macromolecules, 2009. 45(1): p. 80-85.
79. Braun, L.J., et al., Development of a freeze-stable formulation for vaccines containing
aluminum salt adjuvants. Vaccine, 2009. 27(1): p. 72-79.
80. Wang, J., Sonocatalytic damage of bovine serum albumin (BSA) in the presence of
nanometer titanium dioxide (TiO~ 2) catalyst. 中國化學快報, 2005. 16(8): p. 1105.
81. Nel, A.E., et al., Understanding biophysicochemical interactions at the nano-bio
interface. Nat Mater, 2009. 8(7): p. 543-557.
82. Griep, M.H., Walczak, K., Winder E., Lueking D. R., Friedrich C.R., Quantum dot
enhancement of bacteriorhodopsin-based electrodes. Biosensors and Bioelectronics,
2010. 25.
83. Alfinito, E.R.L., Charge Transport in bacteriorhodopsin monolayers: The contribution of
conformational change to current-voltage characteristics. European Physica Lett., 2009.
85.
84. Layne, E., Spectrophotometric and Turbidimetric Methods of Measuring Proteins.
Methods in Ezymology, 1957. 3.
85. Stoscheck, C., Qantitation of Protein Methods in Ezymology, 1990. 182.
86. Jiang, L.C., et al., Self-assembly behaviour of alkylpolyglucosides (APG) in mixed
surfactant-stabilized emulsions system. Journal of Molecular Liquids, 2011. 158(3): p.
175-181.
87. Totland, C., R.T. Lewis, and W. Nerdal, 1H NMR relaxation of water: A probe for
surfactant adsorption on kaolin. Journal of Colloid and Interface Science, 2011. 363(1):
p. 362-370.
88. Totland, C., et al., Water Structure and Dynamics at a Silica Surface: Pake Doublets in 1H
NMR Spectra. Langmuir, 2011. 27(8): p. 4690-4699.
89. Tehrani Bagha, A.R., et al., Interactions of gemini cationic surfactants with anionic azo
dyes and their inhibited effects on dyeability of cotton fabric. Dyes and Pigments, 2007.
72(3): p. 331-338.
90. Chen, Z.-J. and R.E. Stark, Evaluating spin diffusion in MAS-NOESY spectra of
phospholipid multibilayers. Solid State Nuclear Magnetic Resonance, 1996. 7(3): p. 239-
246.
91. Pisliakov, A.V., et al., Enzyme millisecond conformational dynamics do not catalyze the
chemical step. Proceedings of the National Academy of Sciences, 2009. 106(41): p.
17359-17364.
92. Kamerlin, S.C.L. and A. Warshel, At the dawn of the 21st century: Is dynamics the missing
link for understanding enzyme catalysis? Proteins: Structure, Function, and
Bioinformatics, 2010. 78(6): p. 1339-1375.
93. Fenimore, P.W., et al., Slaving: Solvent fluctuations dominate protein dynamics and
functions. Proceedings of the National Academy of Sciences, 2002. 99(25): p. 16047-
16051.
94. Frauenfelder, H., et al., Protein folding is slaved to solvent motions. Proceedings of the
National Academy of Sciences, 2006. 103(42): p. 15469-15472.
95. Collins, K.D., Ions from the Hofmeister series and osmolytes: effects on proteins in
solution and in the crystallization process. Methods, 2004. 34(3): p. 300-311.
159
96. Collins, K.D., Why continuum electrostatics theories cannot explain biological structure,
polyelectrolytes or ionic strength effects in ion–protein interactions. Biophysical
Chemistry, 2012. 167(0): p. 43-59.
97. Balzer, D., Cloud point phenomena in the phase behavior of alkyl polyglucosides in
water. Langmuir, 1993. 9(12): p. 3375-3384.
98. Tóth, K., et al., Flexibility and enzyme activity of NADH oxidase from Thermus
thermophilus in the presence of monovalent cations of Hofmeister series. Biochimica et
Biophysica Acta (BBA) - Proteins and Proteomics, 2008. 1784(5): p. 789-795.
99. Pal, S.K., J. Peon, and A.H. Zewail, Biological water at the protein surface: Dynamical
solvation probed directly with femtosecond resolution. Proceedings of the National
Academy of Sciences, 2002. 99(4): p. 1763-1768.
100. Hamm, P., M. Lim, and R.M. Hochstrasser, Structure of the Amide I Band of Peptides
Measured by Femtosecond Nonlinear-Infrared Spectroscopy. The Journal of Physical
Chemistry B, 1998. 102(31): p. 6123-6138.
101. Paci, E. and M. Marchi, Intrinsic compressibility and volume compression in solvated
proteins by molecular dynamics simulation at high pressure. Proceedings of the National
Academy of Sciences, 1996. 93(21): p. 11609-11614.
