University of Southern California Dissertations and Theses

Enhancement of biofuel enzyme activity and kinetics with azoTAB surfactants

(USC Thesis Other)

Enhancement of biofuel enzyme activity and kinetics with azoTAB surfactants

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

ENHANCEMENT OF BIOFUEL ENZYME
ACTIVITY AND KINETICS
WITH AZOTAB SURFACTANTS

by

Zumra Peksaglam Seidel

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(CHEMICAL ENGINEERING)

August 2021

Copyright 2021 Zumra Peksaglam Seidel

ii

ACKNOWLEDGMENTS
Firstly, I would like to express my deepest appreciation to Dr. C. Ted Lee, Jr. who gave
me this great opportunity to work in his laboratory. He always supported and encouraged me over
the years. He always provided his intellectual insights and guided me how to be a better scientist
in every meeting.

Next, I would like to thank Dr. Katherine Shing, Dr. Aiichiro Nakano, Dr. Nicholas A.
Graham, Dr. Pin Wang, Dr. Noah Malmstadt, Dr. Richard Roberts, Dr. Robert Young and Dr.
Mohammad Sahimi who were always there for me to broaden my vision, motivate and give
inspiring guidance. I would like to express my sincere gratitude to Tina Silva, Shokry Bastorous,
Jivin Seward and my colleagues for their immediate reachability, assistance and suggestions. I also
would like to thank Dr. George Tolomiczenko, Dr. Terry Sanger, Nadine Afari and Health,
Technology and Engineering program for enlarging my perspective and their support and
encouragements. I would like to acknowledge the financial support of the Ministry of National
Education of Turkish Government and USC Graduate School through Turkish Government
Scholarship, Viterbi-Graduate School and Research Enhancement Fellowships. I appreciate the
training, beam time opportunities, precious advises and instructions of their scientists at National
Institute of Standards and Technology, Oak Ridge National Laboratory and Argonne National
Laboratory.

Finally, I am extremely grateful for my family and friends for their unlimited love, support,
encouragement and understanding throughout my USC journey.

iii

Contents
ACKNOWLEDGMENTS .............................................................................................................. ii
LIST OF FIGURES ........................................................................................................................ v
LIST OF TABLES .......................................................................................................................... x
LIST OF SCHEMES...................................................................................................................... xi
ABSTRACT .................................................................................................................................. xii
CHAPTER 1: INTRODUCTION ................................................................................................... 1
1.1. Surfactants ........................................................................................................................ 1
1.1.1. Stimuli-Responsive Surfactants ................................................................................ 3
1.2. Proteins ............................................................................................................................. 5
1.2.1. Enzymes .................................................................................................................... 8
1.2.2. Biofuel Enzymes ..................................................................................................... 12
1.2.3. TIM-Barrel Fold...................................................................................................... 15
1.2.4. Surfactant-Protein Interactions ............................................................................... 16
1.3. Catanionic Surfactant Systems and Their Application .................................................. 18
1.3.1. SiRNA Delivery ...................................................................................................... 21
1.4. Small-angle Neutron Scattering (SANS) ....................................................................... 23
1.5. Overview ........................................................................................................................ 24
CHAPTER 2: SUPERACTIVITY OF THE CELLULASE ENZYME β-GLUCOSIDASE
ACHIEVED THROUGH CONFORMATIONAL CHANGES IN THE PRESENCE OF A
PHOTORESPONSIVE SURFACTANT ...................................................................................... 26
2.1. Abstract .......................................................................................................................... 26
2.2. Introduction .................................................................................................................... 27
2.3. Experimental Section ..................................................................................................... 30
2.4. Results and Discussion ................................................................................................... 36
2.5. Conclusions .................................................................................................................... 58
CHAPTER 3: ENHANCEMENT OF ENDOGLUCANASE BIOFUEL ENZYME KINETICS
IN THE PRESENCE OF AZOTAB SURFACTANT .................................................................. 60
3.1. Abstract .......................................................................................................................... 60
3.2. Introduction .................................................................................................................... 61
3.3. Experimental Section ..................................................................................................... 63
3.4. Results ............................................................................................................................ 70
3.5. Conclusion ...................................................................................................................... 87

iv

CHAPTER 4: PHOTO-CONTROLLED INHIBITION OF BREAST CANCER CELL
GROWTH VIA CO-DELIVERY OF BCL-2 SIRNA AND PACLITAXEL WITH
CATANIONIC VESICLES .......................................................................................................... 89
4.1. Abstract .......................................................................................................................... 89
4.2. Introduction .................................................................................................................... 90
4.3. Experimental Section ..................................................................................................... 94
4.4. Results and Discussion ................................................................................................. 103
4.5. Conclusions .................................................................................................................. 126
CHAPTER 5: VISIBLE LIGHT-INDUCED CRYSTALLIZATION OF CATANIONIC
VESICLES AND CELLULASE ENZYME ENCAPSULATION APPLICATION ................. 128
5.1. Abstract ........................................................................................................................ 128
5.2. Introduction .................................................................................................................. 128
5.3. Experimental Section ................................................................................................... 131
5.4. Results .......................................................................................................................... 134
5.5. Conclusion .................................................................................................................... 156
CHAPTER 6: FUTURE WORK ................................................................................................ 157
6.1. Biofuel Enzymes ........................................................................................................ 157
6.1.1. Biofuel Enzymes-Introduction ............................................................................ 157
6.1.2. Structural Studies of Endoglucanase ................................................................... 158
6.1.3. Structural Effect of AzoTAB into Cellulase Multienzyme Complex .................. 159
6.1.4. Structure-Function Relationship of Other Biofuel Enzymes ............................... 159
6.1.5. Conclusion ........................................................................................................... 160
6.2. Catanionic Surfactant Systems ................................................................................. 162
6.2.1. Catanionic Surfactant Systems - Introduction ...................................................... 162
6.2.2. Experimental Section ............................................................................................ 164
6.2.3. Preliminary Results ............................................................................................... 166
6.2.4. Light-responsive Drug Release ............................................................................ 171
6.2.5. Photoresponsive Membrane Protein Isolation ...................................................... 172
6.2.6. Light-induced Genetic Disorder Erasability ......................................................... 172
6.2.7. Conclusion ............................................................................................................ 176
BIBLIOGRAPHY ....................................................................................................................... 177

v

LIST OF FIGURES
Figure 1. UV-vis absorbance spectra of azoTAB surfactant in water under dark, visible and UV
light (path length=2 mm, [azoTAB]= 0.4 mM) .............................................................................. 5
Figure 2. A simplified cartoon of the enzymatic degradation of cellulose. ................................. 14
Figure 3. A top-down, schematic view of the TIM-barrel fold (from Nagarajan et al.
83
) ........... 16
Figure 4. Light-induced self-assembly structure changes in azoTAB-SDBS catanionic solutions.
....................................................................................................................................................... 21
Figure 5. Gene silencing mechanism of siRNA. .......................................................................... 22
Figure 6. SDS-PAGE of the crude glucosidase product (lane 2) and purified β-glucosidase (lane
3) versus standard protein markers (lane 1). ................................................................................. 37
Figure 7. (a) Relative activity of β-glucosidase with cellobiose as a substrate in the presence of
azoTAB (under visible or UV light), SDS, and DTAB surfactants. Lines are drawn to guide the
eyes. (b) and (c) Absorbance of the hydrolysis product from the model substate. ...................... 38
Figure 8. (a) Hydrodynamic diameter of β-glucosidase (1 mg/mL) in the presence of azoTAB
under different light conditions. The hatched regions denote surfactant concentrations beyond
which large-scale protein aggregation was observed (i.e., diameters > 100 nm). ........................ 40
Figure 9. (a) Small-angle neutron scattering (SANS) data and corresponding (b) pair-distance
distribution functions and (c) modified Guinier analyses of β-glucosidase (1 mg/mL) at low
azoTAB concentrations versus light conditions. The inset in (a) gives I(0)/c versus MW for a variety
of proteins, including lysozyme,
40
lactalbumin,
186
α-chymotrypsin,
143
carbonic anhydrase,
38
and
bovine serum albumin.
37
The inset in (b) displays the P(r) functions at low values of r. In part (c),
the 0.05 mM azoTAB data under visible and UV light are successively offset by –0.5 for clarity,
while the reported values are the corresponding cross-sectional radius of gyration. ................... 43
Figure 10. (a) Illustration of the method used to obtain the monomer-only SANS data. The raw
spectrum for the system containing 1 mM azoTAB under visible light (a mixture of monomers and
fibrils) is subtracted by the scaled pure β-glucosidase spectrum (containing only fibrils). The red
curve indicates the best-fit Guinier approximation to the monomer-only data used to obtain the
scaling factor (inset: Guinier plots using various scaling factors). Also shown are the
corresponding (b) pair-distance distribution functions and (c) Guinier analyses of the
deconvoluted, monomer-only data at each condition. In part (a), the P(r) values for the
crystallographic monomer and dimer (PDB: 4IIB) are shown for comparison. In part (c), the data
with 1 mM azoTAB (UV), 0.25 mM azoTAB (visible), and 0.25 mM azoTAB (UV) light are
successively offset by –1.0 for clarity, while the reported values are the corresponding radius of
gyration. ........................................................................................................................................ 46
Figure 11. (a) Results of the shape-reconstruction analysis of the monomer-only SANS data for
β-glucosidase in the presence of 0.25 mM azoTAB under visible light. The overlaid ribbon
diagrams are from the crystal structure of monomeric β-glucosidase from A. aculeatus (PDB:
4IIB), (TIM barrel-like domain, blue; α/β sandwich domain, green; fibronectin type III domain,
yellow); linker regions, dark gray; and insertion region, red). (b) The crystal structure of the β-
glucosidase dimer shown for comparison. (c) Results of the shape-reconstruction analysis as a
function of azoTAB concentration and light conditions. .............................................................. 50

vi

Figure 12. Circular dichroism spectroscopy measurements of β-glucosidase (1 mg/mL) in the
presence of various azoTAB concentrations under visible and UV light, with the latter being
shifted to the right by 20 nm for clarity. ....................................................................................... 52
Figure 13. Intrinsic fluorescence of β-Glucosidase (90 μg/mL) with various concentrations of
azoTAB under (a) visible and (b) UV light. See text for a description of the absorbance correction
procedure. The inset in (a) demonstrates the shift in the measured fluorescence independent of the
absorbance. The inset in (b) shows the maximum emission wavelength of β-glucosidase in the
presence of azoTAB under visible and UV light, SDS, and DTAB. The enclosed image illustrates
the location of tryptophan residues in β-glucosidase from A. aculeatus (PDB: 4IIB). ................ 53
Figure 14. Reaction velocities obtained as a function of substrate (cellobiose) concentration both
with and without azoTAB. Solid curves correspond to nonlinear fits of the data to the Michaelis-
Menten equation, while dashed curves represent fits to a modified equation accounting for
substrate inhibition. ....................................................................................................................... 56
Figure 15. SDS-PAGE of crude cellulase (lane 2) and purified endoglucanase after anion
exchange column Sephadex (lane 4 and 5) versus molecular weight protein ladders (lane 1 and 3)
....................................................................................................................................................... 71
Figure 16. Relative specific activity of endoglucanase (100 μg/mL) as a function of azoTAB
(under visible or UV light), SDS, SDBS and DTAB concentration towards 0.01 mg/mL avicel. pH
= 5, T = 25 C. Lines are drawn to guide the eyes. ...................................................................... 74
Figure 17. Light-responsive p-nitrophenol enzymatic cleavage product release from 4-nitrophenyl
β-D-cellobioside substrate (5 mM) with 0.4 mM azoTAB and 0.3 mg/mL endoglucanase. pH = 5,
T = 37 °C. In situ light changes are pointed by arrows. The reaction rates are written in parenthesis.
For reference, the product concentration in the presence of 0.4 mM azoTAB visible and UV light
versus time are drawn with dashed lines. ...................................................................................... 75
Figure 18. Reaction rates acquired as a function of Avicel microcrystalline substrate concentration
in the presence of azoTAB (0.4 mM) under visible and UV light. Solid curves display the nonlinear
regression fits of Michaelis-Menten equation. The inlet graph represents Hanes-Woolf plot, and
the dashed lines show the fits to calculate kinetics constants. ...................................................... 76
Figure 19. Adsorption isotherm of endoglucanase from Aspergillus niger on Avicel (20mg/mL)
in addition of azoTAB (0.4 mM) under visible and UV light, SDS (1.5 mM), SDBS (1.5mM) and
DTAB (1.5 mM) surfactants (pH 5, T=25 °C). 30 min incubation time was used. The lines are
drawn according to the individual fit of Langmuir isotherm (to obtain Emax and Kad values). ..... 79
Figure 20. Percentage of the adsorbed endoglucanase (1 mg/mL) on Avicel (20 mg/mL) as a
function of glucose and cellobiose end-products (inhibitors) with and without azoTAB (0.4 mM),
SDS (1.5 mM), SDBS (1.5 mM) and DTAB (1.5 mM) surfactants. ............................................ 82
Figure 21. Optical microscopy images of Avicel as a function of time. AzoTAB surfactant effect
on Avicel hydrolysis is observed with and without end-product glucose inhibitor. The scale bar
shows 100 µm. .............................................................................................................................. 83
Figure 22. Relative activity of various enzyme mixtures of endoglucanase (62.5 μg/mL),
cellobiohydrolase (120 μg/mL) and β-glucosidase (30 μg/mL) with addition of azoTAB (0.4 mM
under visible or UV light), SDS (1.5 mM), SDBS (1.5 mM) and DTAB (1.5 mM) towards Avicel
(0.01 mg/mL), pH = 5, T = 25 C. (a) Initial, (b) and (c) long-term relative saccharification rate of
Avicel. ........................................................................................................................................... 86

vii

Figure 23. A simplified cartoon to show increase of adsorption of endoglucanase enzymes on
Avicel which results in endoglucanase activity enhancement. ..................................................... 87
Figure 24. (a) The structure of azoTAB and SDBS surfactants, as well as the trans-to-cis
photoisomerization of azoTAB upon exposure to visible (434 nm)  UV (350 nm) light. The
length of the alkyl spacer in azoTAB is n = 2 (2-azo-2), n = 4 (2-azo-4), or n = 6 (2-azo-6). (b)
Illustration of the reversible transition from catanionic vesicles to free surfactant monomers that
can occur upon azoTAB photoisomerization. ............................................................................... 94
Figure 25. The cartoon shows catanionic surfactant vesicles and containing the siRNA and the
Paclitaxel (PTX), with the finally formed “2-in-1 nanocarrier” undergoing endocytosis by breast
cancer cell, then irradiated with ultraviolet light (UVA) causing disruption and emptying the
nanocarrier contents into the cell where the siRNA matches with Bcl-2 producing complementary
mRNA produced by the nucleus and inhibits formation of the Bcl-2 protein. ........................... 104
Figure 26. Critical aggregation concentrations (CAC) of azoTAB/SDBS (7/3 mol/mol) mixtures
in pH 7.4 PBS buffer at 37 °C determined using the micropolarity indicator dye Nile red (2.5 μM)
to detect the onset of bilayer formation. The arrows denote the CAC values at each condition
expressed as the concentration of azoTAB. ................................................................................ 106
Figure 27. Hydrodynamic diameters of azoTAB/SDBS (7/3 mol/mol) vesicles over the course of
50 days as a function of temperature. [azoTAB] = 50 μM in pH 7.4 PBS buffer. .................... 107
Figure 28. Hydrodynamic diameters of catanionic vesicles formed from 7/3 mixtures of 2-azo-4
and various sodium alkyl sulfates, SCnS, (e.g., SC12S corresponds to sodium dodecyl sulfate, SDS).
..................................................................................................................................................... 108
Figure 29. Small-angle neutron scattering (SANS) data of azoTAB/SDBS (7/3 mol/mol) vesicles
under visible light at a total surfactant concentration of 0.1 wt %. The slope of –2 indicates
scattering from large, locally-flat entities (i.e., vesicle bilayers). Oscillations about this slope are
a result of finite vesicle size. Inset: Modified Guinier plots with corresponding bilayer thicknesses,
with solid data points corresponding to those used in the linear fits. ......................................... 109
Figure 30. The percentage of bulk paclitaxel capable of being solubilized within azoTAB/SDBS
(50 μM/21.4 μM) vesicles as a function of the paclitaxel/surfactant (wt/wt) ratio. The upper
abscissa indicates the bulk paclitaxel concentration, while the numbers above each bar correspond
to the paclitaxel concentration (µM) solubilized within the vesicles. ........................................ 111
Figure 31. (a) Electrophoretic mobility shift assays of siRNA upon the addition of azoTAB/SDBS
(7/3 mol/mol) surfactant at a fixed siRNA concentration (30 nM) to achieve various N/P (azoTAB-
to-nucleotide) ratios. (b) Zeta potentials of siRNA-loaded and siRNA-paclitaxel co-loaded vesicles
upon the addition of siRNA at fixed surfactant concentration ([azoTAB]/[SDBS] = 50 μM/21.4
μM) and paclitaxel/surfactant ratio (1/8 wt/wt). Arrows indicate the approximate points where
complete mobility shifts were attained in (a). The inset replots the zeta potential change as a
function of the P/N ratio, where N is the azoTAB concentration minus the corresponding CAC.
..................................................................................................................................................... 113
Figure 32. Viability of MDA-MB-231 human breast cancer cells after treatment with empty,
siRNA-loaded (N/P = 19), paclitaxel-loaded (1/8 wt/wt), and siRNA-paclitaxel co-loaded
azoTAB/SDBS vesicles (7/3 mol/mol) under visible and UV light. .......................................... 119

viii

Figure 33. Uptake of Bcl-2 siRNA encapsulated in azoTAB/SDBS (50 μM/21.4 μM) vesicles in
MDA-MB-231 human breast cancer cells. Numbers correspond to the percentage of cells
exhibiting fluorescence from Cy5.5-labeled Bcl-2 siRNA. ........................................................ 122
Figure 34. Confocal fluorescence images of MDA-MB-231 human breast cancer cells following
treatment with Cy5.5-conjugated Bcl-2 siRNA (red) at N/P = 4.8 and 4 μM Coumarin 6
hydrophobic dye (green) co-loaded in azoTAB/SDBS (50 μM/21.4 μM) vesicles for 4 hours. The
nuclei were stained with DAPI (blue). ........................................................................................ 123
Figure 35. Endosomal escape of siRNA following to transfection of MCF-7 cells for 4 hours with
siRNA loaded 2-azo-6/SDBS vesicles (N/P=19) and 60 seconds exposure of 358 nm. ............ 124
Figure 36. Western blot assay (run #1) following transfection of MDA-MB-231 human breast
cancer cells with naked Bcl-2 siRNA and siRNA loaded in azoTAB/SDBS (50 μM/21.4 μM, N/P
= 19) vesicles for 4 hours. β-actin was used as an equal loading control. .................................. 125
Figure 37. Quantification of Western blots showing the effect of Bcl-2 siRNA loaded in 2-azo-
2/SDBS and 2-azo-6/SDBS (50 μM/21.4 μM, N/P = 19) vesicles on MDA-MB-231 human breast
cancer cells with and without UV exposure. The expression of Bcl-2 was normalized to that of
the equal loading control β-actin. Western blots are shown in Figures S6-S10. n = 4 except 2-azo-
2/SDBS -UV, where n = 3. * p = 0.0282, ** p = 0.016. ............................................................ 126
Figure 38. UV-vis spectra of 0.1 wt% 30-70 azoTAB-SOBS and azoTAB-SdecS crystal formation
under visible light (a and d), dark (b and e) and UV light (c and f). (g) and (h) azoTAB-SOBS and
azoTAB-SdecS catanionic sample pictures at different times inside of Eppendorf tubes,
respectively. ................................................................................................................................ 135
Figure 39. Catanionic salt formation under different light wavelength. The lines show the fits
calculated with Avrami equation (Eqn 27). ................................................................................ 137
Figure 40. Physical appearance of 0.1 wt% 30-70 and 70-30 azoTAB-SOBS and azoTAB-SdecS
samples. ....................................................................................................................................... 139
Figure 41. Schematic presentation of liquid phase, solid and liquid crystal orientations. ......... 139
Figure 42. (a) and (b) parallelogram-like smectic liquid crystals of 0.1 wt% 30-70 azoTAB-
azoTAB-SOBS under bright field and polarized optical microscopy. (c) and (d) schlieren texture
of nematic phase of 0.1 wt% 30-70 azoTAB-SdecS under bright field and polarized optical
microscopy. (e) and (f) needle-like solid crystals of 0.1wt% 50-50 azoTAB-SOBS under bright
field and polarized optical microscopy. (g) and (h) schlieren texture of nematic phase of 0.1 wt%
50-50 azoTAB-SdecS under bright field and polarized optical microscopy. (i) and (j) rectangular-
like smectic liquid crystals of 0.1 wt% 70-30 azoTAB- azoTAB-SOBS under bright field and
polarized optical microscopy. (k) and (l) smectic liquid phase of 0.1 wt% 52.5-47.5 azoTAB-
azoTAB-SOBS under bright field and polarized optical microscopy. ....................................... 141
Figure 43. Ternary diagram of (a) azoTAB-SOBS and (b) azoTAB-SdecS under visible and UV
light. ............................................................................................................................................ 143
Figure 44. Two-dimensional (2D) Small-Angle Neutron Scattering (SANS) patterns of 0.1 wt%
azoTAB-SOBS and azoTAB-SdecS 70-30 ratio samples with consequentially exposure to visible-
UV-visible light. ......................................................................................................................... 146
Figure 45. I, intensity versus Q, scattering vector profile of 0.1 wt% azoTAB-SOBS (a) 30-70, (b)
70-30 and azoTAB-SdecS (c) 30-70, (d) 70-30 using Small-Angle Neutron Scattering. The

ix

samples were prepared under visible light and the light condition was switched to UV light and
then visible light in situ. .............................................................................................................. 147
Figure 46. UV-vis spectra of UV-light induced crystal to vesicles formation of 0.1 wt% azoTAB-
SOBS (a, b) and azoTAB-SdecS (c, d) at a ratio of 30-70 ......................................................... 151
Figure 47. Eppendorf pictures of UV-light induced crystal to vesicles formation of 0.1 wt%
azoTAB-SOBS and azoTAB-SdecS at a ratio of 30-70. ............................................................ 152
Figure 48. Native-PAGE of (1) 1 mg/mL bovine serum albumin (BSA) (2) 0.1 wt%, (3) 0.25 wt%
and (4) 0.5 wt% azoTAB-SdecS (70-30) vesicles under UV light, (5) 24 hours visible light exposed
0.5 wt% azoTAB-SdecS (70-30) vesicles with 1 mg/mL bovine serum albumin. ..................... 152
Figure 49. Relative activity of β–glucosidase and endoglucanase with 0.05 wt% 30-70 azoTAB-
SOBS and azoTAB-SdecS under visible and UV light as a function of (a) β–glucosidase and (b)
endoglucanase concentration. ..................................................................................................... 154
Figure 50. The cartoon shows the enzymes can be encapsulated and the enzymes show no activity.
Upon visible light exposure, the enzymes are released, and the reaction occurs while the catanionic
salt settles at the bottom. ............................................................................................................. 155
Figure 51. Small-angle neutron scattering (SANS) data for mixtures of cationic (azoTAB analog
2-azo-4) and anionic (sodium alkyl sulfate) surfactants at molar ratios of 30/70 and 70/30 at 0.1
wt % overall surfactant concentration and 25 °C. ...................................................................... 168
Figure 52. UV-vis spectroscopy data of azoTAB (2-azo-4)/sodium alkyl sulfate catanionic
mixtures (30/70 and 70/30) at an overall surfactant concentration of 0.05 wt % at 25 °C. ........ 171

x

LIST OF TABLES
Table 1. Reaction kinetics parameters of nonlinear regression fits of Michaelis-Menten equation
and linear fit of Hanes-Woolf equation. The data points and fits are represented in the Figure 18.
....................................................................................................................................................... 77
Table 2. Comparison of adsorption parameters of endoglucanase on Avicel (20 mg/mL) with and
without azoTAB, SDS, SDBS and DTAB surfactants. Emax and Kad values are calculated with
Langmuir isotherm equation and the data points with the fits are represented in Figure 19. ....... 80
Table 3. Fractal kinetics parameter of endoglucanase on Avicel (0.01 mg/mL) with and without
surfactants. Values of (*) are calculated Xu and Ding, 2007 and Bommarius equation, 2008,
whereas (**) numbers are determined Wang and Feng, 2010 model. The values of h and k were
calculated with 62.5 µg/mL endoglucanase enzyme towards 0.01mg/mL Avicel whereas f values
were determined with 62.5 µg/mL and 100 µg/mL endoglucanase enzyme towards 0.01mg/mL
Avicel. ........................................................................................................................................... 81
Table 4. Encapsulation and photo-initiated release percentages of Bcl-2 siRNA (N/P = 4.8) and
paclitaxel (1/8 drug/surfactant wt/wt) loaded in azoTAB/SDBS (50 μM/21.4 μM) vesicles at 37
°C in pH 7.4 PBS buffer. ............................................................................................................ 118
Table 5. Comparison of crystallization rate parameters of azoTAB-SOBS and azoTAB-SdecS 30-
70, 0.1wt% under different light conditions. Avrami exponent, n and overall crystallization rate
constant, k were calculated with Avrami equation (equation 27) and the fits were shown in Figure
39................................................................................................................................................. 136
Table 6. Hydrodynamic diameters of azoTAB-SOBS and azoTAB-SdecS at 0.1 wt%
concentration. SLC: smectic liquid crystals, NLC: nematic liquid crystals, m: micelle, L: lamellar,
Average of NNLS first peak of 10 different runs are used. ........................................................ 138
Table 7. The degree of alignment of 0.1 wt% azoTAB-SOBS and azoTAB-SdecS at 30-70 and
70-30 ratio samples with consequentially exposure to visible-UV-visible light. Integral method
(equation 25) and average method (equation 26) are used. ........................................................ 145
Table 8. Results of Guinier, PolyCore and Paracrystalline SANS models and their comparison
with light scattering data. ............................................................................................................ 148
Table 9. Biofuel enzymes of lignocellulosic plant-cell wall natural substrates ......................... 161
Table 10. Molecular formula, critical micelle concentration (CMC), and Krafft temperatures of
sodium alkyl sulfate surfactants .................................................................................................. 164
Table 11. DLS-determined hydrodynamic diameters (value in nm given in parenthesis) of
catanionic mixtures of the azoTAB-analog 2-azo-4 (labeled “S1” below for brevity) with various
sodium alkyl sulfate surfactants as a function of the cationic-to-anionic surfactant molar ratio under
both visible and UV light (0.1 wt% total surfactant concentration, 25 °C). Microstructure
designations are V = vesicles, L = lamellar, C = crystals, M = micelles, and T = white powder
(temperature below the Krafft temperature). .............................................................................. 167
Table 12. Results of Guinier and model fit analysis of Small-Angle Neutron Scattering catanionic
systems under visible light data in Figure 51 with comparison to light scattering results. ........ 169
Table 13. Results of Guinier and model fit analysis of Small-Angle Neutron Scattering catanionic
systems under UV light data in Figure 36 with comparison to light scattering results. ............. 170
Table 14. Various methods to deliver the CRISPR/Cas9 system. ............................................. 174

xi

LIST OF SCHEMES
Scheme 1. Chemical structure of the 4-ethyl-4’(trimethylaminobutoxy) azobenzene bromide
(“azoTAB”) photoresponsive surfactant. ........................................................................................ 4

xii

ABSTRACT
Azobenzene trimethyl ammonium bromide (azoTAB), a photoresponsive analog of the
common dodecyltrimethylammonium bromide (DTAB) surfactant, has been widely studied in
areas ranging from biotechnology, nanotechnology, and pharmacology. AzoTAB undergoes a
reversible photoisomerization from the relatively-hydrophilic cis form under UV light (max = 350
nm) to the relatively-hydrophobic trans form under visible light (max = 434 nm). This switchable
hydrophobicity of the surfactant has been used to photoreversibly control the structure-function
relationship of enzymes. Specifically, the more hydrophobic trans photoisomer of azoTAB has a
greater binding affinity to proteins relative to the cis form, thereby resulting in a greater degree of
protein unfolding. In some cases, this unfolding occurs at the active site, leading to a “switched
off” enzyme that can be switched back “on” upon UV illumination. In other cases, this unfolding
occurs distal to the active site in regions that impart greater flexibility to the enzyme, allowing the
enzyme-substrate complex to more easily traverse the activation energy barrier and leading to a
“superactive” enzyme.
In the present work, photocontrol of the structure and function of cellulase enzymes will
be explored. Cellulases are generally multienzyme complexes containing endocellulase,
exocellulase, and -glucosidase that work in a synergistic manner to convert cellulose, the most
abundant organic molecule on Earth, into fermentable sugars (i.e., glucose). Bioethanol obtained
in this manner thus represents an abundant and sustainable energy source to minimize the
dependence on fossil fuels. The process begins with endocellulase or endoglucanase randomly
cleaving internal glycoside bonds to convert crystalline cellulose into individual polymer chains.
Exocellulase or cellobiohydrolase then liberates cellobiose (two linked glucose units) from the
terminal ends of these chains, while -glucosidase hydrolyzes cellobiose into two glucose

xiii

molecules. Notably, cellobiose is an inhibitor of endo- and exo-cellulase, thus, the last enzymatic
step can be rate limiting. Addition of azoTAB to -glucosidase leads to a dimer → monomer
transition of the enzyme and a corresponding 50% increase in catalytic activity. Endoglucanase
activity increase of 45% with azoTAB addition towards Avicel crystalline substate will be
correlated with endoglucanase adsoption increase on the crystalline substrate. Furthermore, 45-
50% activity enhancement via azoTAB surfactant preserved for all three cellulase enzyme mixture
of endoglucanase, cellohydrolase and β-glucosidase. This superactivity or activity enhancement of
cellulase enzymes could have profound impacts on the economic viability of bioethanol (e.g.,
about one-third of the costs of every gallon of bioethanol are enzyme-related).
As another application of azoTAB surfactants, photocontrol of the spontaneous self-
assembly of “catanionic” surfactant systems (i.e., aqueous solutions of cationic and anionic
surfactants) will be discussed. Catanionic systems can form various microstructures depending on
the conditions, including spherical and cylindrical micelles, vesicles, multilamellar assemblies and
crystalline structures in addition to the monomeric state. As a result, azoTAB-based catanionic
systems allow for photoreversible transitions including vesicle  monomer, vesicle  micelle,
vesicles  multilamellar and vesicles  liquid crystals transitions. Notably, azoTAB-based
catanionic vesicles will be utilized to facilitate co-delivery of siRNA (i.e., small interfering RNA)
and a hydrophobic chemotherapy drug into breast cancer cells, with the photo-initiated vesicle-to-
monomer transition with UV illumination used to release the siRNA and drug from the carrier once
entered into the cells. Additionally, light-induced vesicle-to-liquid crystal transition will be
optimized for cellulase enzyme encapsulation, controlling the reaction as well as recycling
enzymes infinitely. Furthermore, the effects of azoTAB chemical structure and solution conditions

xiv

on vesicle properties will be examined using a range of techniques as a means to optimize the
bilayer thickness of vesicles for membrane protein isolation with vesicle micelle phototransition.

1

CHAPTER 1: INTRODUCTION
1.1. Surfactants
Surfactants (i.e., surface active agents) are amphiphilic compounds applied to lower the
surface or interfacial tension. Their dispersion, emulsion, cleansing (i.e., solubilization of
immiscible substances), stabilization, and wetting abilities have been used in pharmaceutical
applications,
1
the food industry,
2
oil spill cleaning,
3
remediation of contaminated soil,
4

transportation of pesticides,
5
enhanced oil recovery,
6, 7
personal care products,
8
asphalt, cement,
fiber, and textile manufacturing.
9

Surfactants can be broadly divided into two categories: natural or biosurfactants and
synthetic surfactants. Natural surfactants derive from plants, animals, or microorganisms and are
obtained through extraction, filtration, precipitation, fermentation, or distillation techniques; or
they may be synthesized from natural raw substances such as amino acids, sugars, sterols, or fatty
alcohols.
10,11
Common natural surfactants are pulmonary surfactants, glycolipids, saponins, and
bile salts.
12,13
The majority of surfactants used industrially are synthetic surfactants, which are
synthesized from fossil-fuel based substances.
14
Alkylates, linear paraffins, ethylene, toluene and
xylene are the most frequently used raw materials for synthetic surfactants.
15

Surfactants owning to their amphiphilic nature consist of a hydrophobic (often lipophilic
or oil soluble) tail, frequently alkyl chains of 8-22 carbons,
16
and a hydrophilic (water-soluble)
head group. Surfactants are classified based on their ionic properties in water: anionic, cationic,
nonionic, and zwitterionic or amphoteric surfactants. Anionic surfactants dissociate in water into
a negatively-charged amphiphile and a positively-charged ion (often Na
+
or K
+
). Typical anionic
surfactants include alkylbenzene sulfonates, alkyl sulfates, alkyl ether sulfates, sulfated
alkanolamides, glyceride sulfates, lignosulfonates, and alkyl carboxylates.
17
Cationic surfactants

2

ionize in water into a positively-charged amphiphile and an anion such as Br
-
or Cl
-
. Commonly,
cationic surfactants are linear alkyl amines, fatty amines, quaternary alkyl ammoniums,
nitrogenated amides, esters, and ether-amides.
15
Nonionic surfactants have head groups with no
electrostatic charge, such as polyethylene oxide, fatty acid esters, glyceryl, and sorbitol groups.
16

Zwitterionic surfactants dissociate in water into an amphiphile with both a positive and negative
charge, often a quaternary ammonium ion paired with a carboxylate, sulfate, sulfonate, or
phosphate ion.
18

Surfactants most typically aggregate into spherical micelles in dilute aqueous solutions
with the associated hydrophobic tails forming the hydrophobic interior and the head groups
forming the hydrophilic exterior of the micelles. The critical micelle concentration (CMC) is the
lowest surfactant concentration beyond which micelles are formed. The Krafft point occurs where
the temperature-dependent solubility curve for the monomeric surfactant intersects with the CMC
curve. As a result, pure surfactants self-associate only above the Krafft temperature and the CMC.
At elevated surfactant concentrations, a variety of microstructures can develop including
cylindrical micelles, and lamellar, cubic or hexagonal phases depending on the surfactant
concentration as well as the surfactant packing parameter.
19
The packing parameter, p = /aƖ, is
the volume of hydrophobic chain divided by the effective area per surfactant head group and
surfactant tail length. Generally, spherical micelles are formed when p < 1/3, while cylindrical or
hexagonal structure are found when 1/3 < p < 1/2, and flat lamellar structures occur when p ~ 1.
When p > 1 inverse spherical or hexagonal micelles are predicted.
20
Based on this rich
microstructure, surfactants can be applied to a wide range of applications, as noted above.

3

1.1.1. Stimuli-Responsive Surfactants
Traditionally, surfactant properties and microstructure are manipulated irreversibly
through the increase in the surfactant concentration, temperature, or electrolyte concentration.
21

Conversely, stimuli-responsive surfactants allow control of surfactant morphology and the
resulting properties (e.g., hydrophobicity, solubility, viscosity, electrical conductivity, interfacial
tension, Krafft temperature, etc.) in a reversible manner. Examples include surfactants responsive
to pH,
22
temperature,
23
light,
24, 25
magnetic field,
26
enzymes,
27
or CO2,
28
as well as combined
stimuli,
29
allowing for both spatial and temporal control over the physico-chemical properties of
surfactant solutions.

1.1.1.1. Photoresponsive Surfactants

Azobenzene is the most commonly used isomerizable unit in photoresponsive
surfactants.
25
The azobenzene moiety exists primarily as the trans form under visible light, while
the cis form dominates under UV illumination (typically UVA light ranging from 330 nm to 380
nm).
30
Thus, the trans  cis isomerization of photoresponsive surfactants containing azobenzene
can be controlled simply and reversibly with light illumination. The photoresponsive cationic
surfactant 4-ethyl-4’(trimethylamine-butoxy) azobenzene bromide (azoTAB or 2-azo-4) shown in
Scheme 1, is a light-responsive analog of the common dodecyltrimethylammonium bromide
(DTAB) surfactant. AzoTAB consists of a trimethylammonium head group and a hydrophobic
tail containing an azobenzene moiety bracketed by an alkoxy spacer and a terminal alkyl chain.
Briefly, azoTAB is prepared in three steps sequentially: azocoupling of alkylaniline with phenol,
alkylation with dibromoalkane, and quaternalization with trimethylamine.
31,32
The hydrophobicity
of the azoTAB tail can be modified with changes in the hydrocarbon tail length of the alkoxy

4

spacer and/or the terminal alkyl chain. The effect of the alkoxy spacer chain length (C2-C6) of
azoTAB on the co-delivery of siRNA and a hydrophobic drug into diseased cells will be examined
in Chapter 4.
Scheme 1. Chemical structure of the 4-ethyl-4’(trimethylaminobutoxy) azobenzene
bromide (“azoTAB”) photoresponsive surfactant.

The UV-vis spectra of azoTAB under various light conditions are displayed in Figure 1.
Four isosbestic points at 222 nm, 246 nm, 307 nm, and 429 nm can be seen, indicating a two-state
trans  cis photoisomerization. Peaks at 350 nm (-* transition of the trans isomer) and 440 nm
(n-* transition of both isomers, but more prevalent for the cis state) are also observed in the
spectra, along with a new band near 310 nm upon conversion to the cis state.
27,33

Photoisomerization involves rotation about the nitrogen double bond (-N=N-) from the planar
trans isomer to a bent and slightly twisted cis conformation. In the dark, the surfactant exists as
100% trans isomers, while under visible or UV illumination a 75/25 or 10/90 trans/cis
photostationary state is adopted, respectively.
33
Photoisomerization is rapid and reversible, while
thermal conversion of this cis isomer back to the trans state (due to the trans isomer being more
stable that the cis conformation) occurs in ~ 24 hours in the dark at 25 C.

5

0
0.4
0.8
1.2
1.6
200 250 300 350 400 450 500 550
visible
dark
UV
Wavelength
Absorbance

Figure 1. UV-vis absorbance spectra of azoTAB surfactant in water under dark, visible
and UV light (path length=2 mm, [azoTAB]= 0.4 mM)
Dipole moment gives an estimation of the polarity of a chemical bond between two atoms
in a molecule. The dipole moment of nitrogen double bond is 0.5 D and 3.1 D for the trans and cis
isomers, respectively.
34
Thus, more polar cis isomer can be solubilized in polar water molecules.
As a result, the trans isomer is more hydrophobic than the cis state, resulting in light-induced
surfactant properties such as CMC, electrical conductivity, and surface tension. Notably, we have
previously utilized azoTAB to allow photo-control of gelation,
24
gene delivery,
35
DNA
condensation,
36
soluble and membrane protein folding,
32, 37, 38
protein dynamics,
39
enzyme
kinetics,
38, 40
and microstructural transitions.
41

1.2. Proteins
Proteins are biomacromolecules that consist of a specific, linear combination of the 20
natural amino acids. Each amino acid contains a central carbon atom covalently bound to an amine
(-NH2), carboxylic acid (-COOH), hydrogen atom, and unique side chain (R group). The protein

6

chain is translated from a messenger RNAs with transfer RNA (tRNA) and ribosome, resulting in
a series of peptide bonds (i.e., amide) from the N-terminus, amine end to the C-terminus, carboxyl
end. Each amino acid side chain can be an acidic, basic, polar, or nonpolar molecule, which
influences the protein characteristics. Generally, the amino acid chain folds into a unique and stable
three-dimensional confirmation (i.e., native structure, which imparts function and dynamics to the
protein), with the hydrophobic amino acids buried in the interior of the protein and the hydrophilic
amino acids located on the surface of the folded protein.
The structure of proteins is classified into four categories: primary, secondary, tertiary and
quaternary structure. The primary structure is the amino acid sequence of a protein. The most
commonly applied technique to determine the primary structure is mass spectrometry, which sorts
the ionized molecules in the gas phase based on their mass-to-charge ratio to identify amino acid
sequencing. The secondary structure refers to local amino acid formations that are stabilized by
hydrogen bonds, such as α-helices and β-sheets. This structure can be determined by using either
circular dichroism (CD), which measures the differential absorption of left- and right-handed
circularly polarized light of a molecule with one or more chiral chromophores, or Fourier-
transform infrared spectroscopy (FTIR), where IR absorption bands can be associated with specific
bonds (e.g., C=O and N-H).
The tertiary structure of a protein is the three-dimensional conformation that results from
secondary structures, hydrophobic interactions, salt bridges, hydrogen bonds, and disulfide bonds.
Current tertiary-structure characterization methods are X-ray crystallography (~90%),
42
nuclear
magnetic resonance (NMR) (~8%),
43
electron microscopy (1.8%),
44
and neutron crystallography
(0.04%),
45
with the given percentages representing the relative proportions of the 150 thousand
proteins structures found in the Protein Data Bank (https://www.rcsb.org/). X-ray crystallography

7

involves the scattering of X-rays from the electrons within the molecules, allowing the atomic and
molecular structure to be determined. Neutron crystallography works similarly except with
neutrons scattering off nuclei, which in addition to the protein molecular structure allows location
of hydrogen atoms and hydrogen bonds.
46
As the names imply, both of these techniques require
crystallization of the protein into a single crystal, which raises doubts as to whether the determined
structures correspond to those actually occurring in vivo. In contrast, NMR, which measures the
local magnetic fields around atomic nuclei to find the distances between atoms and hence the
protein conformation, allows structural determination in vitro, but is generally limited to relatively-
small proteins (< 50 kDa) due to signal overlap. Cryogenic electron microscopy (cryo-EM) has
been used to visualize protein structures, especially membrane proteins that are difficult to
crystalize, with electron microscopy images again collected in a non-dynamic state after an
aqueous protein solution has been cooled to cryogenic temperatures (e.g., -180 C in liquid
nitrogen).
The quaternary structure of a protein refers to the arrangement of multiple polypeptide
chains (i.e., protein subunits) within an oligomeric structure. Determining the exact quaternary
structure of a protein is challenging, as crystal-packing constraints often lead to oligomeric states
in the crystal that do not exist in vivo. Current experimental methods of assessing the quaternary
structure include amino acid sequencing with mass spectrometry, measuring the molecular weight
with polyacrylamide gel electrophoresis (i.e. native-PAGE), gel filtration chromatography, light
scattering, and small-angle X-ray or neutron scattering.
47
As described in further detail below, the
current work involves the control of protein structure and function with photoresponsive
surfactants. Thus, protein conformations will be determined with small-angle neutron scattering
(SANS), which allows the conformation of large, partially-unfolded and/or oligomeric proteins

8

(i.e., states that could not be crystallized due to rapid hydrophobic aggregation) to be determined
in vitro.
In addition to the characterization methods mentioned above, other specialized techniques
can be applied to elucidate the local protein structure. The three aromatic amino acids,
phenylalanine (Phe), tryptophan (Trp), and tyrosine (Tyr), are naturally fluorescent;
48
thus, the
intrinsic fluorescence of a protein can be used to determine the micropolarity surrounding these
amino acids (i.e., buried in the protein core versus exposed to aqueous solvent). Additional
methods include hydrogen/deuterium (H/D) exchange, covalently-bonded chromophores, non-
covalent bonded labels, and cross-linking, which can be used to examine protein dynamics,
orientational changes, structural modifications, folding/unfolding status, and ligand binding
interactions.
1.2.1. Enzymes
Enzymes are biological catalysts that accelerate the rate of biochemical reactions by
lowering the activation energy. Most enzymes are proteins except ribonucleic acid enzymes (i.e.,
ribozymes) that are catalytic RNA molecules used in gene expression and cut larger RNAs.
49

Enzymes have active sites where the catalytic reactions take place. The reaction substrate binds
specifically to the active site of an enzyme through van der Waals forces, hydrophobic interactions,
hydrogen bonds, etc., to create an enzyme-substrate (ES) complex. Following the reaction, the
product releases from the enzyme active site and the enzyme is then free to bind a new substrate
molecule. Enzymes can interact with only a few specific substrates in what is often described with
“lock and key” analogy. However, enzymes are dynamic entities and undergo conformational
changes as the enzymes bind to the substrates. The continuous alteration in the confirmation of an
enzyme responding to substrate binding is known as “induced fit”. The induced fit of enzymes

9

enhances enzyme catalytic activity and hence, lowers the activation energy barrier. In other words,
it should be noted that this static view of enzymes does not properly account for the enzyme
dynamics and flexibility that are key to understanding the reaction pathway (e.g., enzymes
frequently must undergo subtle conformational changes to surpass the activation energy barrier).
50,
51
Enzyme conformational alterations and associated activity changes have been studied at
different conditions such as pH
52
and temperature
53
as well as in the presence of ligands,
54

denaturants,
55
and inhibitors.
56
Furthermore, enzyme binding sites can be selectively mutated to
change the enzymatic activity.
57

Leonor Michaelis and Maud Menten established the following reactions to explain enzyme
mechanism
58

E + S

𝑘 1
⇌

𝑘 −1
ES
𝑘 cat
→ E + P, (1)
where the enzyme, E, and substrate, S, reversibly bind and form an enzyme-substrate complex,
ES, with association and dissociation rate constants of k1 and k-1, respectively. Subsequently, the
enzyme-substrate complex converts with rate constant kcat to the product, P, and dissociated free
enzyme, E. Using the steady-state approximation, the Michaelis-Menten equation, namely

𝑑 [P]
𝑑𝑡 𝑉 =
𝑉 max
[S]
𝐾 M
+[S]
, (2)
can be obtained, where V is the rate of product formation, Vmax is the maximum reaction velocity,
[S] is the substrate concentration, and KM is the Michaelis constant. Briefly, this equation can be
derived as follows. The rate of product formation, V, can determined with the equation

𝑑 [P]
𝑑𝑡 =𝑉 =𝑘 cat
[ES]. (3)
To calculate V, the time rate of change of the enzyme-substrate complex concentration

10

𝑑 [ES]
𝑑𝑡 =𝑘 1
[E][S]−(𝑘 −1
+𝑘 𝑐𝑎𝑡 )[ES] (4)
is assumed to quickly level off at steady state and remain constant until the substrate concentration
decreases dramatically. Through this steady-state approximation, setting equation (4) to zero gives
𝑘 1
[E][S]=(𝑘 −1
+𝑘 cat
)[ES]. (5)
Next, the free enzyme concentration can be calculated by subtracting the enzyme-substrate
complex concentration [ES] from the total enzyme concentration [E]T, namely
[E]=[E]
T
−[ES]. (6)
Substituting equation (5) into equation (4) gives
𝑘 1
([E]
T
−[ES])[S]=(𝑘 −1
+𝑘 cat
)[ES], (7)
which can be rearranged to give
[ES]=
[E]
T
[S]
(
𝑘 −1
+𝑘 cat
𝑘 1
)+[𝑆 ]
. (8)
Substituting equation (8) into equation (2) then gives
𝑉 =
𝑘 cat
[E]
T
[S]
(
𝑘 −1
+𝑘 cat
𝑘 1
)+[S]
, (9)
where the Michaelis constant, KM, is defined as
𝐾 M
=
𝑘 −1
+𝑘 cat
𝑘 1
. (10)
Note that when the kinetic step (kcat) is slow compared to dissociation of the ES complex (k–1),
then KM is approximately equal to KD = k–1/k1 = [E][S]/[ES], namely the dissociation constant.
The maximum velocity, Vmax is the reaction rate at substrate saturating concentration and can be
determined with the assumption of the total enzyme concentration is equal to enzyme-substrate
complex concentration.

11

𝑉 max
=𝑘 cat
[ES]=𝑘 cat
[E]
T
(11)
After substituting equation (10) and (11) into equation (9), the Michaelis-Menten equation (2) is
obtained.
Enzyme kinetics are often measured to find the optimum conditions such as temperature,
pH, and concentration. It is usually found that the initial rate (Vi) is proportional to the substrate
concentration until the point is reached where all enzyme active sites are occupied by a substrate
molecule (i.e., the enzyme saturation point). After this substrate concentration is reached, the
reaction rate is limited by the rate of conversion of ES → E + P and reaches a maximum value
called Vmax (see Eqn. 2 and 11).
The Michaelis constant, KM, corresponds to the substrate concentration at which the
reaction rate is half of the Vmax. Enzyme-substrate dissociation constant, KD is defined with the
ratio of enzyme-substrate complex dissociation and formation rate constant (KD=k-1/k1). According
to the rapid equilibrium approximation, the enzyme-substrate production and separation are faster
than product generation; thus, rate constant of product formation value can be neglected compared
to the rate constants of enzyme-substrate formation and dissociation (k1 >> kcat and k-1 >> kcat) and
the Michaelis constant become equal to the enzyme-substrate dissociation constant (KM  KD =
k-1/k1). KM value shows the affinity of an enzyme for its substrate to reach half of the enzyme
maximum rate and the lower KM generally indicates higher enzyme binding affinity to its substrate.
The catalytic constant or the turnover number of an enzyme, kcat, is the maximum number
of substrate molecules converted to product per enzyme active site per unit time. The turnover
number is found from the following equation
𝑘 cat
=
𝑉 max
[E]
T
. (12)

12

The specificity constant or catalytic activity, kcat/KM, indicates the enzyme relative
efficiency. When substrate concentration is small compared to KM ([S] << KM), the Michaelis-
Menten equation (2) with substitution of equation (11) becomes
𝑉 =
𝑑 P
𝑑𝑡 =
𝑉 max
[S]
𝐾 M
=
𝑘 cat
[E]
T
[S]
𝐾 M
. (13)
Thus, kcat/KM would be the pseudo-second order rate constant for the corresponding
bimolecular reaction. While the specificity constant is only a pseudo rate constant, it can be used
effectively to examine a range of substrates for a given enzyme, with higher value of the specificity
constant indicating a better substrate for the enzyme. Similarly, the specificity constant, kcat/KM,
for an enzyme with a given substrate can be examined to optimize the conditions. Since it is often
difficult to measure Vmax and KM (and hence kcat) directly, these numbers are often estimated by
fitting rate data various linearized versions of equation (2), such as Lineweaver-Burk (1/V vs
1/[S]), Eadie-Hofstee (V vs V/[S]), and Hanes-Woolf (V vs V/[S]) plots.
59,60

1.2.2. Biofuel Enzymes
Sustainable energy sources are needed to meet current and quickly-expanding future
energy demands without harmful effects to the planet.
61
Biomass, which includes plants, crops,
corn stovers,
62
grass, wastes from lumber industry,
63
and agricultural and forest residues,
64
is the
most ubiquitous renewable energy source on Earth
65
and can be converted into biofuels to meet
the energy demand. Biofuels are vegetable oils, biodiesel, bio-oil, biochar, biohydrogen, bio-
synthetic gas (i.e., bio-syngas) and bio-alcohols (e.g., biomethanol, bioethanol, and biopropanol).
Ethanol (C2H5OH) is a promising biofuel with 34.7% oxygen content by mass, which
minimizes greenhouse gas generation and increases air quality.
66
For example, the combustion of
1 kg of ethanol results in the production of only 1.9 kg of CO2 versus 3.1 kg for 1 kg of octane, a

13

representative gasoline molecule for octane (C8H8). The lower heating values (LHV, i.e., the heat
produced by combustion of a fuel at constant pressure) of ethanol and octane are 26.95 MJ/kg and
43.45 MJ/kg, respectively. Although the LHV values of ethanol 37% lower than octane, the release
of greenhouse gas of ethanol is 61% lower than octane. Consequently, ethanol is an environment
friendly alternative sustainable energy source to gasoline.
Bioethanol is produced in four steps: (1) pretreatment, (2) hydrolysis of the biomass into
fermentable sugars, (3) bioconversion of the fermentable sugars into ethanol by microbial
fermentation, and (4) ethanol separation via distillation.
67,62
Lignocellulosic plant-cell walls
consists of four major components in a complex, semi-rigid structure: cellulose (15-50% of
lignocellulose by mass), hemicellulose (20-40%), pectin polysaccharides (30-35%), and the
aromatic polyphenol polymer lignin, each described separately below.
63,68,69
Pretreatment is
necessary to separate the components of the plant walls and prepare the constituent for enzymatic
saccharification. Lignocellulosic biomass is highly heterogeneous often requiring several enzymes
used simultaneously and synergistically for effective biomass saccharification.
Cellulose is packed tightly in a crystalline lattice structure that is stabilized by hydrogen
bonds. Thus, cellulose degradation into glucose requires three enzymes: endoglucanase
(endocellulase, EC 3.2.1.4), cellobiohydrolase (exocellulase, EC 3.2.1.91), and β-glucosidase
(cellobiase, 1,4-β-D-glucoside glucohydrolase, EC 3.2.1.21), as represented in Figure 2.
Endocellulase and exocellulase act simultaneously to break crystalline cellulose into cellobiose
(i.e., two linked glucose molecules), with endocellulase randomly breaking β-1,4 linkages and
exocellulase cleaving the ends of cellulose chains. Subsequently, cellobiose is degraded to two
glucose molecules by β-glucosidase. Generally, a mixture of all three types of biofuel enzymes is
applied to obtain higher amounts of reducing sugar since the presence of cellobiose inhibits endo-

14

and exo-cellulase. Thus, our initial efforts at increasing the activity of biofuel enzymes with
photoresponsive surfactants focuses on β-glucosidase and endoglucanase, as will be discussed in
Chapter 2 and 3, respectively.

Figure 2. A simplified cartoon of the enzymatic degradation of cellulose.

Hemicellulose is a heterogeneous polysaccharide containing xylan, glucuronoxylan,
arabinoxylan, xyloglucan, mannan, galactomannan, glucomannan in an amorphous structure.
69

Although hemicellulose is composed of different sugars, each residue is bonded with β-1,4
linkages to the backbone. Hemicellulose does not have hydrogen bonds between chains and, thus,
cannot form self-aggregates or long crystalline fibrils like cellulose and instead binds to the cell
wall through hydrogen bonds.
70
Xylan, the main component of hemicellulose, is composed of
xylose subunits and can be degraded with by a xylanase multi-enzyme consisting of endo-1,4-β-
xylanase (EC 3.2.1.8), β-xylosidase (EC 3.2.1.37), α-arabinofuranosidase (EC 3.2.1.55), and
acetyl xylan esterase (EC 3.1.1.72).
71
Mannan, another main component of hemicellulose, is
comprised of D-mannose, D-galactose, and D-glucose molecules. A mixture of β-mannosidase
(1,4-β-D-mannopyranoside hydrolase, EC 3.2.1.25), β-mannanase (1,4-β-D-mannan
mannohydrolase, EC 3.2.1.78), β-glucosidase (EC 3.2.1.21), α-galactosidase (1,6-α-D-galactoside

15

galactohydrolase, EC 3.2.1.22), and acetyl mannan esterase (EC 3.1.1.6) enzymes are used to
obtain monomeric sugars from mannan.
72

Pectin is composed of α-1,4-linked D-galacturonic acid colloidal polysaccharides such as
galacturonan, rhamnogalacturonan I and II, arabinan, galactan, D-galacturonan, and
arabinogalactan.
73
Pectinase, which contains pectate lyase (EC 4.2.2.2), pectin esterase (EC
3.1.1.11), and polygalacturonase (EC 3.2.1.15) enzymes, are used to degrade pectins into reduced
sugar molecules.
74
Hemicellulose and pectin are differentiated by their extraction method from the
cell wall. Hemicellulose, which is attached to cellulose and pectin fibrils, is dissolved in a soluble
alkaline solution after lignin separation, while pectin heterogenous fibrils are extracted with hot
water or chelating solvents such as oxalates.
Lignin is a complex aromatic polymer of polyphenols, mainly coumaryl, conferyl, and
sinapyl alcohols.
74
Lignin can be degraded into small molecules with phenol oxidases (EC
1.14.18.1), peroxidases, laccase (EC 1.10.3.2), lignin peroxidase (EC 1.11.1.14), manganase
peroxidase (EC 1.11.1.13), and versatile peroxidase (EC 1.11.1.16) enzymes.
75

1.2.3. TIM-Barrel Fold
The β-glucosidase enzyme discussed in Chapter 2 contains the common TIM-barrel fold,
which is composed of eight alternating α-helices and eight parallel β-strands closed into a toroidal
shape with loops connecting the α-helix and β-strand segments, as shown in Figure 3. The parallel
β-strands form the internal core of the fold, while the α-helices are located on the periphery of the
fold. Although there is a hole in the scheme, the core of proteins is tightly packed with mostly
hydrophobic amino acids. It is well-known that identical sequences fold into similar shapes and
the differences in the sequence gives differentiated shapes. Although TIM-barrel proteins fold into
(α/β)8 structure, they have divergent sequences.
76

16

The TIM-barrel structure, first observed in triosephosphate isomerase, is a commonly-
occurring enzyme framework that has been found in ten percent of the known protein structures.
77
All TIM-barrel folded proteins have catalytic activities with the C-terminal β-strand and adjacent
N-terminal α-helix forming the active site for substrate binding.
78
TIM-barrel folded enzymes are
typically isomerases, lyases,
76
oxidoreductases,
79
transferases,
77
aldolases,
77
or hydrolases such β-
glucosidase,
80
endoglucanase
81
, and endo-1,4-β-xylanase.
82

Figure 3. A top-down, schematic view of the TIM-barrel fold (from Nagarajan et al.
83
)

1.2.4. Surfactant-Protein Interactions
A protein typically folds into a compact, three-dimensional arrangement, the conformation
of which is responsible for the protein function (e.g., the TIM-barrel fold discussed just above).
Protein denaturation (i.e., unfolding), commonly observed upon increases in temperature (e.g.,
cooking an egg) or changes in pH, results in the loss of this native secondary, tertiary, and for
oligomeric proteins quaternary structure due to a combination of external (e.g., heat, electric or
magnetic field) and internal (e.g. high concentrations of proteins/molecular crowding, denaturants,
detergents, salts) factors. Similarly, protein unfolding can be achieved with surfactant addition
with the surfactant tails binding with the normally-buried hydrophobic amino acids.
84
The most

17

common example of this phenomenon is SDS-PAGE, where proteins are completely denatured in
a high-concentration sodium dodecyl sulfate (SDS) surfactant solution to separate proteins based
on their molecular weights. SDS micelles bind to proteins and break secondary, tertiary and
quaternary structure and produce linear amino acid chains. The applied electric field in SDS-PAGE
gives molecular weight-associated mobility to the linear amino acid chains. A tracking dye
attached to the proteins help to identify molecular weight of proteins. Therefore, protein-surfactant
interactions have been extensively studied to understand protein structure-property relationship.
85

At concentrations lower than the critical micelle concentration (CMC), the monomeric surfactant
molecules interact with proteins by a combination of electrostatic and hydrophobic forces. The
degree of protein unfolding is thus determined by the charge, hydrophobicity, carbon length, and
concentration of surfactants. In general, cationic surfactant head groups attach to acidic amino
acids such as aspartic acid (Asp) and glutamic acid (Glu), while anionic surfactant head groups
bind to basic amino acids like lysine (Lys), histidine (His), and arginine (Arg).
86
Furthermore, the
surfactant hydrophobic tails interact primarily with the hydrophobic amino acids valine (Val),
alanine (Ala), glycine (Gly), leucine (Leu), isoleucine (Ile), methionine (Met), proline (Pro),
phenylalanine (Phe), and tryptophan (Trp). Conversely, surfactant-protein interaction become
more complicated at surfactant concentrations higher than the CMC, as now entire micelles can
bind to the protein chain (i.e., a “beads on a string” model).
87

Azobenzene moiety is the most common used photoisomerizable unit in photoregulation
of biologics.
88, 89
Azobenzene has been applied to trigger protein unfolding,
90
, protein refolding,
91

conjugation to protein backbone,
92
replacement of an amino acid in protein with a
phenylazophenylalanine (PAP) amino acid.
93
Based on this nature of protein-surfactant
interactions, and the ability to control azoTAB hydrophobicity of with light illumination discussed

18

above, photo-induced changes in protein folding, function, and dynamics can be achieved. The
relatively-hydrophobic trans azoTAB isomer (75% of azoTAB under visible light) has a higher
affinity to bind and unfold proteins compared to the relatively-hydrophilic cis form (10% of
azoTAB under UV light). This in turn allows easily and reversibly control of the conformation of
the protein of interest. Additionally, the azobenzene portion of azoTAB, which is planar under
visible light and bent and twisted under UV light) might be expected to preferentially bind to
aromatic amino acid side chains (i.e., phenylalanine, tyrosine, and tryptophan) through -
stacking interactions, giving rise to localized and selective protein unfolding. Conversely, with
traditional surfactants such as SDS, protein-surfactant interactions are non-specific hydrophobic
interactions occurring throughout the protein, which further cannot be controlled with light
exposure. The ability to induce localized unfolding in -glucosidase and, as a result, enhanced
enzymatic activity with azoTAB is discussed in Chapter 2.

1.3. Catanionic Surfactant Systems and Their Application
When a cationic surfactant is mixed with an anionic surfactant in an aqueous solution,
spontaneous self-assemblies, most notably vesicles, can be formed in the catanionic solution. The
vesicles form spontaneously without the need for external energy such as shearing (recall G =
H – TS + A, where  is the interfacial tension and A in the increase in surface area upon
vesicle formation), a result of favorable electrostatic and hydrophobic interactions between the
oppositely-charged surfactants. Hence, the vesicles, once formed, are infinitely stable as the
system is in thermodynamic equilibrium. Aqueous mixtures of anionic and cationic surfactants
display synergistic behavior and create various microstructures at concentrations orders of
magnitude lower than the pure-component critical micelle concentrations. These self-assemblies

19

can be spherical micelles, cylindrical rod-like micelles, disk-like micelles, vesicles (i.e., hollow
spherical structures with a closed single bilayer), multilamellar (layered flat bilayers), or liquid
crystalline structures depends on the cationic-anionic surfactant ratio, temperature, and pH.
94,95

Generally, equimolar concentrations of catanionic solutions precipitate or form lamellar structures
owing to the offsetting positive and negative charges.
Catanionic vesicles have received considerable attention in nanotechnology, chemical and
pharmaceutical applications such as membrane protein folding/solubilization,
32
gene therapy,
35

genetic research, drug delivery,
96
separation,
97
vaccines,
98
cosmetic probe and molecule
encapsulation
99
, with long-term stability,
100
and cost-effectiveness. These vesicles are formed at
very low concentrations; thus, they can be used for health products, cosmetics, and environmental
applications with low toxicity.
101
In addition, catanionic-vesicle formulations are stable for years
when the solution and the conditions are kept undisturbed,
102,103
owning to the aforementioned
equilibrium nature. Vesicle properties such as size, surface charge, and permeability can be
modified by adjusting the temperature, pH, concentration, cationic-to-anionic surfactant molar
ratio, and surfactant hydrophobic tail length.
104
For example, catanionic vesicles formed with short
alkyl-based surfactants have higher membrane permeabilities, whereas vesicles formed with
longer chain lengths are generally more stable.
105
The bilayer thickness changes of catanionic
vesicles formed with the increasing hydrocarbon tail length of sodium dodecyl sulfate (SDS)
anionic surfactant analogs (C8-C 18) and azoTAB photoresponsive cationic surfactant interactions
will be suggested to explore in Chapter 7.
Common experimental methods used to characterize catanionic vesicle systems includes
dynamic light scattering, fluorescence spectroscopy, cryo-EM, and small-angle neutron or X-ray
scattering (SANS or SAXS). Dynamic light scattering measures oscillations of scattered light

20

intensity due to Brownian motion of microstructures, allowing the diffusion coefficient to be
measured and the corresponding size distribution calculated from the Stokes-Einstein equation.
The critical aggregation concentration (CAC), the lowest surfactant concentration where vesicle
first appears, is often detected using fluorescence spectroscopy of hydrophobic fluorescent probes
solubilized into bilayers. Cryogenic-electron microscopy (cryo-EM) provides a means to directly
visualize the self-assemblies at high resolutions, albeit following flash freezing at cryogenic
temperatures (i.e., -150 C to -273 C). Small-angle neutron scattering (SANS), discussed and used
extensively below, allows estimation for the size, shape, and correlations between catanionic
microstructures.
Photoresponsive catanionic surfactant mixtures consisting of aqueous solutions of azoTAB
mixed with the conventional anionic surfactant sodium dodecyl benzene sulfonate (SDBS) allows
microstructural changes to be initiated with light.
41
As shown in Figure 4, photo-induced micelle
↔ vesicle, vesicle ↔ micelle, vesicle ↔ multilamellar, and vesicle ↔ monomer transitions were
observed.
41
These light-controlled transitions could potentially allow for the triggered release of
biomolecules, therapeutics, peptides, proteins, and drugs. In previous studies, the photoreversible
vesicle-to-micelle transition were used to control membrane protein folding (i.e., folded-to-
unfolded, respectively),
32
while the vesicle-to-monomer transition was utilized for gene delivery
via light-triggered DNA release,
35
siRNA and paclitaxel release
96
from the vesicle once the carrier
passed the cell membrane.
35
In Chapter 4, the vesicle-to-monomer transition will be used to allow
for co-delivery of siRNA and a hydrophobic anti-cancer drug into breast cancer cells.

21

Figure 4. Light-induced self-assembly structure changes in azoTAB-SDBS catanionic
solutions.
1.3.1. SiRNA Delivery
SiRNA is a promising therapeutic genetic material due to the ability to knock down specific
genes that causes life-threatening diseases such as cancer, viral infections, and so-called
undruggable diseases such as cystic fibrosis,
106
sickle cell anemia,
107
autism spectrum disorder,
108

Gaucher disease,
109
hemophilia
110
, etc. SiRNA inhibits only one specific target mRNA to suppress
the undesired genes. This is known as “sequence-specific gene silencing” and is illustrated in
Figure 5. The formation of siRNA begins with double-stranded RNA (dsRNA) or hairpin RNA
(shRNA) that are cut by DICER enzyme to create small interfering RNA (siRNA) generally 21-
23 nucleotides long. Once siRNA enters into a cell, the strands are separated and associate with
AGO proteins to form RNA Induced Silencing Complex (RISC). In RISC, the sequence of single-
stranded siRNA matches with the complementary mRNA responsible for the production of the
disease-related protein. Subsequently, the mRNA is sliced when the mRNA leaves RISC, causing
the broken mRNA to not be able to synthesize the disease-associated protein.

22

Figure 5. Gene silencing mechanism of siRNA.

Significantly, a single-strand siRNA has the ability to engage with several complementary
mRNAs.
111
Although siRNA is a promising genetic strategy to inhibit cancerous cell growth,
siRNA delivery is complicated by the negative charge, low cellular uptake through cellular
membranes,
112
and degradation by RNases (i.e., serum nucleases) of unprotected siRNA before
reaching to the target cells.
113
Conversely, a complex that protects and allows for the uptake of
siRNA is more often than not too stable to allow release of siRNA into the cytoplasm. Photo-
induced vesicle disruption of the type shown in Figure 4 provides a unique opportunity to address
this challenge, as will be demonstrated in Chapter 4.
Nanoparticles of various types have been used for the effective delivery siRNA into tumor
cells and tissues,
114
as siRNA encapsulated in a nanoparticle is protected from endo- and exo-
nucleases. The size, shape, stability, and surface charge/chemistry of nanoparticles are important
factors for cellular uptake.
112, 113
The long blood-circulation times (half-life >2 h) of nanocarriers,

23

which depend on the nanoparticle size, increase the chance of diffusion into tumor cells.
113
For
example, positively-charged nanoparticles approximately 100 nm in diameter are able to enter into
tumor angiogenic endothelial cells.
115
Smaller particles (<80 nm) are clarified by liver, while
particles larger than about 250 nm are filtered by the spleen.
116,117

One obstacle of using siRNA-loaded nanoparticle is often an uncontrollable release profile,
including “burst release” and “delayed release”. Conversely, triggered release can be achieved by
using stimuli-responsive nanoparticles, which have unique properties such as pH-, enzyme-,
magnetic field-, ultrasound-, or light- responsiveness, which can solve the siRNA release problem
and increase the localization of the loaded therapeutics.

1.4. Small-angle Neutron Scattering (SANS)
Neutron scattering techniques have been used for several decades for structural analysis of
polymers, colloidal solutions, macromolecules, biomaterials, and nanomaterials. As will be
discussed in greater detail in the chapters that follow, SANS data access length scales L = 2/Q
ranging from 1 nm – 1000 nm, where the scattering vector Q = [4sin/],  is the neutron
wavelength (~ 6 A), and  is the scattering angle. Often the scattering intensity, I(Q), is fit to a
model of the form
𝐼 (𝑄 )=𝜙𝑉𝛥 𝜌 2
𝑃 (𝑄 )𝑆 (𝑄 ) (14)
where ϕ is the particle volume fraction, V is the particle volume, 𝛥𝜌 is the scattering length density
difference between the particle and the solvent, P(Q) is the form factor accounting for the size and
shape of the particle, and S(Q) is the structure factor, which is correlated with particle interactions
and the distribution function of particle distances.

24

Small-angle neutron scattering is used to illustrate the tertiary and quaternary structure of
native and denatured (non-native) protein. The unfolded region or adjusted configuration of the
denatured protein in the presence of the azoTAB surfactant under visible and UV light could be
specified. We have previously utilized SANS to determine the native and the denatured
confirmation of a variety of proteins in solution.
24, 37-39, 41
Native state and localized azoTAB-
promoted unfolding of the enzyme of β-glucosidase will be examined in Chapter 2.
SANS technique is utilized to obtain better estimations for the shape, physical appearance,
and structure of catanionic microstructures in solution. The average core radius and the constant
shell thickness of catanionic vesicles could be characterized. Additionally, effective radius and
bilayer thickness calculation and, hence, accurate understanding of self-assembly structure could
be estimated. Furthermore, the transition of light-responsive micelle ↔ vesicle, vesicle ↔ micelle,
vesicle ↔ multilamellar, and vesicle ↔ monomer could be observed. We have used SANS to
characterize the self-assembled structures of surfactants, including micelles, vesicles, and
multilamellar structures.
41, 118
Bilayer thickness changes of catanionic vesicles of azoTAB
(cationic) and SDS analogs (anionic, C8-C18) will be suggested to investigate in Chapter 7.

1.5. Overview
In Chapter 2, the photocontrol of β-glucosidase (cellobiase from Aspergillus niger) activity
and conformation as a function of azoTAB concentration under visible and UV light will be
discussed. In Chapter 3, the improvement of endoglucanase (from Aspergillus niger) kinetics with
addition of a light responsive azoTAB surfactant will be explored. In Chapter 4, the unique photo-
assisted transition of vesicles to free monomers will be utilized to co-deliver both siRNA and a
chemotherapeutic drug, paclitaxel, to cancer cells via light-responsive catanionic vesicles. In

25

Chapter 5, a unique photo-initiated transition of vesicles to a crystalline state of azoTAB with
sodium decyl sulfate, SdecS and sodium 4-octylbenzenesulfonate, SOBS and their enzyme
recycling application will be inquired. In Chapter 6, the interaction of azoTAB surfactant and other
biofuel enzymes and their potential usage in industry will be discussed. Additionally, several new
catanionic surfactant mixtures formed from a variety of azoTAB and sodium dodecyl sulfate (SDS)
analogs will be introduced, and plans will be put forward to measure the measure the
microstructure of these systems.

26

CHAPTER 2: SUPERACTIVITY OF THE CELLULASE ENZYME β-
GLUCOSIDASE ACHIEVED THROUGH CONFORMATIONAL CHANGES IN THE
PRESENCE OF A PHOTORESPONSIVE SURFACTANT

2.1. Abstract
β-Glucosidases catalyze the hydrolysis of cellobiose to glucose, which is often the rate-limiting
step in the conversion of cellulose into fermentable sugars and subsequent bioethanol for use as a
readily-available and sustainable energy source. Thus, the effect of a photoresponsive azobenzene-
based surfactant azoTAB on the structure and function of β-glucosidase from Aspergillus niger
was examined as a means to explore potentially superactive enzyme conformations. AzoTAB
undergoes a reversible photoisomerization from a relatively-hydrophobic trans form under visible
light to a relatively-hydrophilic cis isomer upon UV illumination, allowing protein-surfactant
interactions, and the protein conformational changes that result, to be controlled with light
illumination. Light scattering and neutron scattering data indicate that pure β-glucosidase exists
as dimers or higher aggregates in solution that are progressively converted to the monomeric state
with increasing azoTAB concentration, which is accompanied by up to a 60% increase in catalytic
activity. In contrast, the enzyme is simply deactivated in the presence of straight-chain
hydrocarbon surfactants. Shape-reconstructed images obtained from small-angle neutron
scattering data suggest that azoTAB causes selective unfolding in the α/β sandwich domain that
comprises the majority of the crystallographic dimer interface, consistent with the observed
transition to monomeric structures. Furthermore, this domain forms one side of a long cleft that
begins at the active site and facilitates the binding of oligosaccharides, which at times can block
the active site. By fitting the kinetic data to a model where binding of a second substrate leads to

27

a catalytically-inactive ES2 complex (i.e., competing pathways ES2
𝑆 ↔ ES

→ E + P), azoTAB-
induced superactivity in β-glucosidase was found to be a result of a 4-fold increase in the ES2
dissociation constant (i.e., diminished substrate inhibition), thus, providing a unique means of
obtaining glucose-tolerant -glucosidases.
2.2. Introduction
Cellulose, the primary component of plant cell walls, is the most abundant organic
molecule on Earth.
119
Thus, bioethanol produced by the conversion of cellulose into fermentable
sugars (i.e., glucose) using cellulase enzymes represents a renewable and readily-available energy
source.
120
Furthermore, the replacement of gasoline with bioethanol could have substantial
environmental benefits, principally due to the fact that plant growth utilizes preexisting
atmospheric CO2 that would simply be re-released upon bioethanol combustion, while conversely
petroleum-based fuels liberate ancient carbon sources. Indeed, relative to the above consideration,
differences in the amounts of CO2 generated from these two fuels by light-duty vehicles are usually
small, indicating a net 10% reduction in CO2 emission for bioethanol versus gasoline.
121
This is
perhaps not surprising given that while the energy density of ethanol is ~63% of the gasoline value
on a per volume basis,
66
or equivalently 24% on a per mole basis using octane as a representative
species, combustion of the former (C2H5OH + 3 O2 → 2 CO2 + 3 H2O) produces four times less
CO2 than the latter (C8H18 +
25
2
O2 → 8 CO2 + 9 H2O). The high oxygen content of ethanol relative
to gasoline does promote complete combustion and, thus, reduce the emission of harmful
particulates.
122
Furthermore, ethanol has negligible sulfur content, thereby minimizing sulfur
dioxide emission.
66

Production of glucose from cellulose requires the coordinated action of three cellulase
enzymes. First, endoglucanase (endocellulase, EC 3.2.1.4) breaks down crystalline cellulose into

28

individual cellulose polymer chains and further breaks internal glycoside bonds resulting in
random chain ends. Next, cellobiohydrolase (exocellulase, EC 3.2.1.91) cleaves off the
disaccharide cellobiose from these exposed chain ends. Finally, β-glucosidase (cellobiase,
gentiobiase, EC 3.2.1.21) breaks the linkage of cellobiose to produce two glucose molecules that
can be subsequently fermented into ethanol. Of these three enzymatic steps, hydrolysis of
cellobiose by -glucosidase is often rate limiting as cellobiose inhibits the activity of
endoglucanase and exoglucanase,
62, 123
while β-glucosidase can be inhibited by both the cellobiose
substrate and glucose product.
124
In general, cellulases are one to two orders of magnitude less
reactive than enzymes used for corn ethanol production (e.g., amylases),
125
with enzymatic
efficiencies (i.e., kcat/KM values) in the bottom 25% of all known enzymes.
126
Thus, 40–100 times
more enzyme can be required to produce the same amount of ethanol from cellulose relative to
corn,
127
impacting the economic feasible of cellulosic ethanol. Specifically, ~100 g of enzyme are
required per gallon of cellulosic ethanol produced
128
(or equivalently ~25 mg per mL of ethanol).
Thus, for the optimistic case of on-site production, the resulting enzyme costs is on the order of
$0.50 per gallon of ethanol
129
(or $0.80 per equivalent gallon of gasoline), while others have
predicted as high as $1.30 per gallon of ethanol.
130
For comparison, costs of a surfactant additive
can be estimated to be ~$0.01 per gallon assuming a surfactant/protein ratio of 10
–3
mol/mg (see
below) and using palm kernel oil as a C12 fatty acid feedstock at a bulk cost of ~$1000/ton.
Surfactants are frequently used during bioethanol production,
131
providing a means to enhance the
surface availability of the biomass (versus acid pretreatment, steam explosion, etc.), while at the
same time minimizing irreversible enzyme binding to lignin that leads to rapid deactivation during
hydrolysis.
130, 132
Furthermore, various surfactants
133-136
have been employed to directly increase
the enzymatic efficiency of cellulases, typically with modest results (i.e., <20% improvement).

29

β-glucosidases are produced by a myriad of microorganisms, plants, and animals.
137
The
catalytic center is generally found in a 15 Å – 20 Å deep cavity located in a TIM-barrel domain,
and consists of an aspartic acid nucleophile (conserved across family 3 glycosidases, GH3)
separated by ~5.5 Å from a (non-conserved) carbocyclic acid which serves as the catalytic
acid/base residue.
138
In the case of -glucosidase from Aspergillus aculeatus (PDB: 4IIB), which
shares 83% sequence identity
139
with -glucosidase from Aspergillus niger employed in this work,
the active site cavity is located at the interface of the TIM barrel and α/β sandwich domains, which
respectively house the catalytic nucleophile and acid/base residues. This domain interface further
forms a long cleft that begins at the active site (i.e., subsites –1/+1) and extends at least past subsite
+4, allowing oligosaccharides to bind and either undergo hydrolysis (when bound to subsites –
1/+1 and beyond) or potentially block the active site (when binding begins at subsite +1).
140

Hence, much work has been devoted to develop so-called glucose-tolerant glucosidases where the
above inhibitory effect is minimized.
141

Azobenzene trimethylammonium bromide (azoTAB) is a light-responsive surfactant that
undergoes a reversible photoisomerization from a planar trans orientation under visible light
(wavelength of maximum absorption, max = 434 nm) to a bent and twisted cis conformation under
UV light (max = 350 nm) light, as shown in Scheme 1. A smaller dipole moment across the
nitrogen double bond causes the trans isomer to be relatively hydrophobic and thus have a higher
binding affinity to proteins compared to the cis isomer.
34
This photoreversible control of
surfactant-protein interactions has been utilized as a means to initiate changes in protein
secondary,
142
tertiary,
40
and quaternary
143, 144
structure; enzymatic kinetics,
38, 145
internal
dynamics,
39
and inhibitor interactions;
146
and membrane protein folding.
32
The azobenzene moiety
in general has been frequently employed to photo-regulate biological macromolecules,
88, 89
often

30

through azobenzene-modified polymers,
90, 91
direct conjugation to the protein backbone,
92, 147, 148

or replacement of an amino acid with phenylazophenylalanine.
149
Notable examples from the
perspective of the current enzymatic study include a 3-fold increase in the activity of azobenzene-
conjugated papain,
150
a 4-fold increase in the reaction rate of ribonuclease S containing
phenylazophenylalanine,
93
an 8-folding increase in the activity of lysozyme upon interaction with
azoTAB,
145
and a 16-fold increase in activity of endonuclease crosslinked to an azobenzene
moiety.
92

In the present study, the influence of azoTAB on the conformation and activity of β-
glucosidase from Aspergillus niger is examined as a means to initiate superactivity in the enzyme.
The catalytic activity is evaluated as a function of both azoTAB and substrate concentrations, and
correlated with changes in the enzyme structure determined from dynamic light scattering,
fluorescence spectroscopy, and circular dichroism spectroscopy. Small-angle neutron scattering
is further employed to develop shape-reconstructed images of the protein in solution at resolutions
on the order of individual protein domains. The results indicate that the observed β-glucosidase
superactivity (up to 60% higher than the native enzyme) is a result of oligomeric dissociation into
a slightly-unfolded monomeric structure where the  sandwich domain, responsible for both
dimer formation and substrate inhibition, has separated from the remaining protein molecule.
2.3. Experimental Section
Materials. 4-ethyl-4’(trimethylamino-butoxy) azobenzene bromide (azoTAB, Mw = 420
g/mol), the photosensitive surfactant shown in Scheme 1, was synthesized according to published
procedures.
31,33,32
Briefly, azoTAB was produced in three steps sequentially: azocoupling,
alkylation, and quaternalization. All chemicals for azoTAB preparation were obtained from
Sigma-Aldrich at the highest purity. AzoTAB photoisomerization to the cis state was typically

31

achieved with an 84-W long-wave (365 nm) UV lamp (Spectroline, Model no. XX-15A), with the
samples kept in the dark during experiments. Conversion to the cis isomer during the small-angle
neutron scattering experiments was attained by using a 200 W Mercury arc lamp (Oriel, 6283)
equipped with a light guide (Oriel, 77557), fiber-bundle focusing assembly (Oriel, 77800), 320 nm
band pass filter (Newport Corporation, FSQ-UG5), and an IR light and heat absorbing filter
(Newport Corporation, FSQ-KG3). Conversion back to the trans state was achieved using 400 nm
longpass filter (Newport Corporation, FSQ-GG400). Nile red (Sigma Aldrich, N3013-), sodium
dodecyl sulfate (J. T. Baker, 4095-04), and dodecyltrimethylammonium bromide (TCI, D1468)
were purchased as indicated.
Purification of β-Glucosidase. Glucosidase from Aspergillus niger was purchased from
Sigma-Aldrich (Cat. no. 49291). The gray-brown crude powder was solubilized in 50 mM Tris-
HCl, pH 7.5 with 100 mM NaCl at a protein concentration of 15 mg/mL. Solid ammonium sulfate
(NH4)2SO4 at 80% saturation was then added into the protein solution, followed by storing the
overnight at 4 C. The protein precipitate was recovered by centrifugation at 12,000 rpm for 15
minutes and washed with an 80% (NH4)2SO4 solution three times. Next, the protein precipitate was
dissolved in 50 mM sodium phosphate, pH 7.2 buffer containing 1 M (NH4)2SO4 and passed
through three HiTrap Phenyl HP (GE Healthcare, 17-1351-01) 1 mL columns to purify β-
glucosidase based on hydrophobic interactions using fast protein liquid chromatography, FPLC
(BIO-RAD, BioLogic DuoFlow 10) at 4 C. Samples were subsequently purified with FPLC using
a 50 mM sodium phosphate, pH 7.2 buffer and (NH4)2SO4 concentrations from 1 M to 0 M.
Fractions of 4 mL were collected at 1 mL/min to achieve the desired resolution. Fractions turning
blue following mixing 10 μL with Bradford reagent (Sigma-B6916) were then tested against p-
nitrophenyl β-D-glucopyranoside (Sigma, N7006) activity to detect the presence of β-glucosidase.

32

Finally, the excess (NH4)2SO4 was removed by concentrating the β-glucosidase fractions using
Amicon Ultra 15 mL centrifuge tubes with 50 kDa membranes (UFC905008). The concentrated
protein solutions were then washed 5-6 times with 50 mM Tris-HCl, 100 mM NaCl buffer, pH 7.5.
Determination of β-Glucosidase Activity. β-glucosidase activity was monitored by
measuring the conversion of cellobiose to glucose using the Somogyi-Nelson method.
151
Briefly,
1 mL of cellobiose (15 mM) was mixed with 1 mL of an azoTAB solution at various concentration
and then with 1 mL of a β-glucosidase solution (90 μg/mL), each in 50 mM sodium acetate buffer,
pH 5 at 25 C. When desired, azoTAB was pre-converted to the cis state by exposing the azoTAB
stock solution separately to UV light, with the reaction then performed under dark conditions to
minimize conversion of azoTAB back to the trans state. Three 50-μL samples from the reaction
solution were separately stopped into 450 μL of a 50 mM phosphate buffer, pH 9.5 following
initial mixing and subsequently every two minutes for twenty minutes. 500 μL of alkaline copper
tartrate was then added to each solution followed by heating for 20 minutes in boiling water to
produce cuprous oxide with the reduced sugar. 500 μL of arsenomolybdic acid was then mixed
into each solution, where molybdic acid was reduced to molybdenum blue in the presence of
cuprous oxide. 200 μL of each solution was then diluted with 2 mL of deionized water and the
absorbance at 760 nm minus the background absorbance at 500 nm was determined using UV-vis
spectroscopy (Agilent, model 8453). The activity of β-glucosidase was calculated from the initial
change in absorbance versus time (converted to units of mM/s using an extinction coefficient of
0.32 mM
–1
cm
–1
determined independently from a glucose concentration curve) by fitting the data
to an exponential of the form (At – A)/(A0 – A) = e
–at
, where At is the absorbance at time t. Error
bars were estimated by averaging the triplicated data. To determine kcat and KM, the cellobiose

33

concentration was varied from 0.5 mM – 7.5 mM, while the enzyme molecular weight was
assumed to be 110 kDa (see Fig. 6) using kcat = Vmax/[E].
Photocontrol of β-glucosidase activity: To monitor changes in β-glucosidase activity
with light conditions, 0.5 mM azoTAB, 5 mM p-nitrophenyl β-D-glucopyranoside, and 10 μg/mL
β-glucosidase were employed in 50 mM sodium acetate pH 5 buffer. At various times, 0.5 mL of
reaction sample were taken and quickly mixed with 0.5 mL of 50 mM phosphate buffer pH 9.5 at
to stop the reactions. The formation of p-nitrophenol was monitored at 429 nm (429 = 13.2 mM
-1

cm
-1
)
152
as opposed to the more traditional 405 nm so as to coincide with the isosbestic point of
azoTAB located between the strong absorption peaks at 350 nm (-* transition band of the trans
isomer) and 434 nm (n-* transitions of both isomers). Following ~12 minutes of reaction under
either visible or UV light (with azoTAB pre-converted to the cis state)”, the reaction mixture was
then exposed to UV or visible (room) light, respectively, and the reaction was followed for 20
minutes total.
Small-Angle Neutron Scattering (SANS). Small-angle neutron scattering experiments
were performed on the 30 m NG7 SANS instrument at NIST.
153
A neutron wavelength of  = 6
Å and a detector offset of 25 cm with two sample-to-detector distances of 1.33 and 7.0 m were
utilized to achieve a Q-range of Q = (4/)sin(/2) = 0.0049 − 0.55 Å
-1
, where  is the scattering
angle. The net intensities were corrected for the background and empty cell (pure D2O), accounting
for the detector efficiency using the scattering from an isotropic scatterer (Plexiglas), and
converted to an absolute differential cross section per unit sample volume (in units of cm
-1
) using
an attenuated empty beam. The SANS data were reduced, analyzed and modeled using the Igor
Pro (WaveMetics) program supplied by NIST.
154
A constant background scattering intensity (i.e.,
the isotropic scattering observed at high Q due to the sample cell, incoherent scattering resulting

34

from the hydrogen atoms, etc.) was subtracted from each sample. β-glucosidase solutions at 1
mg/mL were prepared in 50 mM sodium acetate buffer, pH 5 using D2O as the solvent. Following
data collection at 25 C under visible light, the same SANS cuvettes were then illuminated with
UV light for at least 2 hours and the SANS data was again collected. A portable UV lamp
(Spectroline, Model no. XX-15A) was also placed in the sample chamber for continuous UV
illumination during SANS data collection to prevent thermal conversion of azoTAB back to the
trans conformation (the half-life of thermal conversion is ~24 hours at 25 C).
The SANS data were analyzed using a range of complementary techniques to determine
the conformation of β-glucosidase in solution as a function of azoTAB concentration and light
conditions. Cross-sectional radii of gyration, Rc, were determined by fitting the data to the modified
Guinier approximation for cylinders, namely 𝑄 𝐼 (𝑄 )≅𝐼 c
(0)exp(−𝑅 c
2
𝑄 2
2 ⁄ ), using QRc < 1.
155

Following deconvolution, globular radii of gyration, Rg, were determined by fitting the data to the
Guinier approximation for spherical scatters, namely 𝐼 (𝑄 )≅𝐼 (0)exp(−𝑅 g
2
𝑄 2
3 ⁄ ), using QRg <
1.3.
155
Pair distance distribution functions, P(r), related to the probability of two scattering centers
(i.e., nuclei for SANS) being within a distance of r apart within the protein particle (single
macromolecule or aggregate), were calculated using the program GNOM
156
through the equation
( ) ( )
( )

=
max
0
sin
4
D
dr
Qr
Qr
r P Q I 
, (15)
where Dmax is the maximum intraparticle distance. Shape reconstructions were performed using
the program GA_STRUCT.
157
Briefly, the program beings with an initial system of randomly-
oriented scattering centers, and uses a genetic algorithm to optimize the positions of the scattering
centers to fit the experimental data. A data range of Q = 0.01 – 0.3Å
-1
was employed to exclude
intermolecular interactions at low Q and avoid length scales too small for protein continuity at

35

high Q. Ten independent runs were performed, and the reconstructed models were compared for
similarity and averaged to give the reported consensus envelopes.
Dynamic Light Scattering (DLS). Dynamics light scattering measurements were
performed at a scattering angle of 90 and temperature of 25 C using a Brookhaven Model BI-
200SM instrument equipped with a 35 mW HeNe (632.8 nm) laser (Melles Griot, model no. 05-
LHP-928), an avalanche photodiode detector (BI-APD), and a BI-9000AT digital correlator. 1
mg/mL β-glucosidase in 50 mM sodium acetate buffer, pH 5 was prepared and filtered through a
100-nm syringe filter (Whatman, Anotop 10, 6809-1012). A proper amount of a 30 mM stock
azoTAB solution was separately filtered and added into the above β-glucosidase solution. Samples
were allowed to reach the respective photostationary state for half an hour under either visible
(room light) or UV light, respectively. Hydrodynamic diameters were obtained using the non-
negative least squares (NNLS) method.
Fluorescence Spectroscopy. Fluorescence measurements were performed using a Quanta-
Master spectrofluorometer (Photon Technology International, model QM-4) with 4 nm excitation
and emission slit widths at 25 C. 90 μg/mL β-glucosidase in 50 mM sodium acetate buffer, pH 5
was used with various concentrations of azoTAB. A quartz cuvette with a volume of 315 μL and
a path length of 3 mm (QS-608, Nova Biotech) was employed. When indicated, Nile red was used
as a hydrophobic and nonionic fluorescence probe using an excitation wavelength of 590 nm with
the emission spectra monitored from 600 nm – 760 nm. β-glucosidase solutions with various
azoTAB concentrations were prepared and remained under laboratory or UV light exposure for 30
min, followed by the addition of 2 μL of Nile red (from a 0.1 mM solution in ethanol). Prior to
data collection, each sample was allowed to reach equilibrium for 20 minutes in the dark. When
indicated, the intrinsic fluorescence spectra of β-glucosidase were recorded from 300 nm – 500

36

nm with an excitation wavelength of 280 nm. The measured fluorescence emission intensities were
corrected for the varying azoTAB absorbance in this region as described in the text. The maximum
emission wavelengths (λmax) were calculated by fitting each curve to a fifth-order polynomial.
Circular Dichroism (CD). The secondary structure of β-glucosidase was measured using
a Jasco J-815 CD spectrometer (Jasco Inc., Easton, MD) at the Nanobiophysics Center of the
University of Southern California. β-glucosidase at a concentration of 0.2 mg/mL was prepared in
50 mM sodium acetate buffer, pH 5 at various surfactant concentrations (< 1 mM due to the strong
absorbance of azoTAB) and light conditions. CD spectra were monitored in the far-UV region
(200 nm – 250 nm) at 25 C using a 1-mm path length quartz cuvette (Agilent Technologies, 5061-
3384). Each measured spectrum was averaged over three scans and corrected by subtraction of
the buffer spectrum. The percentages of the secondary structure corresponding to α helices, β
sheets, and random coils were quantified with analysis according to the K2D3 method.
158

2.4. Results and Discussion
Glucosidase derived from the fungus Aspergillus niger is commonly employed due to the
relatively-high kinetic activity observed with this particular isozyme.
159
As seen in Figure 6, the
commercially-available crude product contains at least four protein fractions in the molecular
weight range studied, not surprising given that over 170 genes in A. niger encode proteins capable
of biomass degradation.
160
To purify the active β-glucosidase fraction, the method employed by
Lima et al.
139
was applied with a few modification. First, the proteins were precipitated as a solid
pellet in an 80 % ammonium sulfate aqueous solution to remove non-protein impurities (notably
the brown pigments found in crude glucosidase). Next, the solid pellet was dissolved in 50 mM
sodium phosphate, pH 7.2 buffer and subjected to hydrophobic-interaction chromatography.
Following desalting using a 50 kDa cutoff membrane, fractions containing protein (via Bradford

37

reagent) and exhibiting hydrolysis activity (with p-nitrophenyl β-D-glucopyranoside as a
substrate) were collected. As seen in Figure 6, the final result was primarily a single β-glucosidase
protein (estimated purity ~95%) with a molecular weight of approximately 110 kDa. For
comparison, the molecular weight of β-glucosidases from similar Aspergillus niger strains have
been reported to be between 90 and 130 kDa.
139, 161-165

Figure 6. SDS-PAGE of the crude glucosidase product (lane 2) and purified β-glucosidase
(lane 3) versus standard protein markers (lane 1).
The activity of purified β-glucosidase was measured as a function of azoTAB concentration
and light conditions, as shown in Figure 7. The activity was found to initially increase upon adding
azoTAB, leading to a maximum of ~40% superactivity near 0.5 mM azoTAB under visible light
and ~55% superactivity near 0.75 mM azoTAB under UV light. Further increases in azoTAB
concentration, however, cause the β-glucosidase activity to decrease from these maxima,
eventually resulting in an activity lower than the native state at elevated azoTAB concentrations.
For comparison, β-glucosidase activity was also measured in the presence of the conventional
surfactants dodecyltrimethylammonium bromide (DTAB, analogous to azoTAB) and sodium
dodecyl sulfate (SDS). Compared to azoTAB, these surfactants show no superactivity and

38

furthermore give rise to more pronounced deactivation at higher concentrations. Similarly, only
modest superactivity (<20%) has been previously reported for β-glucosidase in the presence of
Tween 80 or a rhamnolipid biosurfactant.
135
Likewise, β-glucosidase activity has been shown to
be inhibited at high concentrations of SDS, Tween 20, β-mercaphtoethanol, Triton X-100, or
sodium lauroyl sarcosinate.
166
Surfactants have also been used with the complete cellulase
complex containing endo- and exo-cellulase as well as β-glucosidase, with the cellulose hydrolysis
rate increasing by 10% – 20% upon the addition of low concentrations (0.5 g/L) of Triton X-100,
Tween 20, or Tween 80, while the activity was inhibited at higher concentrations (5 g/L).
133

50
100
150
0 0.5 1 1.5 2 2.5 3
azoTAB (vis)
azoTAB (UV)
SDS
DTAB
Surfactant Concentration (mM)
Relative Activity (%)
(a)

Figure 7. (a) Relative activity of β-glucosidase with cellobiose as a substrate in the
presence of azoTAB (under visible or UV light), SDS, and DTAB surfactants. Lines are drawn to
guide the eyes. (b) and (c) Absorbance of the hydrolysis product from the model substate.
0.4
0.5
0.6
0.7
0 5 10 15 20 25
Absorbance
UV (0.019 M/s)
visible
(0.013)
(b)
0.4
0.5
0.6
0.7
0 5 10 15 20 25
Time (min)
Absorbance
vis (0.013 M/s)
UV (0.019)
(c)

39

Notably, the maximum activities in Figure 7 are found to occur at azoTAB concentrations
where the hydrodynamic diameters of β-glucosidase are a minimum (Figure 8a), having decreased
from dimeric (dH ~13 nm) to monomeric (dH = 9  1 nm) values but prior to the observation of
large, and presumably inactive, protein aggregates (denoted by the hatched regions in Figure 8a).
The reason for the assignment of these dH values to the dimer and monomer is as follows (and
further supported with SANS below). For a wide range of globular proteins the diameter has been
shown
167
to vary with MW
0.369
, slightly larger than the MW
1/3
dependence expected for a perfectly
compact conformation. Therefore, as a first approximation the hydrodynamic diameter of a dimer
should be ~30% larger than that of the corresponding monomer, consistent with the values above
given that the dimer is likely less globular than the monomer. For comparison, the reported
hydrodynamic diameters of various monomeric β-glucosidases follow the same scaling, namely
dH = 6 nm for Humicola insolens (54 kDa)
168
and Trichoderma harzianum (55 kDa),
62
while dH =
10 nm for Exiguobacterium antrcticum (113 kDa),
169
similar to the assumed monomeric values of
dH = 9  1 nm in Figure 3b for Aspergillus niger (110 kDa from Figure 6). The reason for the
wide range of molecular weights above is that the term -glucosidase refers generally to any of
the various isozymes that catalyze the hydrolysis of the terminal residue of -glucosides to form
glucose.
170
Even A. niger can produce different β-glucosidases depending on the production
conditions.
171, 172

The light-scattering data above suggest that the addition of azoTAB causes β-glucosidase
to undergo a dimer → monomer dissociation, with the monomer initially being more active than
the dimer until elevated azoTAB concentrations cause the progressively-unfolded monomers to
aggregate into large inactive species. The relatively-hydrophobic trans form of azoTAB causes
this process to occur at lower surfactant concentrations compared to the relatively-hydrophilic cis

40

isomer, likely due to enhanced protein-surfactant interactions in the former case. Dissociation of
the β-glucosidase dimer in the presence of azoTAB resembles the thermal denaturation of endo-β-
1,4-glucanase from Aspergillus niveus, where the highest reaction rate was correlated with the
appearance of monomers.
173
In addition, the monomer aggregation seen at elevated azoTAB
concentrations is similar to the effect of guanidine hydrochloride on monomeric β-glucosidase,
where protein swelling was reported to lead to eventual aggregation and a loss in activity.
174
Note
that the critical micelle concentrations (CMCs) determined for azoTAB using the same buffer and
temperature employed were 1.5 mM and 7.5 mM under visible and UV light, respectively,
175

indicating that the observed superactivity is a result of protein-monomer as opposed to protein-
micelle interactions.
7
8
9
10
11
12
13
14
15
visible
UV
0 0.5 1 1.5 2
Hydrodynamic Diameter (nm)
AzoTAB Concentration (mM)
(a)
645
650
655
0 0.5 1 1.5 2
visible
UV
AzoTAB Concentration (mM)
Maximum Emission Wavelength (nm)
(b)

Figure 8. (a) Hydrodynamic diameter of β-glucosidase (1 mg/mL) in the presence of
azoTAB under different light conditions. The hatched regions denote surfactant concentrations
beyond which large-scale protein aggregation was observed (i.e., diameters > 100 nm).
Evidence of β-glucosidase monomer unfolding at elevated azoTAB concentrations can be
seen in the Nile red fluorescence measurements in Figure 8b. Nile red is a hydrophobic dye that
exhibits an emission peak at 655 nm in water that shifts to shorter wavelengths in increasingly
hydrophobic microenvironments.
176
Thus, protein unfolding and particularly the exposure of the

41

hydrophobic protein core to the dye can be readily detected.
40
As seen in Figure 8b, blueshifts in
the emission peak of Nile red begin to occur at ~0.5 mM azoTAB under visible light and ~1.0 mM
azoTAB under UV light, corresponding to regions in Figure 8a where the respective hydrodynamic
diameters have reached a minimum (i.e., after dissociation of the dimer but prior to monomer
aggregation). These azoTAB concentrations are also near where the activities in Figure 7 reached
maximum values under visible and UV light. This combined evidence again supports the
conclusion that β-glucosidase undergoes a transition from dimers → superactivated-monomers →
deactivated-aggregates with azoTAB addition.
The ability to control -glucosidase activity with in-situ light illumination is illustrated in
Figures 7b and 7c using p-nitrophenyl β-D-glucopyranoside as the enzyme substrate (as opposed
to cellobiose in Figure 7a). Compared to the rate that the pure enzyme is able to generate the p-
nitrophenol hydrolysis product (0.011 M/s, data not shown), the addition 0.5 mM azoTAB pre-
converted to the cis state leads to ~68% superactivity in Figure 7b (i.e., 0.019 M/s), in general
agreement with Figure 7a. In situ illumination with visible light to convert to trans azoTAB lowers
the reaction rate to 0.013 M/s (12% superactivity). Essentially identical rates were achieved with
the experiment performed in reverse order (Figure 7c), namely using in situ UV illumination to
convert azoTAB from the trans to the cis state. In addition to demonstrating photo-initiated
changes in -glucosidase activity, the results in Figures 7b and 7c further serve to illustrate that
azoTAB-induced superactivity is not substrate specific. However, the magnitude of superactivity
does depend on the choice of either natural (cellobiose) or artificial (p-nitrophenyl β-D-
glucopyranoside) substrate, as has been reported by others.
177
Due to complications arising from
azoTAB absorbance in the region of p-nitrophenol detection at azoTAB concentrations higher than

42

those employed in Figures 7b and 7c, cellobiose was used as the substrate for the more systematic
studies in Figure 7a.
To obtain precise information on the protein structural changes responsible for the
superactivity observed in Figure 7, small-angle neutron scattering data were collected on systems
containing β-glucosidase and azoTAB, as seen in Figure 9a. SANS data probe conformational
changes on length scales L = 2/Q, here ranging from 12 Å to 1200 Å, which is ideally suited to
examine the variety of sizes expected based on the transitions suggested above. Specifically, since
light scattering intensity is proportional to the square of particle volume (i.e., diameter to the power
of six),
178
light scattering inherently requires that any large species (particularly the dH > 100 nm
aggregates in Figure 8a) be filtering out of the mixture in order to accurately study the smaller
monomeric and dimeric species (dH ~10 nm). In contrast, neutron scattering from these disparate
species occur in separately-resolvable Q ranges. As a result, the SANS data in Figure 9 were
collected without filtration (unlike Figure 8a) to provide a complete accounting of the concentrated
protein solutions common in practical application.
172, 179
For example, given the large size and
high proportion of hydrophobic amino acids found in β-glucosidase, aggregation is commonly
observed.
180,181
Thus, not surprisingly the pure -glucosidase spectrum and the visible- and UV-
light spectra containing 0.05 mM azoTAB all exhibit significant low-Q scattering, consistent with
a considerable fraction of the protein existing as high-MW aggregates. Specifically, the zero-angle
scattering intensity, I(0), is proportional to the weight-average molecular weight, MW, via the
equation 𝐼 (0)=𝑐 𝑀 W
𝜐 ̅ 2
(Δ𝜌 )
2
1000𝑁 A
⁄ , where c is the protein concentration (in units of mg/mL),
𝜐 ̅ is the protein specific volume (cm
3
/g),  is the scattering length density difference between
solvent and protein (cm
-2
), and NA is Avogadro’s number.
182
Thus, the plot of I(0)/c versus MW
shown in Figure 9a-inset is a straight line as expected for a range of monomeric proteins.
37, 38, 40,

43

143
Conversely, the low-Q scattering for these aggregated solutions in Figure 9 have yet to plateau,
and further exhibit a Q
-1
dependence at low Q consistent with the formation of long, cylindrical
species,
183
a common occurrence during protein aggregation (notoriously amyloid fibrils).
98, 184, 185

Figure 9. (a) Small-angle neutron scattering (SANS) data and corresponding (b) pair-
distance distribution functions and (c) modified Guinier analyses of β-glucosidase (1 mg/mL) at
low azoTAB concentrations versus light conditions. The inset in (a) gives I(0)/c versus MW for a
variety of proteins, including lysozyme,
40
lactalbumin,
186
α-chymotrypsin,
143
carbonic
anhydrase,
38
and bovine serum albumin.
37
The inset in (b) displays the P(r) functions at low values
of r. In part (c), the 0.05 mM azoTAB data under visible and UV light are successively offset by
–0.5 for clarity, while the reported values are the corresponding cross-sectional radius of gyration.
In contrast, at 1 mM azoTAB under both visible and UV light, a prominent shoulder
develops in the SANS spectra that when visually extrapolated suggests an I(0) value on the order

44

of 0.1 cm
-1
, corresponding to a Mw  100 kDa (i.e., the monomeric value in Figure 6). This
suggests that a significant fraction of the protein remains in the monomeric state in the presence
of azoTAB. Note that a more accurate estimate of this I(0) value, as well as conclusive
demonstration that the protein remains in the monomeric state, will be shown below by
deconvolution of the SANS data. The spectra obtained at intermediate azoTAB concentrations
(0.25 mM) exhibit a less-prominent shoulder, indicating less suppression of the monomer →
aggregate transition. Interestingly, superactivity in Figure 7 is a maximum near 1 mM azoTAB,
supporting the conclusion that the monomer is indeed the most-active conformation. Although it
must be noted that the higher protein concentrations employed in Figure 9 (1 mg/mL, necessary to
maintain reasonable count times; e.g., lysozyme at this concentration would require ~10 h on the
same beam line to achieve reasonable S/N) relative to Figure 7 (30 μg/mL) likely influence the
relative amounts of the protein existing in the monomeric versus aggregated states, but again this
former concentration is more representative of typical industrial applications.
172, 179

Further evidence of azoTAB constraining the monomer-to-aggregate transition can be seen
in the pair distance distribution functions obtained from the SANS data, shown in Figure 9b.
Recall P(r) is the probability that two scatterers (atoms for SANS) are a distance r apart within the
overall particle (i.e., monomers or aggregates), with the location where P(r) returns to zero giving
the maximum intraparticle distance (i.e., Dmax). The shape of the P(r) curves in Figure 9b,
particularly the slow and nearly linear decrease out to large Dmax values of ~1150 Å, is an
indication of cylindrical particles of length L  Dmax, which is frequently observed during protein
aggregation. Furthermore, as noted by Glatter,
187
the point of inflection at rI (see Figure 9b-inset),
which denotes the transition from an inverse parabolic shape of P(r) at low r (accounting for
intraparticle distances in both the radial and axial directions) and the linear P(r) behavior at high r

45

(where intraparticle distances are primarily in the axial direction), provides a reasonable estimate
of the maximum cross-sectional dimension Dc, here ~70 Å. This indicates that the protein fibrils
are about as thick as a single β-glucosidase molecule when compared to the monomeric dH values
in Figure 8a, indicating that aggregation occurs in an end-to-end fashion. This also agrees with
the modified Guinier analysis shown in Figure 9c, where the cross-sectional radius of gyration,
𝑅 c
=𝐷 c
√8 ⁄ , was determined by fitting the data to the Guinier approximation for cylinders,
namely 𝑄 𝐼 (𝑄 )≅𝐼 c
(0)exp(−𝑅 c
2
𝑄 2
2 ⁄ ), using QRc < 1 since the percent error of the Guinier
approximation is 2.7(QRc)
4
for cylinders.
155
The Rc values determined in Figure 9c give Dc  65
Å, in reasonable agreement with Figure 9b-inset. The inset in Figure 9b also highlights the
increasing relative contribution of radial intraparticle distances in the P(r) curves at low r values
with increasing azoTAB concertation, which is accompanied by a loss of definition in the linear,
axial contribution of P(r) at high r values. This further indicates that the presence of azoTAB
protects the protein from undergoing a transition from monomers with radial P(r) curves to
elongated aggregates with axial P(r) curves. Indeed, the modified Guinier plots for [azoTAB] 
0.25 mM (not shown) were clearly nonlinear, an indication of the presence of small, non-
cylindrical species (e.g., monomers).

46

Figure 10. (a) Illustration of the method used to obtain the monomer-only SANS data. The
raw spectrum for the system containing 1 mM azoTAB under visible light (a mixture of monomers
and fibrils) is subtracted by the scaled pure β-glucosidase spectrum (containing only fibrils). The
red curve indicates the best-fit Guinier approximation to the monomer-only data used to obtain the
scaling factor (inset: Guinier plots using various scaling factors). Also shown are the
corresponding (b) pair-distance distribution functions and (c) Guinier analyses of the
deconvoluted, monomer-only data at each condition. In part (a), the P(r) values for the
crystallographic monomer and dimer (PDB: 4IIB) are shown for comparison. In part (c), the data
with 1 mM azoTAB (UV), 0.25 mM azoTAB (visible), and 0.25 mM azoTAB (UV) light are
successively offset by –1.0 for clarity, while the reported values are the corresponding radius of
gyration.
To conclusively demonstrate that monomers are the predominant low-MW protein
conformation in the presence of azoTAB, the SANS data at elevated azoTAB concentrations were
deconvoluted to allow the low-MW-only scattering data to be obtained. For this process, SANS
has three distinct advantages. First, SANS is an absolute technique since the weight-average

47

molecular weight can be determined directly from I(0), as discussed above. Second, SANS is
additive with the scattering for a non-interacting mixture of two scattering species A and B given
by the scaled sum of the two individual species. Finally, SANS data from scatterers of different
sizes are resolved into separate Q-ranges according to the respective length scales L = 2/Q
(ranging from 12 Å to 1200 Å in Figure 9), as discussed above. The deconvolution process is
illustrated in Figure 10, where the raw scattering data obtained for -glucosidase in the presence
of 1 mM azoTAB under visible light displays both a Q
-1
decay at low Q (particularly evident at Q
< 0.01 Å
-1
) and a shoulder centered around Q = 0.05 Å
-1
, indicating the presence of both large
cylindrical aggregates and a low-MW species, respectively. To obtain the scattering due solely to
this low-MW species, the pure -glucosidase spectrum (which results almost entirely from
cylindrical aggregates, see Fig. 9b) was subtracted from the raw spectrum containing 1 mM
azoTAB using a scaling factor f (i.e., Ilow-MW = Iraw – fIpure). The value of f was determined by
optimizing the resulting Ilow-MW to be most consistent with the scattering from a strictly globular
species, namely the Guinier approximation for spherical scatters
155
𝐼 (𝑄 )≅𝐼 (0)exp(−𝑅 g
2
𝑄 2
3 ⁄ ),
accurate to within 2% for QRg < 1.3 (i.e., the red curve in Figure 10a). As seen in Figure 10-inset,
too low of a scaling factor gives in a steep upturn in the data at low Q and thus a deviation from
the linear Guinier behavior, a result of non-subtracted aggregate scattering still being present in
the deconvoluted data. Conversely, too large of a scaling factor leads to a downturn at low Q, a
result of over correction. Thus, the optimum value of the scaling factor could be obtained at each
condition, allowing the scattering from the low-MW species alone to be determined.
Pair distance distribution functions obtained from the deconvoluted low-MW scattering at
each condition are shown in Figure 10b. The resulting P(r) curves are largely consistent
(especially given the uncertainties in the subtraction technique), indicating that the low-MW

48

species adopts nearly identical conformations independent of the conditions. Specifically, each
P(r) curve exhibits an inverse parabolic shape with only a modest tail at high r values, which is
consistent with a predominantly globular structure (except perhaps for 1 mM azoTAB sample
under visible light, where the burgeoning tail may indicate slight monomer unfolding as suggested
in Figure 8, which may further explain the slightly lower peak superactivity seen under visible
versus UV light in Figure 7). Furthermore, the nearly uniform Dmax value of ~85 Å agrees with
the monomer size of dH = 9  1 nm found in Figure 8a. Additionally, these low-MW-only P(r)
curves are nearly indistinguishable from the P(r) values calculated
188
from the crystal structure of
monomeric -glucosidase from Aspergillus aculeatus (PDB: 4IIB), which shares 83% sequence
identity
139
with the -glucosidase from Aspergillus niger employed in this work. For comparison,
the P(r) calculated for dimeric -glucosidase is shown in Figure 10b, which clearly demonstrates
that the deconvoluted SANS data are inconsistent with dimeric species. Furthermore, the radii of
gyration obtained from the resulting monomer-only P(r) curves using the equation 𝑅 𝑔 2
=
∫ 𝑟 2
𝑃 (𝑟 )𝑑𝑟 𝐷 𝑚𝑎𝑥 0
2∫ 𝑃 (𝑟 )𝑑𝑟 𝐷 𝑚𝑎𝑥 0
⁄ are Rg = 26 Å  1 Å, largely in agreement with Guinier fits of
the monomer-only data shown in Figure 10c. These Rg values correspond to a globular diameter
of 𝑑 =2√5 3 ⁄ 𝑅 𝑔  7 nm, in reasonable agreement with Figure 8a. Finally, the I(0) values
determined from Figure 10c are 0.032 cm
-1
(1 mM azoTAB, visible light), 0.030 cm
-1
(1 mM
azoTAB, UV light), 0.026 cm
-1
(0.25 mM azoTAB, visible light), and 0.026 cm
-1
(0.25 mM
azoTAB, UV light), indicating that while light conditions have only a modest effect, increasing
the surfactant concentration does cause a slight increase in the fraction of protein remaining in the
monomeric as opposed to aggregated state (i.e., 29% with 1 mM azoTAB versus 23% with 0.25
mM azoTAB, calculated using Figure 9a-inset).

49

To determine the conformation of the -glucosidase monomer found in the presence of
elevated azoTAB concentrations, shape-reconstruction analysis was performed on the
deconvoluted monomer-only data, as shown in Figure 11. Briefly, this method involves
approximating the protein as a collection of scattering centers, the positions of which are optimized
to best fit the scattering data. The GA_STRUCT program
157
employed here utilizes a genetic
algorithm for the fits, with 10 independent runs for each sample averaged to obtain consensus
envelops at each condition to overcome the ill-posed nature of the inverse scattering problem (i.e.,
fitting of positions of thousands of scatterers with only a few hundred data points). From the results
displayed in Figure 11, the shape-reconstructed in vitro monomer (gray space-filled structure) is
found to be quite similar to the X-ray crystallographic structure of the -glucosidase monomer
from A. aculeatus (overlaid ribbon diagram, PDB: 4IIB). The A. aculeatus monomer consists of
three domains, a catalytic TIM barrel-like domain (colored in blue), an α/β sandwich domain
(green), and a fibronectin type III domain (yellow), as well as an insertion region (red) and two
linker regions (dark gray).
189
The SANS data is of sufficient resolution (generally 2/Qmax = 11
Å, but see Rambo and Tainer
190
) such that the protruding FnIII domain can be clearly seen in the
shape-reconstructed images, which allows the overlaid ribbon diagrams to be properly oriented.
Once again it is clear from these shape reconstructed images that the low-MW scattering species
cannot be a dimer, which for comparison is shown in Figure 11b for the crystallographic dimer.
The dimeric interface is composed of 25 hydrogen bonds primarily between the α/β sandwich
domains of adjacent monomers. Interestingly, this domain is misalign in the shape-reconstructed
in vitro images relative to the crystallographic monomer, suggesting that in the presence of
azoTAB the α/β sandwich domain partially unfolds and separates from the rest of the molecule.
This could explain why monomers are observed for -glucosidase in the presence of azoTAB as

50

opposed to pure -glucosidase, which has been shown in several studies to adopt a dimeric
structure in vitro.
189, 191-193
In colloquial terms, the enzyme resembles a cupped hand with the
active-site palm partially covered by curled fingers representing the α/β sandwich domain (the
front view in Figure 6a is looking directly into the TIM-barrel active site shown in blue). The
addition of azoTAB appears to unfurl the α/β sandwich domain allowing unhindered access to the
active site, suggesting a potential root cause for the observed superactivity. To our knowledge, this
is the first report indicating that unfolding of the  sandwich domain of -glucosidase is
correlated with superactivity.

Figure 11. (a) Results of the shape-reconstruction analysis of the monomer-only SANS
data for β-glucosidase in the presence of 0.25 mM azoTAB under visible light. The overlaid ribbon
diagrams are from the crystal structure of monomeric β-glucosidase from A. aculeatus (PDB:
4IIB), (TIM barrel-like domain, blue; α/β sandwich domain, green; fibronectin type III domain,
yellow); linker regions, dark gray; and insertion region, red). (b) The crystal structure of the β-
glucosidase dimer shown for comparison. (c) Results of the shape-reconstruction analysis as a
function of azoTAB concentration and light conditions.

51

To augment the tertiary- and quaternary-structure characterization provided above, the
secondary structure of β-glucosidase was examined using circular dichroism (CD) spectroscopy
as a function of azoTAB concentration, as seen in Figure 12. In general, it can be seen that the
presence of azoTAB (up to 0.5 mM) and the light conditions have only a mild effect on the
observed CD spectra. This is not surprising given the shape-reconstructed in vitro conformations
in Figure 6, which indicated that only localized unfolding presumably of the α/β sandwich domain
occurs at these relatively-low azoTAB concentrations. Indeed, deconvolution of the spectra in
Figure 12 support this conclusion. Specifically, the pure β-glucosidase spectrum consists of 16%
α-helical, 23% β-sheet, and 61% coil structures, similar to values found from the crystal structure
of -glucoside from A. aculeatus (22% α-helix, 17% β-sheet, and 57% random). With the addition
of azoTAB these relative contributions changed only slightly, giving 18.01.1% α-helical,
22.40.4% β-sheet, and 59.60.8% coil structures under visible light and 18.60.7% α-helical,
22.00.6% β-sheet, and 59.40.3% coil structures under UV light (average  one standard
deviation). Here the method of deconvolution and particularly complications that arise from the
presence of azoTAB deserve special mention. AzoTAB is a strong absorber of UV light, especially
below 200 nm, with the absorbance in a 1-mm cell exceeding the rule-of-thumb limit
194
of A > 1.5
when [azoTAB] > 0.5 mM. This leads to unacceptably high noise levels at low wavelengths,
particularly noticeable in the 0.5 mM azoTAB sample under UV light in Figure 12. Thus, only the
wavelength range from 200 nm – 250 nm was used to fit the data. Thus, the K2D3 algorithm
158

was used to perform the deconvolutions, which has been reported to lead to more accurate results
when this limited wavelength range is employed (as opposed to the more standard 190 – 250 nm).
Despite these limitations, it can be safely concluded from the results in Figure 7 that the transition

52

to monomeric species reported above is accompanied by at most only minor changes in the
secondary structure, consistent with Figure 11. Furthermore, the illumination conditions (i.e.,
visible versus UV light) were found to have only a very modest effect on the CD spectra (not
shown), indicating that the primary difference between the trans and cis isomers of azoTAB is
observed in the quaternary as opposed to the secondary structure (i.e., the ability to protect the
protein from aggregation, see Figure 9). For comparison, the addition of SDS has been reported
to decrease the α-helical content of β-glucosidase, while the presence of sodium deoxycholate
(DOC) or Tween 20 led to an increase in α-helical or β-sheet content, respectively.
130

Figure 12. Circular dichroism spectroscopy measurements of β-glucosidase (1 mg/mL) in
the presence of various azoTAB concentrations under visible and UV light, with the latter being
shifted to the right by 20 nm for clarity.
To gain further insight into β-glucosidase conformational changes in response to azoTAB,
intrinsic fluorescence spectra were recorded as shown in Figure 13. The wavelength of maximum
emission of tryptophan in general depends on the local polarity, exhibiting a redshift to higher

53

wavelengths in relatively-hydrophilic microenvironments (e.g., protein unfolding) versus a
blueshift to lower wavelengths in relatively-hydrophobic locations (e.g. ligand binding).
48,195
As
seen in Figure 13a, increasing concentrations of primarily trans azoTAB under visible light leads
to a small but perceptible redshift in the emission spectra, again consistent with slight unfolding
of the monomer relative to the crystallographic dimer seen in Figure 11. Conversely, in the
presence of increasing concentrations of primarily cis azoTAB, a blueshift in the emission spectra
was observed. These opposite trends are in line with the relative hydrophobicity of the trans and
cis azoTAB isomers. Conversely, the wavelength of maximum emission remains constant at ~344
nm even at elevated DTAB and SDS concentrations (up to 3 mM), illustrating the unique response
of -glucosidase to azoTAB.

Figure 13. Intrinsic fluorescence of β-Glucosidase (90 μg/mL) with various concentrations
of azoTAB under (a) visible and (b) UV light. See text for a description of the absorbance
correction procedure. The inset in (a) demonstrates the shift in the measured fluorescence
independent of the absorbance. The inset in (b) shows the maximum emission wavelength of β-
glucosidase in the presence of azoTAB under visible and UV light, SDS, and DTAB. The enclosed

54

image illustrates the location of tryptophan residues in β-glucosidase from A. aculeatus (PDB:
4IIB).
Once again complications arising from the presence of azoTAB in measuring the
fluorescence in Figure 13 deserve special discussion. Specifically, since azoTAB absorbs light
throughout the excitation and emission wavelengths, the measured fluorescence intensity steadily
decreased with increasing concentration of azoTAB. This effect is both distinct from and dominant
to any changes in the fluorescence intensity that would be expected to occur due to exposure of
the tryptophan residues to solvent (e.g., quenching). For example, the absorbance of 0.1 mM
azoTAB at the excitation wavelength is A280 = 0.133 under visible light, meaning that the excitation
intensity was effectively reduced by a factor of 10
−𝐴 280
2 ⁄
= 0.86 relatively to the pure protein (the
absorbance is divided by two to account for the reduction in the average path length of light in
fluorescence versus absorbance measurements). Additionally, the varying azoTAB absorbance
reduces the emission intensity by a factor of 10
−𝐴 𝜆 2 ⁄
depending on the emission wavelength (e.g.,
by a factor of 0.38 at 344 nm for 0.1 mM azoTAB under visible light). Nevertheless, by measuring
the absorbance immediately following fluorescence detection using the identical sample and
cuvette, both of these effects could be accurately offset. Furthermore, any uncertainties in this
correction procedure would primarily influence the relative emission intensities as opposed to the
wavelength of maximum emission, thus, only the later values were used above to infer changes in
the local microenvironment of tryptophan. To demonstrate that the shifts in the locations of the
emission maximum reported in Figure 13 are not simply an artifact of this correction procedure,
the inset in Figure 13a shows the ratio of the uncorrected emission intensities measured at two
wavelengths equally spaced on either side of the absorbance maximum of azoTAB (ca. 348 nm).
As seen in the inset, the raw emission intensity at 372 nm indeed grows relative that that at 324

55

nm with increasing azoTAB concentration (i.e., a redshift in the emission spectra), while the ratio
of the respective absorbance values remains essentially constant (i.e., the correction factor is the
same at both wavelengths).
For reference, the locations of the tryptophan residues in -glucoside from A. aculeatus are
displayed in Figure 13. Of the 20 tryptophan residues present in the protein, half are located either
in or adjacent to the  sandwich region (i.e., five in each location). Thus, the results in Figure
13 further support the conclusion that azoTAB facilitates the unfurling of the  sandwich domain
away from the rest of the protein, which in this case is manifested by an increase in the exposure
of tryptophan residues to the solvent and/or azoTAB ligand.
To further examine the mechanistic reason for -glucosidase superactivity in the presence
of azoTAB, the enzyme activity was measured as a function of substrate concentration as shown
in Figure 14a. A surfactant concentration of 0.75 mM was utilized, near where peak superactivity
was observed in Figure 7. The data in Figure 14a demonstrate a consistent 30% – 40%
superactivity in the presence of azoTAB, similar to Figure 7a where the cellobiose concentration
was 5 mM. Notably, the superactivity in Figure 14a further increased to ~60% with 15 mM
substrate. To explain this phenomenon, the data were first fit by using a nonlinear regression of
the Michaelis-Menten equation, namely,  = Vmax[S]/(KM + [S]), where  is the reaction velocity,
Vmax = kcat[E] is the maximum reaction rate, [S] is the substrate concentration, and KM is the
Michaelis constant. Using this approach, the turnover number was found to increase by 90% from
kcat = 17 s
–1
for pure β-glucosidase to 32 s
–1
with 0.75 mM azoTAB, while KM increased by 70%
from 9.5 mM for the pure enzyme to 16 mM. The net effect is an ~12% increase in the apparent
bimolecular reaction rate constant kcat/KM from 1.8 mM
–1
s
–1
to 2.0 mM
–1
s
–1
, somewhat low relative
the 40% superactivity seen in Figure 7. Furthermore, this simplified analysis seems to indicate

56

that azoTAB hinders association of the substrate to the enzyme (here it can be safely assumed that
KM  KD = [E][S]/[ES], namely the enzyme-substrate dissociation constant, due to the given the
low kcat values), which runs counter to the slight unfolding seen in Figures 11–13. For comparison,
kcat/KM values of -glucosidase have been reported to vary between 0.68 mM – 43.1 mM with
cellobiose as the substrate.
177, 196, 197

Figure 14. Reaction velocities obtained as a function of substrate (cellobiose)
concentration both with and without azoTAB. Solid curves correspond to nonlinear fits of the data
to the Michaelis-Menten equation, while dashed curves represent fits to a modified equation
accounting for substrate inhibition.
However, it is clear that the data in Figure 14 do not follow simple Michaelis-Menten
kinetics. Instead, the data suggest that substrate inhibition may be responsible for the leveling off
of the reaction velocities seen at elevated substrate concentrations (especially noticeable for pure
-glucosidase). Specifically, the data are consistent with the modified equation  = Vmax[S]/(KS +
[S](1 + [S]/𝐾 𝑆 ′
), where KS represents the dissociation of the standard enzyme-substrate ES complex
while 𝐾 𝑆 ′
accounts for binding of an additional substrate molecule forming a catalytically-inactive

57

ES2 complex (i.e., competing pathways ES2
𝑆 ↔ ES

→ E + P).
198
To illustrate this effect, Hanes-
Woolf plots of the data are shown in the inset in Figure 14a, which serve to highlight the deviation
from the linear behavior typical of Michaelis-Menten kinetics (i.e., when the ES2 dissociation
constant 𝐾 𝑆 ′
→ ). Instead, the data are consistent with a quadratic dependence on [S] as expected
for finite values of 𝐾 𝑆 ′
from the above equation (i.e., as [S]/ = KS/Vmax + [S]/Vmax + [S]
2
/Vmax𝐾 𝑆 ′
).
Nonlinear fits of the data in Figure 14a to the modified equation above indicate that kcat decreases
from 84 s
–1
for pure β-glucosidase to 67 s
–1
with 0.75 mM azoTAB, while KS simultaneously
decreases from 53 mM to 35 mM, leading to a modest 20% increase in the catalytic efficiency.
Most telling, however, is the dramatic 4-fold increase in the 𝐾 𝑆 ′
dissociation constant from 3.3 mM
for pure -glucosidase to 13 mM with the addition of azoTAB, indicating that azoTAB-induced
superactivity (40% – 55% in Figure 2, up to 60% in Figure 14a) is primarily a result of diminished
substrate inhibition. For comparison, KI values for -glucosidases with the monosaccharide
inhibitor glucose generally range between 1 M – 14 mM using cellobiose as the substrate.
199, 200

The underlying cause of substrate inhibition is illustrated in Figure 14b, where the crystal
structure of -glucosidase from A. aculeatus complexed with glucose is shown (PDB: 4IIG).
189

Two glucose molecules are located in the active-site cavity at subsites –1 and +1 (following the
suggested
201
numbering sequence where the scissile bond is between these subsites). A long cleft
begins at subsite +1 and extends past subsite +4 (seen bound to the cryoprotectant molecule 2-
methyl-2,4-pentanediol), which explains the high activity of this -glucosidase towards longer
oligosaccharides.
140
However, this promiscuous nature of the active site can also lead to
nonproductive substrate binding to subsites +1/+2,
124
as is frequently observed for oligomer-active
enzymes in the presence of disaccharide substrates.
202
This appears to be the case in Figure 14a
where cellobiose concentrations greater than 𝐾 𝑆 ′
= 3.3 mM lead to noticeable substrate inhibition

58

for pure -glucosidase. In the presence of azoTAB, however, this inhibition is markedly reduced,
perhaps not surprising given that the  sandwich domain, which forms one wall of the cleft in
Figure 14b, was in Figure 11 postulated to partially separate from the rest of the protein
macromolecule. This aforementioned unfurling could also explain the slight reduction in kcat from
84 s
–1
for pure β-glucosidase to 67 s
–1
with 0.75 mM azoTAB, as the nucleophile and the acid/base
catalyst residues (i.e., Asp280 in the TIM barrel domain and Glu509 in the  sandwich domain,
respectively) generally must remain with ~5.5 Å to be effective,
138
as well as the greater affinity
of the substrate for the actual active site (i.e., KS decreasing from 53 mM to 35 mM), which is
partially blocked by the  sandwich domain in the native enzyme.
203, 204

During the cellulose saccharification process, inhibition of -glucosidase, endoglucanase,
and exoglucanase by glucose and cellobiose is a common problem.
124,205
Thus, much work has
been devoted to develop so-called glucose-tolerant cellulases. The present work demonstrates that
azoTAB holds promise as an inhibition-protectant additive for bioethanol production.
2.5. Conclusions
The structure and activity of β-glucosidase from Aspergillus niger has been examined as a
function of an azobenzene-based photoresponsive surfactant (azoTAB). Analysis of small-angle
neutron scattering allows the in vitro protein conformation to be determined at a resolution on the
order of the individual protein domains. Upon addition of the surfactant, the enzyme is found to
undergo a transition from dimers/higher aggregates to monomers, an apparent response to slight
unfolding of the  sandwich domain responsible for dimer formation. Furthermore, the
separation of this domain from the remaining protein molecule leads to a 4-fold decrease in
substrate inhibition and a net 60% increase in the reaction rate relative to the native enzyme. In
contrast, generic hydrophobic interactions exhibited by traditional surfactants lead to simple

59

enzyme deactivation. As a result, azoTAB could potentially find use as a novel additive imparting
inhibition protection during cellulose hydrolysis.
Acknowledgements. We acknowledge the support of the National Institute of Standards
and Technology, U. S. Department of Commerce, in providing the neutron research facilities used
in this work. This material is based upon work supported by the National Science Foundation
under Grant 1758225. Any opinions, findings, and conclusions or recommendations expressed in
this material are those of the author(s) and do not necessarily reflect the views of the National
Science Foundation. We would also like to acknowledge support from USC’s Research
Enhancement Fellowship program. We thank Dr. Terry Torao Takahashi and Prof. Rich Roberts
for their assistance with purifying β-glucosidase. We further acknowledge Dr. Shuxing Li of
USC’s Center of Excellence in NanoBiophysics for his aid with the circular dichroism studies.

60

CHAPTER 3: ENHANCEMENT OF ENDOGLUCANASE BIOFUEL ENZYME
KINETICS IN THE PRESENCE OF AZOTAB SURFACTANT
3.1. Abstract
Endoglucanases degrade -1,4-glycosidic bonds of crystalline cellulose (i.e., the most
common polysaccharide on Earth) into insoluble or soluble cellooligosaccharides at the solid-
liquid interface. The improvement of endoglucanase (from Aspergillus niger) kinetics with
addition of a light responsive azobenzene trimethyl ammonium bromide (azoTAB) surfactant is
studied. AzoTAB is a photoresponsive surfactant that exists as a relatively-hydrophobic trans
isomer under visible light (434 nm) and a relatively-hydrophilic cis isomer under UV light (350
nm). Endoglucanase catalytic activity can be controlled with light illumination slightly for
microcrystalline cellulose natural substate(~15%) and significantly for p-nitrophenol-based model
substrate(~2-fold) by switching between the trans (higher enzyme binding affinity, resulting
higher enzyme unfolding) and cis form of azoTAB. Endoglucanase activity increases 45% towards
Avicel crystalline substate and 4-fold towards 4-nitrophenyl β-D-cellobioside substrate in the
presence of 0.4 mM azoTAB under UV light (90% cis and 10% trans isomers). In comparison,
endoglucanase catalytic activity increases 5-10% towards crystalline cellulose substrate with the
addition of benzene-based surfactant sodium dodecyl benzene sulfonate (SDBS), and common
linear chain surfactants sodium dodecyl sulfate (SDS) and dodecyl trimethyl ammonium (DTAB).
AzoTAB addition led to an increase of maximum endoglucanase adsorption (Emax) from 7.89 mg
enzyme/g Avicel to 12.92 mg enzyme/g Avicel and catalytic enzyme efficiency (kcat/KM) from
0.031 L/(mg.s) to 0.061 L/(mg.s). The correlation between adsorbed enzyme concentration on
substrate and enzyme specific activity is observed, and 40-50% activity enhancement or increased
bound enzyme to substrate is detected with azoTAB addition at different enzyme, substrate, and

61

inhibitor concentrations. Improvement of substrate properties in addition of azoTAB surfactant is
associated with lower reciprocal terms of Michaelis constant, KM, the adsorption coefficient, Kad
and fractal parameter, h values. Furthermore, 45-50% activity enhancement via azoTAB surfactant
preserved for all three cellulase enzyme mixture of endoglucanase, cellohydrolase and β-
glucosidase. Consequently, azoTAB can be applied as profitable substance for the heterogenous
enzymatic cellulose hydrolysis, resulting in a 30% decrease in the enzyme load based on the
specific activity, adsorption and fractal kinetics results.
Keywords: photosurfactant, azobenzene-based surfactants, enzyme surfactant interaction,
cellulase enzymes, surfactant-added heterogenous reaction
3.2. Introduction
Cellulose, a homopolymer of -1,4-linked glucose monomers, is the main constituent of
plant cell walls and one of the primary polysaccharides in biomass.
206
Cellulose can be hydrolyzed
by the cellulase multienzyme complex into fermentable sugars (i.e., glucose) and hence fermented
into bioethanol as a readily-available sustainable energy source. Enzymatic saccharification of
cellulose into glucose requires three enzymes: random breakage of internal glycoside bonds by
endoglucanase (endocellulase, -1,4-endoglucan hydrolase, -1,4--D-glucanase, EC 3.2.1.4),
cellobiose (i.e., two linked glucose molecules) cleavage from the terminal ends of cellulose chains
via cellobiohydrolase (exocellulase, EC 3.2.1.91), and hydrolysis of cellobiose and short-chain
cellooligosaccharides into fermentable glucose by -glucosidase (cellobiase, EC 3.2.1.21).
One of the major challenges of the above process is the high cost of saccharification
enzymes due to the need for high enzyme concentrations combined with irreversible enzyme-
substrate binding.
128
The enzymatic cellulose hydrolysis process is slow, and the catalytic activity
of cellulase enzymes decreases quickly as a result of this nonproductive binding and end-product

62

inhibition.
207
The enzymes bound to the substrate surface cannot be recovered and hence high
concentrations of enzymes are required.
As an illustration, 30 grams of enzyme is needed per liter of bioethanol.
208
Higher
fermentable sugars could be obtained at lower enzyme loading with combination of potential
additives. Enhancements in the process yield have been observed by upon surfactant addition in
cellulose hydrolysis.
209, 130
The enzymatic conversion of cellulose is a heterogeneous reaction that
occurs at solid-liquid interface with soluble enzymes and crystalline cellulose. Two expected
effects of surfactants on cellulose degradation: easy to access to the active sites of cellulose and
prevent deactivation of adsorbed enzyme.
210
Surfactants adsorb on surfaces of crystalline cellulose
to modify cellulose hydrolysis.
132
The adsorption of surfactants on substrate surfaces lowers
irreversible enzyme binding to the substrate. Additionally, enzyme-surfactant interaction increases
enzyme stability.
211
The most commonly applied surfactants are polyethylene glycol (PEG)
134
and
Tween
133
for the heterogenous enzymatic hydrolysis.
In this work, we study enhancement of endoglucanase activity and kinetics in the addition
of a photoresponsive surfactant, azobenzene trimethylammonium bromide (azoTAB) that
undergoes photoisomerization based on nitrogen double bond rotation upon exposure to different
wavelengths of light. AzoTAB exists primarily as the trans isomer (75/25 trans/cis) with planar
structure and lower dipole moment of the nitrogen bond under visible light (434 nm), whereas
under UV light (350 nm) the cis form (10/90 trans/cis) of azoTAB is predominant (see Scheme
1).
33
The lower dipole moment of the trans photoisomer leads to the trans form of azoTAB being
relatively hydrophobic compared to the cis isomer. The more hydrophobic trans isomer of azoTAB
has greater tendency to bind and unfold proteins compared to the cis form due to the
hydrophobicity differences of the photoisomers. The reversible hydrophobicity changes of the

63

surfactant has been applied to control protein unfolding, activity, and dynamics.
37, 38, 40
In our
earlier study, we previously reported that -glucosidase and azoTAB interaction led to dimer-to-
monomer transition associated with 50% superactivity of β-glucosidase towards cellobiose natural
substrate.
175
All in all, azoTAB can be used as an additive to lower enzyme saccharification cost
of bioethanol production.
3.3. Experimental Section
Materials. The cationic azobenzene trimethyl ammonium bromide surfactant (MW=420
g/mol) shown in Scheme 1 was synthesized according to previous studies.
31, 33
Briefly, 4-
ethylaniline was azo-coupled with phenol; subsequently alkylated by 1,4-dibromobutane and
followed by quaternalized with trimethylamine. All chemicals were purchased from Millipore
Sigma at the highest purity unless otherwise stated. The trans-to-cis photoisomerization of
azoTAB was obtained with an 84-W long-wave (365 nm) UV lamp (Spectroline, Model no. XX-
15A). The experiments of UV light samples were performed in dark due to the fact that the thermal
conversion of cis-to-trans azoTAB is ~24 hours in dark at 25 C.
Purification of endoglucanase. Insoluble crystalline cellulose has limited accessibility to
soluble enzymes; thus, endoglucanase is a necessary enzyme to break random -1,4-glycosidic
bonds at the solid-liquid interface. Endoglucanases have been produced mainly by diverse bacteria,
fungi, archaea, protozoan.
128, 212
Aspergillus niger fungi has been selected as a source of
commercial enzyme preparations due to its high activity compared to other endoglucanases.
213

Cellulase from Aspergillus niger was acquired from Sigma-Aldrich (Cat. no. 22178). 4 grams of
the off-white crude powder was dissolved in 100 mL of protein buffer (25 mM sodium phosphate
buffer pH 6.5, 50 mM NaCl, 10% glycerol) resulting in 40 mg/mL of protein solution. The protein
solution was filtered and applied to a 5 mL HiTrap Q HP anion exchange column (GE Healthcare,

64

Cat. no. 17115301) and later washed with 30 column volume with the protein buffer. The protein
was eluted by gradient against the elution buffer (25 mM sodium phosphate buffer pH6.5, 550 mM
NaCl, 10% glycerol) and eluted at a flow rate of 2 mL/min using fast protein liquid
chromatography, FPLC (ÄKTA Pure from GE Healthcare) at 4 °C and the eluted fractions were
run on SDS-PAGE. All fractions were tested using Nanodrop 2000 (Thermo Scientific, USA) at
280 nm to detect the presence of protein and provide a rough estimate of protein concentration.
Equal concentrations of fractions were applied for SDS-PAGE and measured against Avicel
activity to identify endoglucanase fraction. The active fractions were pooled, concentrated using
Amicon Ultra 4 mL centrifuge tubes with 10 kDa cutoff membranes (UFC801024), and
concentrated protein solutions were kept at 4 °C for further studies.
Measurement of Endoglucanase Activity. Endoglucanase activity was detected by
measuring the reduced ends of Avicel, a natural microcrystalline cellulose, upon conversion to
insoluble or soluble cellooligosaccharides applying Somogyi-Nelson procedure.
151
Briefly, the
reaction took place in 2 mL vials (1.1 mL reaction sample) at a final concentrations of 0.01 mg/mL
(0.001%, w/v) Avicel and 100 μg/mL purified endoglucanase fraction and various azoTAB
concentrations in 50 mM sodium acetate buffer, pH 5 at 25 C. The reaction vial was constantly
stirred at 240 rpm with a magnetic stir bar to prevent microcrystalline cellulose settling at the
bottom of the vial. When desired, azoTAB was pre-converted to the cis state under UV light and
then mixed into the reaction solution with the reaction performed under dark to minimize photo-
transition to trans isomer. At the initial time and every ten minutes for two 90 minutes, 50 μL of
reaction mixture was diluted ten-fold and stopped with 450 μL of 50 mM phosphate buffer, pH
9.5. This stopped sample solution was combined with 500 μL of alkaline copper tartrate and heated
for 20 minutes in boiling water to induce cuprous oxide with the reduced ends of Avicel.

65

Thereafter, 500 μL of arsenomolybdic acid was added into the samples to allow molybdenum blue
to form for five minutes at room temperature from the interaction between molybdic acid and
cuprous oxide. The resultant solutions were diluted eleven-fold in deionized water and the
absorbance at 760 nm was monitored with UV-vis spectroscopy (Agilent, model 8453). The
absorbance at 760 nm was background corrected by subtracting the absorbance at 500 nm, where
the absorbance of both azoTAB and molybdenum blue is essentially zero. The endoglucanase
catalytic activity was determined from the initial rate of increase in the corrected absorbance
applying data fitting to (At-A∞)/(A0-A∞) =e
-at
, where At is the absorbance at time t. The optimum
surfactant concentrations (0.4 mM for azoTAB under visible and UV light; 1.5 mM for SDS,
SDBS and DTAB) were used for the further studies unless otherwise stated.
The enzyme catalytic rate kinetic constants, kcat and KM were calculated using 0.001-40
mg/mL Avicel with 100 μg/mL endoglucanase. The nonlinear regression model with Michaelis-
Menten equation (equation 15) was applied, Vmax and KM values are calculated by minimizing the
residual sum of squared errors method as shown in equation 16:

V=
V
max
[S]
K
M
+[S]
(15)
m=∑ (V
e
-V
c
)
2
n
i=1
(16)
where Ve is the experimental velocity and Vc is the calculated velocity with equation (15).
Additionally, Hanes-Woolf linear equation (equation (17)
214
) was used to calculate Vmax and KM.

[𝑆 ]
𝑣 =
1
𝑉 𝑚𝑎𝑥 [𝑆 ]+
𝐾 𝑀 𝑉 𝑚𝑎𝑥 (17)

66

Photoreversible Endoglucanase Activity. Light responsive endoglucanase activity was
detected with 4-nitrophenyl β-D-cellobioside model substrate (Millipore Sigma-N5759) under
visible and UV light. The reaction took place at a final concentration of 5 mM substrate, 0.3 mg/mL
enzyme and 0.4 mM pre-converted azoTAB solution in 50 mM sodium acetate (pH 5) buffer at 37
°C. Every 2 minutes, 500 µL of the reaction sample was stopped into 500 µL of 50 mM phosphate
(pH 9.5) buffer. The p-nitrophenol product (p-NP) was monitored at 429 nm with extinction
coefficient of 13.2 mM
-1
cm
-1
to avoid strong absorption peak effect of azoTAB and similar to our
previous study.
175
Subsequent to 10 minutes of reaction under visible and UV light, the light
conditions were switched to record light-induced enzyme catalysis rate.
Endoglucanase Adsorption on Avicel. Twenty-two milligrams of Avicel (2 wt%) were
placed in 2-mL screwcap tubes and filled with 1.1 mL of reaction solution in 50 mM sodium
acetate buffer at pH 5 with a total endoglucanase concentration between 0.1-2 mg/mL for
adsorption assays. The optimum activity associated surfactant concentrations were selected: 0.4
mM azoTAB visible and UV; 1.5 mM SDS, SDBS and DTAB. The reaction vials were stirred at
300 rpm with a magnetic stir bar for 40 minutes to allow the enzyme to bind to the substrate
surface. Thereafter, 400 µL reaction sample was taken and centrifuged at 13,000 rpm for a couple
of seconds to settle down microcrystalline substrate. The enzyme content in the supernatant was
measured at 280 nm with a subtraction of 600 nm wavelength using UV-vis spectroscopy. The
same procedure including magnetic stir bars without the substrate were repeated to calculate
calibration lines for pure and surfactant added solutions separately for each condition due to 350nm
strong wavelength of azoTAB or 260 nm peak of SDBS, etc. The adsorbed endoglucanase was
determined from the difference between the initial enzyme concentration and the calculated
supernatant concentration value. Simply, the calibration lines were drawn for each condition: pure,

67

azoTAB-vis, azoTAB-UV, SDS, SDBS and DTAB. The measured data points were fitted into
each calibration lines to find hydrolysate liquid phase enzyme concentration. These values
represented as free enzyme concentration. The difference between the initial enzyme concentration
and the free enzyme concentration was calculated to find adsorbed enzyme concentrations. Four
batches of cellulase adsorption experiments were conducted to acquire the adsorption isotherms.
Langmuir isotherm has been widely used cellulase adsorption model to determine
adsorption parameters with the assumptions of a monolayer and reversible adsorption on cellulose
binding sites without interactions with other enzymes. The enzyme adsorption was assumed to
follow Langmuir isotherm as in equation (18)
62, 215
:
E
B
=
E
max
K
ad
E
F
S
1+K
ad
E
F
(18)
where EB is the adsorbed enzyme concentration (mg/mL), Emax is the maximum adsorbed enzyme
concentration (mg enzyme/g substrate), EF is the free enzyme in solution (mg/mL), S is the
substrate concentration (g/mL) and Kad is the adsorption coefficient. Emax and Kad were interpreted
with linearized form of equation (18)
62,216
:

S
E
B
=
1
E
max
K
ad
1
E
F
+
1
E
max
(19)
with S/EB on the y-axis and 1/EF on the x-axis.
The same experiment protocol with a constant endoglucanase concentration of 1 mg/mL
and substrate concentration of 20mg/mL was used to measure end-product inhibition effect with
0-200 mM of glucose or cellobiose with and without surfactant solutions. The same solutions
without substrates were run in parallel and the average absorbance of them was used as initial
absorbances. The difference between the supernatant of glucose or cellobiose added samples and

68

initial absorbance of control solutions was determined to find surfactant effect on the adsorbed
enzyme concentrations as a function of glucose or cellobiose concentration.
Optical microscopy. Final concentrations of 0.5 mg/mL endoglucanase, 0.03 mg/mL, 0.4
mM azoTAB and 200 mM glucose were mixed in 8 mL screwcap tubes and filled with 5 mL
solution in 50 mM sodium acetate, pH 5 buffer. The reaction vials were mixed at 240 rpm for six
days. 300 µL reaction samples were taken at time 0 min, 3
rd
hour, 6
th
hour, 2
nd
, 4
th
, and 6
th
days
and terminated into 300 µL 50 mM phosphate (pH 9.5) buffer. The samples were observed using
an Olympus IX71 inverted microscope equipped with a 10 objective lens. Optical microscopy
images were captured with a CCD digital camera (Hamamatsu, model no. C4742-95).
Fractal kinetics. Fractal parameter, f, was calculated with fractal Michaelis kinetics in
equation (20)
62, 217,

218
:

log([E]
a
/[E]
b
)
log(t
b
/t
a
)
=1-f (20)
where [E]a and [E]b were the enzyme concentrations and ta and tb were the reaction times until 10%
of conversion was succeeded. A final concentration of 0.01 mg/mL Avicel, 100 µg/mL and 62.5
µg/mL endoglucanase were mixed with and without surfactant solutions.
Additionally, fractal exponent h and rate constant k were determined using equation (21)
219
:
Cp=[S]*100(1-exp(-k(1+
t
(1-h)
-1
1-h
) (21)
where Cp was the product concentration and [S] was the substate concentration(10mg/L). Cpe,
experimental product concentrations were calculated with the measured glucose extinction

69

coefficient of 0.24744L mg
-1
cm
-1
at 760 nm minus 500 nm wavelength values. Residual sum of
squared errors method (equation 22) was applied to quantify h and k.
m=∑ (Cp
e
-Cp
c
)
2
n
i=1
(22)
Surfactant effect on enzyme mixtures. Using combination of enzymes, the hydrolysis
rate was monitored in the absence and presence of surfactants.
Glucosidase from Aspergillus niger (Sigma-Aldrich cat. no. 49291) was purchased as a
source of β-glucosidase and was purified as previously reported.
175
Briefly, the brown powder
solubilized at pH 7.5 and saturated with 80% ammonium sulfate overnight at 4°C. Then, the
protein precipitate washed 3 times with ammonium sulfate solution and resuspended in 50 mM
sodium phosphate (pH 7.2) buffer including 1 M ammonium sulfate. Subsequently, HiTrap Phenyl
HP (GE Healthcare, 17-1351-01) hydrophobic was used to obtain β-glucosidase fraction. The
fraction was desalted using 50 kDa membrane centrifuge tubes.
Cellobiohydrolase from Hyrpocrea jerorina (Trichoderma reesei) (Sigma-Aldrich cat. no.
E6412) which is the most common component of enzyme mixtures for cellulose hydrolysis,
220,221

was purchased, aliquoted and kept at -20 °C until usage. Enzyme solution was washed in 10 kDa
membrane centrifuge tubes 4-5 times with 50 mM sodium acetate pH 5 buffer before adding to
reaction tubes.
Optimal β-glucosidase and cellobiohydrolase enzyme ratios were selected and final
enzyme concentrations of 60 μg/mL of cellobiohydrolase, 62.5 μg/mL endoglucanase and 30
μg/mL β-glucosidase were used towards 0.01 mg/mL Avicel. The surfactant effects on initial rate
and long-term behaviors were examined with collecting samples every 10 minutes for the first

70

hour and two samples at 5
th
hour and 24
th
hour, respectively. Somogyi-Nelson method was applied
to measure the samples similar to endoglucanase activity measurement. The rate was calculated
with data fitting and the difference between 1 hour, 5 hours and 24 hours with time zero.
3.4. Results
Endoglucanase from Aspergillus niger, the most common industrial fungi,
222
was purified
from commercially available cellulase enzyme. The crude enzyme contains at a minimum of seven
protein fractions as shown in Figure 15 (lane 2). The dissolved crude enzyme solution was applied
into strong anion exchange column to separate the endoglucanase enzyme fraction from the other
fractions. Positively-charged proteins are expected to bind to the strong anion exchange column,
while negatively-charged proteins and impurities would pass through the column without binding.
According to SDS-PAGE, the flow through did not contain any protein fraction and further showed
no activity against Avicel substrate. Yet surprisingly none of the purified protein fractions
displayed endoglucanase activity without the addition of flow through, while returning and mixing
the flow through with all of the purified protein fractions at a ratio between 20/80 an 80/20 showed
endoglucanase catalytic activity. All purified fractions were catalytically active after mixing with
flow through (150 mL flow through in 1100 mL reaction mixture). The flow through itself and
mixture of bovine serum albumin, lysozyme and β-glucosidase three different enzymes did not
show any endoglucase catalytic activity.

71

Figure 15. SDS-PAGE of crude cellulase (lane 2) and purified endoglucanase after anion
exchange column Sephadex (lane 4 and 5) versus molecular weight protein ladders (lane 1 and 3)
Endoglucanases from Aspergillus niger species are reported to have various isozymes
having identical catalytic functions with divergent molecular weights.
223, 224
Commercial enzyme
preparations combine these diverse isoenzymes to provide diverse associations with the insoluble
substrate, thus, break glycosidic bonds more efficiently. We speculate that the flow through of the
anion exchange column possibly contains previously unreported cofactors of endoglucanase that
keep the enzymes active. Interestingly, prior endoglucanase purification studies did not mention a
loss of the catalytic activity of the purified fractions in the absence cofactors in the literature. The
major protein fraction (with ImageJ calculation: 35%) with a molecular weight of 48kDa was
chosen, pooled, concentrated and used for further studies.
As the surfactants are more cost-effective than the biofuel enzymes,
225
the most favorable
enzyme to surfactant ratio was studied. The catalytic activity of purified endoglucanase was
measured using Avicel substrate as a function of azoTAB concentration under visible and UV
light, as displayed in Figure 16. As the concentration of azoTAB initially increases up until 0.4

72

mM, the endoglucanase activity is enhanced to a maximum of 30% and 45% activity enhancement
under visible and UV light, respectively. Although azoTAB surfactant photoisomers have a
different hydrophobicity and hence binding affinity towards proteins, photoresponsive
endoglucanase relative activity changes were minute. Similarly, the increase of surfactant tail
length from C12 to C14 or C16 showed minimal activity changes on cellulose hydrolysis with
endoglucanase enzyme.
217
Additional increases in azoTAB concentration, however, led to a
decrease in endoglucanase activity. Similar to our previous work with β-glucosidase enzyme,
175

~50% superactivity was observed for the purified endoglucanase in the addition of azoTAB
surfactant under UV light.
For comparison, SDS (sodium dodecyl sulfate), SDBS (sodium dodecyl benzene sulfonate,
benzene containing common used surfactant) and DTAB (dodecyl trimethyl ammonium bromide,
no-stimuli responsive analog or straight hydrophobic tail analog or TAB analog) surfactants were
applied to compare benzene and TAB (trimethyl ammonium bromide) containing surfactants on
endoglucanase activity. Maximum of 5-12% endoglucanase activity increase was observed at
relatively higher concentrations (i.e. 1.5-2 mM common surfactant concentration versus 0.4 mM
azoTAB concentration) of these common surfactants. Similarly, SDS addition increased the
endoglucanase activity at low concentrations and above 2 mM addition of the surfactant
diminished the catalytic activity of endoglucanase.
226
The 5-12% increase of catalytic activity
could be associated with the general surfactant application on cellulose hydrolysis due to
minimization irreversible substrate-enzyme binding.
209
At the 1.5-2 mM surfactant concentration
range, azoTAB addition showed 20-30% of endoglucanase catalytic activity enhancement.
Furthermore, Tween 80, Triton X-100 PEG 4000 surfactants and saponin biosurfactants have been
reported to result in 25% – 60% increases in endoglucanase activity.
227, 228, 229

73

Additionally, surfactant and Avicel substrate incubated overnight before adding
endoglucanase to understand if there was a change in the endoglucanase activity enhancement.
Avicel contains a homogenous ~50 µm particles (according to the manufacturer) of parallel
glucose chains bound by hydrogen bonding. Not surprisingly, no significant endoglucanase
activity change was observed towards uniform Avicel microcrystalline substrate (data not shown).
Likewise, surfactants has been mixed right before addition of enzymes or at the beginning of the
reaction to alter the microcrystalline substrate surface or enzyme confirmation.
210
Preincubation
with surfactants has been applied as pretreatment of untreated biomass sources or lignin containing
substrates.
208
As an example, preincubation of nonylphenol 20 nonionic surfactant and the recycled
newspapers increased the enzymatic hydrolysis.
230

The optimal surfactant concentrations of 0.4 mM azoTAB under visible (1.5 mM
175
) and
UV light (7.5 mM
175
); and 1.5 mM SDS (8 mM
231
), SDBS (2.5 mM
232
) and DTAB (15mM
233
)
were selected for the following experiments. The CMC (critical micelle concentration) points were
specified in parenthesis. All of the conditions would characterize surfactant monomer and enzyme-
substrate interactions rather than micelle and enzyme-substrate association.

74

50
100
150
0 1 2 3 4 5
azoTAB-vis
azoTAB-UV
SDS
SDBS
DTAB
Relative Specific Activity (%)
Surfactant Concentration (mM)

Figure 16. Relative specific activity of endoglucanase (100 μg/mL) as a function of
azoTAB (under visible or UV light), SDS, SDBS and DTAB concentration towards 0.01 mg/mL
avicel. pH = 5, T = 25 C. Lines are drawn to guide the eyes.

75

In situ light illumination is represented in Figure 17a and 17b towards 4-nitrophenyl β-D-
cellobioside model substrate as opposed to Avicel in Figure 16. The addition of pre-converted
visible and UV azoTAB solutions led to 2-fold (i.e. 0.0010 µmol/s) and 3.8-fold (i.e. 0.0019
µmol/s) reaction rate compared to the pure enzyme hydrolysis rate (i.e. 0.0005 µmol/s).
Endoglucanase activity enhancement with azoTAB towards p-nitrophenol-based substrate was
significantly higher than Avicel microcrystalline substrate. The degree of superactivity of
endoglucanase is substrate specific as reported by others ,
212, 227
and our study also showed the
difference on a soluble artificial substrate and an insoluble natural substrate. Nearly identical
reaction velocities were obtained with in situ light illumination from visible to UV light and in
reverse order.
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
0 5 10 15 20
UV (0.38)
visible (0.20)
Time(min)
[p-NP](mM)
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
0 5 10 15 20
Time(min)
UV
(0.38 mol/min/mg)
visible (0.20)
[p-NP](mM)
pure (0.10)
Figure 17. Light-responsive p-nitrophenol enzymatic cleavage product release from 4-
nitrophenyl β-D-cellobioside substrate (5 mM) with 0.4 mM azoTAB and 0.3 mg/mL
endoglucanase. pH = 5, T = 37 °C. In situ light changes are pointed by arrows. The reaction rates
are written in parenthesis. For reference, the product concentration in the presence of 0.4 mM
azoTAB visible and UV light versus time are drawn with dashed lines.

76

Although Michaelis-Menten kinetics is an oversimplified method to analyze two-phase
enzymatic reactions, it has been still a common applied model to identify catalytic constants and
provide comparative results.
234
The azoTAB effect on kinetic parameters, Michaelis constant (KM)
and maximum velocity (Vmax), of endoglucanase was determined using Michaelis-Menten non-
linear regression method (Figure 18) and listed in Table1. The turnover number, kcat increased by
36% and 39% with azoTAB from 0.00033 s
-1
to 0.00045s
-1
and 0.00046s
-1
under visible and UV
light, respectively. Michaelis-Menten constant, KM as a reciprocal term, lowered 3% and 30% with
addition of 0.4 mM azoTAB under visible and UV light, respectively. Overall, the total influence
of azoTAB to catalytic efficiency (kcat/kM) of endoglucanase for Avicel showed ~40% and ~100%
enhancement from 0.03082 L*mg
-1
s
-1
to 0.04322 L*mg
-1
s
-1
and 0.06145 L*mg
-1
s
-1
under visible
0
0.01
0.02
0.03
0.04
0.001 0.01 0.1 1 10
pure
azoTAB-vis
azoTAB-UV
Reaction Velocity (mg/L.s)
Avicel Conc. (mg/mL)
0
500
1000
1500
0 5 10 15 20 25 30 35 40
[S]/v x10
-3
(s)
[S](mg/mL)
Figure 18. Reaction rates acquired as a function of Avicel microcrystalline substrate
concentration in the presence of azoTAB (0.4 mM) under visible and UV light. Solid curves display
the nonlinear regression fits of Michaelis-Menten equation. The inlet graph represents Hanes-
Woolf plot, and the dashed lines show the fits to calculate kinetics constants.

77

and UV light, respectively. Hanes-Woolf plot was also used to acquire enzyme kinetic terms and
similar results to nonlinear regression values were obtained.

Table 1. Reaction kinetics parameters of nonlinear regression fits of Michaelis-Menten
equation and linear fit of Hanes-Woolf equation. The data points and fits are represented in the
Figure 18.

Soluble endoglucanase adsorbs onto the surface of the cellulosic substrate and diffuse into
the inside through pores and then converts into soluble sugars. AzoTAB and other surfactants
effect on endoglucanase adsorption on Avicel was assessed. As amphiphilic molecules, surfactants
have a tendency to adsorb onto surfaces that provides modification on the cellulose surfaces.
Adsorbed enzyme versus total enzyme concentration plot (Figure 19) showed that endoglucanase
followed a Langmuir-type adsorption pattern with the assumptions of homogenous adsorption
sites, one molecule binds to each adsorption site, the solutes on the surface do not interact with
each other, reversible binding and unbinding of the solute. As the endoglucanase concentration
increased, the adsorbed protein on the solid surface raised. AzoTAB-vis and azoTAB-UV show

nonlinear regression Hanes-Woolf constants
k cat(s
-1
) K M(mg/L) k cat /K M (L/mg.s) k cat(s
-1
) K M(mg/L) k cat /K M (L/mg.s)
pure 0.00027 0.03015 0.00891 0.00027 0.10425 0.00256
azoTAB-vis 0.00036 0.03463 0.01042 0.00037 0.12446 0.00294
azoTAB-UV 0.00036 0.03043 0.01181 0.00036 0.06884 0.00528

78

40-60 % higher amount of enzyme adsorption compared to the pure endoglucanase reaction,
whereas 10-20% decrease in enzyme adsorption was observed for SDS, SDBS and DTAB
containing samples. Overall, the adsorption results were correlated with relative hydrolysis rate
outcomes. Similarly, the rate of cellulose hydrolysis is associated with the adsorbed enzyme onto
the substrate surface.
210,

216

The binding affinity is specific to enzyme properties and it was reported only insignificant
amounts of BSA or β-glucosidase were adsorbed onto cellulose substrate.
62, 216
Enzyme adsorption
on solid substrate is affected by features of the enzyme and cellulose such as concentration,
temperature, pH, structural properties or crystallinity of substrate, presence of inhibitors or
enhancers.
208, 216
Tyrosine (Tyr), tryptophan (Trp), Glutamic acid (Glu) and aspartic acid (Asp)
amino acids on the surface of cellulase contributes to cellulase binding to the substrate.
130
The
cellulose binding affinity was modified with single amino acid substitutions, specifically tyrosine
to tryptophan increased the binding affinity.
235
The high adsorbed endoglucanase on Avicel could
be associated with special azoTAB interaction to tyrosine and tryptophan aromatic amino acids in
the binding sites.
The linearized Langmuir isotherm equation was used to calculate maximum adsorbed
enzyme concentration Emax and adsorption constant Kad and parameters were listed in Table 2. The
maximum enzyme concentration Emax increased from 7.89 mg enzyme/g substate to 11.52 mg
enzyme/g substate (46% increase) and 12.92 mg enzyme/g substate (64% increase) with 0.4 mM
azoTAB addition under visible and UV light, respectively. Only 3-10% increase of maximum
enzyme adsorption was determined for SDS, SDBS and DTAB containing reactions. The
adsorption constant or dissociation constant (Kad), as a reciprocal term for enzyme adsorption
affinity, decreased 20-30% for all surfactant added solutions. Similarly, increase of Emax and

79

decrease of Kad were correlated with catalytic activity enhancement with different pretreatment
conditions.
62

Figure 19. Adsorption isotherm of endoglucanase from Aspergillus niger on Avicel
(20mg/mL) in addition of azoTAB (0.4 mM) under visible and UV light, SDS (1.5 mM), SDBS
(1.5mM) and DTAB (1.5 mM) surfactants (pH 5, T=25 °C). 30 min incubation time was used.
The lines are drawn according to the individual fit of Langmuir isotherm (to obtain Emax and K ad
values).
0
1
2
3
4
5
0.1 0.2 0.4 0.6 0.8 1 1.5 2
pure
azoTAB-vis
azoTAB-UV
SDS
SDBS
DTAB
Total Enzyme Conc.(mg/mL)
Absorbed Enzyme/Avicel
conc. (mg/g)

80

Table 2. Comparison of adsorption parameters of endoglucanase on Avicel (20 mg/mL)
with and without azoTAB, SDS, SDBS and DTAB surfactants. Emax and Kad values are calculated
with Langmuir isotherm equation and the data points with the fits are represented in Figure 19.
E max(mg/g
substrate)
K ad(mg
enzyme/mL)

R
2

pure 7.89±0.89 0.33±0.06 0.9998
azoTAB-vis 11.52±0.18 0.27±0.10 0.9996
azoTAB-UV 12.92±1.13 0.26±0.04 0.9999
SDS 8.12±0.76 0.27±0.03 0.9998
SDBS 8.38±0.93 0.28±0.04 0.9998
DTAB 8.68±0.10 0.24±0.01 0.9999

Fractal kinetic models have been applied to understand further information for the
surfactant effect on enzyme-cellulose surfaces at the heterogeneous reaction. Enzymatic hydrolysis
in time was fitted into equation (20) and (21) to determine fractal parameters and kinetic constants
for different surfactant solutions and the control sample. Rate constant k was correlated with the
initial rate of the enzyme concentration as well as the substrate accessibility while fractal exponent
gives the time effect of diminished rate.
236
An addition of azoTAB led to an increase of rate
constants while conventional surfactant inclusion caused a slight decrease in the rate constants.
This suggested slightly high surfactant concentrations (1.5 mM) altered endoglucanase
confirmation and hence the rate constant decreased. At the same time, the fractal parameter f and
fractal exponent h of all surfactant added solutions lessened compared to the control samples. The
combination of lower fractal exponent and fractal parameter and higher rate coefficient showed

81

the agreement on the optimum azoTAB concentrations of 0.4 mM both under visible and UV light.
The results also showed the optimal SDS, SDBS and DTAB concentrations were at 1.5 mM range,
the endoglucanase activity was dropped at the further surfactant concentrations (Figure 16).
Table 3. Fractal kinetics parameter of endoglucanase on Avicel (0.01 mg/mL) with and
without surfactants. Values of (*) are calculated Xu and Ding, 2007 and Bommarius equation,
2008, whereas (**) numbers are determined Wang and Feng, 2010 model. The values of h and k
were calculated with 62.5 µg/mL endoglucanase enzyme towards 0.01mg/mL Avicel whereas f
values were determined with 62.5 µg/mL and 100 µg/mL endoglucanase enzyme towards
0.01mg/mL Avicel.
f
(*)
h 2
(**)
k 2(x10
6
)
(**)

pure 0.398 0.405 3.842
azoTAB-vis 0.362 0.329 4.380
azoTAB-UV 0.359 0.330 4.683
SDS 0.283 0.314 3.679
SDBS 0.294 0.301 3.270
DTAB 0.252 0.323 3.572

Accumulation of glucose and cellobiose inhibits endoglucanase catalytic activity.
207
The
adsorption of cellulases and hence cellulase catalytic activity decrease as the solid content
increases.
237
Product inhibition was correlated with solid content increase rather than end product
binding to catalytic sites of the enzyme or substrate.
238
Endoglucanase enzyme adsorption on the
solid substrate as a function of inhibitor concentrations has been studied in the literature.
239, 240
In
order to understand the surfactant effect on endoglucanase adsorption, and, thus; cellulose
hydrolysis rate at high end-product concentrations, adsorption assay was studied a constant

82

cellulose (20 mg/mL) and endoglucanase (1 mg/mL) concentrations with 0-200 mM glucose and
cellobiose (Figure 20). In here, Somogyi-Nelson method could not be applied due to the high
concentrations of end-products. The reaction rate changes were minute and hence making a
judgement was trivial. Cellobiose found to be stronger inhibitor compared to glucose
concentrations similar to the earlier study.
238
Again, this proves the solid content and cellulase
activity since cellobiose has twice molecular weight of glucose. Similarly, adsorbed endoglucanase
amount was found almost identical for 50 mM cellobiose-100 mM glucose and 100 mM
cellobiose-200 mM glucose. The relative increment of enzyme activity by addition of all of the
surfactants are similar throughout different inhibitor concentrations. Furthermore, 1 mg/mL BSA
addition decreased 60% endoglucanase activity as well as surfactant-added enzyme catalytic
activity (data is not shown).

0
40
80
120
160
0 50 100 200
pure
azoTAB-vis
azoTAB-UV
SDS
SDBS
DTAB
Relative Adsorbed Enzyme (%)
Glucose Conc.(mM)
0
40
80
120
160
0 50 100 200
pure
azoTAB-vis
azoTAB-UV
SDS
SDBS
DTAB
Cellobiose Conc.(mM)
Relative Adsorbed Enzyme (%)
Figure 20. Percentage of the adsorbed endoglucanase (1 mg/mL) on Avicel (20 mg/mL)
as a function of glucose and cellobiose end-products (inhibitors) with and without azoTAB (0.4
mM), SDS (1.5 mM), SDBS (1.5 mM) and DTAB (1.5 mM) surfactants.

83

Avicel, a microcrystalline cellulose displayed no structural changes with the addition of
surfactants as shown in Figure 21. Gel or slurry-like structure which was observed with Tween
20
211
was not observed with addition of azoTAB. Figure 21 shows the azoTAB addition enhanced
Avicel saccharification compared to the pure enzyme condition with and without addition of 200
mM glucose.

Figure 21. Optical microscopy images of Avicel as a function of time. AzoTAB surfactant
effect on Avicel hydrolysis is observed with and without end-product glucose inhibitor. The scale
bar shows 100 µm.
The fusion protein of β-glucosidase and endoglucanase was designed and applied for
carboxymethyl cellulose hydrolysis.
234
Endoglucanase and cellobiohydrolase mixtures have
always been used synergistically towards microcrystalline cellulose avicel.
221, 241
Therefore, four
different enzyme mixtures of endoglucanase, endoglucanase and β-glucosidase, endoglucanase
and cellobiohydrolase, and all three enzymes of endoglucanase, β-glucosidase and

84

cellobiohydrolase have selected to study the effect of surfactants on synergistic activity of cellulase
enzyme mixtures for initial and long-term behavior (Figure 22). As the reaction time or conversion
increases, the reaction rate slowed down as reported in the literature.
242
AzoTAB-UV addition
showed the highest enzyme activity enhancement (i.e. 25-55% of catalytic activity increase) at
four different enzyme formulations and different time points. While 6-15 % light-responsive
relative activity was observed throughout endoglucanase and a mixture of endoglucanase and β-
glucosidase at different timelines, the light-responsivity was the lowest at cellobiohydrolase
enzyme containing solutions. Cellobiohydrolase addition lowered the enzyme activity
enhancement with azoTAB surfactant both under visible and UV light compared to the pure
enzyme solution. In a similar fashion, poly ethylene glycol (PEG) addition increased
cellobiohydrolase activity while showed no effect on endoglucanase activity,
221
and Tween 20
increased cellulase adsorption while it had no effect on β-glucosidase activity.
211
It is also reported
that while Tween 80 was an activator for cellobiohydrolase, it behaved like an inhibitor for
endoglucanase or β-glucosidase.
243
Presence of 0.4 mM azoTAB became more announced with
the addition of β-glucosidase with 5-25% relative catalytic activity enhancement of the enzyme
mixtures. Likewise, our previous study reported 30% and 60% relative activity increase in β-
glucosidase activity at 0.4 and 0.75 mM azoTAB addition under UV light, respectively.
175
Our
current results showed the enhancement of β-glucosidase activity seemed conserved in the
cellulase enzyme mixtures. β-glucosidase activity could be increased with addition of azoTAB
either isolated β-glucosidases
175
or mixtures of endoglucanases and cellohydrolases as shown in
Figure22. Overall, 50% of activity enhancement was obtained with the addition of 0.4 mM of
azoTAB independent of light wavelength for the combination of all three enzymes.

85

On the other hand, cellobiohydrolase addition showed minimal change for other surfactants
while β-glucosidase addition slightly increased the activity enhancement. Traditional surfactant
addition showed increased activity in long-term (i.e. 5h and 24h) while azoTAB conserved the
constant superactivity of enzymes. Same activity enhancement throughout 24 h for azoTAB while
other surfactants show activity increase in longer time. This could be explained with the main
reason of surfactant addition into the cellulose degradation: (1) making the substrate surface better,
(2) diminishing irreversible binding of enzyme to substrate.
209, 229
The explanation of why common
surfactants showed increased activity rate in long-term could be the surfactant concentrations
differences with azoTAB (azoTAB is 0.4 mM while other common surfactants are at 1.5 mM.) All
in all, azoTAB has been shown to significantly improve (25-50% increment) both the initial rate
and the long-term enzyme reaction rate in cellulose degradation.

86

0
50
100
150
endo endo+BGL endo+exo
azoTAB-vis
azoTAB-UV
SDS
SDBS
DTAB
endo+exo
+BGL
Enzyme Mixtures
Relative Specific Activity(%)
(a) 60 min
0
50
100
150
endo endo+BGL endo+exo
Enzyme Mixtures
(b) 5 h
Relative Specific Activity(%)
endo+exo
+BGL
0
50
100
150
endo endo+BGL endo+exo
Relative Specific Activity(%)
(c) 24 h
Enzyme Mixtures
endo+exo
+BGL
Figure 22. Relative activity of various enzyme mixtures of endoglucanase (62.5 μg/mL),
cellobiohydrolase (120 μg/mL) and β-glucosidase (30 μg/mL) with addition of azoTAB (0.4 mM
under visible or UV light), SDS (1.5 mM), SDBS (1.5 mM) and DTAB (1.5 mM) towards Avicel
(0.01 mg/mL), pH = 5, T = 25 C. (a) Initial, (b) and (c) long-term relative saccharification rate of
Avicel.

87

3.5. Conclusion
The reaction activities and kinetics of endoglucanase at the solid-liquid interface have been
investigated with of azoTAB photosurfactant under visible and UV light. The enzyme specific
activity was enhanced 45% with the addition of 0.4 mM azoTAB under UV light. Light-induced
endoglucanase activity range became narrower (~15%) for natural substrate Avicel while the
endoglucanase activity could be controlled between 2-fold to 4-fold toward p-nitrophenol-based
model substrate (4-nitrophenyl β-D-cellobioside). The underlying reason of the 45% specific
activity enhancement was correlated with the ~40-50% increased adsorbed enzyme content and
catalytic enzyme efficiency with lower fractal parameters (f and h). In comparison, 5-10% slight
catalytic activity increase was observed in addition of sodium dodecyl sulfate (SDS), sodium
dodecyl benzene sulfonate (SDBS) and dodecyl trimethyl ammonium (DTAB) traditional
surfactants. Lastly, azoTAB addition maintained 45-50% enzyme catalytic rate increase on all
three cellulase enzyme mixture of endoglucanase, cellobiohydrolase and β-glucosidase both for
initial rate and long-term behavior. Overall, azoTAB could be applied as a promising enhancer of
cellulase enzymes for biomass hydrolysis and this study can be summarized as shown in Figure
23.

Figure 23. A simplified cartoon to show increase of adsorption of endoglucanase enzymes
on Avicel which results in endoglucanase activity enhancement.

88

Acknowledgement. This material is based upon work supported by the National Science
Foundation under Grant 1758225. Any opinions, findings, and conclusions or recommendations
expressed in this material are those of the author(s) and do not necessarily reflect the views of the
National Science Foundation. The authors acknowledge Dr. Fariborz Nasertorabi of USC’s Bridge
Institute for his assistance with enzyme purification studies.

89

CHAPTER 4: PHOTO-CONTROLLED INHIBITION OF BREAST CANCER
CELL GROWTH VIA CO-DELIVERY OF BCL-2 SIRNA AND PACLITAXEL WITH
CATANIONIC VESICLES
4.1. Abstract
Localized drug delivery holds great promise as a means of circumventing traditional
chemotherapy side effects associated with high toxicity and extended treatment durations. The use
of various nanocarriers loaded with chemotherapeutics has been shown to increase cancer
treatment efficiency since the nanoparticle size can be designed to facilitate preferential
accumulation and penetration into tumor cells as opposed to normal, healthy cells. Yet the
challenge remains that designing a carrier that is both durable enough to survive the bloodstream,
while at the same time allowing release of the payload once inside the cell, can be mutually
exclusive. In the present study, we have developed a method to co-deliver siRNA and the
hydrophobic drug paclitaxel into cancer cells via photoresponsive catanionic vesicles. The
vesicles with diameters of ~100 nm form spontaneously upon simple mixing of an azobenzene-
based cationic surfactant and a conventional anionic surfactant. Upon photoisomerization of the
azobenzene moiety with UV light, the vesicles disassociate into free surfactant monomers. This
allows for both delivery and triggered release of the co-loaded vesicle payload into tumor cells.
Small-angle neutron scattering, fluorescence spectroscopy, dynamic light scattering, and zeta
potential measurements are utilized to determine the optimal vesicle size, charge, and bilayer
thickness for the encapsulation and cellular uptake of co-delivered Bcl-2 siRNA and paclitaxel
into MDA-MB-231 human breast cancer cells. Cell viability measurements combined with flow
cytometry and confocal microscopy were used to determine the safe and effective dosage range of
the catanionic vesicles, while successful knockdown of Bcl-2 protein suppression was confirmed

90

by using a western blot. Photo-triggered release of siRNA and paclitaxel from the azobenzene-
based catanionic vesicles following cellular uptake is shown to enhance the therapeutic efficacy.
4.2. Introduction
Breast cancer is a ubiquitous cancer type and a leading cause of death of women throughout
the world.
244,245
It is predicted almost 300,000 people will develop breast cancer and 45,000 deaths
will be attributed to the disease.
246
Chemotherapy is the most common method used to control or
inhibit breast cancer cell growth,
247
with chemotherapy drugs typically selected to induce apoptosis
of cancer cells. The microtubule-targeting drug paclitaxel has been one of the most widely used
chemotherapy drug for breast cancer treatment over the last three decades.
248
Paclitaxel suppresses
the rapid microtubule dynamics required during mitosis, thereby preventing the dividing cancer
cells from progressing past metaphase
249
and leading to apoptosis.
250-252
Unfortunately, paclitaxel
has poor solubility in the blood stream, which can cause adverse reactions during breast cancer
treatment.
253-255
Moreover, the existence of apoptotic-related gene alterations can lower the
effectiveness of paclitaxel and lead to tumor recurrence.
256
Thus, combining paclitaxel usage with
silencing of anti-apoptotic genes can be a promising treatment strategy.
257, 258
Furthermore, such a
combination of chemotherapeutics offers the ability to inhibit diverse disease pathways that result
from the complexity of different mechanisms in breast tumor cells.
259, 260
For example, the
apoptotic B-cell lymphoma-2 (Bcl-2) gene has been reported to be associated with paclitaxel
resistance.
261,262
Bcl-2 protein expression prevents cell death, with high amounts of Bcl-2 in cancer
cells leading to cell survival.
263
Notably, overexpression of Bcl-2 protein is observed in
approximately 75% of breast cancers.
264
Thus, the co-delivery of Bcl-2 siRNA and the
chemotherapy drug paclitaxel has the potential to increase the inhibition of cancer cell growth and
improve the efficacy of cancer treatment. Unfortunately, co-delivery delivery can be challenging

91

due to differences in physical, chemical, and biological properties of the two therapeutics, such as
molecular weight, hydrophobicity/hydrophilicity, stability, and solubility.
265

Nanoparticles have been widely used as a means of increasing the loading capacity of
poorly water-soluble molecules, providing safe and effective drug and gene delivery into tumor
cells while also preventing the degradation of loaded molecules before reaching the target cells.
266-
268
Passive-targeting nanoparticle delivery strategies have been developed that use positively-
charged nanocarriers on the order of 100 nm in diameter to preferentially accumulate in tumor
tissue rather than healthy cells. This enhanced permeability and retention (EPR) effect
113, 269

purportedly occurs due to the leaky microvessels of tumors. Note, however, that the EPR effect is
likely more complicated and heterogeneous in the human body compared to mouse and rats where
it was first observed. This is because murine tumors grow relatively quickly (leading to much
larger tumor-to-body ratios), while human tumors are more densely-packed and lack leaky
microvessels.
270
Nevertheless, nanoparticles have shown promise as a means to specifically target
drugs to cancerous cells. The size, shape, composition, and surface charge of nanoparticles are
important parameters for cellular internalization.
271
For example, Choi et al. reported that
poly(lactic-co-glycolic acid) (PLGA) nanoparticles with sizes less than 200 nm had less
cytotoxicity, higher tumor-cell interaction, and greater cellular uptake compared to larger PLGA
nanoparticles.
272
Likewise, nanoparticles smaller than 200 nm exhibit a higher possibility to enter
tumor cells through endocytosis and also result in less toxicity.
273, 274
Spherical nanoparticles with
size distributions generally have higher tendency for vascular permeability, tumor penetration, and
cellular uptake compared to elongated nanoparticles.
275, 276
The nanoparticle composition can also
dramatically affect cellular uptake and independent of particle size.
276

92

Another issue limiting the effectiveness of drug and gene delivery is the often mutually
exclusive design parameters requiring that the carriers be durable enough to survive the
bloodstream, while at the same time releasing their payload once inside the cell cytoplasm. For
example, release of the encapsulated therapeutics following delivery can take from days to weeks
in extreme cases.
277
Thus, stimuli-responsive nanoparticles, including pH-,
278, 279
enzyme-,
280, 281

magnetic field-,
282, 283
ultrasound-,
284, 285
redox potential-,
286, 287
or light-responsive
nanoparticles,
35, 288
are attractive carriers that potentially allow for the localization and controlled
release of chemotherapeutics. Light-responsive nanoparticles in particular have been applied in
drug delivery to enhance treatment efficiency as a result of photo-triggered drug release in target
tissues.
289
The simultaneous long-term protection and nearly-instantaneous light-triggered release
of unstable and/or poorly water soluble payloads could increase the efficiency and safety of the
treatment. Technical issues regarding the exposure of diseased cells to light have been largely
resolved, notably with conventional photodynamic therapy for the treatment of certain cancers,
where a photosensitizer drug, localized in tumors, is illuminated to generate reaction products that
are toxic to tumor cells.
290,

291

Herein we investigate the use of photoresponsive vesicles to allow co-delivery and
triggered release of paclitaxel and Bcl-2 siRNA. These so-called “catanionic” vesicles form
spontaneously upon simple mixing of an azobenzene-based cationic surfactant (azoTAB) with a
traditional alkylbenzene-based anionic surfactant (SDBS). AzoTAB undergoes a reversible
photoisomerization from a planar and relatively-hydrophobic trans conformation under visible
light to a bent and relatively-hydrophilic cis structure under UV illumination, as seen in Figure
15a (more precisely, azoTAB exhibits a 75/25 trans/cis or 10/90 trans/cis photostationary state
under visible or UV light, respectively).
41
As a result, a variety of photoreversible microstructural

93

changes can be initiated with light, including vesicle  micelle, vesicle  lamellar, and vesicle
 free–monomer transitions.
41
Notably, the photo-initiated transition from vesicles to free
surfactant monomers (see Figure 15b) was used in our previous work to deliver and then release
e-GFP DNA into cells.
35
In the present study, light-responsive vesicles are used to co-deliver the
chemotherapy drug paclitaxel with Bcl-2 siRNA to MDA-MB-231 human breast cancer cells. The
effects of varying the azoTAB hydrophobic tail on the vesicle surface charge, size, and bilayer
thickness was studied to optimize the delivery vector. Vesicles were characterized by a
combination of small-angle neutron scattering, dynamic light scattering, fluorescence
spectroscopy, and zeta potential measurements. The encapsulation of siRNA and paclitaxel in
azoTAB/SDBS vesicles was investigated through agarose gel and UV-vis spectroscopy,
respectively. Vesicles co-loaded with paclitaxel and Bcl-2 siRNA that pass through MDA-MB-
231 human breast cancer cell membranes can be subsequently exposed to UV light to cause vesicle
rupture and release of the payload into the cell interiors of cells, where siRNA reduces Bcl-2
protein production allowing paclitaxel to induce cell death.

94

Figure 24. (a) The structure of azoTAB and SDBS surfactants, as well as the trans-to-cis
photoisomerization of azoTAB upon exposure to visible (434 nm)  UV (350 nm) light. The
length of the alkyl spacer in azoTAB is n = 2 (2-azo-2), n = 4 (2-azo-4), or n = 6 (2-azo-6). (b)
Illustration of the reversible transition from catanionic vesicles to free surfactant monomers that
can occur upon azoTAB photoisomerization.

4.3. Experimental Section
Materials. Surfactants: Three photoresponsive azoTAB analogs were prepared as
previously reported,
31,33
namely 4-ethyl-4’(trimethylamino-ethoxy) azobenzene bromide (“2-azo-
2”), 4-ethyl-4’(trimethylamino-butoxy) azobenzene bromide (“2-azo-4”) and 4-ethyl-
4’(trimethylamino-hexyloxy) azobenzene bromide (“2-azo-6”). Briefly, these surfactants were
synthesized by azocoupling, alkylation (with the corresponding dibromoalkane), and
quaternalization steps. Sodium dodecylbenzene sulfonate and all chemicals for azoTAB synthesis
were obtained from Sigma-Aldrich at the highest purity available and used as received.
Chemotherapeutic agents: Custom Cy5.5-labeled Bcl-2 siRNA (Sense: GUA CAU CCA UUA
UAA GCU G-dTdT; Anti-Sense: CAG CUU AUA AUG GAU GUA C-dTdT) was ordered from

95

Dharmacon, GE Life Sciences. Cyanine 5.5 (Cy5.5) is a near-infrared dye with excitation
maximum at 688 nm and emission maximum at 703 nm, chosen so as to not be influenced by
azoTAB absorption, which occurs below ~550 nm.
24
Dilutions of siRNA were done by addition
of nuclease-free molecular biology grade water (Thermo Fisher Scientific). Paclitaxel was
purchased from Sigma-Aldrich and stock solutions of the hydrophobic drug were prepared in
ethanol. Cell culture: The MDA-MB-231 human breast cancer cell line was obtained from
American Type Culture Collection (ATCC) and cultured in Dulbecco’s modified Eagle’s medium
(DMEM, Mediatech, Inc.) with 10% fetal bovine serum (FBS, Sigma-Aldrich) in a humidified
incubator containing 5% CO2 at 37 C. Cells were passaged with trypsin solution every 3-4 days
to keep the cells subconfluent. Cell concentrations were quantified in trypsin solution with a
hemocytometer.
Preparation and characterization of azoTAB-based catanionic vesicles. Preparation of
photoresponsive catanionic vesicles: Catanionic vesicles were formed by simply mixing separate
stock aqueous solutions of photoresponsive azoTAB (cationic) and SDBS (anionic) at a molar
ratio of 7/3, followed by one minute of gentle vortexing. This cationic-to-anionic surfactant ratio
was selected to result in a net positive charge of the nanoparticles, similar to our earlier study.
35

Trans-to-cis photoisomerization of nitrogen-nitrogen double bond of azoTAB was achieved by
illumination with UV light from an 84 W long-wavelength 365 nm UV lamp (XX-15A-
Spectroline), which in turn causes a disruption of catanionic vesicles as illustrated in Figure 24.
This allows delivery of genetic material and hydrophobic drug after entering into the cell, as Bcl-
2 siRNA and Paclitaxel are co-encapsulated under visible light and released under UV light.
Bilayer Thickness Determination: Small-angle neutron scattering measurements were performed
at 25 C on the 30-meter NG7 SANS instrument at the National Institute of Standards and

96

Technology (NIST).
153
Two sample-to-detector distances of 1.33 m and 7 m with a detector offset
of 25 cm and a neutron wavelength (λ) of 6 Å were used to acquire a Q range of 0.00494 − 0.5453
Å
-1
(Q = 4λ
-1
sin(θ/2), where θ is the scattering angle). The background and empty cell (pure
D2O) were subtracted from the raw data to obtain the net intensities with consideration of the
detector efficiency from an isotropic scatterer (Plexiglas). The units were changed to cm
-1
as an
absolute differential cross section per unit sample volume by utilization of an attenuated empty
beam. Separate 0.1 wt % azoTAB and SDBS surfactant solutions were prepared in D2O, mixed at
a molar ratio of 7/3, and then loaded into a 2 mm quartz cell. The SANS data were analyzed by
using the Guinier approximation appropriate for lamellar particles,
182

𝐼 (𝑄 )=𝐼 (0)𝑒𝑥𝑝 (−𝑄 2
𝑅 𝑡 2
)/𝑄 2
(23)
where I(Q) is the scattering intensity with extrapolated intensity I(0) at Q = 0, while Rt is the
thickness radius of gyration. Thus, from modified Guinier plots of ln(IQ
2
) versus Q
2
, generally
valid for QRt < 1,
41
Rt values were calculated and then converted to bilayer thicknesses dL = Rt√12.
Critical Aggregation Concentration (CAC) Calculation: The fluorescent spectra of the
hydrophobic and nonionic fluorescence probe Nile red (N3013-Sigma-Aldrich) was measured
using a Quanta-Master spectrofluorometer (Photon Technology International, model QM-4) at 37
C. Catanionic surfactant solutions with a molar ratio of 7/3 azoTAB/SDBS were prepared over a
range of surfactant concentrations in pH 7.4 PBS buffer. To convert azoTAB to the cis isomer,
samples were exposed to UV light (XX-15A-Spectroline) for at least 30 minutes prior to addition
of Nile red to avoid photobleaching of the dye. 2 mL of surfactant solution (in pH 7.4 PBS buffer)
and 5 μL of Nile red solution (1 mM concentration in ethanol) were mixed for data collection. The
samples were allowed to reach steady state for 20 min in the dark prior to loading into a quartz
cuvette with a path length of 10 mm (QS-554-Nova Biotech). Fluorescence emission spectra were

97

collected between 600 and 800 nm using an excitation wavelength of 590 nm with excitation and
emission slit widths of 4 nm. The wavelength of maximum emission (λmax) of each sample was
calculated by fitting the data near the peak to a fifth-order polynomial. Hydrodynamic Diameter
Measurements: Dynamic light scattering measurements were performed on a Brookhaven
instrument (Brookhaven Instrument Corp., Holtsville, NY, model BI-200SM) with a digital
correlator (BI-9000AT) and a 35 mW He-Ne (wavelength = 632.8 nm) laser (Melles Griot, model
05-LHP-928) at 37 C. The data were collected at a scattering angle of 90. AzoTAB and SDBS
aqueous solutions were prepared independently in pH 7.4 PBS buffer and then mixed by passing
through separate 100-nm syringe filters (Whatman, Anotop 10, 6809-1012) to a molar
concentration ratio of 50 μM/21.3 μM, followed by vortexing for one minute. The samples were
then allowed to reach equilibrium for one hour under visible (i.e., lab) light. Following subsequent
light scattering measurements, the samples were exposed to UV light for at least 30 min and then
re-measured via DLS. Size distributions were acquired periodically over the course of 50 days
with the samples stored at either room temperature, 4 C, or 37 C. Hydrodynamic radii were
determined from the measured diffusion coefficients according to the Stokes-Einstein equation
(RH = kT/6D, where k is Boltzmann’s constant, T is temperature,  is the solvent viscosity, and
D is the diffusion coefficient). Seven separate 5-minute runs were measured for each sample and
condition. The NNLS (non-negative least squares) calculation method was used to analyze the
hydrodynamic diameters, with the size reported as the predominant NNLS peak. These reported
diameters were consistent with the effective diameters calculated from the method of cumulants,
with polydispersity values varying between 0.1 – 0.3 (generally indicating a unimodal size
distribution).

98

Preparation of siRNA, PTX and co-loaded Photoresponsive Catanionic Vesicles.
Small (~1 L) amounts of an siRNA stock solution (100 μM in nuclease-free molecular biology
grade water) were mixed with 1.99 mL of pH 7.4 PBS buffer and 5 μL each of a 20 mM azoTAB
solution and 8.6 mM SDBS solution (to achieve a 7/3 azoTAB/SDBS molar ratio). Subsequently,
a small amount (< 10 L) of a concentration paclitaxel solution (13.4 mM) in ethanol was added
to the solution. The vesicles were then left to encapsulate the siRNA and paclitaxel and reach
steady state for 20 – 30 minutes. Paclitaxel Encapsulation Efficiency and Loading Capacity: The
ability of azoTAB-SDBS vesicles to solubilize paclitaxel was assessed by adding varying amounts
of the drug from the stock ethanol solution to a 50 μM azoTAB and 21.4 μM SDBS vesicle
solutions. After gently mixing with repeated pipetting, the samples were allowed to rest for thirty
minutes at room temperature prior to filtering through a 1 μm polytetrafluoroethylene (PTFE)
syringe filter (Midland Scientific Inc. #NSC F2500-13) to remove the unencapsulated paclitaxel.
UV-vis spectroscopy and light scattering confirmed that empty vesicles were able to pass through
the filter without loss. The encapsulated paclitaxel concentrations were then determined with UV-
vis spectroscopy (Agilent Technologies, Inc., Santa Clara, CA, model 8453) at a wavelength of
230 nm using an extinction coefficient of 19.3 mM
-1
cm
-1
(linearity coefficient of 0.9993)
determined independently in 1% ethanol pH 7.4 PBS buffer, similar to reported values in a
methanol solution.
292, 293
The encapsulation efficiency and paclitaxel loading capacity were then
calculated as the mass of encapsulated paclitaxel divided by either the total amount of added drug
or mass of surfactant, respectively. Preparation of siRNA-loaded nanoparticles and electrophoretic
mobility shift assay: Agarose gel (1%) electrophoresis was utilized to analyze the loading of
siRNA into catanionic vesicles. Each sample was prepared by mixing 1 μL of a stock siRNA
solution with various amounts of azoTAB and SDBS stock solutions (20 mM and 8.57 mM,

99

respectively, in pH 7.4 PBS buffer) to arrive at a final siRNA concentration of 30 nM and nitrogen-
to-phosphorous (N/P) ratios (i.e., azoTAB-to-nucleotide ratios) ranging from 2.5 to 40. The
samples were incubated at room temperature for 30 minutes prior to running the gel to allow for
siRNA encapsulation. 4 μL of nucleic acid loading dye (Thermo Fisher R0611) was added to track
siRNA migration throughout electrophoresis. The experiments were repeated in the presence of
paclitaxel to ensure siRNA and paclitaxel co-encapsulation by the catanionic vesicles. Agarose gel
images were taken under UV light to obtain the siRNA and paclitaxel-siRNA loading capacity of
the vesicles. Surface Charge and Size of Loaded Catanionic Vesicles: The zeta potential and z-
averaged hydrodynamic diameter (i.e., the effective diameter determined by the method of
cumulants) of empty and loaded vesicles were measured on Malvern instrument (Malvern
Instruments Inc., Westborough, MA, model Zen 3690) with a detection angle of 90 and
wavelength of 632 nm at 25 C by using a folded capillary cell (DTS1070-Malvern). Catanionic
vesicles were prepared in pH 7.4 PBS buffer, followed by the addition of various amounts (from
30 nM – 500 nM) of Bcl-2 siRNA at room temperature when desired. For co-delivery
measurements, an appropriate amount of paclitaxel was subsequently added to achieve a 1/8 (w/w)
paclitaxel/surfactant ratio. The vesicles were allowed to encapsulate either siRNA or siRNA with
paclitaxel for 30 minutes before measurements. Photo-controlled Release Behavior from
Catanionic Vesicles: UV light-induced paclitaxel and siRNA release experiments were performed
by preparing stock 50 μM azoTAB and 21.4 μM SDBS solutions in pH 7.4 PBS buffer under
visible light mixed to achieve the desired 7/3 azoTAB/SDBS ratio. Subsequently, paclitaxel (at a
1/8 paclitaxel/surfactant weight ratio) or siRNA (N/P = 2.5) was added to the samples as described
above, followed by incubation for 30 minutes in the dark at 37 C to complete the encapsulation
process. Half of each sample was then placed in separate microcentrifuge tubes, with one tube

100

exposed to UV light for 10 minutes. Each sample was then centrifuged at 10,000 rpm for 30
minutes using a Microfuge® 18 centrifuge (Beckman Coulter Inc., Brea, CA). The percentage of
paclitaxel released (i.e., returning to the supernatant) following UV-induced vesicle rupture was
detected with UV-vis spectroscopy at a wavelength of 230 nm, while siRNA release was detected
using fluorescence spectroscopy to detect Cy5.5 at an emission wavelength of 703 nm. The
absorbance was measured before and after centrifugation (i.e., vesicles and supernatant,
respectively), allowing the percent release to be calculated from the relative vesicle-minus-
supernatant differences under UV and visible light, namely
Percentage Release = (1−
(𝑣𝑒𝑠𝑖𝑐𝑙𝑒𝑠 −𝑠𝑢𝑝𝑒𝑟𝑛𝑎𝑡𝑎𝑛𝑡 )𝑈𝑉
(𝑣𝑒𝑠𝑖𝑐𝑙𝑒𝑠 −𝑠𝑢𝑝𝑒𝑟𝑛𝑎𝑡𝑎𝑛𝑡 )𝑣𝑖𝑠 )100.
Cytotoxicity. MDA-MB-231 human breast cancer cells were seeded in a 96-well plate at
a density of 5  10
3
cells/well with 10% FBS media and incubated overnight at 37 C to allow cell
attachment. The following day the medium was removed, and the cells were treated with either
pure medium (control), different concentrations of blank vesicles, siRNA-loaded vesicles,
paclitaxel-loaded vesicles, or siRNA-paclitaxel co-loaded vesicles under visible and UV light. In
the latter case, the plates were exposed to UV light for 10-20 minutes every two hours. After 24
hours, the XTT assay (Cell Proliferation Kit II, Roche, Catalog Number: 11 465 015 001) was
applied according to the manufacturer’s protocol to test cell viability, with a UV-vis plate reader
(Molecular Devices) utilized to read the absorbance of 450 nm after two hours. The cell viability
was calculated as the difference of the absorbance of the sample minus the absorbance obtained
for blank vesicles in pure medium without cells, normalized by the absorbance of the cells treated
with pure medium.
Flow Cytometry. Transfection efficiency was analyzed quantitatively using a MACS
Quant flow cytometer (Miltenyi Biotech Inc., San Diego, CA). MDA-MB-231 human breast
(24)
4)

101

cancer cells were seeded in 6-well plates at a density of 6  10
5
cells/well and incubated for two
days at 37 C with 5% CO2 environment until 70% – 80% confluency was reached. The cell
medium was exchanged with either pure medium (control) or medium containing either naked
siRNA or siRNA-loaded catanionic vesicles. After transfection for four hours at 37 C, the cells
were washed with PBS and collected with trypsin. The cells were then centrifuged at 10,000 rpm
for 5 minutes, after which the medium was removed, and the cells were re-suspended in pH 7.4
PBS buffer. A 635-nm red laser was used to excite Cy5.5-labeled siRNA using a Miltenyi Biotec
flow cytometer to detect transfected cells, with FlowJo software used to analyze the data. The cells
were also measured without siRNA treatment to measure the background fluorescence intensity of
cells, which was subtracted from the siRNA-treated samples to identify Bcl-2 siRNA delivered
cells.
Confocal Microscopy. Confocal laser scanning microscopy was applied to visualize
cellular uptake. The cells were seeded in 6-well plates at a density of 6  10
5
cells/well and allowed
to attach to coverslips and reach 70% – 80% confluency in a humidified incubator for two days at
37 C. Cells were then treated with a medium containing the hydrophobic dye Coumarin 6 (C6,
ex/em: 443/494-505 nm) and Cy5.5-conjugated siRNA (ex/em: 685/703 nm) co-loaded into
azoTAB/SDBS vesicles as described above. After 3.5 hours of incubation at 37 C, the cells were
stained with DAPI (ex/em: 340/480 nm) for an additional 30 minutes. The cells were then washed
with PBS buffer, fixed with 4% paraformaldehyde for 20 minutes at room temperature, and placed
on a confocal microscopy dish. A Nikon TI Eclipse inverted microscope (Nikon Instruments Inc.,
Melville, NY) attached to a spinning disk CSUX confocal scanning unit (Yokogawa Electric
Corp., Tokyo, Japan) was used to visualize intracellular-localized co-loaded catanionic vesicles.
The images were captured using a 16-bit Cascade II 512 EMCCD camera (Photometrics Inc.,

102

Huntington Beach, CA) with a Nikon Plan-Apo 40 oil immersion objective through 405 nm, 491
nm, and 640 nm laser lines.
Endosomal Escape. Nikon TI Eclipse inverted microscope (Nikon Instruments Inc.) was
used to visualize endosomal escape of siRNA. MCF-7 breast cancer cells were seeded in 6-well
plates at a density of 6 × 10
5
cells/well and incubated for 2 days at 37 °C with 5% CO2 environment
until 70−80% confluency was reached. The cells were then treated with a medium containing
azoTAB/SDBS vesicles loaded with the Cy5.5-conjugated siRNA (N/P = 19). After transfection
for 4 h at 37 °C in the dark, Cy5.5-siRNA-transfected cells were detected using a Cy5 filter set
before and after 350 nm exposure for 60 s. The images were captured using a 16-bit Cascade II
512 EMCCD camera (Photometrics Inc.) with a Nikon Plan-Apo 40× objective through Cy5 filter
set. The data were analyzed using Fiji software.
In Vitro Release Behavior - Western Blot Analysis. The siRNA efficiency of naked Bcl-
2 siRNA and Bcl-2 siRNA loaded catanionic vesicles were performed using a western blot. MDA-
MB-231 human breast cancer cells were cultured in 6-well plates at a density of 6  10
5
cells/well
and allowed to attach to the well and reach 70% – 80% confluency in a humidified incubator for
two days at 37 C and 5% CO2 environment. The cells were then treated with medium containing
naked siRNA and siRNA-loaded catanionic vesicles for four hours, followed by illumination with
UV light for one hour at a distance of 20 cm and then incubation in the dark for an additional 43
hours, all at 37 °C. The cells were then washed with PBS buffer, collected, and lysed on ice with
protease-inhibitor containing lysis buffer. Bradford reagent was used to verify the presence of
protein. Laemmli buffer and protein solutions were mixed at a 1/1 (v/v) ratio and heated for five
minutes at 95 °C. SDS-PAGE gels were utilized to separate lysates with the proteins transferred
onto polyvinylidene fluoride (PVDF) membranes blocked in Tris-buffered saline with a 0.1%

103

Tween 20 (TBST) solution and incubated overnight with the antibodies for human Bcl-2
(VMA00017, Bio-Rad) and β-actin at 4 °C. β-actin was used to verify that the protein loading was
the same in each gel, and to show that only Bcl-2 protein was suppressed. After the membranes
were washed 3-4 times with TBST, the images were obtained using an Odyssey infrared
fluorescent imager (LI-COR BioSciences, Lincoln, Nebraska).
Statistical Analysis. All experiments were collected at least three times to present the
mean and calculate the standard deviation. The results were considered statistically significant
with a value of p < 0.05.
4.4. Results and Discussion
Different hydrophobic tail length photo-surfactants were synthesized as C2H5-
Azobenzene-O(CH2)2-, C2H5-Azobenzene-O(CH2)4- and C2H5-Azobenzene-O(CH2)6- with same
hydrophilic head groups of N
+
(CH3)3Br
-
as described in the experimental methods. The molecular
weights of 2-Azo-2, 2-Azo-4 and 2-Azo-6 were confirmed by utilizing H-NMR 392, 420 and 448
g/mol, respectively.
33
Dipole moment of nitrogen-nitrogen double bond is about 0.5 D for trans
azo-surfactant whereas it is 3.1 D for cis isomer that leads trans isomer to have higher
hydrophobicity than cis isomer.
34,288
Hydrophobicity differences of 2-Azo-X photo-isomers in the
introduction of an anionic surfactant produce light-sensitive self-assemblies which can be switched
from vesicles to micelles, lamellar structures or free surfactant monomers by simple light
illumination.
35

Cationic to anionic surfactants ratio was selected 7:3 to retain positive charge in the
nanoparticles similar to our earlier study.
35
Catanionic samples were prepared simply by mixing a
7/3 molar ratio of 2-azo-X and SDBS solutions. Photo-induced trans to cis isomerization of
nitrogen-nitrogen double bond of 2-Azo-X caused by irradiation with UV light, that in turn causes

104

disruption of catanionic vesicles that is illustrated in Figure 25. Light-responsive vesicle to free
monomer transition was utilized to deliver a genetic material and a hydrophobic drug. Thus, the
genetic material (hydrophilic, Bcl-2 siRNA) and a hydrophobic chemotherapy drug (i.e.
Paclitaxel) are co-encapsulated under visible light and disrupted under UV light to release
encapsulated materials controllably as shown in the Figure 25.

Figure 25. The cartoon shows catanionic surfactant vesicles and containing the siRNA and
the Paclitaxel (PTX), with the finally formed “2-in-1 nanocarrier” undergoing endocytosis by
breast cancer cell, then irradiated with ultraviolet light (UVA) causing disruption and emptying
the nanocarrier contents into the cell where the siRNA matches with Bcl-2 producing
complementary mRNA produced by the nucleus and inhibits formation of the Bcl-2 protein.
The critical aggregation concentration (CAC) of azoTAB/SDBS (7/3 mol/mol) mixtures
using each of the three azoTAB analogs was determined by using the hydrophobic fluorescence
dye Nile red as a micropolarity indicator, as shown in Figure 26. The dye exhibits a wavelength of

105

maximum emission λmax  655 nm in water, while a blueshift in this emission is observed when
the probe becomes solubilized in a hydrophobic medium.
294, 295
In pure pH 7.4 PBS buffer, an
average λmax value of 655.7±0.2 was found, in agreement with the literature. The onset of vesicle
formation (i.e., the CAC) was then estimated as the azoTAB concentration where the λmax value
first decreases (denoted by arrows in Figure 26), indicating the point where Nile red begins to
partition into the vesicle bilayers. Under visible light, the CAC values of 25 μM (2-azo-2/SDBS),
15 μM (2-azo-4/SDBS), and 5 μM (2-azo-6/SDBS) decrease as the azoTAB tail becomes longer
and more hydrophobic, as demonstrated by the decreasing pure-surfactant critical micelle
concentrations of 9.5 mM (2-azo-2),
31
5.3 mM (2-azo-4),
24
and 1.6 mM (2-azo-6).
31
Likewise, the
CAC values are observed to be higher under UV light (50 μM, 40 μM, and 20 μM, respectively),
in agreement with the increased hydrophilicity of the cis azoTAB conformation, as well as the
planar trans form being less sterically hindered to participate in bilayer formation relative to the
bent cis state. For comparison, the CAC values obtained for 2-Azo-4/SDBS are in reasonable
agreement with those previously reported in buffer at 37 C (visible: 10 μM, UV: 60 μM)
35
as
well as in pure water at 25 C (visible: 8 μM, UV: 32 μM).
41
As a result of the large differences in
CAC values obtained under visible versus UV light, operating at an azoTAB concentration
between these respective CAC values allows for vesicle to free monomer transitions to be initiated
with light, which would enable light-triggered release of the vesicle payload. Importantly, the low
(i.e., micromolar) CAC values overall also allow for siRNA and paclitaxel delivery with low
cytotoxicity (discussed below with Figure 33).

106

630
635
640
645
650
655
660
vis
UV
0 20 40 60 80 100

max
(nm)
vis
UV
2-Azo-2/SDBS
azoTAB Concentration (M)

630
635
640
645
650
655
660
vis
UV
0 20 40 60 80 100
vis UV
2-Azo-4/SDBS
azoTAB Concentration (M)

630
635
640
645
650
655
660
0 20 40 60 80 100
vis
UV
vis
UV
2-Azo-6/SDBS
azoTAB Concentration (M)

Figure 26. Critical aggregation concentrations (CAC) of azoTAB/SDBS (7/3 mol/mol)
mixtures in pH 7.4 PBS buffer at 37 °C determined using the micropolarity indicator dye Nile red
(2.5 μM) to detect the onset of bilayer formation. The arrows denote the CAC values at each
condition expressed as the concentration of azoTAB.
Dynamic light scattering was utilized to measure the size of azoTAB/SDBS vesicles, as
seen in Figure 27. These diameters are similar to those reported for other catanionic systems (e.g.,
vesicle sizes in cetyltrimethylammonium chloride (CTAC)/SDBS mixtures range from 60 to 140
nm for surfactant concentrations < 1 wt %),
296
and are within the size regime suitable for cellular
uptake by endocytosis.
297
Furthermore, these relatively-small sizes suggest catanionic vesicles
may exhibit the enhanced permeability and retention (EPR) effect, which is the aforementioned
and somewhat controversial hypothesis that sub-200 nm particles accumulate at tumor sites rather
than healthy cells.
298
Interestingly, vesicles formed with the shortest 2-azo2 surfactant are
significantly larger than those formed with the two longer azoTAB analogs. Although the
correlation between surfactant tail length and catanionic vesicle size varies,
176
others have found
a similar trend as that reported in Figure 27,
299-301
which may be a result of the greater mismatch
in tail lengths of SDBS and the corresponding azoTAB surfactant.
302
Hydrodynamic diameters of
catanionic vesicles formed upon mixing 2-azo-4 with sodium alkyl sulfates (SCnS, where n = 12

107

corresponds to the common sodium dodecyl sulfate) are shown in Figure 28. Similar trend of
catanionic vesicle size decrease as the hydrophobic tail length increase was observed.
Furthermore, the light-scattering intensity was found to significantly decrease in all samples
following UV exposure, as suggested by the CAC measurements in Figure 26. Specifically, no
vesicles were observed in 2-Azo-2/SDBS system under UV light at 37 C (i.e., the light scattering
intensity was essentially equivalent to that of pure buffer, with I/Ibuffer = 1.05), as 50 M is below
the respective CAC value under UV light. For the 2-Azo-4/SDBS system following UV exposure,
the aggregate size could not be reliably determined as the light-scattering intensity was only
slightly higher than pure pH 7.4 PBS buffer (I/Ibuffer = 2.4), perhaps not surprising as 50 M
azoTAB is just beyond the respective CAC under UV light (40 M azoTAB from Figure 25).
Finally, for the 2-Azo-6/SDBS system, small hydrodynamic diameters of dH = 30 nm – 50 nm
were observed following UV exposure, as 50 μM azoTAB well exceeds the corresponding CAC
value of 20 M from Figure 26.

Figure 27. Hydrodynamic diameters of azoTAB/SDBS (7/3 mol/mol) vesicles over the
course of 50 days as a function of temperature. [azoTAB] = 50 μM in pH 7.4 PBS buffer.

108

40
45
50
55
60
65
70
75
80
6 8 10 12 14 16 18 20
Hydrodynamic Diameter (nm)
Alkyl Chain Length (n)

Figure 28. Hydrodynamic diameters of catanionic vesicles formed from 7/3 mixtures of 2-
azo-4 and various sodium alkyl sulfates, SCnS, (e.g., SC12S corresponds to sodium dodecyl sulfate,
SDS).
The long-term stability of azoTAB/SDBS catanionic vesicles was also demonstrated in
Figure 27 at several temperatures. Only modest changes, generally within the error, in the vesicle
sizes were observed over the 50-day time period. Similarly, the size of 2-Azo-4/SDBS catanionic
vesicles have been observed to be stable for over two years,
41
while catanionic vesicles made with
an azobenzene-based surfactant and SDS had negligible size variation over three months.
303
This
long-term stability could be advantageous for various drug-delivery applications by providing a
long shelf life.
304

Small-angle neutron scattering (SANS) measurements were performed on azoTAB/SDBS
(7/3 mol/mol) mixtures, as seen in Figure 29. The scattering intensities for each mixtures exhibit
a Q
-2
dependence in the low-Q region, which indicates scattering from planar structures.
305
Such
would be the case for the scattering from the bilayers of relatively-large vesicles, which is
consistent with the vesicle sizes determined with dynamic light scattering above. Specifically,

109

SANS probes length scales L = 2/Q. Thus vesicles with diameters greater than approximately
2/Qmin = 125 nm would appear flat (hence the Q
-2
dependence), while smaller vesicles would
display oscillations from this Q
-2
dependence due to the finite vesicle size (i.e., boundary losses).
187

For example, oscillations near Q  0.01 Å
–1
(L  600 Å) are seen in the 2-azo-4/SDBS and 2-azo-
6/SDBS scattering data (consistent with dH  80 nm obtained for these systems in Figure 27), while
the 2-azo-2/SDBS system (dH  130 nm) exhibits a more monotonous Q
-2
decay.

Figure 29. Small-angle neutron scattering (SANS) data of azoTAB/SDBS (7/3 mol/mol)
vesicles under visible light at a total surfactant concentration of 0.1 wt %. The slope of –2 indicates
scattering from large, locally-flat entities (i.e., vesicle bilayers). Oscillations about this slope are
a result of finite vesicle size. Inset: Modified Guinier plots with corresponding bilayer thicknesses,
with solid data points corresponding to those used in the linear fits.
Modified Guinier plots of ln(IQ
2
) versus Q
2
are shown in the inset of Figure 29 (see
Equation 23). In this manner, the thickness radius of gyration Rt and the bilayer thickness dL =
Rt√12 were calculated from the slope of each plot. The Guinier approximation is generally valid
182

110

for QRt < 1, equating to Q
2
< 0.01 Å
-2
for a bilayer thickness of 30 Å. Deviations from the linear
Guinier behavior at low Q are due to the aforementioned effects of finite vesicle size. As expected,
the bilayer thickness steadily increases as the length of the azoTAB hydrophobic tail increases,
ranging from 30 Å to 36 Å. These dL values are approximately the sum of the tail lengths of SDBS
(15 Å) and the corresponding 2-azo-2 (15 Å), 2-azo-4 (18 Å), or 2-azo-6 (20 Å) surfactant, each
estimated as 82% (i.e., sin(109.5/2)) of the sum of the bond lengths,
306
suggesting only modest
interdigitation. To a first approximation, thicker bilayers would be expected to provide better hosts
for large hydrophobic drugs (see below) and result in less permeable membranes.
The encapsulation efficiency of paclitaxel in azoTAB-SDBS catanionic vesicles (i.e., the
percentage of added paclitaxel that can be solubilized in 50 μM azoTAB/21.3 μM SDBS vesicle
bilayers) is shown in Figure 30 as a function of the bulk paclitaxel/surfactant mass ratio. A
maximum encapsulation efficiency of ~80% can be achieved, which appears to be limited by the
aqueous solubility of paclitaxel (reported to range between 0.3 M – 0.5 M).
307-309
For example,
at the lowest bulk paclitaxel concentration of 2.1 M, approximately 1.6 M paclitaxel is
solubilized within the vesicles, with the remaining 0.5 M apparently dissolved in water. As the
paclitaxel concentration is increased, the vesicle bilayers appear to become saturated at ~3 M
paclitaxel. This corresponds to a paclitaxel loading capacity (i.e., the amount of encapsulated
paclitaxel divided by the total amount of surfactant) of ~6% (wt/wt) or ~4% (mol/mol). For
comparison, the loading capacity of paclitaxel in polymersomes and liposomes generally ranges
between 0.5% and 7% (mol/mol).
310-312

111

0
20
40
60
80
100
2.1 4.1 8.2 16.4 33 66 130
2-azo-2/SDBS
2-azo-4/SDBS
2-azo-6/SDBS
Encapsulation Efficiency (%)
1:16 1:8 1:4 1:2 1:1 1:0.5 1:0.25
Paclitaxel/Surfactant (wt/wt)
Paclitaxel Bulk Concentration (M)
2.9
3.4
3.1
3.4
3.4
2.3
3.3
3.0
2.3
3.9
2.3
2.1
3.3
2.3
2.3
3.6
1.7
2.3
1.6
1.6
1.7

Figure 30. The percentage of bulk paclitaxel capable of being solubilized within
azoTAB/SDBS (50 μM/21.4 μM) vesicles as a function of the paclitaxel/surfactant (wt/wt) ratio.
The upper abscissa indicates the bulk paclitaxel concentration, while the numbers above each bar
correspond to the paclitaxel concentration (µM) solubilized within the vesicles.
Interestingly, increasing the length of the hydrophobic tail of azoTAB at low paclitaxel
concentrations results in slightly higher paclitaxel encapsulation. This could be a result of the
increased hydrophobicity of the azoTAB analogs with increasing tail length, allowing for greater
partitioning of the hydrophobic drug into the bilayers (see also the larger blueshifts in Nile red
emission with increasing azoTAB tail length in Figure 26, indicating a more hydrophobic bilayer).
This effect could also originate from the lower CAC values found with the longer azoTAB analogs,
which would cause a greater amount of surfactant to participate in bilayer formation. For example,
for azoTAB to a first approximation this amount is presumably the bulk concentration of 50 M
minus the corresponding CAC (but note the zeta potential discussions below). Furthermore, the
increase in bilayer thickness with increasing azoTAB tail length (from 30 Å to 36 Å in Fig. 29)

112

would be expected to lower the steric hindrance of paclitaxel solubilization in the bilayers (e.g.,
paclitaxel adopts a T-shape when bound to microtubules, spanning ~10 Å in both the horizontal
and vertical directions).
313
Similarly, higher encapsulation of hydrophobic drugs has been reported
with an increase of hydrophobic tail length in liposomes.
300
Curiously, at high paclitaxel
concentrations the shortest azoTAB analog (2-azo-2) provides the highest paclitaxel
encapsulation, although admittedly this is only a small portion of the total drug available. Overall,
the data in Figure 30 suggest that a paclitaxel/surfactant ratio of 1/8 appears to combine both
efficient drug encapsulation and high loading capability and, thus, was selected for co-
encapsulated delivery experiments below.
The ability of azoTAB/SDBS vesicles to encapsulate Bcl-2 siRNA was assessed using
agarose gel electrophoresis with various N/P (i.e., azoTAB-to-nucleotide) ratios, as shown in
Figure 31a. Here the siRNA concentration was kept constant at 30 nM to allow uniform imaging,
while the azoTAB and SDBS concentrations were varied from 50 μM – 800 μM and 21.4 μM –
342.4 μM, respectively. Naked siRNA would be expected to migrate through the gel as opposed
to siRNA encapsulated in relatively large azoTAB/SDBS vesicles with a net positive charge (i.e.,
constituted with a 7/3 ratio of cationic/anionic surfactant; see also the zeta potential measurements
below). Indeed, a complete mobility shift of Bcl-2 siRNA was achieved at N/P = 4.8 (2-azo-
6/SDBS), 9.5 (2-azo-4/SDBS), or 19 (2-azo-2/SDBS), indicating full encapsulation into the
corresponding vesicles. The difference between the bulk N/P values required for full encapsulation
is likely a result of the lower CAC values of the longer azoTAB analogs (5 μM, 15 μM, and 25
μM for 2-azo-6/SDBS, 2-azo-4/SDBS, and 2-azo-6/SDBS vesicles, respectively; see Figure 26)
relative to the surfactant concentrations employed in Figure 31a, meaning that more catanionic
vesicles exist in systems comprised of the longer azoTAB analogs. Similarly, Xu et al. reported

113

that an increase of the hydrophobic chain length of cationic alkyltrimethylammonium bromide
surfactants mixed with anionic sodium laurate led to enhancements in DNA encapsulation in
catanionic vesicles.
314
The encapsulation of hydrophilic materials into vesicles is, in general,
affected by the bilayer thickness, hydrophobicity of the vesicle components, size (multilamellar >
large unilamellar > small unilamellar vesicles), and charge (positive > negative > neutral).
315-317

For reference, results nearly identical to Figure 31 were obtained with siRNA-paclitaxel co-
encapsulation (not shown) at a 1/8 paclitaxel/surfactant ratio, indicating that the catanionic vesicles
tolerate the presence of the hydrophobic drug.

Figure 31. (a) Electrophoretic mobility shift assays of siRNA upon the addition of
azoTAB/SDBS (7/3 mol/mol) surfactant at a fixed siRNA concentration (30 nM) to achieve
various N/P (azoTAB-to-nucleotide) ratios. (b) Zeta potentials of siRNA-loaded and siRNA-
paclitaxel co-loaded vesicles upon the addition of siRNA at fixed surfactant concentration
([azoTAB]/[SDBS] = 50 μM/21.4 μM) and paclitaxel/surfactant ratio (1/8 wt/wt). Arrows indicate
the approximate points where complete mobility shifts were attained in (a). The inset replots the
2-azo-6
/SDBS
[azoTAB] = 0 50 100 200 400 600 800 M
2-azo-2
/SDBS
2-azo-4
/SDBS
N/P = 0 4.8 9.5 19 38 57 76
-60
-40
-20
0
20
40
0 20 40 60 80
Zeta Potential (mV)
N/P ratio
8
(b) [azoTAB] = 50 M, [SDBS] = 21 M
2-azo-2/SDBS
w/ paclitaxel
2-azo-4/SDBS
w/ paclitaxel
w/ paclitaxel
2-azo-6/SDBS
(a) [siRNA] = 30 nM
0 0.25 0.5
-50
-25
0
( − 

) (mV)
P/N'

114

zeta potential change as a function of the P/N ratio, where N is the azoTAB concentration minus
the corresponding CAC.
The results of zeta potential measurements for azoTAB/SDBS vesicles as a function of
siRNA concentration with and without co-loaded paclitaxel are shown in Figure 31b. In the
absence of siRNA and paclitaxel, zeta potentials of 37.2 mV, 19.5 mV, and –13.2 mV were found
for vesicles made with 2-azo-6, 2-azo-4, and 2-azo-2, respectively. The positive surface charge for
2-azo-6/SDBS and 2-azo-4/SDBS vesicles is as expected given the 7/3 cationic/anionic surfactant
molar ratio. Somewhat surprisingly, however, the 2-Azo-2/SDBS vesicles were observed to be
negatively charged despite this 7/3 bulk ratio. Here it is important to note that bilayer
concentration in cationic vesicles does not necessarily reflect the bulk concentration for several
reasons. First, decreasing the azoTAB hydrophobicity by decreasing the hydrocarbon tail length
would be expected to decrease the partitioning of azoTAB into the bilayers, which seems to be the
trend in Figure 31b. For example, prior studies have shown that in catanionic vesicles formed with
cationic cetylpyridinium chloride (CPC) and four different sodium N-alkanoyl-L-alaninate (C8 –
C14) anionic surfactants, increases in the alkanoyl chain length led to decreases of the zeta potential
from –19.2 mV to –37.2 mV (i.e., more of the anionic surfactant partitioning into the bilayers with
increasing hydrophobicity).
318
Second, the bilayers in 2-azo-4/SDBS vesicles were previously
found to be composed of SDBS mixed primarily with trans azoTAB as a result of the steric
hindrance of incorporating the bent cis isomer into bilayers,
41
which given the 75/25 trans/cis
photostationary state under visible light would be expected to make the vesicles less positively
charged that the bulk ratio suggests. Finally, in general it has been observed that the surfactant in
excess in catanionic mixtures often exhibits orders of magnitude higher free surfactant
concentrations than the minor component.
319
Combined, these effects give rise to negatively-

115

charged 2-azo-2/SDBS vesicles even at a bulk 7/3 cationic/anionic surfactant ratio. This
consequence has potential benefits, as positively-charged carriers have been found to interact with
negatively-charged serum proteins leading to rapid clearance from the body.
320, 321
Furthermore,
negatively-charged nanoparticles have been reported to more effectively pass through cancer
cells,
322
potentially through clustering of the nanoparticles on cationic sites of the plasma
membrane.
266
Not surprisingly, the situation is far more complex than implied by the brief
discussion above. For example, the cytotoxicity of as the polyethyleneimine (PEI), one of the
more effective delivery vectors, has been found to correlate with the magnitude of the positive
charge.
323
Even more striking, many others have reported that negative liposomes are less stable
when injected into the blood circulation.
324
Given these uncertainties, being able to control the
charge of azoTAB/SDBS vesicles with changes in composition (Figure 31b) or light conditions
41

could provide a unique way to optimize the effectiveness and safety of drug/gene delivery.
As the siRNA concentration initially increases (i.e., as the N/P ratio decreases at the
constant 50 μM azoTAB concentration in Figure 31b), the surface charge at first slowly decreases
until an N/P generally below 19 is reached, beyond which a steep decrease in the zeta potential is
observed. Specifically, the arrows in Figure 31b correspond to the N/P ratios in Figure 31a where
siRNA was observed to become completely entrapped within vesicles (i.e., a complete mobility
shift), which should equate to regions where the complexes are rapidly becoming more negative.
Note that the N/P ratios in the two figures are not directly comparable since the experiments were
performed at constant siRNA concentration (480 nM) or constant surfactant concentration (50 mM
azoTAB), respectively, but they do generally occur near the same range (e.g., [siRNA] = 120 nM
at N/P = 19 in Figure 31b). It should also be noted that following the standard protocol of plotting
the zeta potential versus the N/P ratio in Figure 31b gives the misleading notion that siRNA is not

116

binding with vesicles at low siRNA concentrations (i.e., high N/P values). For example, after
correcting for the concentration of surfactant free in solution (i.e., the CAC, with now N =
[azoTAB] – CAC), the zeta potential can then be replotted versus the P/N ratio (commonly referred
to as r = [base]/[surfactant] for nucleic acids mixed with cationic surfactant) as shown in the inset
of Figure 31b. Here the zeta potential is found to indeed have a linear response to siRNA addition
up to P/N = 1 (akin to comparing plots of 𝑓 (𝑥 )=−1 𝑥 ⁄ and 𝑔 (𝑥 )=−𝑥 ). For comparison, when
nucleic acids are mixed with unilamellar cationic vesicles, the nucleic acids are frequently
intercalated within a multilamellar lipid structure,
325
while the zeta potential has been found to be
a linearly decreasing function of the r value.
314

The presence of paclitaxel (1/8 paclitaxel/surfactant bulk ratio) is seen to have little effect
on the zeta potential of 2-azo-4/SDBS and 2-azo-6/SDBS vesicles, both with and without siRNA,
as expected since the drug is nonionic. Conversely, the surface charge of 2-azo-2/SDBS vesicles
is markedly less negative in the presence of paclitaxel, which suggests that 3 mol % drug in bilayer
(based on the loading capacity in Figure 30b) causes a greater fraction of the 2-azo-2 surfactant to
partition into the bilayer relative the paclitaxel-free case. This agrees with the fact that while
paclitaxel is hydrophobic, the drug exhibits a very low solubility (< 0.1 wt %) in straight-chain
hydrocarbons such as hexane and mineral oil,
326
and would thus be expected to exhibit a low
solubility in the C12 chains of SDBS alone. Finally, DLS found no marked changes in the vesicle
size with either siRNA or paclitaxel loading, with diameters ranging from 60 nm – 130 nm with a
polydispersity index (PDI) < 0.45, similar to Figure 26. Thus, based on the results in Figure 31,
the in vitro experiments below were performed at N/P = 19 with a 1/8 ratio of paclitaxel/surfactant,
which allows complete entrapment of siRNA in either positive (2-Azo-4/SDBS and 2-Azo-
6/SDBS) or slightly negative (2-Azo-2/SDBS) vesicles.

117

The UV light-induced release of siRNA and paclitaxel encapsulated in azoTAB/SDBS
catanionic vesicles was investigated by monitoring (via UV-vis and fluorescence spectroscopy,
respectively) the fraction of each species remaining within spun-down vesicles or released into the
supernatant, as shown in Table 4. Under visible light, the percentage of the bulk siRNA and
paclitaxel concentrations encapsulated in vesicles increased with increasing length of the azoTAB
analog, as expected based on the relative zeta potentials and CACs, respectively. Following UV
illumination, a substantial portion of each encapsulated species was then released into the
supernatant. This can be understood by comparing the employed azoTAB concentration (c = 50
M) to the relative CAC values obtained under UV light, namely 50 μM for 2-Azo-2/SDBS, 40
μM for 2-Azo-4/SDBS, and 20 μM for 2-Azo-6/SDBS (versus 25 μM, 15 μM, and 5 μM under
visible light). In other words, UV illumination is expected to result in complete or nearly-complete
vesicle rupture for the 2-Azo-2 and 2-Azo-4 systems, while the majority of vesicles should remain
in the 2-Azo-6 system. Note, however, that the release percentages in Table 4 do not match the
hypothetical values predicted by assuming that the initial and final vesicle concentrations correlate
simply with (c – CACvis) and (c – CACUV), which would give release percentages of 100%, 71%,
and 33%, respectively. Rather, the release values in Table 4 better correlate with the percentage
loss of trans surfactant upon going from visible → UV illumination, which based on the
aforementioned photostationary states of 75/25 and 10/90 trans/cis, respectively, gives (0.75 –
0.1)/0.75 = 87% loss. This suggest that both siRNA and paclitaxel may be incorporated within
assemblies composed primarily of trans azoTAB regardless of the light conditions. This is in
agreement with our previous observation that azoTAB/SDBS bilayers are comprised of a large
fraction of trans azoTAB even under UV illumination (e.g., evidence of H-aggregates, presumably
due to π-π stacking of the planar trans isomers).
41
Additionally, cryo-TEM images captured during

118

UV-induced vesicle rupture have shown detached bilayer patches that persist for several minutes
after light exposure and are likely comprised of locally H-aggregated regions within the bilayer.
35

Table 4. Encapsulation and photo-initiated release percentages of Bcl-2 siRNA (N/P = 4.8)
and paclitaxel (1/8 drug/surfactant wt/wt) loaded in azoTAB/SDBS (50 μM/21.4 μM) vesicles at
37 C in pH 7.4 PBS buffer.
2-azo-2/SDBS 2-azo-4/SDBS 2-azo-6/SDBS
siRNA
encapsulation (vis)
encapsulation (UV)
release (vis → UV)

22% ± 2%
3% ± 1%
85% ± 3%

60% ± 3%
12% ± 2%
81% ± 3%

89% ± 2%
27% ± 3%
70% ± 3%

paclitaxel
encapsulation (vis)
encapsulation (UV)
release (vis → UV)

58% ± 3%
10% ± 2%
84% ± 3%

75% ± 4%
12% ± 2%
84% ± 2%

84% ± 2%
22% ± 2%
74% ± 1%

The cytotoxicity of siRNA and paclitaxel co-loaded vesicles was investigated using the
XTT cell viability assay, as shown in Figure 32. Cancer cells were incubated for 24 hours with
various final concentrations of vesicles (at a 7/3 azoTAB/SDBS molar ratio) that were either
empty, siRNA-loaded (N/P = 19), paclitaxel-loaded (1/8 drug/total surfactant mass ratio), or
siRNA-paclitaxel co-loaded under visible or UV light. The dotted lines in each plot denote the
CAC values under visible and UV light (i.e., at concentrations below these values, dilution of the
vesicles into the cell medium is expected to result in vesicle dissociation). For the empty vesicles
or vesicles loaded with only siRNA, viabilities of nearly 100% were observed at low azoTAB
concentrations, while the viabilities dropped to 80% at ~70 M azoTAB and 20% at ~700 M
azoTAB. Furthermore, illumination with a low-intensity UVA light source was found to have no
marked effect on the observed viabilities in these samples. These results using MDA-MB-231

119

breast cancer cells are consistent with our previous study with 2-Azo-4/SDBS vesicles in NIH 3T3
cells,
35
demonstrating an upper limit of IC90  50 M for the safe use the catanionic vesicles.

0
20
40
60
80
100
0.1 1 10 100 1000
empty (vis)
empty (UV)
siRNA
paclitaxel
siRNA+PTX (vis)
siRNA+PTX (UV)
0.01 0.1 1 10
Cell Viability (%)
AzoTAB Concentration (M)
Paclitaxel Concentration (M)
2-azo-2/SDBS
0
20
40
60
80
100
0.1 1 10 100 1000
0.01 0.1 1 10
empty (vis)
empty (UV)
siRNA
paclitaxel
siRNA+PTX (vis)
siRNA+PTX (UV)
Cell Viability (%)
Paclitaxel Concentration (M)
AzoTAB Concentration (M)
2-azo-4/SDBS
0
20
40
60
80
100
0.1 1 10 100 1000
0.01 0.1 1 10
empty (vis)
empty (UV)
siRNA
paclitaxel
siRNA+PTX (vis)
siRNA+PTX (UV)
Cell Viability (%)
2-azo-6/SDBS
Paclitaxel Concentration (M)
AzoTAB Concentration (g/mL)

Figure 32. Viability of MDA-MB-231 human breast cancer cells after treatment with
empty, siRNA-loaded (N/P = 19), paclitaxel-loaded (1/8 wt/wt), and siRNA-paclitaxel co-loaded
azoTAB/SDBS vesicles (7/3 mol/mol) under visible and UV light.
The presence of paclitaxel within azoTAB/SDBS vesicles, either alone or co-loaded with
siRNA, was found to result in more complex viability profiles depending on the conditions. For
reference, IC50 values for cells treated with paclitaxel for 72 h are generally in the 10 nM range,
while the cytotoxicity at a fixed time (e.g., 24 h) does not vary with paclitaxel concentrations
beyond 50 nM.
327, 328
This is consistent with the data for cells exposed to paclitaxel for 24 h in
Figure 32, which in general appear shifted downward by a nearly constant amount relative to the
paclitaxel-free samples except at low ( 0.1 M) paclitaxel concentrations. The lower viabilities
observed for the 2-azo-4 and 2-azo-6 systems at these low concentrations of paclitaxel (i.e., below
the aqueous solubility of ~0.4 M) and surfactant (i.e., below the CAC values) suggests that these
more hydrophobic azoTAB analogs may be forming small, non-vesicle aggregates with paclitaxel

120

in this concentration range,
320, 329
similar to the self-aggregation of dye molecules in aqueous
solutions that begin to form at micromolar concentrations.
330
Furthermore, the viability is
generally slightly lower for the paclitaxel-containing samples when exposed to UV light,
particularly for the 2-azo-4 and 2-azo-6 systems at azoTAB concentrations between the visible-
and UV-light CAC values where the illumination conditions would be expected to have the largest
effect on the vesicle concentration. This effect is particularly pronounced for the 2-azo-4/SDBS
system, where the cell viability consistently drops by ~50% when exposed to UV versus visible
light, suggesting this system could provide localized cytotoxicity only in regions illuminated with
UV light.
To achieve siRNA-based gene knockdown, the siRNA must first penetrate the cell
membrane and then escape from the vesicle complex. Thus, the cellular uptake of siRNA delivered
with catanionic vesicles was examined using Cy5.5-labeled Bcl-2 siRNA via flow cytometry, as
shown in Figure 33. Uptake values in excess of the control of 36.8%, 33.2% and 43.9% were
obtained for 2-azo-2/SDBS, 2-azo-4/SDBS and 2-azo-6/SDBS vesicles, respectively, while a
value of 8.1% was obtained for naked siRNA. Electrostatic interactions between vesicles and cell
membranes could facilitate uptake via endocytosis, which could explain the slightly higher siRNA
internalization for 2-azo-6/SDBS vesicles due to the more positive zeta potential observed in
Figure 33b. Note, however, that while cancer cell membranes are highly negatively charged and
thus positively-charged nanoparticles can be more readily engulfed, there are a small number of
cationic locations on the membranes where negatively-charged nanoparticles can bind.
331
For
example, it has been reported that an increase in the absolute value of the surface charge of
carboxymethyl chitosan-grafted (negatively charged) and chitosan hydrochloride-grafted
(positively charged) nanoparticles was correlated with 10% – 30% improvement in uptake in a

121

range of cell lines.
332
Similarly, nanoparticle uptake in macrophages correlates with the increasing
surface charge densities, either positive or negative.
333
Furthermore, the higher uptake of siRNA
using 2-azo-6/SDBS vesicles in Figure 33 may be a result more vesicles being present in this
system due to the lower CAC value of this surfactant pair compared to the 50 M azoTAB
concentration employed in Figure 33, as discussed above. To examine the intracellular
accumulation of siRNA and paclitaxel co-loaded in azoTAB/SDBS vesicles, confocal laser
scanning microscopy was employed as shown in Figure 34. Here the hydrophobic dye Coumarin
6 (C6), which is frequently used to model poorly water-soluble drugs,
334
was used as a fluorescent
mimic of paclitaxel, while DAPI (4’,6-diamidino-2-phenylindole) was used as a fluorescent stain
for the nuclei of the MDA-MB-231 cells. The appearance of the merged yellow (i.e., red plus
green) fluorescence in relation to the blue nuclei indicates successful co-localization of siRNA and
the hydrophobic-drug mimic in the cytoplasm of breast cancer cells with azoTAB/SDBS vesicles.
Some of the C6 dye appears to have also migrated to the nucleus in the mint green (green + blue)
regions, as noted by others.
335, 336

122

Figure 33. Uptake of Bcl-2 siRNA encapsulated in azoTAB/SDBS (50 μM/21.4 μM)
vesicles in MDA-MB-231 human breast cancer cells. Numbers correspond to the percentage of
cells exhibiting fluorescence from Cy5.5-labeled Bcl-2 siRNA.

123

Figure 34. Confocal fluorescence images of MDA-MB-231 human breast cancer cells
following treatment with Cy5.5-conjugated Bcl-2 siRNA (red) at N/P = 4.8 and 4 μM Coumarin 6
hydrophobic dye (green) co-loaded in azoTAB/SDBS (50 μM/21.4 μM) vesicles for 4 hours. The
nuclei were stained with DAPI (blue).

Endosomal escape of siRNA was detected after 60 seconds of UV light exposure
subsequent to transfection. After the first minute, the Cy5.5 conjugated-siRNA diffused into the
cells as shown in Figure 35.

124

Figure 35. Endosomal escape of siRNA following to transfection of MCF-7 cells for 4
hours with siRNA loaded 2-azo-6/SDBS vesicles (N/P=19) and 60 seconds exposure of 358 nm.
In order to demonstrate in vitro gene silencing, MDA-MB-231 cells were treated with Bcl-
2 siRNA-loaded azoTAB/SDBS vesicles for four hours, followed by UV illumination for 10
minutes. A surfactant concentration of 50 M azoTAB at a 7/3 azoTAB/SDBS molar ratio was
employed, chosen to allow release of siRNA from the vesicle carrier upon UV illumination (recall
the CAC values in Figure 26) and to minimize cytotoxicity (see Figure 32b). A western blot
representative (Figure 36) and the average of five western blot results in Figure 37 demonstrate
the effective knockdown of the Bcl-2 gene using azoTAB/SDBS vesicles, while the cells exposed
simply to naked siRNA continue to express the Bcl-2 protein. In other words, Bcl-2 siRNA
encapsulated in azoTAB/SDBS catanionic vesicles appears capable of passing through breast

125

cancer cell membranes through endocytosis, while subsequent UV illumination leads to
disassociation of the carrier and thus release of siRNA into the cytoplasm. In truth, the situation
is somewhat more complicated than the latter statement implies, since the endocytic pathway
results in encapsulation of the carrier in an endosome. It is the release of siRNA from these
endosomes and not the carrier per se that must be achieved for gene silencing to occur.
Transfection studies with conventional cationic lipids indicate that significant “lipid mixing”
occurs between the carrier vesicles and the endosomes.
337-339
Following this intracellular mixing,
azoTAB has been shown to allow UV-induced endosome disruption,
35
with effects similar to those
seen in photochemical internalization, where an amphiphilic photosensitizer is added to the
delivery vehicle to produce reactive oxygen species.
340
While the distinctions between UV-
induced carrier and endosome disruption are subtle, they do deserve note in context of the observed
gene silencing. Regardless, the results in Figure 37 clearly show Bcl-2 siRNA release into the
cytoplasm, which ultimately results in gene silencing.

Figure 36. Western blot assay (run #1) following transfection of MDA-MB-231 human
breast cancer cells with naked Bcl-2 siRNA and siRNA loaded in azoTAB/SDBS (50 μM/21.4
μM, N/P = 19) vesicles for 4 hours. β-actin was used as an equal loading control.

126

0
20
40
60
80
100
Percentage of Bcl-2 /-actin
control siRNA 2-azo-2/SDBS 2-azo-6/SDBS
+UV -UV +UV -UV +UV -UV +UV -UV
*
**

Figure 37. Quantification of Western blots showing the effect of Bcl-2 siRNA loaded in
2-azo-2/SDBS and 2-azo-6/SDBS (50 μM/21.4 μM, N/P = 19) vesicles on MDA-MB-231 human
breast cancer cells with and without UV exposure. The expression of Bcl-2 was normalized to
that of the equal loading control β-actin. Western blots are shown in Figures S6-S10. n = 4 except
2-azo-2/SDBS -UV, where n = 3. * p = 0.0282, ** p = 0.016.

4.5. Conclusions
Photoresponsive catanionic vesicles have been shown capable of co-delivery of siRNA and
paclitaxel into MDA-MB-231 human breast cancer cells. By operating at concentrations
intermediate the respective critical aggregation concentrations under visible and UV light, the
photo-initiated vesicle → free monomer transition can be utilized as a means of releasing these
therapeutics from the vesicles following cellular uptake. Increasing the tail length and thus
hydrophobicity of the azoTAB photoresponsive surfactant allows the resulting vesicle size (from

127

140 nm down to 60 nm), charge (from slightly negative to highly positive), and bilayer thickness
(from 30 Å – 36 Å) to be tuned. Furthermore, vesicles formed from longer azoTAB analogs
showed higher siRNA and paclitaxel encapsulation and were more effective at transfection. Light-
controlled co-delivery of siRNA and a hydrophobic drug with azoTAB-based vesicles could
potentially provide a rapid, localized, and safe method of treating complex diseases with minimal
side effects.
Acknowledgements. We acknowledge the support of the National Institute of Standards
and Technology, U. S. Department of Commerce, in providing the neutron research facilities used
in this work. This material is based upon work supported by the National Science Foundation
under Grant 1758225. Any opinions, findings, and conclusions or recommendations expressed in
this material are those of the author(s) and do not necessarily reflect the views of the National
Science Foundation. We would also like to acknowledge support from USC’s Research
Enhancement Fellowship program. We thank Dr. Kwang-Jin Kim for help with zeta potential and
light scattering measurements. We would also like to gratefully acknowledge Prof. Noah
Malmstadt and Mr. Ahmed Elbaradei for their assistance with confocal laser microscopy.

128

CHAPTER 5: VISIBLE LIGHT-INDUCED CRYSTALLIZATION OF
CATANIONIC VESICLES AND CELLULASE ENZYME ENCAPSULATION
APPLICATION

5.1. Abstract
Photoswitchable vesicle formation and crystallization in aqueous mixtures of light-
responsive cationic surfactant (4-ethyl-4’(trimethylamino-butoxy) azobenzene bromide, azoTAB)
and commercially available anionic surfactants (sodium decyl sulfate, SDecS and sodium 4-
octylbenzenesulfonate, SOBS) are studied. The hydrophobicity similarity of azoTAB with SdecS
and SOBS lead liquid and solid crystal formation at a wide range of cationic to anionic surfactant
ratio under visible light. On the other hand, mostly vesicles with a narrow range micelle and
multilamellar vesicles region are observed under UV light. Vesicle to crystal transition can be
controlled by light wavelength from fast 3-4 hours to slow 3-4 days. The unique light-induced
crystal to vesicle transition with UV light illumination is achieved with 70-80%. Furthermore, the
encapsulation of cellulase enzymes on enzyme reaction is studied and the reaction rate of cellulase
can be controlled with light wavelength.

5.2. Introduction
Catanionic mixtures of positively-charged (cationic) and negatively-charged (anionic)
surfactants have been widely studied
341
due to the synergistic behavior and create various
microstructures at concentrations orders of magnitude lower than the pure-component critical
micelle concentrations. These microstructures can be micelles (spherical, rod-like, disc-like),
vesicles, multilamellar, and liquid-crystalline structures based on the cationic-to-anionic surfactant

129

molar ratio, overall surfactant concentration, alkyl chain length, temperature, and pH.
94,95,104

Generally, equimolar concentrations of catanionic solutions precipitate or form lamellar structures
owing to the offsetting positive and negative charges. Additionally, catanionic salt or crystalline
phase is formed if the tail lengths of surfactants are similar. These crystals are packed into solid or
liquid crystalline form.
The catanionic vesicles form spontaneously as a result of favorable electrostatic and
hydrophobic interactions between the oppositely-charged surfactants without the need for external
energy such as shearing. Hence, the vesicles, once formed, are infinitely stable as the system is in
thermodynamic equilibrium.
102,103
Catanionic vesicles, also known as “equilibrium vesicles”, have
received special interest in nanotechnology, chemical and pharmaceutical applications such as
membrane protein folding/solubilization,
32
gene therapy,
35
drug delivery, separation,
97
vaccines,
98

cosmetic probe and molecule encapsulation
99
as a result of the long-term stability,
102
cost-
effectiveness,
101
and ease of preparation
100
of these vesicles. Vesicles can encapsulate hydrophilic
compounds in the aqueous core while hydrophobic materials can be solubilized in vesicle bilayers.
Encapsulation of biomolecules, therapeutics, peptides, proteins, and drugs is important to keep the
biological activity of the encapsulated materials until it reaches its target.
Triggered control of the vesicles by external stimuli has received significant attention in
controlled drug or therapeutics delivery. Azobenzene moiety is the commonly used light
responsive molecule in vesicles for external controllability.
89
Vesicle disruption and reconstruction
of aqueous mixture of azobenzene-based cationic surfactant AZTMA and SDBS anionic
surfactants was able to be controlled by light illumination.
342
Aqueous solutions of azoTAB
(azobenzene trimethylammonium bromide, a photoresponsive cationic surfactant) mixed with the
conventional anionic surfactant sodium dodecyl benzene sulfonate (SDBS) allows reversible

130

vesicle  micelle, vesicle  lamellar, and vesicle  free–monomer transitions to be initiated
with light.
41
In previous studies, the photoreversible vesicle-to-micelle transition were used to
control membrane protein folding (i.e., folded-to-unfolded, respectively),
32
while the vesicle-to-
monomer transition was utilized for gene delivery via light-triggered release of DNA,
35
siRNA
and paclitaxel drug
96
from the vesicle once the carrier passed the cell membrane.
In this work, catanionic solutions of a photoresponsive cationic surfactant (4-ethyl-
4’(trimethylamino-butoxy) azobenzene bromide, azoTAB) with alkyl (sodium decyl sulfate,
SdecS) and alkylbenzene sulfonate (sodium 4-octylbenzenesulfonate, SOBS) anionic surfactants
are examined. AzoTAB undergoes a reversible photoisomerization from a planar and relatively-
hydrophobic trans conformation under visible light to a bent and twisted relatively-hydrophilic cis
structure under UV light illumination, as shown in Scheme 1. As a result of dipole moment
difference of azoTAB and hence photoisomerization of azoTAB, unique reversible crystal to
vesicle switch has obtained by light illumination. Whereas vesicle disruption can be controlled
easily, the reformation of the vesicles would be unfeasible/impractical. On the other hand,
formation and disruption of vesicles reversibly and spontaneously could be achieved using photo
switchable cationic surfactant and conventional anionic surfactants. The unique reversible light-
triggered azobenzene photoisometization effect on vesicle-to-crystal is investigated with
microscopic observations, spectroscopic analysis, light and neutron scattering techniques. The
encapsulation efficiency of the vesicles is explored with endoglucanase (48 kDa), bovine serum
albumin (66 kDa) and β-glucosidase (180-220 kDa) Lastly, encapsulated endoglucanase and β-
glucosidase enzymes in encapsulation of the vesicles and their effect into the vesicle-to-crystal
transition is studied.

131

5.3. Experimental Section
Materials. The photoresponsive azoTAB surfactant of the form shown in Scheme 1 (4-
ethyl-4’(trimethylamino-butoxy) azobenzene bromide, 2-azo-4) was synthesized as previously
reported.
31,33
Briefly, azoTAB is prepared by three sequential steps: azocoupling of 4-ethylaniline
with phenol, alkylation with 1,4-dibromobutane, and finally quaternalization with trimethylamine.
All chemicals for azoTAB synthesis, sodium decyl sulfate and sodium 4-octylbenzene sulfonate
surfactants were purchased from Sigma-Aldrich at the highest purity and used as received. Trans-
to-cis photoisomerization of azoTAB was achieved using an 84 W long-wavelength UV lamp
(XX-15A-Spectroline) unless otherwise stated.
Bilayer thickness and degree of alignment calculation. Small-angle neutron scattering
measurements were performed on the 30-meter NG7 SANS instruments at 25 °C at NIST.
153
A
neutron wavelength () of 6 Å and a detector offset of 25 cm with two sample-to-detector distances
of 1.33 and 7 m were used to obtain a Q-range of Q = (4/)sin(/2) = 0.0049 − 0.55 Å
-1
, where
 is the scattering angle. The net intensities were obtained after deduction of the background and
empty cell (pure D2O) scattering from the raw data, followed by accounting for local differences
in the detector efficiency using an isotropic scatterer (Plexiglas). The units were converted to an
absolute differential cross section per unit sample volume (cm
-1
) with an attenuated empty beam.
The SANS data were reduced and examined with utilizing the Igor Pro (WaveMetrics) software
supplied by NIST.
154
Separate 0.1 wt% azoTAB and anionic surfactant solutions were prepared as
stock solutions. Samples were mixed at cationic/anionic molar ratios of 30/70 and 70/30 at room
temperature. A transportable visible and UV lamp (Spectroline, Model no. XX-15A) was located
inside the sample chamber for in situ (visible-to-UV light and UV-to-visible light) experiments.
The degree of alignment is calculated from the following equations
343
:

132

A
f
(q)=
∫ I(q,Φ) cos(2Φ) dΦ
2
0
∫ I(q,Φ)dΦ
2
0
(25)
Additionally, equation 26 was used to compare the degree of alignment results
344
:
A

(q)=
𝐼 𝐻 − 𝐼 𝑉 𝐼 𝐻 + 𝐼 𝑉 (26)
IH was calculated with drawing two lines 5° above and below the horizontal line which
intersects with the center of the 2D SANS graphs. Similarly, IH was determined with the 5° above
and below the vertical line which intersects to the center of the SANS pattern.
Hydrodynamic diameter determination. Dynamic light scattering experiments were
performed using a Brookhaven Model BI-200SM instrument (Brookhaven Instrument Corp.)
equipped with a 35 mW HeNe (632.8 nm) laser (Melles Friot, model no. 05-LHP-928), an
avalanche photodiode detector (BI-APD), and a BI-9000AT digital correlator at a scattering angle
or 90° and temperature of 25 °C. Stock 0.1 wt % solutions of azoTAB and anionic surfactants
were prepared and separately passed through 100-nm syringe filters (Anatop) into sample vials,
vortexed for a minute, and then allowed to reach the equilibrium microstructure for one hour under
either visible or UV light. The azoTAB solution was kept either under visible or UV light prior to
mixing with the anionic surfactants. At least eight separate 5-minute runs were collected and
averaged to determine the hydrodynamic diameter at each condition. The data reported are the first
peak present in the size distribution obtained by fitting the correlation function using the non-
negative least squares (NNLS) routine provided by Brookhaven. The data were collected
consequence 5 days and after a month and 6 months for stability experiments.
Bilayer characterization. Absorption experiments were measured using an Agilent model
8453 UV-vis spectrophotometer. Separate 0.05 wt% and 0.1 wt% stock solutions of azoTAB and

133

anionic surfactant were prepared, mixed at molar ratios between 1/99 and 99/1 (cationic/anionic
surfactant), and vortexed for thirty seconds. The azoTAB solution was stored for an hour under
either visible or UV light prior to mixing with the anionic surfactant solutions. Following mixing,
the samples were kept under visible or UV light for an hour, then placed in 0.1 cm cuvettes for
measurement. The samples are measured at the first five days and one month and six months to
detect vesicle stability and crystal formation rate.
Optical and Polarized Microscopy. Catanionic salt phases were identified using an
optical microscope under brightfield and/or polarized light. Morphologies of crystalline phase
were observed under an Olympus IX71 inverted fluorescence microscope equipped with 10x oil-
immersed objective lens and a Hamamatsu digital CCD camera (model no. C4742-95) at room
temperature.
Melting Point Determination of Crystallites. The samples as suspension (after quick
vortex) were loaded 1 mm cuvettes and 0.2 C/min were increased using temperature controller.
When spectrum of azoTAB increased, the temperature kept constant until the whole spectrum has
obtained.
Crystallization Rate Calculation. 0.1 wt% 30-70 azoTAB-SOBS and azoTAB-SdecS
samples were prepared with pre-converted azoTAB-UV solutions. The samples were placed under
UV light for 20 minutes to form vesicles and reach steady state. Subsequent to UV light
illumination, the samples were left under visible light, dark and UV light. Avrami equation
345

(equation 27) is used with the assumption of single crystallization type to determine crystallization
kinetics at different light conditions:
𝑋 (𝑡 )=1−𝑒𝑥𝑝 (−𝑘 𝑡 𝑛 ) (27)

134

X(t) is the relative crystallinity; k is the overall crystallization rate constant and n is the Avrami
exponent. Ln (t) versus ln(-ln(1-X(t))) plots were drawn; the slope gave the Avrami exponent n,
while the intercept gave the natural logarithm of overall crystallization rate constant, ln(k).
5.4. Results
Catanionic vesicles are formed spontaneously azoTAB-UV/SdecS or azoTAB-UV/SOBS
aqueous solutions in both azoTAB-rich and SdecS- or SOBS-rich compositions under UV light.
Upon visible light exposure, azoTA
+
decS
-
and azoTA
+
OBS
-
catanionic salt or crsytals are formed.
The larger difference of alkyl chain length of catanionic pairs has shown higher stability.
346

Additionally, our previous study of azoTAB (18 Å trans isomer) and SDBS (15 Å) catanionic
vesicles were stable up to 2.5 years.
41, 96
On the other hand, when the tail lengths of cationic and
anionic surfactants are similar, the crystalline-like salts are formed.
104
Beginning with the 2-azo-
4/SdecS system, crystalline salts of the neutral azoTA
+
–decS
-
complex were observed under
visible light. It has been previously reported that similar tail lengths of the cationic and anionic
surfactants leads to favorable hydrophobic interactions causing the surfactant tails to pack into a
crystalline lattice.
347,348
Although here the crystallization phenomenon appears more complex, as
the hydrophobic tail lengths of 2-azo-4 (lt = 18 Å) and SdecS (lt = 13 Å) are different. In general,
aromatic rings in surfactant tails lower the hydrophobicity.
105

AzoTAB has 90% cis and 10% trans isomers under UV light.
33
As the samples aged for
days, the catanionic crystals are formed very slowly compared to the visible light (%75 trans and
25% cis) due to the fact that azoTAB trans-cis balance under UV light. Figure 38a and d represent
the catanionic salt formation under visible light for azoTAB-SOBS and azoTAB-SdecS,
respectively. At first, the 350 nm peak appears with higher intensity, then the 350 nm peak
diminishes at the end of 24 hours and 48 hours. Under dark, the catanionic salt formation process

135

is slower compared to the visible light samples as shown Figure 38 b and e. The vesicle peak at
330 nm decreases as the thermal conversion of trans-to-cis happens in 24 hours at dark at room
temperature. When the samples are placed under UV light, the vesicle to-crystal transition is the
slowest as shown in Figure 38 c and f. The Eppendorf tubes at Figure 38g and h illustrate the
catanionic salt formation under different light conditions.

Figure 38. UV-vis spectra of 0.1 wt% 30-70 azoTAB-SOBS and azoTAB-SdecS crystal
formation under visible light (a and d), dark (b and e) and UV light (c and f). (g) and (h) azoTAB-

136

SOBS and azoTAB-SdecS catanionic sample pictures at different times inside of Eppendorf tubes,
respectively.
The Avrami equation (equation 27) was applied to calculated Avrami exponent n and
overall crystallization rate constant k and they are listed in Table 5. The relative crystallinity, X(t)
versus time plots are shown in Figure 39 with the calculated fits. The crystallization rates under
UV light were clearly the slowest compared to the crystallization rate under visible light.

Table 5. Comparison of crystallization rate parameters of azoTAB-SOBS and azoTAB-
SdecS 30-70, 0.1wt% under different light conditions. Avrami exponent, n and overall
crystallization rate constant, k were calculated with Avrami equation (equation 27) and the fits
were shown in Figure 39.

n k R2

azoTAB-SOBS
vis 0.59 0.3372 0.9718
dark 1.47 0.0066 0.9971
UV 0.34 0.0784 0.9695
azoTAB-SdecS vis 0.75 0.3682 0.9726
dark 0.60 0.0599 0.9098
UV 0.30 0.0993 0.9208

137

Similarly, crystals were observed in mixtures of 2-azo-4 and SOBS (lt = 12 Å) at all 2-azo-
4/SOBS molar ratios between 1/99 and 99/1 at 0.05 and 0.1wt%. In both of these systems, a
crystalline precipitate (alongside an upper vesicle phase, confirmed with DLS-Table 6) was
observed on the first day of storage, as seen in Table 6. The coexistence of catanionic crystalline
salts and vesicles is well known,
20
but the long-term stability of these systems is frequently
overlooked. For example, the upper vesicle phase in the above systems were found to not be
thermodynamically (i.e., indefinitely) stable, in contrast to single-phase catanionic vesicles.
Specifically, when the stored samples were re-measured after one month, no vesicles could be
detected in the supernatant with light scattering, a result of progressed crystalline phase separation.
Furthermore, the anionic surfactant-rich samples (i.e., 2-azo-4/SdecS and 2-azo4/SOBS mixtures
0
0.2
0.4
0.6
0.8
1
0 8 16 24 32 40 48
vis
dark
UV
0.1 wt% 30-70 azoTAB-SOBS
X(t)
time(hours)
0
0.2
0.4
0.6
0.8
1
0 8 16 24 32 40 48
0.1 wt% 30-70 azoTAB-SdecS
time(hours)
X(t)
Figure 39. Catanionic salt formation under different light wavelength. The lines show the fits
calculated with Avrami equation (Eqn 27).

138

at molar ratios of 1/99 – 40/60) exhibited a yellowish (i.e., azoTAB-containing) solid phase at the
bottom of the sample tubes, while the upper liquid phase was clear (i.e., containing only miniscule
amounts of azoTAB, with the remaining excess anionic surfactant existing as free monomers).
Conversely, the upper liquid phase in the azoTAB-rich samples (i.e., 2-azo-4/SdecS and 2-
azo4/SOBS mixtures at molar ratios of 60/40 – 99/1) maintained a yellow color, indicating the
continued presence of azoTAB (i.e., at these ratios the anionic surfactant is the limiting reagent for
formation of the neutral crystalline salt). Representative pictures of 30/70 mixtures of 2-azo-
4/SdecS and 2-azo4/SOBS are shown in Figure 40.

Table 6. Hydrodynamic diameters of azoTAB-SOBS and azoTAB-SdecS at 0.1 wt%
concentration. SLC: smectic liquid crystals, NLC: nematic liquid crystals, m: micelle, L: lamellar,
Average of NNLS first peak of 10 different runs are used.

azoTAB-SOBS azoTAB-SdecS

vis UV vis UV
1 99 42+SLC m 52+NLC 47
3 97 87+SLC m 92+NLC 61
5 95 98+SLC 69 92+NLC 77
7 93 95+SLC 112 77+NLC 98
10 90 83+SLC 40 79+NLC 169
20 80 66+SLC 74 80+NLC 167
30 70 65+SLC 98 92+NLC 151
40 60 53+SLC 29-126 90+NLC 62
60 40 63+SLC 123 SLC L
70 30 50+SLC L 93+NLC L
80 20 62+SLC L 81+NLC 76
90 10 151+SLC L 66+NLC 69
93 7 48+SLC 74 89+NLC 49
95 5 26+SLC 51 114+NLC 42
97 3 26+SLC 41 86+NLC 67
99 1 13+SLC 35 62+NLC 46

139

Figure 40. Physical appearance of 0.1 wt% 30-70 and 70-30 azoTAB-SOBS and azoTAB-
SdecS samples.
To understand the composition of azoTAB-SOBS and azoTAB-SdecS catanionic salt, the
crystals were observed under brightfield and polarized microscopy. Disorder of molecules
increases when the phase of the system changes from solid to liquid as shown in Figure 41. Solid
crystals are highly ordered and has single reflection whereas liquid crystals are anisotropic and
hence they can doubly reflect. As a result of this the solid crystals can be seen under brightfield
easily while polarized microscopy with cross polarizers is used to detect liquid crystals.

Figure 41. Schematic presentation of liquid phase, solid and liquid crystal orientations.

140

Oppositely charged surfactant molecules form a variety of self-assemblies in solution with
hydrophobic and electrostatic interactions and they are stable for a long time. Catanionic surfactant
salt is commonly formed at equimolar concentrations and it can be lamellar, liquid crystalline such
as hexagonal, cubic, etc. The shape and size of catanionic salt depends on surfactant molecular
structure such as hydrophobic chains and the head groups. On the other hand, the vesicles are
formed in a wide range of compositions as the asymmetry of hydrophobic tail lengths increased.
20

Generally, symmetric alkyl chains pack tightly into crystalline lattices for a broad ratios of
surfactant compositions.
349
Needle shape solid crystals have been observed at equimolar
concentrations of sodium dodecyl sulfate and n-alkyltrimethylammonium bromide homologs.
350

Additionally, the linear increase of needle-shaped crystals of sodium dodecyl sulfate and n-
alkyltrimethylammonium bromide homologs were reported with 55, 118, 110, 150 and 185 µm
when n=10, 12, 14, 16 and 18 respectively.
350
Similar tail length showed larger crystal length at
equimolar ratio due to the favorable interaction between them.
350
Electrostatic and hydrophobic
forces between the head groups and hydrophobic tails show importance on the self-assembly
formation.
Representative solid, smectic and nematic liquid crystal pictures of azoTAB-SOBS and
azoTAB-SdecS were shown in Figure 42. The collected motifs were compared with the observed
liquid or solid crystals of other surfactants.
351
There are two phases in azoTAB-vis/SdecS or
azoTAB-vis/SOBS, while there are three regions (micelles, vesicles and lamellar phases) under
UV light. Rectangular smectic liquid crystalline pattern has observed under crossed polarizers as
shown in Figure 42. The azoTAB-UV/SdecS or azoTAB-UV/SOBS vesicles were observed as
small dots by light microscopy due to their small size (80-100nm). In conclusion, aqueous mixtures
of cationic azoTAB and anionic SdecS or SOBS surfactants represented liquid and solid crystals

141

under visible light while micelles, vesicles, lamellar phases were observed with UV light
illumination. The nano to micrometer scale self-assemblies were formed as a result of
hydrophobicity, geometry of the hydrophobic chain of cationic and anionic surfactants and the
attractive forces between the head groups. Similarly, catanionic salt formation of equal
hydrophobic tail length oppositely charged surfactants; DTAB (C12) and SDS (C12) were reported
with a broad range composition.
346
As a result of charge neutralization of oppositely charged
surfactants, catanionic precipitation or lamellar phases observation is common.

Figure 42. (a) and (b) parallelogram-like smectic liquid crystals of 0.1 wt% 30-70
azoTAB- azoTAB-SOBS under bright field and polarized optical microscopy. (c) and (d) schlieren
texture of nematic phase of 0.1 wt% 30-70 azoTAB-SdecS under bright field and polarized optical

142

microscopy. (e) and (f) needle-like solid crystals of 0.1wt% 50-50 azoTAB-SOBS under bright
field and polarized optical microscopy. (g) and (h) schlieren texture of nematic phase of 0.1 wt%
50-50 azoTAB-SdecS under bright field and polarized optical microscopy. (i) and (j) rectangular-
like smectic liquid crystals of 0.1 wt% 70-30 azoTAB- azoTAB-SOBS under bright field and
polarized optical microscopy. (k) and (l) smectic liquid phase of 0.1 wt% 52.5-47.5 azoTAB-
azoTAB-SOBS under bright field and polarized optical microscopy.

The summary of the solid crystals, nematic and smectic liquid crystals were represented in
the ternary diagrams of the catanionic pairs in Figure 43. AzoTAB-SOBS pair showed more stable
crystalline phase compared to azoTAB-SdecS catanionic system. The - stacking interaction
between the benzene rings of the anionic and cationic surfactant could be the underlying reason.
Majority of the azoTAB-SOBS samples contains smectic liquid crystals. Solid crystals were
observed when the anionic and cationic surfactant ratio was 45-55 and 55-45. On the other hand,
nematic liquid crystals were detected almost at every ratio for azoTAB-SdecS except a narrow
region 52.5-47.5 to 55-45 cationic-to-anionic surfactants. As shown, 30-70 azoTAB-SOBS gives
rectangular shaped liquid crystals under polarized microscopy while needle like solid crystals were
observed at 50-50 azoTAB-SOBS. 2 days-aged 30-70 rectangular crystals did not convert to solid
crystals when extra azoTAB added. The shape of rectangles was rearranged into random shape of
liquid crystals; however, solid crystals did not form.

143

Figure 43. Ternary diagram of (a) azoTAB-SOBS and (b) azoTAB-SdecS under visible
and UV light.
The azoTAB/SdecS and azoTAB/SOBS catanionic mixtures were also prepared with
azoTAB pre-converted to the 10% trans/90% cis photostationary state with UV illumination. Here
is may be suspected that increased asymmetry due to the bent-and-twisted cis azoTAB
conformation would cause self-assembly into a fluid as opposed to crystalline microstructure, as
has been reported with traditional surfactants.
104, 346
Indeed, single-phase vesicles were initially
observed over a wide range of cationic-to-anionic surfactant molar ratios with both light (Table 5)
and neutron (Figure 43, with the Q
–2
dependence at low-Q indicative of scattering from bilayers)
scattering. After four days, vesicles could still be detected along with a small amount of crystalline
salts, suggesting a two-phase equilibrium (although longer time studies are proposed to ensure that
this is the true equilibrium state of these systems). Interestingly, the vesicle (UV) → crystalline
(visible) phase transition could be quickly photo-initiated with in situ visible-light illumination,
suggesting that the payload release rate could be tuned from “fast” (~3 hours) to “slow” (~4 days
above when relying on the thermal cis-to-trans conversion, which has a half-life of ~24 hours in
the dark). Figure 39 represents the crystal formation rate of 30-70 azoTAB-SdecS and azoTAB-

144

SOBS at 0.1wt% at dark and under visible and UV light. (1) azoTAB-SOBS sample forms
catanionic salt faster than azoTAB-SdecS. This could be associated with - interaction between
the benzene rings of azoTAB and SOBS. Additionally, more stable catanionic salt formation (i.e.
formation of solid crystals and smectic liquid crystals compared to nematic salt formation) of
azoTAB-SOBS compared to azoTAB-SdecS was explained above. (2) After visible light exposure,
the azoTAB in vesicles turned from cis-to-trans form as the 330 nm peak still observed. These
vesicles are the intermediary state and catanionic crystal formation was completed in a period of
24 hours. (3) The vesicles converted to catanionic crystals in ~48 hours at dark.
The two-dimensional (2D) Small-Angle Neutron Scattering (SANS) images of
azoTAB/SdecS and azoTAB/SOBS catanionic surfactant pairs using small-angle neutron
scattering are shown in Figure 44. The degree of alignment of them were calculated with equation
(25) and equation (26) and summarized in Table 7. The samples were initially formed under visible
light and formed catanionic crystals. While the crystals were formed and settled at the bottom of
the SANS cuvette, the measurement took place. As the particles move down vertically, they
created horizontal asymmetry due to the 90° shift of SANS instrument. The underlying reason of
this asymmetry is the anisotropy of the sample. After shining UV light, the system became more
isotropic and the 2D images showed symmetrical circular shape with very low Af(q) and A(q)
numbers as shown in Table 7. Subsequently, the samples were converted to crystal formation phase
with visible light. The slight horizontal asymmetry of the samples could be still seen. Due to the
fact that, most of the catanionic crystals were formed and settled at the bottom, the degree of
alignment numbers was diminished although they were higher than the UV-exposed samples.

145

Table 7. The degree of alignment of 0.1 wt% azoTAB-SOBS and azoTAB-SdecS at 30-
70 and 70-30 ratio samples with consequentially exposure to visible-UV-visible light. Integral
method (equation 25) and average method (equation 26) are used.

146

Figure 44. Two-dimensional (2D) Small-Angle Neutron Scattering (SANS) patterns of
0.1 wt% azoTAB-SOBS and azoTAB-SdecS 70-30 ratio samples with consequentially exposure
to visible-UV-visible light.

The 2D images of SANS data was reduced to 1D I versus Q plots (Figure 45), and Guinier
approximation was used to understand the bilayer thickness and radius of gyration of the vesicles
and how they change with light illumination. Unilamellar PolyCore and Multilamellar
ParaCrystalline SANS models were applied to compare the Guinier results. The results of all three
methods and comparison with light scattering data were listed in Table 8. The instable vesicle
formation with the catanionic salt formation under visible light were seen. After UV light
illumination, the vesicles were formed. Subsequently, the crystals continue to form with visible
light exposure.

147

0.0001
0.001
0.01
0.1
1
10
100
0.01 0.1
(b) 0.1wt% 70-30 azoTAB-SOBS
I (cm
-1
)
Q (Å
-1
)
0.0001
0.001
0.01
0.1
1
10
100
0.01 0.1
1-vis
2-UV
3-vis
(a) 0.1wt% 30-70 azoTAB-SOBS
I (cm
-1
)
Q (Å
-1
)
10
-5
0.0001
0.001
0.01
0.1
1
10
100
0.01 0.1
I (cm
-1
)
Q (Å
-1
)
(c) 0.1wt% 30-70 azoTAB-SdecS
0.0001
0.001
0.01
0.1
1
10
100
0.01 0.1
Q (Å
-1
)
I (cm
-1
)
(d) 0.1wt% 70-30 azoTAB-SdecS
Figure 45. I, intensity versus Q, scattering vector profile of 0.1 wt% azoTAB-SOBS (a)
30-70, (b) 70-30 and azoTAB-SdecS (c) 30-70, (d) 70-30 using Small-Angle Neutron Scattering.
The samples were prepared under visible light and the light condition was switched to UV light
and then visible light in situ.

148

Table 8. Results of Guinier, PolyCore and Paracrystalline SANS models and their
comparison with light scattering data.
Guinier DLS PolyCore Model Paracrystalline Model
t G(Å) R gG(Å) D(nm) t P(Å) R core(Å) χ
2
t L(Å) N L spacing χ
2

azoTAB-SOBS

(30-70)

1-vis 37 250 65+SLC no fit no fit no fit 83 13.24 22.25 1.47
2-UV 28 174 98 27 161 1.04 27 1.18 35.32 1.04
3-vis 32 162 31 273 1.09 30 >100 127.13 1.10

azoTAB-SOBS
(70-30)
1-vis 40 284 50+SLC no fit no fit no fit no fit no fit no fit no fit
2-UV 35 242 lamellar no fit no fit no fit 21 >100 7.25 0
3-vis 35 257 no fit no fit no fit 20 17.73 40.79 1.13

azoTAB-SdecS
(30-70)
1-vis 32 173 92+NLC no fit no fit no fit no fit no fit no fit no fit
2-UV 28 178 151 26 181 1.09 26 1.32 45.36 1.09
3-vis 34 172 32 215 1.17 33 1.28 40.49 1.54

azoTAB-SdecS
(70-30)
1-vis 39 239 93+LC no fit no fit no fit 50 28.61 12.86 1.20
2-UV 59 205 lamellar no fit no fit no fit 26 3.73 29.56 1.32
3-vis 31 180 25 414 1.43 27 1.28 393 1.11

One month old 0.1wt% 30/70 azoTAB/SOBS and azoTAB/SdecS solutions initially have
light-yellow color solid phase at the bottom and transparent liquid (See Figure 46 and 47).
Subsequent to the samples in duplicates were vortexed for 30 seconds, they were placed under
visible or UV light for an hour. The same procedure has repeated in the next day. The pictures and
UV/vis spectra were shown in Figure 46 and 47. After vortexing and leaving under visible light,
the crystals settled at the bottom and the solvent color did not change. On the other hand, after 1h

149

of UV light irradiation, the samples reformed H aggregated (UV/vis maximum peak at 330nm)
vesicles with hydrodynamic diameter of 100-200 nm by partially solubilizing of the solid phase.
The solubilized azoTAB amount was measured using UV/vis spectroscopy and ~83% and 68% of
the azoTAB-SOBS and azoTAB-SdecS crystals were solubilized under UV light, respectively. In
our previous study, we have observed 80% cis and 20% trans in association with an anionic
surfactant under UV light
41
(90% cis and 10% trans; aqueous solution of single azoTAB
component). This ~80% of azoTAB photoisomerization suggested that the azoTAB is still in
interaction with the anionic surfactants at the solid phase. Additionally, the bright yellow color of
the crystals under visible light became dark yellow or light orange under UV light. The melting
temperature of azobenzene-containing crystals were measured 45 °C and 37 °C for azoTAB/SdecS
under visible and UV light, respectively. The melting temperature of azoTAB/SOBS solid part
was found 82°C (visible light) and 65 °C (UV light). The difference of the melting temperature
suggests the composition of the crystals are distinct and can still undergo reversible trans-to-cis
photoisomerization. The photo-transition of the azobenzene-based crystals with the characteristic
changes have been reported in the literature.
89, 352, 353
These results demonstrated ~80% reversible
vesicle-to-crystal phototransition with the repeated UV and visible light illumination in both
catanionic azoTAB/SOBS and azoTAB/SdecS system. The stability of catanionic vesicles is
longer as the difference of cationic and anionic surfactant alkyl chain length is higher.
346

Oppositely, when the alkyl tail length of the anionic and cationic surfactants are similar, the
crystalline phase or catanionic salt is formed. The melting temperature of azoTAB-SOBS (82 °C)
was higher than azoTAB-SdecS (45 °C) which agrees with azoTAB-SOBS crystals (i.e. smectic
liquid crystals) were more ordered than azoTAB-SdecS (i.e. nematic liquid crystals) as discussed
above. Additionally, the earlier study showed as the tail length of same chain length cationic and

150

anionic surfactant isomers increased, the crystal length of solid crystals and the phase transition
temperature increased.
354
Furthermore, electrostatic attraction is the origin of the spontaneous
formation of catanionic system rather than stronger bonds such as covalent bonds. Consequently,
unique light-triggered crystal to vesicle transition was observed in azoTAB-SOBS and azoTAB-
SdecS catanionic systems. Although catanionic vesicles are spontaneously formed in solution, the
biggest drawback of them is the unexpected catanionic salt formation. Here, we report ~80%
stimuli-responsive catanionic vesicle formation from catanionic salt crystals. To our knowledge,
this is the first report indicating catanionic crystals can be converted to vesicles with light
illumination.
To further investigate application of vesicle to crystal transition, we used bovine serum
albumin protein to examine encapsulation efficiency of azoTAB-SOBS and azoTAB-SdecS
vesicles and their release. Protein encapsulation and release efficiency of positively charged
vesicles (70-30) were measured with 1 mg/mL BSA protein. As the concentration of vesicles
increase from 0.1wt% to 0.5 wt%, the encapsulation efficiency raised from 8% to 46% as shown
in Figure 48. When we draw the encapsulated BSA concentration with the wt% of surfactants,
linear line was seen. (not shown) When the weight percentage of surfactants increased, more
vesicles are formed. This shows, the encapsulated amount of protein is linearly correlated with the
weight percentage of surfactants. As the crystals formed and precipitated at the bottom, the
supernatant had almost all of the protein released (%94).

151

0
0.2
0.4
0.6
0.8
1
1.2
280 320 360 400 440 480 520
(d)
30-70 azoTAB-SdecS
0.1 wt%
Wavelength(nm)
68% vesicle
re-formation
0
0.2
0.4
0.6
0.8
1
1.2
280 320 360 400 440 480 520
Absorbance
Wavelength(nm)
(c)
30-70 azoTAB-SdecS
0.1 wt%
0
0.2
0.4
0.6
0.8
1
1.2
280 320 360 400 440 480 520
Wavelength (nm)
(b)
30-70 azoTAB-SOBS
0.1 wt%
83% vesicle
re-formation
0
0.2
0.4
0.6
0.8
1
1.2
280 320 360 400 440 480 520
Sample1
Sample2
UV light(control)
(a)
30-70
azoTAB-SOBS
0.1 wt%
Absorbance
Wavelength(nm)
1 hour
old
1 month
old
+1 hour
UV
Figure 46. UV-vis spectra of UV-light induced crystal to vesicles formation of 0.1 wt% azoTAB-
SOBS (a, b) and azoTAB-SdecS (c, d) at a ratio of 30-70

152

Figure 47. Eppendorf pictures of UV-light induced crystal to vesicles formation of 0.1
wt% azoTAB-SOBS and azoTAB-SdecS at a ratio of 30-70.

Figure 48. Native-PAGE of (1) 1 mg/mL bovine serum albumin (BSA) (2) 0.1 wt%, (3)
0.25 wt% and (4) 0.5 wt% azoTAB-SdecS (70-30) vesicles under UV light, (5) 24 hours visible
light exposed 0.5 wt% azoTAB-SdecS (70-30) vesicles with 1 mg/mL bovine serum albumin.

Enzymes have been applied for industrial and pharmaceutical usage with their specific
catalytic activities.
57
Enzyme stability can be protected by encapsulation of enzymes in biomimetic
inorganic matrices or nanoparticles.
355
The enzyme hosting capability of vesicles were measured

153

as a concentration of endoglucanase and β-glucosidase enzymes in Figure 49. Constant relative
activity of 1 was seen as a function of β-glucosidase concentration while 0-150% of relative
activity was seen encapsulated and free enzymes. This 0-150% of relative activity increase can be
divided into three parts. Enzyme concentration between 0.04-0.08 µg/mL showed the first part
with all of the enzymes were encapsulated and the encapsulated enzymes showed no activity. The
encapsulated β-glucosidase was calculated as 0.1 µg/mL, up to this enzyme concentration, there is
no reaction under UV light and once the samples were exposed visible light, the reaction started.
The encapsulated enzymes showing no activity and visible light-induced vesicle dissociation into
catanionic salts inducing normal activity (~100%) are summarized in Figure 50. In the second part,
as the enzyme concentration increased, the enzyme activity under UV light increased from 0 to
150%. In the third part, constant 150% activity of β-glucosidase was correlated with the free β-
glucosidase enzyme-azoTAB interactions with negligible amount of β-glucosidase encapsulation.
AzoTAB-associated 50% increase of β-glucosidase relative activity was reported at our earlier
study
175
and Chapter 2.
Similarly, 20 µg/mL endoglucase was encapsulated as shown in Figure 49b. Under visible
light, 140% and 160% activities of endoglucanase enzymes were seen due to excess SdecS and
SOBS free monomers, respectively. As the molecular weight of enzyme decreased from ~200 kDa
(β-glucosidase) to 48kDa (endoglucanase), the encapsulation efficiency increased 200fold. On the
other hand, endoglucanase activity increased to 280% and 320% with azoTAB-SdecS and
azoTAB-SOBS under UV light. This represented again the association of free surfactant
monomers and the surfactants on the vesicles with endoglucanase enzyme. Although
endoglucanase activity could be slightly increased (~5%) towards Avicel microcrystalline substate
with the common surfactants, the traditional surfactants addition increased 2-3.2-fold of

154

endoglucanase activity increased towards p-nitrophenol-based model substrate as discussed in
chapter 3. Additionally, azoTAB addition led to 2-fold to 4-fold endoglucanase activity towards
4-nitrophenyl β-D-cellobioside substrate under visible and UV light, respectively.

Figure 49. Relative activity of β–glucosidase and endoglucanase with 0.05 wt% 30-70
azoTAB-SOBS and azoTAB-SdecS under visible and UV light as a function of (a) β–glucosidase
and (b) endoglucanase concentration.

155

Figure 50. The cartoon shows the enzymes can be encapsulated and the enzymes show no
activity. Upon visible light exposure, the enzymes are released, and the reaction occurs while the
catanionic salt settles at the bottom.

These biofuel enzymes can be encapsulated and released by using light illumination. In the
biofuel production process, the highest cost is the enzymatic hydrolysis due to substrate and end-
product inhibition of biofuel enzymes. Encapsulation of enzymes could potentially decrease the
enzymatic hydrolysis cost with recycling the biofuel enzymes by encapsulating the enzymes before
and after using them. The enzymes can be encapsulated after their use and 80-100 nm vesicles can
be separated from the fermentable sugar end-products. After addition of new substates, the
enzymes can be released with light illumination to catalyze the new substates. Additionally,
encapsulation of enzymes can protect the enzymes from end-product inhibitions after the enzyme
and product separation. This way, the enzymes can be used more than one time and number of
cycles that one enzyme can catalyze in substates can be increased.

156

Apart from the protein encapsulation and release with vesicle to crystal transition,
azobenzene-based crystals have future applications in molecular machines, flexible utensils,
medical apparatuses and flexible robotics based on their bending capacity, response level to light
illumination, rate of thermal relaxation.
356,

357

5.5. Conclusion
Visible light irradiation subsequent to the UV light illumination starts the azoTA
+
decS
-
and
azoTA
+
OBS
-
salt or crystal formation. The crystallization rate of catanionic salts can be controlled
with light wavelength. For crystallization in 3-4 hours, visible light can be used while UV light
can be used for slower crystallization rates (i.e. 3-4 days). Unique 70-80% photo-induced liquid
crystalline to isotropic phase transition was observed. The loading efficiency of enzymes were
studied and as the molecular weight of enzymes increases, the loading capacity of the vesicles
decreased. Bioactive molecules can be encapsulated for preservation and released controllably.
Their activity can be protected and used when the wavelength of light changed.

157

CHAPTER 6: FUTURE WORK
6.1. Biofuel Enzymes
Lignocellulosic plant-cell wall natural substates would be most promising cost effective
natural substates to obtain fermentable sugars and consequently acquire biofuel as an alternative
energy source. Lignocellulosic plant-cell wall consists of cellulose, hemicellulose, pectin and
lignin. The structure-function relationship of β-glucosidase (i.e. one of the cellulase enzymes) as
a function of azoTAB concentration under visible and UV light was studied in Chapter 2. The
endoglucanase (i.e. one of the cellulase enzymes) kinetics with and without azoTAB, SDS, SDBS
and DTAB was assessed in Chapter 3. In this chapter, confirmation-function relationship of other
biofuel enzymes individually and in a mixture will be explained.

6.1.1. Biofuel Enzymes-Introduction
Cellulose is the most abundant polymer with being the main constituent of plant cell walls
and one of the primary polysaccharides in biomass.
206
Cellulose can be hydrolyzed by the cellulase
multienzyme complex into fermentable sugars (i.e., glucose) and hence fermented into bioethanol
as a readily-available sustainable energy source. Enzymatic saccharification of cellulose into
glucose requires three enzymes: random breakage of internal glycoside bonds by endoglucanase
(endocellulase, -1,4-endoglucan hydrolase, -1,4--D-glucanase, EC 3.2.1.4), cellobiose (i.e.,
two linked glucose molecules) cleavage from the terminal ends of cellulose chains via
cellobiohydrolase (exocellulase, EC 3.2.1.91), and hydrolysis of cellobiose and short-chain
cellooligosaccharides into fermentable glucose by -glucosidase (cellobiase, EC 3.2.1.21).

158

Hemicellulose is the next common polymer in plant cell walls and contains xylan and
mannan. The other components of plant cell walls are lignin and pectin. The composition of
hemicellulose, lignin and pectin are summarized at Table 9. The enzymes of these natural
substrates with enzyme commission numbers are listed at Table 9.
6.1.2. Structural Studies of Endoglucanase
Endoglucanases generally fold into one of the three different structures: ()8 TIM barrel
fold, -jelly or Swiss roll fold (-strands), and (/)6 fold (six inner and six outer -helices).
212,358

Endoglucanases have been produced mainly by diverse bacteria, fungi, archaea, protozoan.
128, 212

Aspergillus niger fungi has been used as a source of commercial enzyme preparations due to its
high activity compared to other endoglucanases.
213
Endoglucanases from Aspergillus niger have
been reported from Glycoside Hydrolase family 5 (GH5) with (/)8 TIM barrel structure
213
and
Glycoside Hydrolase family 12 (GH12) jelly-roll fold with two antiparallel sheets (six inner
strands and nine outer strands)
359
.
The structure-activity relationship of endoglucanase from Aspergillus niger would be
studied as a function of azoTAB surfactant under visible and UV light. Endoglucanase catalytic
activity in Chapter 3 and endoglucanase confirmation would be controlled with light illumination
by switching between the trans (higher enzyme binding affinity, resulting higher enzyme
unfolding) and cis form of azoTAB. Structural changes of endoglucanase in the presence of
azoTAB surfactant would be acquired by using small-angle neutron scattering (SANS),
fluorescence spectroscopy, dynamic light scattering, and circular dichroism.

159

6.1.3. Structural Effect of AzoTAB into Cellulase Multienzyme Complex
The activity of cellulase multienzyme (i.e., combined endoglucanase, cellobiohydrolase,
and -glucosidase enzymes) for the conversion of insoluble crystalline cellulose into soluble
glucose was studied in the presence of azoTAB surfactant under visible and UV light in Chapter
3. The photosurfactant effect on multienzyme kinetics were specified with the comparison of the
Michaelis constant, KM, maximum velocity, Vmax, turnover number of an enzyme, kcat, as well as
adsorption constants with Langmuir isotherm in absence and presence of azoTAB. The promising
50% cellulase enzyme mixture would be further studied to understand the tertiary structural
changes. Dynamic light scattering and neutron scattering techniques would be used to predict how
the hydrodynamic diameter and radius of gyration changes with azoTAB surfactant addition.
Additionally, each enzyme would be expressed by using only deuterated water-based
buffers. Two of the deuterated enzymes would be added into the non-deuterated enzyme (i.e.
enzymes which are expressed in H2O based buffers). Tertiary structural changes of individual
enzymes in the cellulase enzyme mixture with and without addition of azoTAB would be studied
with neutron scattering techniques.
Furthermore, the enzyme mixture with azoTAB would be loaded to Size Exclusion
Chromatography-Small-Angle X-Ray Scattering (SEC-SAXS) with and without azoTAB
surfactant in the buffer solution. Individual enzymes would be separated by the SEC column and
azoTAB effect would be obtained.
6.1.4. Structure-Function Relationship of Other Biofuel Enzymes
Complete enzymatic hydrolysis of lignocellulosic plant-cell walls into fermentable sugars
are obtained with the enzymes listed in Table 9. The structure-function relationship of -

160

glucosidase in the presence of azoTAB was studied in Chapter 2. The interactions with azoTAB
of the remaining enzymes listed in Table 9, and particularly regions of superactivity, could be
studied in the future. Light scattering experiments would be applied to examine the effects of
azoTAB surfactant and light illumination on endoglucanase tertiary structure. Small-angle neutron
scattering (SANS) along with shape-reconstruction analysis would be utilized to locate regions of
unfolding, as well as to demonstrate light-responsive structural changes of biofuel enzymes.
Secondary structural changes would be obtained with circular dichroism spectroscopy as a
function of azoTAB concentration under visible and UV light. Finally, enzyme structure would be
correlated with catalytic activity in the presence of photosurfactant.

6.1.5. Conclusion
Combination of the low catalytic activity of lignocellulosic biofuel enzymes with the
irreversible enzyme-substrate binding and hence high enzyme loadings are the main reason of high
bioethanol costs. In this dissertation, β-glucosidase, and endoglucanase enzymes in addition of
azoTAB light-responsive surfactant were studied. 50% of individual enzyme and cellulase
multienzyme mixture were obtained. Thus, this part of the chapter suggests the possible biofuel
research-based future project ideas with azoTAB surfactant.

161

Table 9. Biofuel enzymes of lignocellulosic plant-cell wall natural substrates
Substrate Enzyme EC Number

cellulose

-1,4-linked D-glucoses
Endoglucanase 3.2.1.4
Cellobiohydrolase 3.2.1.91
-glucosidase 3.2.1.21

hemicellulose
xylan; -1,4-linked D-
xylose backbone and L-
arabinose, D-galactose,
acetyl, glucuronic acid side
chains
endo-1,4--xylanase 3.2.1.8
-xylosidase 3.2.1.37
-arabinofuranosidase 3.2.1.55
acetyl xylan esterase 3.1.1.72

mannan; -1,4-linked D-
mannose, D-glucose and D-
galactose,
-mannosidase 3.2.1.25
β-mannanase 3.2.1.78
-glucosidase 3.2.1.21
α-galactosidase 3.2.1.22
acetyl mannan esterase 3.1.1.6

pectin
α-1,4-linked D-galacturonic
acid backbone and L-
arabinose, D-galactose side
chains
pectate lyase 4.2.2.2
pectin esterase 3.1.1.11
Polygalacturonase 3.2.1.15

lignin

complex aromatic polymer
of polyphenols, mainly
coumaryl, conferyl, and
sinapyl alcohols
phenol oxidase 1.14.18.1
Laccase 1.10.3.2
lignin peroxidase 1.11.1.14
manganase peroxidase 1.11.1.13
versatile peroxidase 1.11.1.16

162

6.2. Catanionic Surfactant Systems
“Catanionic” aqueous solutions of a photoresponsive cationic surfactant (azoTAB) and six
sodium alkyl sulfate anionic surfactants (C8-C18) have been investigated. The photoisomerization
of azoTAB from the relatively-hydrophobic planar trans form under visible light to the relatively-
hydrophilic bent and twisted cis isomer under UV light has been found to lead to light-induced
microstructure transitions. Two important results include (1) a unique photo-initiated transition of
vesicles to a crystalline state was observed, and studied in Chapter 5, and (2) catanionic vesicles
can be formed at temperatures below the respective Krafft temperature of sodium tetradecyl sulfate
(30 °C), sodium hexadecyl sulfate (45 °C), and sodium octadecyl sulfate (56 °C) as a result of
associating with azoTAB. Spherical catanionic vesicles are the predominate microstructure
according to small-angle neutron scattering, while the bilayer thickness of the catanionic vesicles
increasing with the length of the hydrophobic tail of the corresponding sodium alkyl sulfate
surfactant.

6.2.1. Catanionic Surfactant Systems - Introduction
Aqueous mixtures of positively-charged (cationic) and negatively-charged (anionic)
surfactants, so-called catanionic mixtures, have been widely studied
341
due to the variety of
microstructures that form spontaneously as a result of favorable electrostatic and hydrophobic
interactions that offset the energy penalty associated with creation of additional interfacial area
(i.e., since G = H – TS +A, where  is the surface tension, traditional vesicles typically
require the input of mechanical energy into the system in the form of shear, etc.). Various
microstructures can be formed in catanionic systems including micelles (spherical, rod-like, disc-

163

like), vesicles, multilamellar, and liquid-crystalline structures based on the cationic-to-anionic
surfactant molar ratio, overall surfactant concentration, alkyl chain length, temperature, and
pH.
94,95,104

Catanionic vesicles, also known as “equilibrium vesicles” as suggested above, have
received special interest due to the long-term stability,
102
cost-effectiveness,
101
and ease of
preparation
100
of these vesicles. Potential applications range from nanotechnology, chemical and
pharmaceutical applications such as membrane protein folding/solubilization,
32
gene therapy,
35

genetic research, drug delivery, separation,
97
vaccines,
98
cosmetic probes, and molecule
encapsulation.
99
The vesicle size, surface charge, and permeability can be controlled with the
molar ratio of cationic-to-anionic surfactant, concentration, alkyl chain length (e.g., adding
aromatic groups), temperature, and pH.
Aqueous solutions of the photoresponsive cationic surfactant azoTAB and six sodium alkyl
sulfate anionic surfactants (C8-C18) were examined. AzoTAB undergoes a reversible
photoisomerization from a planar and relatively-hydrophobic trans conformation under visible
light to a bent and twisted relatively-hydrophilic cis structure under UV illumination, as shown in
Scheme 1. In our previous work, light-induced reversible vesicle  micelle, vesicle  lamellar,
and vesicle  free–monomer transitions have been observed.
41
The effect of the hydrophobic tail
length of the sodium alkyl sulfate surfactants on bilayer thickness of catanionic vesicles at 0.1 wt
% concentration and 30-70 or 70-30 ratios of cationic-to-anionic surfactant are examined using
small-angle neutron scattering (SANS). With the consideration of Krafft temperature differences
of sodium alkyl sulfate surfactants (as reported in Table 10), 1-99 to 99-1 ratios of cationic-to-
anionic surfactant mixtures are also studied to map the ternary diagram using complementary light
scattering and UV-vis spectroscopy.

164

Table 10. Molecular formula, critical micelle concentration (CMC), and Krafft temperatures of
sodium alkyl sulfate surfactants
Name Chemical Structure MW
(g/mol)
CMC
(mM)
360

TKrafft
(°C)
Tail
Length (Å)
Sodium Octyl Sulfate (SOS) CH 3(CH 2) 7OSO 3
-
Na
+
232.27 133 10
Sodium Decyl Sulfate (SdecS) CH 3(CH 2) 9OSO 3
-
Na
+
260.32 33.5 13
Sodium Dodecyl Sulfate (SDS) CH 3(CH 2) 11OSO 3
-
Na
+
288.37 8.47 16 15
Sodium Tetradecyl Sulfate (STS) CH 3(CH 2) 13OSO 3
-
Na
+
326.43 2.08 30
361
18
Sodium Hexadecyl Sulfate (SHS) CH 3(CH 2) 15OSO 3
-
Na
+
344.49 0.56 45
362
20
Sodium Octadecyl Sulfate (SodS) CH 3(CH 2) 17OSO 3
-
Na
+
372.54 0.19 56
363
23

6.2.2. Experimental Section
Materials. The photoresponsive azoTAB surfactant of the form shown in Scheme 1 (4-
ethyl-4’(trimethylamino-butoxy) azobenzene bromide, 2-azo-4) was synthesized as previously
reported.
31,33
Briefly, azoTAB is prepared by three sequential steps: azocoupling of 4-ethylaniline
with phenol, alkylation with 1,4-dibromobutane, and finally quaternalization with trimethylamine.
All chemicals for azoTAB synthesis and sodium alkyl sulfate surfactants were purchased from
Sigma-Aldrich at the highest purity and used as received. Trans-to-cis photoisomerization of
azoTAB was achieved using an 84 W long-wavelength UV lamp (XX-15A-Spectroline) unless
otherwise stated. STS, SHS and SodS stock solutions were prepared at 35, 50 and 65°C due to
their respective Krafft temperatures. The Krafft temperature is the intersection point between
temperature-dependent solubility curve for monomeric surfactant and the critical micelle
concentration (CMC) curve. Therefore, pure surfactants self-aggregate only above the Krafft point
and the CMC.

165

Small-angle Neutron Scattering (SANS). Small-angle neutron scattering measurements
were performed on the 30-meter NG7 SANS instruments at 25 °C at NIST.
153
A neutron
wavelength () of 6 Å and a detector offset of 25 cm with two sample-to-detector distances of 1.33
and 7 m were used to obtain a Q-range of Q = (4/)sin(/2) = 0.0049 − 0.55 Å
-1
, where  is the
scattering angle. The net intensities were obtained after deduction of the background and empty
cell (pure D2O) scattering from the raw data, followed by accounting for local differences in the
detector efficiency using an isotropic scatterer (Plexiglas). The units were converted to an absolute
differential cross section per unit sample volume (cm
-1
) with an attenuated empty beam. The SANS
data were reduced and examined with utilizing the Igor Pro (WaveMetrics) software supplied by
NIST.
154
Separate 0.1 wt % azoTAB and sodium alkyl sulfate solutions were prepared as stock
solutions. Samples were mixed at cationic/anionic molar ratios of 30/70 and 70/30 at room
temperature. A transportable visible and UV lamp (Spectroline, Model no. XX-15A) was located
inside the sample chamber for in situ (visible-to-UV light and UV-to-visible light) experiments.
Dynamic Light Scattering. Dynamic light scattering experiments were performed using
a Brookhaven Model BI-200SM instrument (Brookhaven Instrument Corp.) equipped with a 35
mW HeNe (632.8 nm) laser (Melles Friot, model no. 05-LHP-928), an avalanche photodiode
detector (BI-APD), and a BI-9000AT digital correlator at a scattering angle or 90° and temperature
of 25 °C. Stock 0.1 wt % solutions of azoTAB and sodium alkyl sulfates were prepared and
separately passed through 100-nm syringe filters (Anatop) into sample vials, vortexed for a minute,
and then allowed to reach the equilibrium microstructure for one hour under either visible or UV
light. The azoTAB solution was kept either under visible or UV light prior to mixing with the
anionic surfactants. At least eight separate 5-minute runs were collected and averaged to determine
the hydrodynamic diameter at each condition. The data reported are the first peak present in the

166

size distribution obtained by fitting the correlation function using the non-negative least squares
(NNLS) routine provided by Brookhaven.
UV-vis spectroscopy. Absorption experiments were measured using an Agilent model
8453 UV-vis spectrophotometer. Separate 0.05 wt % and 0.1 wt % stock solutions of azoTAB and
sodium alkyl sulfates were prepared, mixed at molar ratios between 1/99 and 99/1 (cationic/anionic
surfactant), and vortexed for one minute. The azoTAB solution was stored for an hour under either
visible or UV light prior to mixing with the anionic surfactant solutions. Following mixing, the
samples were kept under visible or UV light for an hour, then placed in 0.1 cm cuvettes for
measurement.
6.2.3. Preliminary Results
Preliminary results of the microstructural characterization of various catanionic surfactant
pairs using dynamic light scattering and small-angle neutron scattering with representative pictures
of 30-70 mixtures of catanionic solutions are shown in Table 11 and Figure 51, respectively. A
direct comparison of the effect of the sodium alkyl sulfate tail length on the resulting
microstructure is challenging due to the fact that not only is the tail hydrophobicity increasing, but
so too is the surfactant Krafft temperature (i.e., the temperature below which the pure surfactant
forms hydrated crystals as opposed to micelles with increasing concentration in aqueous
solutions).
364
Thus, it would not be surprising to observe pure anionic-surfactant crystals (a
different phenomenon than the above azoTA–decS crystalline salts) in catanionic mixtures formed
at 25 °C with azoTAB mixed with STS, SHS or SodS (with Krafft temperatures of 30 °C, 45 °C,
and 56 °C, respectively, see Table 10). To make these mixtures, the respective stock solutions were
prepared at 35 °C, 50 °C, and 65 °C, respectively, with the resulting catanionic mixtures then
allowed to reach room temperature. Indeed, a white solid precipitated from these anionic

167

surfactant-rich samples as expected, while surprisingly stable vesicles were detected in the
azoTAB-rich samples with both light (Table 11) and neutron (Figure 51) scattering. The 30-70
azoTAB-SdecS, azoTAB-SDS and azoTAB-SOBS catanionic mixtures show a crystalline
precipitate at the bottom of the DLS tubes (Figure 51). The catanionic crystalline was widely
discussed in Chapter 5.

Table 11. DLS-determined hydrodynamic diameters (value in nm given in parenthesis) of
catanionic mixtures of the azoTAB-analog 2-azo-4 (labeled “S1” below for brevity) with various
sodium alkyl sulfate surfactants as a function of the cationic-to-anionic surfactant molar ratio under
both visible and UV light (0.1 wt% total surfactant concentration, 25 °C). Microstructure
designations are V = vesicles, L = lamellar, C = crystals, M = micelles, and T = white powder
(temperature below the Krafft temperature).

168

Figure 51. Small-angle neutron scattering (SANS) data for mixtures of cationic (azoTAB
analog 2-azo-4) and anionic (sodium alkyl sulfate) surfactants at molar ratios of 30/70 and 70/30
at 0.1 wt % overall surfactant concentration and 25 °C.
The results of SANS data in Figure 51 is summarized in Table 12 for visible light samples
and Table 13 for UV light samples. Guinier method is the first method to try to understand the
SANS data. The comparison between Guinier bilayer thickness, radius of gyration and dynamic
light scattering data represented a more comprehensive understanding of each sample. The
catanionic systems at the ratio of 30-70 and 70-30 are mostly far from the micelle regions at 0.1
wt%. Thus, the SANS data was fitted to most common used vesicles models of PolyCore and
Paracrystalline Lamellar models. While PolyCore model is frequently used model for unilamellar
vesicles, Paracystalline Lamellar model is used for multilamellar vesicles. A good SANS fit should
have reasonable parameters and χ
2
results are close to 1.

169

Table 12. Results of Guinier and model fit analysis of Small-Angle Neutron Scattering
catanionic systems under visible light data in Figure 51 with comparison to light scattering results.

Guinier DLS PolyCore Model Paracrystalline Model

30-70
(vis)

t G(Å) R gG(Å) D(nm) t P(Å) R core(Å) χ
2
t L(Å) N L spacing χ
2

azoTAB-SOS
- - 79 - - - - - - -
azoTAB-SOBS
37 250 68+LC no fit no fit no fit 83 13.24 22.25 1.47
azoTAB-SdecS
32 173 92+LC no fit no fit no fit no fit no fit no fit no fit
azoTAB-SDS
40 112-312 62 38 273 1.55 36 292 1030.4 0
azoTAB-SHS
43 109-912 22-111
+SHS crystal

40 258 1.98 39 1.9e9 114 0

70-30
(vis)
azoTAB-SOS
33 164 76 32 1082 2.82 32 1.22 598 1.78
azoTAB-SOBS
40 284 50+LC no fit no fit no fit no fit no fit no fit no fit
azoTAB-SdecS
39 239 93+LC no fit no fit no fit 50 28.61 12.86 1.20
azoTAB-SDS
40 217 71 56 127 2.17 52 6.02 91.95 1.57
azoTAB-SHS
46 123 56 38 18 1.96 41 5.7 >100 0

The combination of PolyCore model failure, Guinier thickness >50Å, number of layers of
Paracystalline model is large and small bilayer thickness showed the presence of multilamellar
vesicles in the system. When PolyCore model fits the SANS data with suitable numbers and low
χ
2
results, most of the time the number of layers is close to 1. In this case, the bilayer thicknesses
of all three fits showed similar results. The most obvious two results are (1) the UV samples have
lower bilayer thickness compared to the visible samples and (2) 70-30 ratio samples showed higher
bilayer thickness than the 30-70 ratio ones. This second trend was seen in our earlier azoTAB-

170

SDBS study.
41
Overall, it can be concluded that as the tail length of anionic surfactant increases,
the bilayer thickness of catanionic vesicles are increased.

Table 13. Results of Guinier and model fit analysis of Small-Angle Neutron Scattering
catanionic systems under UV light data in Figure 36 with comparison to light scattering results.
Guinier DLS PolyCore Model Paracrystalline Model

30-70
(UV)

t G(Å) R gG(Å) D(nm) t P(Å) R core(Å) χ
2
t L(Å) N L spacing χ
2

azoTAB-SOS
55 222 98 no fit no fit no fit 17 17.6 26 1.19
azoTAB-SdecS
28 178 98 26 181 1.09 26 1.32 45.36 1.09
azoTAB-SOBS
28 174 68 27 161 1.04 27 1.18 35.32 1.04
azoTAB-SDS
32 201 55 30 269 1.00 30 2.03 213.91 1.00
azoTAB-SHS
40 135 15-94 35 12 2.95

70-30
(UV)
azoTAB-SOS
18 170 86 18 404 1.02 16 1.13 5.37 1.08
azoTAB-SdecS
59 205 lamellar no fit no fit no fit 26 3.73 29.56 1.32
azoTAB-SOBS
35 242 lamellar no fit no fit no fit 21 >100 7.25 0
azoTAB-SDS
119 213 LC 188 119 1.83 17 12.44 28.01 1.17
azoTAB-SHS
64 228 75+LC no fit no fit no fit 16 >100 11.86 2.78

The UV-vis absorbance spectra of various catanionic surfactant mixtures were also
measured, as shown in Figure 52. Monomeric azoTAB has a strong absorbance peak at 350 nm
under visible light, which when incorporated into bilayers can be observed to shift to either lower

171

wavelengths (~334 nm, indicating so-called H-aggregates in a face-to-face or deck-of-cards
alignment) or higher wavelengths (~377 nm, indicating J-aggregates in a slipped staircase or
brickwork alignment). Interestingly, the sodium alkyl sulfate-rich samples all exhibited blue-
shifted spectra (H-aggregates), while the spectra were redshifted in azoTAB-rich samples,
suggesting that the dominate surfactant species dictates the bilayer molecular arrangement.

Figure 52. UV-vis spectroscopy data of azoTAB (2-azo-4)/sodium alkyl sulfate catanionic
mixtures (30/70 and 70/30) at an overall surfactant concentration of 0.05 wt % at 25 °C.
6.2.4. Light-responsive Drug Release
The bilayer thicknesses of azoTAB and SOS, SdecS, SOBS, SDS and SHS at 0.1wt% with
ratios of 30-70 and 70-30 were calculated. AzoTAB and the other two sodium alkyl sulfate
surfactants STS and SodS catanionic vesicles could be determined using small-angle neutron

172

scattering. Permeability assays with a model hydrophilic or hydrophobic drug compound would
be explored to examine the effects of bilayer thickness on drug release. Vesicle stability with and
without the drug over time could also be measured using light scattering and UV-vis spectroscopy
to understand the shelf life.
6.2.5. Photoresponsive Membrane Protein Isolation
Although membrane proteins are the 20-30% of the cell proteins, only 0.3% of their
structure is known. Light-triggered membrane protein isolation with vesicle-to-micelle could help
to increase membrane protein purification efficiency and assist to interpret structural and
mechanistic knowledge of membrane proteins. Our previous study showed 10% of
photoresponsive unfolding and refolding of Bacteriodopsin with the azoTAB-SOS(C8) vesicle to
micelle transition.
32
As the bilayer thickness increases, the light-responsive membrane protein
folding status could be controlled with higher efficiency. After finding the promising vesicle-to-
micelle transition regions, bilayer thickness effect on membrane protein isolation could be studied.
6.2.6. Light-induced Genetic Disorder Erasability
The CRISPR (Clustered Regularly Interspaced Short Palindromic Repeats)/Cas9
(CRISPR-associated protein/nuclease) system was adapted from the antiviral immune mechanism
of bacteria (prokaryotic cells). CRISPR/Cas9 consists of a specific target gene sgRNA (i.e., single
guide RNA; a combination of crRNA and tracrRNA) and Cas9 endonuclease. The single guide
RNA recognizes a target and Cas9 endonuclease cleave DNA double strands. Therefore, Cas9
nuclease is able to target any gene sequence complementary to the single guide RNA. The same
Cas9 nuclease can be targeted for monogenic diseases such as cystic fibrosis, haemophilia,
Huntington’s disease, sickle cell anemia, etc., with only the sgRNA sequence differentiation.

173

Additionally, the CRISPR/Cas9 complex can be applied as combination therapy for drug
resistance-related genes using permanent gene editing in cancer treatment.
There are several forms of CRISPR/Cas9 system: (a) the Cas9 endonuclease protein and
sgRNA, (b) a plasmid DNA that encodes both the Cas9 protein and sgRNA, (c) Cas9 mRNA and
sgRNA, and (d) Cas9/sgRNA complex in ribonucleoprotein (RNP). Each CRISPR/Cas9 form has
advantages and challenges. For example, the plasmid DNA form of CRISPR has the longest
expression time and off-target effect compared to the other CRISPR modes due to the transcription
of mRNA and translation of the Cas9 protein. Conversely, mRNAs are less stable than plasmid
DNAs and proteins and are prone to degradation by ribonucleases. Overall, Cas9 protein delivery
is considered the most challenging due to the positive charge and large size (~160 kDa) of the
protein, versus sgRNA which is negatively charged. Thus, co-encapsulation of positively-charged
Cas9 protein and negatively-charged sgRNA remains a challenge.
Different methods have been applied to deliver the CRISPR/Cas9 system, as summarized
in Table 14. Adenoviruses have limited packaging capacity (~4.7 kb)
365
for the Cas9 and sgRNA
expression plasmids, which are 4.5 kb – 9.3 kb long double-stranded DNAs.
366,367
Hence smaller
plasmids are delivered by adenoviruses. Electroporation using powerful electric fields to create
pores in cell membranes often leads to ~50 percent cell death even in vitro experiments and, thus,
cannot be applied for clinical translation.
368
Microinjection damages cells long-term and requires
professional skills.
369
Hydrodynamic injection through veins is limited to use in small animals such
as mice and rats.
367
Thus, although viral and physical methods have been widely studied, the lack
of efficient in vivo transition of the viral vectors and physical methods leaves nonviral delivery as
the preferred method for clinical transition.
370

174

Table 14. Various methods to deliver the CRISPR/Cas9 system.

As noted above, so-called “catanionic” vesicles form spontaneously upon simple mixing
of a cationic and anionic surfactant. In our previous and suggested work, we have use azoTAB-
based photoresponsive surfactants paired with conventional anionic surfactants to initiate photo-
triggered microstructural transitions from catanionic vesicles to micelles, multilamellar structures,
or free surfactant monomers (or vice versa).
41
Notably, the photo-initiated transition from
vesicles to free surfactant monomers was used in our previous work to deliver and then release e-
GFP DNA into cells.
35
Furthermore, in a recent study (Chapter 4) light-responsive vesicles were
used to co-deliver the chemotherapy drug paclitaxel and Bcl-2 siRNA into MDA-MB-231 human
Packing
Capacity
Cell
Viability
in vivo transition Cost Preparation
Method
Manufacturability
Viral methods
Adenovirus limited ✓ immunogenicity
concerns
$$$$ Hard difficult
Physical Methods
Electroporation N/A ~50% cell
death
not suitable $$$$$
$
Medium one-time cost
Microinjection N/A damages
cells in
long-term
not suitable $$$$ Very hard applied manually
Nano-/Macro-particles
Lipids large ✓ ✓ $$$ Medium ✓
Polymers large ✓ generally, toxic $$ Medium ✓
azoTAB large ✓ ✓ $ Easy ✓

175

breast cancer cells.
96
The effects of varying the azoTAB hydrophobic tail on the vesicle surface
charge, size, and bilayer thickness were studied to optimize the delivery vector. Vesicles co-loaded
with paclitaxel and Bcl-2 siRNA that pass through MDA-MB-231 human breast cancer cell
membranes can be subsequently exposed to UV light to cause vesicle rupture and release of the
payload into the cell interiors, where siRNA reduces Bcl-2 protein production allowing paclitaxel
to induce cell death.
Catanionic vesicle-to-free monomer transition could be used to investigate the use of
photoresponsive vesicles to allow delivery and triggered release of CRISPR/Cas9 in Cas9 protein
and sgRNA complex with two different approaches. (1) Using knowledge gained from the
microstructure phase diagrams generated in the suggested work above, negatively-charged
azoTAB-based vesicles would be identified that form under UV light but are disrupted into free
surfactant monomers under visible light. First, positively-charged Cas9 protein and negatively-
charged sgRNA will be encapsulated in these vesicles. Note that it was shown in our previous
study that siRNA was still encapsulated with high efficiency by our negatively-charged catanionic
vesicles (recall that while the bilayers may have a net negative charge, they still contain a
significant amount of positively-charged azoTAB, e.g., see Figure 28).
96
Following this
encapsulation, the vesicles will be transfected in the dark, leading to a slow thermal conversion of
azoTAB to the trans state (half-life ~24 hours) and hence release of the payload into target cells.
This passive “dark release” approach has clear advantages over our previous UV-initiated release
studies, mimicking the slow payload release seen in pH-responsive nanoparticles upon
encapsulation by endosomes. (2) Here negatively-charged vesicles will be prepared to first allow
encapsulation of positively-charged Cas9 protein. Subsequently, additional azoTAB surfactant
will be added to produce positively-charged vesicles for sgRNA encapsulation. All of this will be

176

done under visible light, allowing UV light-triggered release similar to our earlier studies.
35, 96

AzoTAB-containing photoresponsive catanionic vesicles are promising non-viral CRISPR/Cas9
delivery vehicles due to their ease of preparation (spontaneous), long-term (thermodynamic)
stability, low cost, large scalability, manufacturability, and intrinsic (visible  UV light) or
extrinsic (UV light → dark) controllability.
6.2.7. Conclusion
Aqueous solutions of the light-responsive cationic surfactant azoTAB and six sodium alkyl
sulfate anionic surfactants (C8-C18) were investigated. The photoisomerization of azoTAB lead to
photo-responsive reversible self-assembly transitions. The bilayer thickness of spherical
catanionic vesicles at the ratios of 30-70 and 70-30 azoTAB- sodium alkyl sulfates were studied.
As the sodium alkyl sulfate tail length increased, mostly the bilayer thickness of the catanionic
vesicles raised. The bilayer thickness effect on membrane protein isolation and controlled drug
release could be studied to show more practical results. Additionally, the vesicle-to-free monomer
regions of these new catanionic pairs could be identified to examine triggered CRISPR delivery.

177

BIBLIOGRAPHY

1. Tenjarla, S., Microemulsions: An Overview and Pharmaceutical Applications. Critical
Reviews in Therapeutics Drug Carrier Systems 1999, 16 (5), 461-521.
2. Kralova, I.; Sjoblom, J., Surfactants Used in Food Industry: A Review. Journal of
Dispersion Science and Technology 2009, 30 (9), 1363-1383.
3. Schramm, L. L.; Stasiuk, E. N.; Marangoni, G., Surfactant and Their Applications. Annu.
Rep. Prog. Chem. 2003, 99, 3-48.
4. Mao, X.; Jiang, R.; Xiao, W.; Yu, J., Use of Surfactants for the Remediation of
Contaminated Soils: A Review. Journal of Hazardous Materials 2015, 285, 419-435.
5. Dollinger, J.; Schacht, V. J.; Gaus, C.; Grant, S., Effect of Surfactant Application Practices
on the Vertical Transport Potential of Hydrophobic Pecticides in Agrosystems. Chemosphere
2018, 209, 78-87.
6. Yegin, C.; Zhang, M.; Suhag, A.; Ranjith, P.; Peksaglam, Z.; Malik, V.; Wijaya, Z.;
Putra, D.; Anggraini, H.; Saracoglu, O.; Dhannoon, D.; Temizel, C., Supramolecular assemblies
as displacement fluids in EOR. In IOR 2017- 19th European Symposium on Improved Oil
Recovery, European Association of Geoscientists & Engineers: 2017; Vol. 2017.
7. Sheng, J. J., Status of Surfactant EOR Technology. Petroleum 2015, 1, 97-105.
8. Lourith, N.; Kanlayavattanakul, M., Natural Surfactants Used in Cosmetics: Glycolipids.
International Journal of Cosmetic Science 2009, 31, 255-261.
9. Banat, I. M.; Makkar, R. S.; Cameotra, S. S., Potential Commercial Applications of
Microbial Surfactants. Appl. Microbiol. Biotechnol. 2000, 53, 495-508.
10. Holmberg, K., Natural Surfactants. Current Opinion in Colloid & Interface Science 2001,
6, 148-159.
11. Pradhan, A.; Bhattacharyya, A., Quest for an Eco-Friendly Alternative Surfactant: Surface
and Foam Characteristics of Natural Surfactants. Journal of Cleaner Production 2017, 150, 127-
134.
12. Otzen, D. E., Biosurfactants and Surfactants Interacting with Membranes and Proteins:
Same but Different? Biochimica et Biophysica Acta 2017, 1859, 639-649.
13. Kulkarni, V. S.; Shaw, C., Chapter 2- Surfactants, Lipids, and Surface Chemistry. In
Essential Chemistry for Formulators of Semisolid and Liquid Dosages, Jones, K., Ed. Elsevier:
2016.
14. Alwadani, N.; Fatehi, P., Synthetic and Lignin-based Surfactants: Challenges and
Opportunities. Carbon Resources Conversion 2018, 1, 126-138.
15. Salager, J.-L., Surfactant Types and Uses. Laboratory of Formulation, Interfaces,
Rheology and Process 2002, E300-A, 1-48.
16. Nakama, Y., Surfactants. In Cosmetic Science and Technology: Theoretical Principles and
Applications, Sakamoto, K.; Lochhead, R.; Maibach, H.; Yamashita, Y., Eds. Elsevier: 2017.
17. Miyake, M.; Yamashita, Y., Molecular Structure and Phase Behavior of Surfactants. In
Cosmetic Science and Technology: Theoretical Principles and Applications, Sakamoto, K.;
Lochhead, R.; Maibach, H.; Yamashita, Y., Eds. Elsevier: 2017.
18. Rapp, B. E., Chapter 20- Surface Tension. In Microfluidics: Modelling, Mechanics and
Mathematics, Rapp, B. E., Ed. Elsevier: United Kingdom, 2017; pp 421-444.
19. Holmberg, K.; Jonsson, B.; Kronberg, B.; Lindman, B., Surfactants and Polymers in
Aqueous Solution. 2nd ed.; John Wiley & Sons: England, 2002; p 39-66.

178

20. Tomasic, V.; Stefanic, I.; N., F.-V., Adsorption, association and precipitation in
hexadecyltrimethylammonium bromide/sodium dodecyl sulfate mixtures. Colloid Polym Sci.
1999, 277, 153-163.
21. Wang, K.; Yin, H.; Sha, W.; Huang, J.; Fu, H., Temperature-Sensitive Aqueous
Surfactant Two-Phase System Formation in Cationic-Anionic Surfactant Systems. J. Phys. Chem.
B 2007, 111, 12997-13005.
22. Goldsipe, A.; Blankschtein, D., Titration of Mixed Micelles Containing a pH-Sensitive
Surfactant and Conventional (ph-Insensitive) Surfactants: A Regular Solution Theory Modelling
Approach. Langmuir 2006, 22.
23. Feng, H.; Verstappen, N. A. L.; Kuehne, A., J. C.; Sprakel, J., Well-defined Temperature-
sensitive Surfactants for Controlled Emulsion Coalescence. Polymer Chemistry 2013, 4, 1842.
24. Lee, C. T.; Smith, K. A.; Hatton, T. A., Photoreversible viscosity changes and gelation in
mixtures of hydrophobically modified polyelectrolytes and photosensitive surfactants.
Macromolecules 2004, 37 (14), 5397-5405.
25. Eastoe, J.; Vesperinas, A., Self-assembly of Light-sensitive Surfactants. Soft Matter 2005,
1, 338-347.
26. Brown, P.; Bushmelev, A.; Butts, C. P.; Cheng, J.; Eastoe, J.; Grillo, I.; Heenan, R. K.;
Schmidt, A. M., Magnetic Control over Liquid Surface Properties with Responsive Surfactants.
Angew. Chem. Int. 2012, 51, 1-4.
27. Brown, P.; Butts, C. P.; Eastoe, J., Stimuli-responsive Surfactants. Soft Matter 2013, 9,
2365-2374.
28. Wu, X.; Zhang, Y.; Sun, X.; Huang, Y.; Dai, C.; Zhao, M., A Novel CO2 and Pressure
Responsive Viscoelastic Surfactant Fluid for Fracturing. Fuel 2018, 229, 79-87.
29. Zhan, Y.; Goncalves, M.; Yi, P.; Capelo, D.; Zhang, Y.; Rodrigues, J.; Liu, C.; Tomas,
H.; Li, Y.; He, P., Thermo/Redox/pH-triple Sensitive Poly(N-isopropylacrylamide-co-acryl acid)
Nanogels for Anticancer Drug Delivery. Journal of Materials Chemistry 2015, 3, 4221-4230.
30. Shin, J. Y.; Abbott, N. L., Using Light to Control Dynamic Surface Tensions of Aqueous
Solutions of Water Soluble Surfactants. Langmuir 1999, 15, 4404-4410.
31. Hayashita, T.; Kurosawa, T.; Miyata, T.; Tanaka, K.; Igawa, M., Effect of structural
variation within cationic azo-surfactant upon photoresponsive function in aqueous solution.
Colloid & Polymer Science 1994, 272, 1611-1619.
32. Zhang, J.; Wang, S. C.; Lee, C. T., Photoreversible Conformational Changes in Membrane
Proteins Using Light-Responsive Surfactants. Journal of Physical Chemistry B 2009, 113 (25),
8569-8580.
33. Zhang, J.; Lee, C. T. J. Controlling Membrane Protein Folding with Light Illumination and
Catanionic Surfactant Systems. University of Southern California, 2008.
34. Shang, T.; Smith, K. A.; Hatton, T. A., Photoresponsive Surfactants Exhibiting Unusually
Large, Reversible Surface Tension Changes under Varying Illumination Conditions. Langmuir
2003, 19, 10764-10733.
35. Liu, Y. C.; Le Ny, A. L. M.; Schmidt, J.; Talmon, Y.; Chmelka, B. F.; Lee, C. T., Photo-
Assisted Gene Delivery Using Light-Responsive Catanionic Vesicles. Langmuir 2009, 25 (10),
5713-5724.
36. Le Ny, A. L. M.; Lee, C. T., Photoreversible DNA condensation using light-responsive
surfactants. Journal of the American Chemical Society 2006, 128 (19), 6400-6408.
37. Lee, J. C. T.; Smith, K. A.; Hatton, T. A., Photocontrol of Protein Folding: The Interaction
of Photosensitive Surfactants with Bovine Serum Albumin. Biochemistry 2005, 44, 524-536.

179

38. Mirarefi, P.; Lee Jr., C. T., Photo-induced Unfolding and Inactivation of Bovine Carbonic
Anhydrase in the Presence of a Photoresponsive Surfactant. Biochimica et Biophysica Acta 2010,
1804, 106-114.
39. Wang, S.-C.; Mirarefi, P.; Faraone, A.; Lee, C. T. J., Light-Controlled Protein Dynamic
Observed with Neutron Spin Echo Measurements. Biochemistry 2011, 50, 8150-8162.
40. Hamill, A. C.; Wang, S. C.; Lee, C. T., Probing lysozyme conformation with light reveals
a new folding intermediate. Biochemistry 2005, 44 (46), 15139-15149.
41. Seidel, Z. P.; Chang, C. H.; Zhang, J.; Chmelka, B. F.; Lee, C. T. J., Light-induced
Morphological Changes in Catanionic Surfactant Systems. Langmuir 2021, (to be submitted).
42. Gutteridge, A. Understanding the Relationship Between Enzyme Structure and Catalysis.
Darwin College, Cambridge, 2005.
43. de Dios, A. C.; Pearson, J. G.; Oldfield, E., Secondary and Tertiary Structural Effects on
Protein NMR Chemical Shifts: and ab Initio Approach. Science 1993, 260 (5113), 1491-1496.
44. Cheng, Y., Membrane Protein Structural Biology in the Era of Single Particle Cryo-EM.
Current Opinion in Structural Biology 2018, 52, 58-63.
45. Nakamura, A.; Ishida, T.; Samejima, M.; Igarashi, K., The Use of Neutron Scattering to
Determine the Functional Structure of Glycoside Hydrolase. Current Opinion in Structural
Biology 2016, 40, 54-61.
46. Blakeley, M. P.; Langan, P.; Niimura, N.; Podjarny, A., Neutron Crystallography:
Opportunities, Challenges, and Limitations. Curr. Opin. Struct. Biol. 2008, 18 (5), 593-600.
47. Poupon, A.; Janin, J., Chapter 20-Analysis and Prediction of Protein Quaternary Structure.
In Data Mining Techniques for the Life Sciences, Methods in Molecular Biology, Carugo, O.;
Eisenhaber, F., Eds. Humana Press: 2010; Vol. 609, pp 349-364.
48. Chan, F. T. S.; Pinotsi, D.; Schierle, G. S. K.; Kaminski, C. F., Structure-Specific Intrinsic
Fluorescence of Protein Amyloids Used to Study their Kinetics of Aggregation In Bio-
nanoimaging -Protein Misfolding and Aggregation, 2014; pp 147-155.
49. Walker, N. G.; Engelke, D. R., Ribozymes: Catalytic RNAs that Cut Things, Make Things,
and Do Odd and Useful Jobs. Biologist (London) 2002, 49 (5), 199-203.
50. Daniel, R. M.; Dunn, R. V.; Finney, J. l.; Smith, J. C., The Role of Dynamics in Enzyme
Activity. Annual Review of Biophysics and Biomolecular Structure 2003, 32.
51. Agarwal, P. K., Enzymes: An Integrated View of Structure, Dynamics and Function.
Microbial Cell Factories 2006, 5:2.
52. DiRusso, N. V.; Estrin, D. A.; Marti, M. A.; Roitberg, A. E., pH-Dependent
Conformational Changes in Proteins and Their Effect on Experimental pKas: The Case of
Nitrophorin 4. PLoS Comput Biol. 2012, 8 (11), e1002761.
53. Fields, P. A.; Dong, Y.; Meng, X.; Somero, G. N., Adaptations of Protein Structure and
Function to Temperature: There is More Than One Way to 'Skin A Cat'. The Journal of
Experimental Biology 2015, 218, 1801-1811.
54. Bhagavan, N. V., Chapter 6 - Enzymes I: General Properties, Kinetics, and Inhibition. In
Medical Biochemistry, Fourth Edition ed.; Hayhurst, J., Ed. Harcort Academic Press: 2002; pp 85-
108.
55. Daniel, R. M.; Dines, M.; Petach, H. H., The Denaturation and Degradation of Stable
Enzymes at High Temperatures. Biochem. J. 1996, 317.
56. Robinson, P. K., Enzymes: Principles and Biotechnological Applications. Essays Biochem.
2015, 59, 1-41.

180

57. Carrillo, N.; Ceccarelli, E. A.; Roveri, O. A., Usefulness of Kinetic Enzyme Parameters in
Biotechnological Practice. Biotechnology and Genetic Engineering Reviews 2010, 27:1, 367-382.
58. Michaelis, L.; Menten, M. L., The Kinetics of the Inversion Effect. Biochem. 1913, Z. 49,
333-369.
59. Dowd, J. E.; Riggs, D. S., Comparison of Estimates of Michaelis-Menten
Kinetic Constants from Various Linear Transformations The Journal of Biological Chemistry
1965, 240, 863-869.
60. Ranaldi, F.; Vanni, P.; Giachetti, E., What students must know about the determination
of enzyme kinetic parameters Biochemical Education 1999, 27, 87-91
61. Bibi, R.; Ahmad, Z.; Imran, M.; Hussain, S.; Ditta, A.; Mahmood, S.; Khalid, A., Algal
bioethanol production technology: A trend towards sustainable development. Renewable and
Sustainable Energy Reviews 2017, 71, 976–985
62. Bommarius, A. S.; Katona, A.; Cheben, S. E.; Patel, A. S.; Ragauskas, A. J.; Knudson,
K.; Pu, Y., Cellulase Kinetics as a Function of Cellulose Pretreatment. Metabolic Engineering
2008, 10, 370-381.
63. Binod, P.; Gnansounou, E.; Sindhu, R.; Pandey, A., Enzymes for Second Generation
Biofuels: Recent Developments and Future Perspectives. Bioresource Technology Reports 2019.
64. Demirbas, A., Biofuels Sources, Biofuel Policy, Biofuel Economy and Global Biofuel
Projections. Energy Conversion and Management 2008, 49, 2106-2116.
65. Payne, C. M.; Knott, B. C.; Mayes, H. B.; Hansson, H.; Himmel, M. E.; Sandgren, M.;
Stahlberg, J.; Beckham, G. T., Fungal Cellulases. Chem. Rev. 2015, 115, 1308-1448.
66. Zabed, H.; Sahu, J. N.; Suely, A.; Boyce, A. N.; Faruq, G., Bioethanol Production from
Renewable Sources: Current Perspectives and Technological Progress. Renewable and
Sustainable Energy Reviews 2017, 71, 475-501.
67. Vohra, M.; Manwar, J.; Manmode, R.; Padgilwar, S.; Patil, S., Bioethanol production:
Feedstock and current technologies. Journal of Environmental Chemical Engineering 2014, 2,
573–584.
68. Keegstra, K., Future Perspectives in Plant Biology-Plant Cell Walls. Plant Physiology
2010, 154, 483-486.
69. Ciolacu, D. E., Biochemical Modification of Lignocellulosic Biomass. In Biomass as
Renewable Raw Material to Obtaoin Bioproducts of High-Tech Value, Popa, V., Ed. Elsevier:
2018; pp 315-350.
70. Smith, H., The Molecular Biology of Plant Cells. University of California Press: Berkeley,
1977; Vol. 14.
71. Bajpai, P., Chapter 4- Structure and Synergism between the Enzymes of the Xylanolytic
Complex. In Xylanolytic Enzymes, Academic Press: London, 2014; pp 37-41.
72. Moreira, L. R. S.; Filho, E. X. F., An Overview of Mannan Structure and Mannan-
Degrading Enzyme Systems. Appl Microbiol Biotechnol 2008, 79, 165-178.
73. Heredia, A.; Jimenez, A.; Guillen, R., Composition of Plant Cell Walls. Z Lebensm Unters
Forsch 1995, 200, 24-31.
74. Chen, H., Chemical Composition and Structure of Natural Lignocellulose. In
Biotechnology of Lignocellulose, Springer: Dordrecht, 2014; pp 25-71.
75. Paramjeet, S.; Manasa, P.; Korrapati, N., Biofuels: Production of Fungal-mediated
Ligninolytic Enzymes and the Modes of Bioprocesses Utilizing Agro-based Residues. Biocatalysis
and Agricultural Biotechnology 2018, 14, 57-71.

181

76. Nagano, N.; Orengo, C. A.; Thornton, J. M., One Fold with Many Functions: The
Evolutionary Relationships between TIM Barrel Families Based on Their Sequences, Structures
and Functions. J. Mol. Biol. 2002, 321, 741-765.
77. Hocker, B.; Jurgens, C.; Wilmanns, M.; Sterner, R., Stability, Catalytic Versatility and
Evolution of the (βα)8-barrel Fold. Current Opinion in Biotechnology 2001, 12, 376-381.
78. Branden, C.-I., The TIM Barrel the most Frequently Occurring Folding Motif in Proteins.
Current Opinion in Structural Biology 1991, 1, 978-983.
79. Rigden, D. J.; Jedrzejas, M. J.; de Mello, L. V., Identification and Analysis of Catalytic
TIM Barrel Domains in Seven Further Glycoside Hydrolase Families. FEBS Letters 2003, 544,
103-111.
80. Wieranga, R. K., The TIM-Barrel Fold: A Versatile Framework for Efficent Enzymes.
FEBS Letters 2001, 492, 193-198.
81. Yang, T. C.; Legault, S.; Kayiranga, E. A.; Kumaran, J.; Ishikawa, K.; Sung, W. L., The
N-Terminal b-Sheet of the Hyperthermophilic Endoglucanase from Pyrococcus Horikoshii is
Critical for Thermostability. Applied and Environmental Microbiology 2012, 3059-3067.
82. Bhardwaj, A.; Mahanta, P.; Ramakumar, S.; Ghosh, A.; Leelavathi, S.; Reddy, V. S.,
Emerging Role of N- and C-terminal Interactions in Stabilizing (β/α)8 fold With Special Emphasis
on Family 1- Xylanases. Computational and Structural Biology Journal 2012, 2 (3).
83. Nagarajan, D.; Deka, G.; Rao, M., Design of Symmetric TIM Barrel Proteins From First
Principles. BMC Biochemistry 2015, 16:18.
84. Patra, A.; Samanta, N.; Das, D. K.; Mitra, R. K., Enhanced Catalytic Activity of
α‑Chymotrypsin in Cationic Surfactant Solutions: The Component Specificity Revisited J. Phys.
Chem. B 2017, 121, 1457−1465
85. Randolph T. W.; Jones, L. S., Surfactant-Protein Interactions. In Rational Design of Stable
Protein Formulations. Pharmaceutical Biotechnology, Carpenter, J. F.; Manning, M. C., Eds.
Springer: Boston, MA, 2002; Vol. 13.
86. Otzen, D., Protein–surfactant interactions: A tale of many states. Biochimica et Biophysica
Acta 2011, 1814 562–591.
87. Chen, S. H.; Teixeira, J., Structural and Fractal Dimension of Protein-Detergent
Complexes. Phys. Rev. Lett. 1986, 57, 2583-2586.
88. Serrano, A. L.; Waegele, M. M.; Gai, F., Spectroscopic Studies of Protein Folding: Linear
and Nonlinear Methods. Protein Science 2012, 21, 157-170.
89. Beharry, A. A.; Woolley, G. A., Azobenzene Photoswitches for Biomolecules. Chem. Soc.
Rev. 2011, 40, 4422-4437.
90. Sun, J.; Ruchmann, J.; Pallier, A.; Jullien, L.; Desmadril, M.; Tricbet, C., Unfolding of
Cytochrome C Upon Interaction with Azobenzene Modified Copolymers. Biomacromolecules
2012, 13, 3736-3746.
91. Kameta, N.; Akiyama, H.; Masuda, M.; Shimizu, T., Effect of Photoinduced Size Changes
on Protein Refolding and Transport Abilities of Soft Nanotubes. Chem. Eur. J. 2016, 22, 7198-
7205.
92. Schierling, B.; Noel, A.-J.; Wende, W.; Hien, L. T.; Volkov, E.; Kubareva, E.;
Oretskaya, T.; Kokkinidis, M.; Andreas, S.; Bernhard, P. A., Controlling the Enzymatic Activity
of a Restriction Enzyme by Light. PNAS 2010, 107 (4), 1361-1366.
93. James, D. A.; Burns, C.; Woolley, G. A., Kinetic Characterization of Ribonuclease S
Mutants Containing Photoisomerizable Phenylazophenylalanine Residues. Protein Engineering
2001, 14 (12), 983-991.

182

94. Zemb, T.; Dubois, M.; Deme, B.; Gulik-Krzvwicki, T., Self-assembly of Flat Nanodiscs
in Salt-free Catanionic Surfactant Solutions. Science 1999, 283, 5403.
95. Zhu, Z.; Xu, H.; Liu, H.; Gonzalez, Y. I.; Kaler, E. W.; Liu, S., Stabilization of Catanionic
Vesicles via Polymerization. J. Phys. Chem. B 2006, 110, 16309-16317.
96. Seidel, Z. P.; Zhang, X.; MacMullan, M. A.; Graham, N. A.; Wang, P.; Lee, C. T. J.,
Photo-Triggered Delivery of siRNA and Pacitaxel into Breast Cancer Cells Using Catanionic
Vesicles. ACS Applied Bio Materials 2020, 3 (11), 7388-7398.
97. Lioi, S. B.; Wang, X.; Islam, M. R.; Danoff, E. J.; English, D. S., Catanionic surfactant
vesicles for electrostatic molecular sequestration and separation. Phys. Chem. Chem. Phys. 2009,
11, 9315-9325.
98. Hsieh, C.-L.; Ptak, C. P.; Tseng, A.; Souza Suguiura, I. M.; McDonough, S. P.; Sritrakul,
T.; Li, T.; Lin, Y.-P.; Gillilan, R. E.; Oswald, R. E.; Chang, Y.-F., Extended Low-Resolution
Structure of a Leptospira Antigen Offers High Bactericidal Antibody Accessibility Amenable to
Vaccine Design. Biophysics and Structural Biology, Microbiology and Infectious Disease 2017,
6.
99. Soussan, E.; Mille, C.; Blanzat, M.; Bordat, P.; Rico-Lattes, I., Sugar-Derived Tricatenar
Catanionic Surfactant: Synthesis, Self-Assembly Properties, and Hydrophilic Probe Encapsulation
by Vesicles. Langmuir 2008, 24, 2326-2330.
100. Hentze, H.-P.; Raghavan, S. R.; McKelvey, C. A.; Kaler, E. W., Silica Hollow Spheres
by Templating of Catanionic Vesicles. Langmuir 2003, 19, 1069-1074.
101. Rajput, S. M.; Kumar, S.; Aswal, V. K.; Seoud, O. A. E.; Malek, N. I.; Kailasa, S. K.,
Drug-Induced Micelle-to-Vesicle Transition of a Cationic Gemini Surfactant: Potential
Applications in Drug Delivery ChemPhysChem 2018, 19, 865-872.
102. Brasher, L. L.; Kaler, E. W., A small-angle neutron scattering (SANS) contrast variation
investigation of aggregate composition in catanionic surfactant mixtures. Langmuir 1996, 12 (26),
6270-6276.
103. Liu, S.; Gonzalez, Y. I.; Kaler, E. W., Structural Fixation of Spontaneous Vesicles in
Aqueous Mixtures of Polymerizable Anionic and Cationic Surfactants. Langmuir 2003, 19, 10732-
10738.
104. Jurasin, D. D.; Segota, S.; Cadez, V.; Selmani, A.; Sikirc, M. D., Recent Advances in
Catanionic Mixtures In Application and Characterization of Surfactants, Najjar, R., Ed. 2017.
105. Khan, A.; Marques, E. F., Synergism and polymorphism in mixed surfactant systems.
Current Opinion in Colloid & Interface Science 2000, 4, 402-410.
106. Manunta, M.; Tagalakis, A. D.; Attwood, M.; Aldossary, A. M.; Barnes, J. L.; Munve,
M. M.; Weng, A.; McAnulty, R.; Hart, S. L., Delivery of ENaC siRNA to Epithelial Cells
Mediated by A Targeted Nanocomplex: A Therapeutic Strategy for Cystic Fibrosis. Scientific
Reports 2017, 7:700.
107. Samakoglu, S.; Lisowski, L.; Budak-Alpdogan, T.; Usachenko, Y.; Acuto, S.; Di Marzo,
R.; Maggio, A.; Zhu, P.; Tisdale, J. F.; Riviere, I.; Sadelain, M., A Genetic Strategy to Treat
Sickle Cell Anemia by Coregulating Globin Transgene Expression and RNA Interference. Nat.
Biotechnol. 2006, 24(1), 89-94.
108. Benger, M.; Kinali, M.; Mazarakis, N. D., Autism Spectrum Disorder: Prospects for
Treatment Using Gene Therapy. Molecular Autism 2018, 9:39.
109. Lu, J.; Chiang, J.; Iyer, R. R.; Thompson, E.; Kaneski, C. R.; Xu, D. S.; Yang, C.;
Chen, M.; Hodes, R. J.; Lonser, R. R.; Brady, R. O.; Zhuang, Z., Decreased Glucocerebrosidase

183

Activity in Gaucher Disease Parallels Quantitative Enzyme Loss due to Abnormal Interaction with
TCP1 and c-Cbl. PNAS 2010, 107:50, 21665-21670.
110. Pasi, K. J.; Rangarajan, S.; Georgiev, P.; Mant, T.; Creagh, M. D.; Lissitchkov, T.;
Beyan, D.; Austin, S.; Hay, C. R.; Hegemann, I.; Kazmi, R.; Chowdary, P.; Gercheva-
Kyuchukova, L.; Mamonov, V.; Timofeeva, M.; Soh, C. H.; Garg, P.; Vaishnaw, A.; Akinc,
A.; Sorensen, B.; Ragni, M. V., Targeting of Antithrombin in Hemophilia A or B with RNAi
Therapy. N. Engl. J. Med. 2017, 377 (9), 819-828.
111. M., S. S.; Nelson, C. E.; Duval, C. L., Technologies for controlled, local delivery of
siRNA. 2015, 218, 94-113.
112. Imani, R.; Shaoa, W.; Taherkhani, S.; Emami, S. H.; Prakash, S.; Faghihi, S., Dual-
functionalized graphene oxide for enhanced siRNA delivery to breast cancer cells. Colloids and
Surfaces B: Biointerfaces 2016, 147, 315-325
113. Kim, H. J.; Kim, A.; Miyata, K.; Kataoka, K., Recent progress in development of siRNA
delivery vehicles for cancer therapy. Advanced Drug Delivery Reviews 2016, 104, 61-77.
114. Yang, Y.; Xie, X.; Xu, X.; Xia, X.; Wang, H.; Li, L.; Dong, W.; Ma, P.; Yang, Y.;
Liu, Y.; Mei, X., Thermal and magnetic dual-responsive liposomes with a cell-penetrating peptide-
siRNA conjugate for enhanced and targeted cancer therapy. Colloids and Surfaces B: Biointerfaces
2016, 146, 607-615.
115. Du, J.; Lane, L. A.; Nie, S., Stimuli-responsive nanoparticles for targeting the tumor
microenvironment. Jourmal of Controlled Release 2015, 219, 205-214.
116. Pan, D. W.; Davis, M. E., Cationic Muric Acid Polymer-Based siRNA Delivery Systems.
Bioconjugate Chem. 2015, 26 (8), 1791-1803.
117. Backer, L. D.; Braeckmans, K.; Stuart, M. C. A.; Demeester, J.; Smedt, S. C. D.;
Raemdonck, K., Bio-inspired pulmonary surfactant-modified nanogels: A promising siRNA
delivery system. Journal of Controlled Release 2015, 206, 177-186.
118. Lee, C. T. J.; Smith, K. A.; Hatton, T. A., Small-Angle Neutron Scattering Study of the
Micellization of Photosensitive Surfactants in Solution and in the Presence of a Hydrophobically
Modified Polyelectrolyte. Langmuir 2009, 25, 24, 13784-13794.
119. Arioli, T.; Peng, L.; Betzner, A. S.; Burn, J.; Wittke, W.; Herth, W.; Camilleri, C.;
Hofte, H.; Plazinski, J.; Birch, R.; Cork, A.; Glover, J.; Redmond, J.; Williamson, R. E.,
Molecular Analysis of Cellulose Biosynthesis in Arabidopsis. Science 1998, 279 (5351), 717-720.
120. Ganner, T.; Roŝ ker, S.; Eibinger, M.; Kraxner, J.; Sattelkow, J. r.; Rattenberger, J.;
Fitzek, H.; Chernev, B.; Grogger, W.; Nidetzky, B.; Plank, H., Tunable Semicrystalline Thin
Film Cellulose Substrate for High- Resolution, In-Situ AFM Characterization of Enzymatic
Cellulose Degradation ACS Appl. Mater. Interfaces 2015, 7, 27900−27909
121. Kar, Y.; Deveci, H., Importance of P-Series Fuels for Flexible-Fuel Vehicles (FFVs) and
Alternative Fuels. Energy Sources 2007, Part A, 28:10, 909-921.
122. Chandel, A. K.; ES, C.; Rudravaram, R.; Narasu, M. L.; Rao, L. V.; Ravindra, P.,
Economics and Environmental Impact of Bioethanol Production Technologies: An Appraisal.
Biotechnology and Molecular Biology Review 2007, 2 (1), 14-32.
123. Wang, B.; Xia, L., High efficient expression of cellobiase gene from Aspergillus niger in
the cells of Trichoderma reesei. Bioresource Technology 2011, 102, 4568–4572
124. Kuusk, S.; Valjamae, P., When Substrate Inhibits and Inhibitor Activates: Implications of
b-glucosidases. Biotechnology for Biofuels 2017, 10 (7).
125. Sainz, M. B., Commercial cellulosic ethanol: the role of plant-expressed enzymes. In
Biofuels, Tomes, D.; Lakshmanan, P.; Songstad, D., Eds. Springer: New York, 2011; pp 237-264.

184

126. Bar-Even, A.; Noor, E.; Savir, Y.; Liebermeister, W.; Davidi, D.; Tawfik, D. S.; Milo,
R., The Moderately Efficient Enzyme: Evolutionary and Physicochemical Trends Shaping
Enzyme Parameters. Biochemistry 2011, 50 (21), 4402-4410.
127. Yang, B.; Wyman, C. E., Pretreatment: the key to unlocking low‐cost cellulosic ethanol.
Biofuels, Bioproducts and Biorefining 2007, 2 (1), 26-40.
128. Zhang, X.-Z.; Zhang, Y.-H. P., Cellulases: Characteristics, sources, production, and
applications. In Bioprocessing Technologies in Biorefi nery for Sustainable Production of Fuels,
Chemicals, and Polymers, Yang, S.-T.; El-Enshasy, H. A.; Thongchul, N., Eds. John Wiley &
Sons, Inc.: New York, 2013.
129. Humbird, D.; Davis, R.; Tao, L.; Kinchin, C.; D. Hsu; Aden, A.; Schoen, P.; Lukas,
J.; Olthof, B.; Worley, M.; Sexton, D.; Dudgeon, D., Process Design and Economics for
Biochemical Conversion of Lignocellulosic Biomass to Ethanol. Laboratory, N. R. E., Ed. Golden,
Colorado, 2011; p 136.
130. Eckard, A. D.; Muthukumarappan, K.; Gibbons, W., A Review of the Role of Amphiphiles
in Biomass to Ethanol Conversion Applied Sciences 2013, 3, 396-419.
131. Rastegari, A. A., Identifying Potential Cationic Surfactant – Cellulase and Computational
Approaches. In New and Future Developments in Microbial Biotechnology and Bioengineering,
2016.
132. Rastegari, A. A.; Bordbarb, A.-K.; Taheri-Kafranib, A., Interaction of cellulase with
cationic surfactants: Using surfactant membrane selective electrodes and fluorescence
spectroscopy Colloids and Surfaces B: Biointerfaces 2009, 73, 132–139
133. Zhou, Y.; Chen, H.; Qi, F.; Zhao, X.; Liu, D., Non-ionic surfactants do not consistently
improve the enzymatic hydrolysis of pure cellulose. Bioresource Technology 2015, 182 136–143.
134. Menegol, D.; Scholl, A. L.; Fontana, R. C.; Dillon, A. J. P.; Camassola, M., Increased
release of fermentable sugars from elephant grass by enzymatic hydrolysis in the presence of
surfactants Energy Conversion and Management 2014, 88, 1252–1256
135. Liu, J.; Yuan, X.; Zeng, G.; Shi, J.; Chen, S., Effect of biosurfactant on cellulase and
xylanase production by Trichoderma viride in solid substrate fermentation Process Biochemistry
2006, 41 (2347–2351 ).
136. Zheng, Y.; Pan, Z.; Zhang, R.; Wang, D.; Jenkins, B., Non-ionic Surfactants and Non-
Catalytic Protein Treatment on Enzymatic Hydrolysis of Pretreated Creeping Wild Ryegrass. Appl.
Biochem. Biotechnol. 2008, 146, 231-248.
137. Pandey, A.; Negi, S.; Soccol, C. R., Current Developments in Biotechnology and
Bioengineering-Production, Isolation and Purification of Industrial Products. 2017.
138. Litzinger, S.; Fischer, S.; Polzer, P.; Diederichs, K.; Welte, W.; Mayer, C., Structural
and Kinetic Analysis of Bacillus subtilis N-Acetylglucosaminidase Reveals a Unique Asp-His
Dyad Mechanism. The Journal of Biological Chemistry 2010, 285, 35675-35684.
139. Lima, M. A.; Oliveira-Neto, M.; Kadowaki, M. A. S.; Rosseto, F. R.; Prates, E. T.;
Squina, F. M.; Leme, A. F. P.; Skaf, M. S.; Polikarpov, I., Aspergillus niger β -Glucosidase Has
a Cellulase-like Tadpole Molecular Shape Journal of Biological Chemistry 2013, 288-46
(32991–33005).
140. Sakamoto, R.; Arai, M.; Murao, S., Enzymic Properties of Three beta Glucosidases from
Aspergillus aculeatus No. F-50. Agricultural and Biological Chemistry 1985, 49 (5), 1283-1290.
141. Salgado, J. C. S.; Meleiro, L. P.; Carli, S.; Ward, R. J., Glucose Tolerant and Glucose
Stimulated b-glucosidases- A Review. Bioresource Technology 2018, 267, 704-713.

185

142. Wang, S.; Lee, J., C. Ted, Protein Secondary Structure Controlled with Light and
Photoresponsive Surfactants. J. Phys. Chem. B 2006, 110 (32), 16117-16123.
143. Hamill, A.; Lee, J., C. Ted, Solution Structure of an Amyloid-Forming Protein During
Photoinitiated Hexamer-Dodecamer Transitions Revealed through Small-Angle Neutron
Scattering. Biochemistry 2007, 46, 7694.
144. Hamill, A. C.; Lee, C. T., Jr., Photocontrol of ß-Amyloid Peptide (1-40) Fibril Growth in
the Presence of a Photosurfactant. J. Phys. Chem. B 2009, 113 (17), 6164-6172.
145. Wang, S.-C.; Lee, J., C. Ted, Enhanced Enzymatic Activity Through Photoreversible
Conformational Changes. Biochemistry 2007, 46 (50), 14557-14566.
146. Mirarefi, P.; Lee, C. T., Jr., Reversible control of enzyme-inhibitor interactions with light
illumination using a photoresponsive surfactant. Proteins 2019, 87 (7), 1-8.
147. Hoppmann, C.; Maslennikov, I.; Choe, S.; Wang, L., In situ formation of an azo bridge
on proteins controllable by visible light. J. Am. Chem. Soc. 2015, 137 (35), 11218-11221.
148. Ihalainen, J. A.; Bredenbeck, J.; Pfister, R.; Helbing, J.; Chi, L.; Stokkum, I. H. M. v.;
Woolley, G. A.; Hamm, P., Folding and unfolding of a photoswitchable peptide from picoseconds
to microseconds. Proc. Nat. Acad. Sci. USA 2007, 104 (13), 5383-5388.
149. Liu, D.; Karanicolas, J.; Yu, C.; Zhang, Z.; Woolley, G. A., Site-Specific Incorporation
of Photoisomerizable Azobenzene Groups into Ribonuclease S. Bioorgan. Med. Chem. Lett. 1997,
7 (20), 2677-2680.
150. Willner, I.; Rubin, S.; Riklin, A., Photoregulation of papain activity through anchoring
photochromic azo groups to the enzyme backbone. J. Am. Chem. Soc. 1991, 113 (9), 3321-3325.
151. Nelson, N., A photometric adaptation of the Somogyi method for the determination of
glucose. J. Biol. Chem. 1944, 153, 375-380.
152. Biggs, A. I., A Spectrophotometric Determination of the Dissociation Constants of p-
Nitrophenol and Papaverine. Transactions of the Faraday Society 1954, 50 (50), 800-802.
153. Glinka, C. J.; Barker, J. G.; Hammouda, B.; Krueger, S.; Moyer, J. J.; Orts, W. J., The
30 m small-angle neutron scattering instruments at the National Institute of Standards and
Technology. Journal of Applied Crystallography 1998, 31, 430-445.
154. Kline, S. R., Reduction and analysis of SANS and USANS data using IGOR Pro. Journal
of Applied Crystallography 2006, 39, 895-900.
155. Feigin, L. A.; Svergun, D. I., Structure Analysis by Small-Angle X-ray and Neutron
Scattering Plenum Press: New York, 1987.
156. Svergun, D. I., Determination of the regularization parameter in indirect-transform
methods using perceptual criteria. J. Appl. Crystallogr. 1992, 25, 495-503.
157. Heller, W. T.; Krueger, J. K.; Trewhella, J., Further insights into calmodulin-myosin light
chain kinase interaction from solution scattering and shape restoration. Biochemistry 2003, 42,
10579- 10588
158. Louis-Jeune, C.; Andrade-Navarro, M. A.; Perez-Iratxeta, C., Prediction of Protein
Secondary Structure from Circular Dichroism using Theoretically Derived Spectra. Proteins 2012,
80 (2), 374-381.
159. Molina, G.; Contesini, F. J.; Melo, R. R. d.; Sato, H. H.; Pastore, G. M., β-Glucosidase
From Aspergillus In New and Future Developments in Microbial Biotechnology and
Bioengineering 2016.
160. Cairns, T. C.; Nai, C.; Meyer, V., How a Fungus Shapes Biotechnology: 100 Years of
Aspergillus niger Research. Fungal Biology and Biotechnology 2018, 5:13.

186

161. Traon-Masson, M.-P. L.; Pellerin, P., Purification and characterization of two b-D-
glucosidases from an Aspergillus niger enzyme preparation: affinity and specificity
toward glucosylated compounds characteristic of the processing of fruits Enzyme and
Microbial Technology 1998, 22, 374–382.
162. Xue, D.-S.; Chen, H.-Y.; Ren, Y.-R.; Yao, S.-J., Enhancing the activity and
thermostability of thermostable β -glucosidase from a marine Aspergillus niger at high
salinity Process Biochemistry 2012, 47, 606–611.
163. Thongpoo, P.; McKee, L. S.; Araujo, A. C.; Kongsaeree, P. T.; Brumer, H., Identification
of the acid/base catalyst of a glycoside hydrolase family 3 (GH3) β-glucosidase from Aspergillus
niger ASKU28 Biochimica et Biophysica Acta 2013, 1830, 2739–2749
164. Shewale, J. G., β-glucosidase: its role in cellulase synthesis and hydrolysis of cellulose.
Int. J. Biochem. 1982, 14, 435-443.
165. Decker, C. H.; Visser, J.; Schreier, P., b-Glucosidases from Five Black Aspergillus
Species: Study of Their Physico-Chemical and Biocatalytic Properties J. Agric. Food Chem.
2000, 48, 4929-4936
166. Chamoli, S.; Kumar, P.; Navani, N. K.; Verma, A. K., Secretory expression,
characterization and docking study of glucose-tolerant β -glucosidase from B. subtilis.
International Journal of Biological Macromolecules 2016, 85, 425–433.
167. Smilgies, D.-M.; Folta-Stogniew, E., Molecular weight–gyration radius relation
of globular proteins: a comparison of light scattering, small-angle X-ray scattering and structure-
based data J. Appl. Cryst. 2015, 48, 1604–1606
168. Souza, F. H. M.; Inocentes, R. F.; Ward, R. J.; Jorge, J. A.; Furriel, R. P. M., Glucose
and xylose stimulation of a b-glucosidase from the thermophilic fungus Humicola insolens: A
kinetic and biophysical study Journal of Molecular Catalysis B: Enzymatic 2013, 94, 119– 128
169. Zanphorlin, L. M.; Gluseppe, P. O.; Honorato, R. V.; Tonoli, C. C. C.; Fattori, J.;
Crespim, E.; Oliveria, P. S. L.; Ruller, R.; Murakam, M. T., Oligomerization as a strategy for
cold adaptation: Structure and dynamics of the GH1 β-glucosidase from Exiguobacterium
antarcticum B7. Scientific Reports 2016, 6-23776.
170. Gunata, Y. Z.; Bayonove, C. L.; Tapiero, C.; Cordonnier, R. E., Hydrolysis of Grape
Monoterpenyl b-D-Glucosides by Various b- Glucosidases J. Agric. Food Chem., 1990, 38, 1236-
1237.
171. Singhania, R. R.; Sukumaran, R. K.; Rajasree, K. P.; Mathew, A.; Gottumukkala, L.;
Pandey, A., Properties of a major b-glucosidase-BGL1 from Aspergillus niger NII-
08121 expressed differentially in response to carbon sources Process Biochemistry 2011, 46,
1521–1524
172. Ahmed, A.; Nasim, F. u.-H.; Batool, K.; Bibi, A., Microbial b-Glucosidase: Sources,
Production and Applications. Journal of Applied & Environmental Microbiology 2017, 5 (1), 31-
46.
173. Damasio, A. R. L.; Ribeiro, L. F. C.; Ribeiro, L. F.; Furtado, G. P.; Segato, F.; Almeida,
F. B. R.; Crivellari, A. C.; Buckeridge, M. S.; Souza, T. A. C. B.; Murakami, M. T.; Ward, R.
J.; Prade, R. A.; Polizeli, M. L. T. M., Functional characterization and oligomerization of a
recombinant xyloglucan-specific endo-β-1,4-glucanase (GH12) from Aspergillus
niveus Biochimica et Biophysica Acta 2012, 1824, 461–467
174. Shah, M. A.; Mishra, S.; Chaudhuri, T. K., Structural stability and unfolding transition of
b-glucosidases: a comparative investigation on isozymes from a thermo-tolerant yeast. Biophysics
of Structure and Mechanism 2011, 40(7), 877-89.

187

175. Seidel, Z. P.; Lee, C. T. J., Enhanced Activity of the Cellulase Enzyme β-Glucosidase upon
Addition of an Azobenzene-Based Surfactant. ACS Sustainable Chem. Eng. 2020, 8, 1751-1761.
176. Feitosa, E.; Alves, F. R.; Niemiec, A.; Oliveira, M.; Castanheira, E. M. S.; Baptista, A.
L. F., Cationic liposomes in mixed didodecyldimethylammonium bromide and
dioctadecyldimethylammonium bromide aqueous dispersions studied by differential scanning
calorimetry, nile red fluorescence, and turbidity. Langmuir 2006, 22 (8), 3579-3585.
177. Chen, H.; Hayn, M.; Esterbauer, H., Purification and Characterization of Two
Extracellular b-glucosidases from Trichoderma reesei. Biochimica et Biophysica Acta 1992, 1121,
54-60.
178. Barnett, C. E., Some Applications of Wave-length Turbidimetry in the Infrared. J. Phys.
Chem. 1942, 46 (1), 69-75.
179. Singh, G.; Verma, A. K.; Kumar, V., Catalytic Properties, Functional Attributes and
Industrial Application of b-glucosidases. 3 Biotech 2016, 6 (1):3.
180. Shah, M. A.; Chaudhuri, T. K.; Mishra, S., Strategy for purification of aggregation prone
b-glucosidases from the cell wall of yeast: a preparative scale approach New Biotechnology
2012, 29:3, 311-320.
181. Ghorai, S.; Chowdhury, S.; Pal, S.; Banik, S. P.; Mukherjee, S.; Khowala, S., Enhanced
activity and stability of cellobiase (b-glucosidase: EC 3.2.1.21) produced in the presence of 2-
deoxy-D-glucose from the fungus Termitomyces clypeatus. Carbohydrate Research 2010, 345,
1015–1022
182. Guinier, A.; Fournet, G., Small-Angle Scattering of X-Rays. Wiley: New York, 1955.
183. Porte, G.; Marignan, J.; Bassereau, P.; May, R., Shape Transformations of the Aggregates
in Dilute Surfactant Solutions: A Small-Angle Neutron Scattering Study. Journal de Physique
1988, 49 (3), 551-519.
184. Norman, A. I.; Ivkov, R.; Forbes, J. G.; Greer, S. C., The Polymerization of Actin:
Structural Changes From Small-angle Neutron Scattering. The Journal of Chemical Physics 2005,
123, 154904.
185. Balu, R.; Mata, J. P.; Knott, R.; Elvin, C. M.; Hill, A. J.; Choudhury, N. R.; Dutta, N.
K., Effects of Crowding and Environment on the Evolution of Conformational Ensembles of the
Multi-Stimuli-Responsive Intrinsically Disordered Protein, Rec1-Resilin: A Small-Angle
Scattering Investigation. J. Phys. Chem. B 120, 120, 6490-6503.
186. Hamill, A. C. Photocontrol of Protein Conformation Through the Use of Photoresponsive
Surfactants, Investigated by Small Angle Neutron Scattering. University of Southern California,
2008.
187. Glatter, O., Evaluation of Small-angle Scattering Data from Lamellar and Cylindrical
Particles by the Indirect Transformation Method. J. Appl. Cryst. 1980, 13, 577-584.
188. Svergun, D. I.; Richard, S.; Koch, M. H. J.; Sayers, Z.; Kuprin, S.; Zaccai, G., Protein
Hydration in Solution: Experimental Observation by X-Ray and Neutron Scattering. Proc. Natl.
Acad. Sci. 1998, 95, 768-773.
189. Suzuki, K.; Sumitani, J.-i.; Nsm, Y.-W.; Nishimaki, T.; Tani, S.; Wakagi, T.;
Kawaguchi, T.; Fushinobu, S., Crystal Structures of Glycoside Hydrolase Family 3 β-glucosidase
1 from Aspergillus Aculeatus. Biochem. J. 2013, 452, 211-221.
190. Rambo, R. P.; Tainer, J. A., Accurate Assessment of Mass, Models and Resolution by
Small-Angle Scattering. Nature 2013, 496 (7446), 477-481.
191. Galas, E.; Romanowska, I., Purification and some properties of beta-glucosidase from
Aspergillus niger IBT-90. Acta Microbiol Pol. 1997, 46 (3), 241-52.

188

192. McAndrew, R. P.; Park, J. I.; Heins, R. A.; Reindl, W.; Friedland, G. D.; D’haeseleer,
P.; Northen, T.; Sale, K. L.; Simmons, B. A.; Adams, P. D., From Soil to Structure, a Novel
Dimeric b-Glucosidase Belonging to Glycoside Hydrolase Family 3 Isolated from Compost Using
Metagenomic Analysis. The Journal of Biological Chemistry 2013, 288-21, 14985-14992.
193. Amouri, B.; Gargouri, A., Characterization of a novel b-glucosidase from a Stachybotrys
strain. Biochemical Engineering Journal 2006, 32, 191–197
194. Li, C. H.; Nguyen, X.; Narhi, L.; Chemmalil, L.; Towers, E.; Muzammil, S.; Gabrielson,
J.; Jiang, Y., Applications of Circular Dichroism (CD) for Structural Analysis of Proteins:
Qualification of Near- and Far-UV CD for Protein Higher Order Structural Analysis. Journal of
Pharmaceutical Sciences 2011, 100 (11), 4642-4654.
195. Zambelli, B.; Cremades, N.; Neyroz, P.; Turano, P.; Uversky, V. N.; Ciurli, S., Insights
in the (un)structural organization of Bacillus pasteurii UreG, an intrinsically disordered GTPase
enzyme. Mol. BioSyst. 2012, 8, 220-228.
196. Gong, G.; Zheng, Z.; Liu, H.; Wang, L.; Diao, J.; Wang, P.; Zhao, G., Purification and
Characterization of a b-Glucosidase from Aspergillus niger and Its Application in the Hydrolysis
of Geniposide to Genipin. J. Microbiol. Biotechnol. 2014, 24 (6), 788-794.
197. Pei, J.; Pang, Q.; Zhao, L.; Fan, S.; Shi, H., Thermoanaerobacterium
Thermosaccharolyticum b-Glucosidase: A Glucose-Tolerant Enzyme with High Specific Activity
for Cellobiose. Biotechnology for Biofuels 2012, 5:31.
198. Fersht, A., Structure and Mechanism in Protein Science. A Guide to Enzyme Catalysis and
Protein Folding. New York, 1999; Vol. 9 (1).
199. Yarema, K. J., Handbook of Carbohydrate Engineering. Taylor & Francis: New York,
2005.
200. Tiwari, P.; Misra, B. N.; Sangwan, N. S., b-Glucosidases from the Fungus Trichoderma:
An Efficient Cellulase Machinery in Biotechnological Applications. BioMed Research
International 2013, 2013 (203735).
201. Davies, G. J.; Wilson, K. S.; Henrissat, B., Nomenclature for Sugar-binding Subsites in
Glycosyl Hydrolases. Biochemical Journal 1997, 321(Pt 2), 557-559.
202. Knott, B. C.; Crowley, M. F.; Himmel, M. E.; Stahlberg, J.; Beckham, G. T.,
Carbohydrate-Protein Interactions That Drive Processive Polysaccharide Translocation in
Enzymes Revealed from a Computational Study of Cellobiohydrolase Processivity. Journal of the
American Chemical Society 2014, 136, 8810-8819.
203. Pang, P.; Cao, L.-C.; Liu, Y.-H.; Xie, W.; Wang, Z., Structures of a glucose-tolerant b-
glucosidase provide insights into its mechanism Journal of Structural Biology 2017, 198 (3),
154-162.
204. Jeng, W.-Y.; Wang, N.-C.; Lin, M.-H.; Lin, C.-T.; Liaw, Y.-C.; Chang, W.-J.; Liu, C.-
I.; Liang, P.-H.; Wang, A. H.-J., Structural and functional analysis of three b-glucosidases from
bacterium Clostridium cellulovorans, fungus Trichoderma reesei and termite Neotermes
koshunensis Journal of Structural Biology 2011, 173 46–56.
205. Sorensen, A.; Lubeck, M.; Lubeck, P. S.; Ahring, B. K., Fungal Beta-Glucosidases: A
Bottleneck in Industrial Use of Lignicellulosic Materials. Biomolecules 2013, 3, 612-631.
206. Mansouri, A.; Rihani, R.; Laoufi, A. N.; Özkan, M., Production of bioethanol from a
mixture of agricultural feedstocks: Biofuels characterization Fuel 2016, 185, 612–621.
207. Teugjas, H.; Valjamae, P., Product inhibition of Cellulases Studied with 14C-labeled
Cellulose Subsrates. Biotechnology for Biofuels 2013, 6: 104.

189

208. Qing, Q.; Yang, B.; Wyman, C. E., Impact of surfactants on pretreatment of corn stover.
Bioresor. Technol. 2010, 101(15), 5941-5951.
209. Eriksson, T.; Borjesson, J.; Tjerneld, F., Mechanism of surfactant effect in enzymatic
hydrolysis of lignocellulose Enzyme and Microbial Technology 2002, 31 (353–364 ).
210. Helle, S. S.; Duff, S. J. B.; Cooper, D. G., Effect of Surfactants on Cellulose Hydrolysis.
Biotechnology and Bioengineering 1993, 42, 611-617.
211. Seo, D.-J.; Fujita, H.; Sakoda, A., Structural changes of lignocelluloses by a nonionic
surfactant, Tween 20, and their effects on cellulase adsorption and saccharification. Bioresource
Technology 2011, 102, 9605-9612.
212. Akram, F.; Haq, I. u.; Ikram, W.; Mukhtar, H., Insight Perspectives of Thermostable
Endoglucanases for Bioethanol Production: A Review. Renewable Energy 2018, 122, 225-238.
213. Yan, J.; Liu, W.; Li, Y.; Lai, H.-L.; Zheng, Y.; Huang, J.-W.; Chen, C.-C.; Chen, Y.;
Jin, J.; Li, H.; Guo, R.-T., Functional and Structural Analysis of Pichia Pastoris-expressed
Aspergillus niger 1,4-b-endoglucanase. Biochemical and Biophysical Research Communications
2016, 475, 8-12.
214. Hanes, C. S., Studies on plant amylases: The effect of starch concentration upon the
velocity of hydrolysis by the amylase f germinated barley. Biochemical Journal 1932, 26(5), 1406-
1421.
215. Kadam, K. L.; Rydholm, E. C.; McMillan, J. D., Development and Validation of a Kinetic
Model for Enzymatic Saccharification of Lignocellulosic Biomass. Biotechnol. Prog. 2004, 20,
698-705.
216. Lee, S. B.; Shin, H. S.; Ryu, D. D. Y., Adsorption of Cellulase on Cellulose: Effect of
Physicochemical Properties of Cellulose on Adsorption and Rate of Hydrolysis. Biotechnology
and Bioengineering 1982, XXIV, 2137-2153.
217. Stoner, M. R.; Dale, D. A.; Gualfetti, P. J.; Becker, T.; Randolph, T. W., Surfactant-
Induced Unfolding of Cellulase: Kinetic Studies. Biotechnol. Prog. 2006, 22, 225-232.
218. Xu, F.; Ding, H., A new kinetic model for heterogeneous (or spatially confined) enzymatic
catalysis: Contributions from the fractal and jamming (overcrowding) effects. Applied Catalysis
A: General 2007, 317 (1), 70-81.
219. Wang, Z.; Feng, H., Fractal kinetic analysis of the enzymatic saccharification of cellulose
under different conditions. Bioresource Technology 2010, 101, 7995-8000.
220. Maurer, S. A.; Bedbrook, C. N.; Radke, C. J., Competitive Sorption Kinetics of Inhibited
Endo- and Exoglucanases on a Model Cellulose Substrate. Langmuir 2014, 28, 14598-14608.
221. Hsieh, C.-w. C.; Cannella, D.; Jorgensen, H.; Felby, C.; Thygesen, L. G.,
Cellobiohydrolase and Endoglucanase Respond Differently to Surfactants During the Hydrolysis
of Cellulose. Biotechnology for Biofuels 2015, 8:52.
222. Vries, R. P.; Visser, J., Aspergillus Enzymes Involved in Degradation of Plant Cell Wall
Polysaccharides. Microbiology and Molecular Biology Reviews 2001, 65 (4), 497-522.
223. Lee, N. E.; Lima, M.; Woodward, J., Hydrolysis of Cellulose by a Mixture of Trichoderma
reesei Cellobiohydrolase and Aspergillus niger Endoglucanase. Biochimica et Biophysica Acta
1988, 967, 437-440.
224. Coral, G.; Arikan, B.; Unaldi, M. N.; Guvenmez, H., Some Properties of Crude
Carboxymethyl Cellulase of Aspergillus niger Z10 Wild-Type Strain. Turk J. Biol. 2002, 26, 209-
213.
225. Wu, J.; Ju, L.-K., Enhancing Enzymatic Saccharification of Waste Newsprint by Surfactant
Addition. Biotechnol. Prog. 1998, 14, 649-652.

190

226. Madsen, J. K.; Pihl, R.; Moller, A. H.; Madsen, A. T.; Otzen, D. E.; Anderson, K. K.,
The anionic biosurfactant rhamnolipid does not denature industrial enzymes Front. Microbiol.
2015, 6: 292, 1-13.
227. Sahin, S.; Ozmen, I.; Biyik, H., Industrial Applications of Endoglucanase Obtained from
Novel and Native Trichoderma atroviride. Chem. Biochem. Eng. Q. 2016, 30 (2), 265-278.
228. Zhang, Y.; Xu, X.; Zhang, Y.; Li, J., Effect of Adding Surfactant for Transforming
Lignocellulose into Fermentable Sugars during Biocatalysing. Biotechnology and Bioprocess
Engineering 2011, 16, 930-936.
229. Ooshima, H.; Sakata, M.; Harano, Y., Enhancement of Enzymatic Hydrolysis of Cellulose
by Surfactant. Biotechnology and Bioengineering 1986, 28, 1727-1734.
230. Kim, H. J.; Kim, S. B.; Kim, C. J., The effects of nonionic surfactants on the pretreatment
and enzymatic hydrolysis of recycled newspaper. Biotechnology and Bioprocess Engineering
2007, 12, 147-151.
231. Naskar, B.; Dey, A.; Moulik, S. P., Counter-ion Effect on Micellization of Ionic
Surfactants: A Comprehensive Understanding with Two Representatives, Sodium Dodecyl Sulfate
(SDS) and Dodecyltrimethylammonium Bromide (DTAB). J Surfact Deterg 2013, 16, 785-794.
232. Hait, S. K.; Majhi, P. R.; Blume, A.; Moulik, S. P., A Critical Assessment of Micellization
of Sodium Dodecyl Benzene Sulfonate (SDBS) and Its Interaction with Poly (vinyl pyrrolidone)
and Hydrophobically Modified Polymers, JR 400 and LM 200. J. Phys. Chem B 2003, 107, 3650-
3658.
233. Bales, B. L.; Zana, R., Characterization of Micelles of Quaternary Ammonium Surfactants
as Reaction Media I: Dodeclytrimethylammonium Bromide and Chloride. J. Phys. Chem. B 2002,
106, 1926-1939.
234. Adlakha, N.; Sawant, S.; Anil, A.; Lali, A.; Yazdani, S. S., Specific Fusion of b-1,4-
endoglucanase and b-1,4-glucosidase enhances celluloytic activity and helps in channeling of
intermediates. Applied and Environmental Microbiology 2012, 78(20), 7447-7454.
235. Linder, M.; Lindeberg, G.; Reinikainen, T.; Teeri, T. T.; Pettersson, G., The difference
in affinity between two fungal cellulose-binding domains is dominated by a single amino acid
substitution. FEBS Letters 1995, 372, 96-98.
236. Wang, Z.; Xu, J.-H.; Feng, H.; Qi, H., Fractal Kinetic Analysis of Polymers/Nonionic
Surfactants to Eliminate Lignin Inhibition in Enzymatic Saccharification of Cellulose. Bioresource
Technology 2011, 102, 2890-2896.
237. Ogeda, T. L.; Silva, I. B.; Fidale, L. C.; Seoud, O. A. E.; Petri, D. F. S., Effect of cellulose
physical characteristics, especially the water sorption value, on the efficiency of its hydrolysis
catalyzed by free or immobilized cellulase. Journal of Biotechnology 2012, 157(1), 246-252.
238. Kristensen, J. B.; Felby, C.; Jorgensen, H., Yield-determining factors in high-solids
enzymatic hydrolysis of lignocellulose. Biotechnology for Biofuels 2009, 2: 11, 1-10.
239. Kristensen, J. B.; Borjesson, J.; Bruun, M. H.; Tjerneld, F.; Jorgensen, H., Use of surface
active additives in enzymatic hydrolysis of wheat straw lignocellulose Enzyme and Microbial
Technology 2007, 40, 888–895
240. Hsieh, C.-w. C.; Cannella, D.; Jorgensen, H.; Felby, C.; Thygesen, L. G., Cellulase
Inhibition by High Concentrations of Monosaccharides. J. Agric. Food Chem. 2014, 62, 3800-
3805.
241. Medve, J.; Karlsson, J.; Lee, D.; Tjerneld, F., Hydrolysis of microcrystalline cellulose by
cellobiohydrolase I and endoglucanase II from Trichoderma reesei: Adsorption, sugar production
pattern, and synergism of the enzymes. Biotechnology and Bioengineering 1998, 59(5), 621-634.

191

242. Bansal, P.; Hall, M.; Realff, M. J.; Lee, J. H.; Bommarius, A. S., Modelling cellulase
kinetics on lignocellulosic substrates. Biotechnology Advances 2009, 27, 833-848.
243. Xin, D.; Yang, M.; Chen, X.; Zhang, Y.; Ma, L.; Zhang, J., Improving the Hydrolytic
Action of Cellulases by Tween 80: Offsetting the Lost Activity of Cellubiohydrolase Cel7A. ACS
Sustainable Chem. Eng. 2017, 5, 11339-11345.
244. Jiang, K.; Li, J.; Yin, J.; Ma, Q.; Yan, B.; Zhang, X.; Wang, L.; Wang, L.; Liu, T.;
Zhang, Y.; Fan, Q.; Yang, A.; Qiu, X.; Ma, B., Targeted delivery of CXCR4-siRNA by scFv for
HER2þ breast cancer therapy Biomaterials 2015, 52, 77-87.
245. Ji, X.; Peng, Z.; Li, X.; Yan, Z.; Yang, Y.; Qiao, Z.; Liu, Y., Neutralization of TNFa in
tumor with a novel nanobody potentiates paclitaxel-therapy and inhibits metastasis in breast
cancer Cancer Letters 2017, 386, 24-34.
246. Siegel, R. L.; Miller, K. D.; Fuchs, H. E.; Jemal, A., Cancer Statistics, 2021. CA: A Cancer
Journal for Clinicians 2021, 71 (1).
247. Gandhi, N. S.; Tekade, R. K.; Chougule, M. B., Nanocarrier mediated delivery of
siRNA/miRNA in combination with chemotherapeutic agents for cancer therapy: Current
progress and advances. Journal of Controlled Release 2014, 194, 238-256.
248. Nemcova-Furstova, V.; Kopperova, D.; Balusikova, K.; Ehrlichova, M.; Brynychova,
V.; Vaclavikova, R.; Daniel, P.; Soucek, P.; Kovar, J., Characterization of acquired paclitaxel
resistance of breast cancer cells and involvement of ABC transporters. Toxicology and Applied
Pharmacology 2016, 310, 215–228.
249. Jordan, M. A.; Wilson, L., Microtubules as a Target for Anticancer Drugs. Nature Reviews
2004, 4, 253-65.
250. Horwitz, S. B., Taxol (paclitaxel): Mechanism of action. Annals of Oncology 1994, 5 Suppl
6, S3-6.
251. Weaver, B. A., How Taxol/paclitaxel kills cancer cells. Molecular Biology of the Cell 2014,
25(18), 2677-2681.
252. Priyadarshini K.; Keerthi, A. U., Paclitaxel Against Cancer: A short Review. Medicinal
Chemistry 2012, 2, 139-141.
253. Zhang, T.; Luo, J.; Fu, Y.; Li, H.; Ding, R.; Gong, T.; Zhang, Z., Novel oral
administrated paclitaxel micelles with enhanced bioavailability and antitumor efficacy for
resistant breast cancer. Colloids and Surfaces B: Biointerfaces 2017, 150, 89-97.
254. Bakrania, A. K.; Variya, B. C.; Patel, S. S., Novel targets for paclitaxel nano formulations:
Hopes and hypes in triple negative breast cancer. Pharmacological Research 2016, 111, 577-591.
255. Zhang, F.; Wang, X.; Xu, X.; Li, M.; Zhou, J.; Wang, W., Reconstituted high density
lipoprotein mediated targeted co-delivery of HZ08 and paclitaxel enhances the efficacy of
paclitaxel in multidrug-resistant MCF-7 breast cancer cells. European Journal of Pharmaceutical
Sciences 2016, 92, 11-21.
256. Li, J.; Wang, Y.; Zhu, Y.; Oupicky, D., Recent advances in delivery of drug–nucleic acid
combinations for cancer treatment. Journal of Controlled Release 2013, 172 589-600.
257. Wang, X.; Li, S.; Shi, Y.; Chuan, X.; Li, J.; Zhong, T.; Zhang, H.; Dai, W.; He, B.;
Zhang, Q., The development of site-specific drug delivery nanocarriers based on receptor
mediation Journal of Controlled Release 2014, 109, 139-153.
258. Eloya, J. O.; Petrilli, R.; Topana, J. F.; Antonio, H. M. R.; Barcellos, J. P. A.; Chesca,
D. L.; Serafini, L. N.; Tiezzi, D. G.; Leec, R. J.; Marchetti, J. M., Co-loaded paclitaxel/rapamycin
liposomes: Development, characterization and in vitro and in vivo evaluation for breast cancer
therapy. Colloids and Surfaces B: Biointerfaces 2016, 141, 74-82.

192

259. Kim, H.; Leal, C., Cuboplexes: Topologically Active siRNA Delivery ACS Nano 2015,
9, 10214–10226.
260. Reddy, T. L.; Garikapati, K. R.; Reddy, S. G.; Reddy, B. V. S.; Yadav, J. S.; Bhadra,
U.; Bhadra, M. P., Simultaneous delivery of Paclitaxel and Bcl-2 siRNA via pH-Sensitive
liposomal nanocarrier for the synergistic treatment of melanoma Scientific Reports 2016, 6:35223.
261. Shia, X.; Dou, Y.; Zhou, K.; Huo, J.; Yang, T.; Qin, T.; Liu, W.; Wang, S.; Yang, D.;
Chang, L.; Wang, C., Targeting the Bcl-2 family and P-glycoprotein reverses paclitaxel resistance
in human esophageal carcinoma cell line. Biomedicine & Pharmacotherapy 2017, 90, 897–905.
262. Akyol, Z.; Çoker-Gürkan, A.; Arisan, E. D.; Obakan-Yerlikaya, P.; Palavan-Ünsal, N.,
DENSpm overcame Bcl-2 mediated resistance against Paclitaxel treatment in MCF-7 breast cancer
cells via activating polyamine catabolic machinery. Biomedicine & Pharmacotherapy 2016, 84,
2029-2041.
263. Akl, H.; Vervloessem, T.; Kiviluoto, S.; Bittremieux, M.; Parys, J. B.; Smedt, H. D.;
Bultynck, G., A dual role for the anti-apoptotic Bcl-2 protein in cancer: Mitochondria versus
endoplasmic reticulum. Biochimica et Biophysica Acta 2014, 1843, 2240-2252.
264. Merino, D.; Lok, S.; Visvader, J.; Lindeman, G., Targeting BCL-2 to enhance
vulnerability to therapy in estrogen receptor-positive breast cancer Oncogene 2016, 35, 1877-
1887.
265. Saraswathy, M.; Gong, S., Recent developments in the co-delivery of siRNA and small
molecule anticancer drugs for cancer treatment. Materials Today 2014, 17:6.
266. Blanco, E.; Shen, H.; Ferrari, M., Principles of nanoparticle design for overcoming
biological barriers to drug delivery. 2015, 33(9), 941-951.
267. Yang, Z.; Gao, D.; Cao, Z.; Zhang, C.; Cheng, D.; Liu, J.; Shua, X., Drug and gene co-
delivery systems for cancer treatment. Biomaterials Science 2015, 3, 1035-1049.
268. Jong, W. H. D.; Borm, P. J. A., Drug Delivery and nanoparticles: Applications and
Hazards. International Journal of Nanomedicine 2008, 3 (2), 133-149.
269. Walde, P.; Cosentino, K.; Engel, H.; Stano, P., Giant Vesicles: Preparations and
Applications. ChemBioChem 2010, 11, 848– 865.
270. Danhier, F., To Exploit the Tumor Microenvironment: Since the EPR effect fails in the
Clinic, What is the Future of Nanomedicine? Journal of Controlled Release 2016, 244, 108-121.
271. Salatin, S.; Dizaj, S. M.; Khosroushahi, A. Y., Effect of the Surface Modification, Size,
and Shape on Cellular Uptake of Nanoparticles. Cell Biol. Int. 2015, 39, 881-890.
272. Choi, J.-S.; Caoa, J.; Naeema, M.; Noh, J.; Hasan, N.; Choi, H.-K.; Yoo, J.-W., Size-
Controlled Biodegradable Nanoparticles: Preparation and Size-Dependent Cellular Uptake and
Tumor Cell Growth Inhibition. Colloids and Surfaces B: Biointerfaces 2014, 122, 545-551.
273. Shang, L.; Nienhaus, K.; Nienhaus, G. U., Engineered Nanoparticles Interacting with
Cells: Size Matters. Journal of Nanobiotechnology 2014, 12:5.
274. Dai, Y.; Xu, C.; Sun, X.; Chen, X., Nanoparticle Design Strategies for Enhanced
Anticancer Therapy by Exploiting the Tumour Microenvironment. Chem. Soc. Rev. 2017, 46,
3830.
275. Verma, A.; Stellacci, F., Effect of Surface Properties on Nanoparticle-Cell
Interaction Small 2010, 6 (1), 12-21.
276. Hoshyar, N.; Gray, S.; Han, H.; Bao, G., The Effect of Nanoparticle Size on In Vivo
Pharmacokinetics and Cellular Interaction. Nanomedicine 2016, 11(6), 673-692.

193

277. Singh, R.; Lillard, J. W. J., Nanoparticle-based Targeted Drug Delivery. Exp Mol Pathol.
2009, 86 (3), 215-223.
278. Lv, Y.; Hao, L.; Hu, W.; Ran, Y.; Bai, Y.; Zhang, L., Novel multifunctional pH-
sensitive nanoparticles loaded into microbubbles as drug delivery vehicles for enhanced tumor
targeting Scientific Reports 2016, 6:29321.
279. Shen, Y.; Tang, H.; Radosz, M.; Kirk, E. V.; Murdoch, W. J., pH-Responsive
Nanoparticles for Cancer Drug Delivery. Drug Delivery Systems 2008, 437, 183-216.
280. Rica, R. d. l.; Aili, D.; Stevens, M. M., Enzyme-responsive nanoparticles for drug release
and diagnostics. Advanced Drug Delivery Reviews 2012, 64, 967-978.
281. Hu, Q.; Katti, P. S.; Gu, Z., Enzyme-Responsive Nanomaterials for Controlled Drug
Delivery. Nanoscale 2014, 6 (21), 12273-12286.
282. McBain, S. C.; Yiu, H. H. Y.; Dobson, J., Magnetic Nanoparticles for Gene and Drug
Delivery. International Journal of Nanomedicine 2008, 3 (2), 169-180.
283. Pankhurst, Q. A.; Connolly, J.; Jones, S.; Dobson, J., Applications of magnetic
nanoparticles in biomedicine. Journal of Physics D: Applied Physics 2003, 36, R167–R181.
284. Mura, S.; Nicolas, J.; Couvreur, P., Stimuli-responsive nanocarriers for drug delivery.
Nature Materials 2013, 12, 991-1003.
285. Paris, J. L.; Cabanas, M. V.; Manzano, M.; Vallet-Regi, M., Polymer-Grafted Mesoporous
Silica Nanoparticles as Ultrasound-Responsive Drug Carriers ACS Nano 2015, 9 (11), 11023-
11033.
286. Yin, T.; Wang, L.; Yin, L.; Zhou, J.; Huo, M., Co-delivery of hydrophobic paclitaxel and
hydrophilic AURKA specific siRNA by redox-sensitive micelles for effective treatment of breast
cancer. Biomaterials 2015, 61, 10-25.
287. Wang, Y.-C.; Wang, F.; Sun, T.-M.; Wang, J., Redox-Responsive Nanoparticles from the
Single Disulfide Bond-Bridged Block Copolymer as Drug Carriers for Overcoming Multidrug
Resistance in Cancer Cells Bioconjugate Chem. 2011, 22 (10), 1939-1945.
288. Alvarez-Lorenzo, C.; Bromberg, L.; Concheiro, A., Light-sensitive Intelligent Drug
Delivery Systems. Photochemistry and Photobiology 2009 85, 848-860.
289. Son, J.; Yi, G.; Yoo, J.; Park, C.; Koo, H.; Choi, H. S., Light-responsive nanomedicine
for biophotonic imaging and targeted therapy. Advanced Drug Delivery Reviews 2019, 138, 133-
147.
290. Tavano, L.; Muzzalupo, R., Multi-functional vesicles for cancer therapy: The ultimate
magic bullet. Colloids and Surfaces B: Biointerfaces 2016, 147, 161-171.
291. Ferenc, P.; Solár, P.; Kleban, J.; Mikes, J.; Fedorocko, P., Down-regulation of Bcl-2 and
Akt induced by combination of photoactivated hypericin and genistein in human breast cancer
cells. Journal of Photochemistry and Photobiology B: Biology 2010, 98, 25-34.
292. Crespi, C. L.; Chang, T. K. H.; Waxman, D. J., Cytochrome P450 Protocols, High-
Performance Liquid Chromatography Analysis of CYPZC8-Catalyzed Paclitaxel 6a-
Hyroxylation. 2nd ed.; Humana Press: Totowa, New Jersey, 2005.
293. Green, H.; Vretenbrant, K.; Norlander, B.; Peterson, C., Measurement of Paclitaxel and
its Metabolites in Human Plasma using Liquid Chromatography/Ion Trap Mass Spectrometry with
a Sonic Spray Ionization Interface Rapid Communications in Mass Spectroscopy 2006, 20, 2183-
2189.
294. Sackett D. L.; J., W., Nile Red as a Polarity-sensitive Fluorescent Probe of Hydrophobic
Protein Surfaces. Anal Biochem. 1987, 167 (2), 228-234.

194

295. Deye, J. F.; Berger, T. A.; Anderson, A. G., Nile Red as a Solvatochromic Dye for
Measuring Solvent Strength in Normal Liquids and Mixtures of Normal Liquids and Mixtures of
Normal Liquids with Supercritical and Near Critical Fluids. Anal Chem. 1990, 62 (6), 615-622.
296. Kaler, E. W.; Murthy, A. K.; Rodriguez, B. E.; Zasadzinski, J. A., Spontaneous vesicle
formation in aqueous mixtures of single-tailed surfactants. Science 1989, 245 (4924), 1371-4.
297. Zauner, W.; Ogris, M.; Wagner, E., Polylysine-based transfection systems utilizing
receptor-mediated delivery. Advanced Drug Delivery Reviews 1998, 30 (1-3), 97-113.
298. Jang, M.; Kim, J. H.; Nam, H. Y.; Kwon, I. C.; Ahn, H. J., Design of a platform
technology for systemic delivery of siRNA to tumours using rolling circle transcription. Nature
Communications 2015, 6:7930, 1-12.
299. Dong, C.; Badea, I.; Poorghorban, M.; Verrall, R.; Foldvari, M., Impact of Phospholipids
on Plasmid Packaging and Toxicity of Gemini Nanoparticles. Journal of Materials Chemistry
2015, B3, 8806.
300. Duangjit, S.; Pamornpathomkul, B.; Opanasopit, P.; Rojanarata, T.; Obata, Y.;
Takayama, K.; Ngawhirunpat, T., Role of the charge, carbon chain length, and content of
surfactant on the skin penetration of meloxicam-loaded liposomes. International Journal of
Nanomedicine 2014, 9, 2005-2017.
301. Mahendra, A.; James, H. P.; Jadhav, S., PEG-grafted Phospholipids in Vesicles: Effect of
PEG chain length and Concentration on Mechanical Properties. Chemistry and Physics of Lipids
2019, 218, 47-56.
302. Yuet, P. K.; Blankschtein, D., Effect of Surfactant Tail-Length Asymmetry on the
Formation of Mixed Surfactant Vesicles. Langmuir 1996, 12, 3819-3827.
303. Geng, S.; Wang, Y.; Wang, L.; Kouyama, T.; Gotoh, T.; Wada, S.; Wang, J.-Y., A
Light-Responsive Self-Assembly Formed by a Cationic Azobenzene Derivative and SDS as a
Drug Delivery System. Scientific Reports 2016, 7, 39202.
304. Rajkhowa, S.; Mahiuddin, S.; Dey, J.; Kumar, S.; Aswal, V. K.; Biswas, R.;
Kohlbrechere, J.; Ismail, K., The Effect of Temperature, Composition and Alcohols on the
Microstructures of Catanionic Mixtures of Sodium Dodecylsulfate and Cetyltrimethylammonium
Bromide in Water. Soft Matter 2017, 13, 3556.
305. Glatter, O.; Kratky, O., Small-Angle X-Ray Scattering. Academic Press: New York, 1982.
306. Tanford, C., The Hydrophobic Effect: Formation of Micelles and Biological Membranes.
Wiley-Interscience: New York, 1973.
307. Sharma, U. S.; Balasubramanian, S. V.; Straubing, R. M., Pharmaceutical and Physical
Properties of Paclitaxel (Taxol) Complexes with Cyclodextrins. Journal of Pharmaceutical
Sciences 1995, 84: 10, 1223-1230.
308. Straubinger, R. M.; Sharma, A.; Balasubramanian, S. V., Taxane Anticancer Agents.
George, G. I., Ed. ACS: Washington DC, 1995; pp 111-123.
309. Ma, P.; Mumper, R. J., Paclitaxel Nano-Delivery Systems: A Comprehensive Review. J
Nanomed Nanotechnol 2013, 4(2).
310. Simon-Gracia, L.; Hunt, H.; Scodeller, P. D.; Gaitzch, J.; Braun, G. B.; Willmore, A.-
M. A.; Ruoslahti, E.; Battaglia, G.; Teesalu, T., Paclitaxel-Loaded Polymersomes for Enhanced
Intraperitoneal Chemotherapy. Molecular Cancer Therapeutics 2016, 15 (4), 670-679.
311. Bailey, M. M.; Berkland, C. J., Modified Release Delivery Systems. CRC Press: Boca
Raton, FL, 2017.
312. Koudelka, S.; Turanek-Knotigova, P.; Masek, J.; Korvasova, Z.; Skrabalova, M.;
Plockova, J.; Bartheldyova, E.; Turanek, J., Liposomes with High Encapsulation Capacity for

195

Paclitaxel: Preparation, Characterization and In Vivo Anticancer Effect. Journal of
Pharmaceutical Sciences 2009, 99 (5), 2309-2319.
313. Paik, Y.; Yang, C.; Metaferia, B.; Tang, S.; Bane, S.; Ravindra, R.; Shanker, N.;
Alcaraz, A. A.; Johnson, S. A.; Schaefer, J.; O'Connor, R. D.; Cegelski, L.; Synder, J. P.;
Kingston, D. G. I., Rotational-Echo Double-Resonance NMR Distance Measurements for the
Tubulin-Bound Paclitaxel Conformation. J. Am. Chem. Soc. 2007, 129, 361-370.
314. Xu, L.; Chen, J.; Feng, L.; Dong, S.; Hao, J., Loading Capacity and Interaction of DNA
binding on Catanionic Vesicles with Different Cationic Surfactants. Soft Matter 2014, 10, 9143-
9152.
315. Monteiro, N.; Martin, A.; Reis, R. L.; Neves, N. M., Liposomes in Tissue Engineering
and Regenerative Medicine. J. R. Soc. Interface 2014, 11: 20140459.
316. Obata, Y.; Saito, S.; Takeda, N.; Takeoka, S., Plasmid DNA-Encapsulating Liposomes:
Effect of a Spacer Between the Cationic Head Group and Hydrophobic Moieties of the Lipids on
Gene Expression Efficiency. Biochimica et Biophysica Acta 2009, 1788, 1148-1158.
317. Rietwyk, S.; Peer, D., Next-Generation Lipids in RNA Interference Therapeutics. ACS
Nano 2017, 11, 7572-7586.
318. Ghosh, S.; Ray, A., Spontaneous Vesicle Based Excipient Formation in Mixtures of
Sodium N-(n-Alkanoyl)-L-alaninate and N-Cetylpyridinium Chloride: Effect of Hydrocarbon
Chain Length. Ind. Eng. Chem. Res. 2015, 54, 1953-1961.
319. Bergstrom, M.; Pedersen, J. S., A small-angle neutron scattering (SANS) study of tablet-
shaped and ribbonlike micelles formed from mixtures of an anionic and a cationic surfactant.
Journal of Physical Chemistry B 1999, 103 (40), 8502-8513.
320. Jiang, Y.; Li, F.; Luan, Y.; Cao, W.; Ji, X.; Zhao, L.; Zhang, L.; Li, Z., Formation of
Drug/Surfactant Catanionic Vesicles and Their Application in Sustained Drug Release.
International Journal of Pharmaceutics 2012, 436, 806-814.
321. Mady, M. M.; Ghannam, M. M.; Khalil, W. A.; Repp, R.; Markus, M.; Rascher, W.;
Muller, R.; Fahr, A., Efficient Gene Delivery with Serum into Human Cancer Cells Using Targeted
Anionic Liposomes Journal of Drug Targeting 2004, 12 (1), 11-18.
322. Lia, X.-X.; Chena, J.; Shena, J.-M.; Zhuang, R.; Zhang, S.-O.; Zhu, Z.-Y.; Ma, J.-B.,
pH-Sensitive Nanoparticles as Smart Carriers for Selective Intracellular Drug Delivery to Tumor.
International Journal of Pharmaceutics 2018, 545, 306-317.
323. Zhang, Y.; Lin, L.; Liu, L.; Liu, F.; Maruyamac, A.; Tiana, H.; Chen, X., Ionic-
Crosslinked Polysaccharide/PEI/DNA Nanoparticles for Stabilized Gene Delivery. Carbohydrate
Polymers 2018, 201, 246-256.
324. Bozzuto, G.; Molinari, A., Liposomes as Nanomedical Devices. International Journal of
Nanomedicine 2015, 10, 975-999.
325. Radler, J. O.; Koltover, I.; Salditt, T., Structure of DNA-cationic Liposome Complexes:
DNA Intercalation in Multilamellar Membranes in Distinct Interhelical Packing Regimes. Science
1997, 275 (5301), 810-4.
326. Kan, P.; Chen, Z.-B.; Lee, C.-J.; Chu, I.-M., Development of Nonionic
Surfactant/Phospholipid o/w Emulsion as a Paclitaxel Delivery System. Journal of Controlled
Release 1999, 58, 271-278.
327. Liebmann, J. E.; Cook, J. A.; Lipschultz, C.; Teague, D.; Fisher, D. T. J.; Mitchell, J. B.,
Cytotoxic Studies of Paclitaxel (Taxol) in Human Tumour Cell Lines. Br. J. Cancer 1993, 68,
1104-1109.

196

328. George, M. S.; Ayoub, A. T.; Banerjee, A.; Churchill, C. D. M.; Winter, P.;
Klobukowski, M.; Cass, C. E.; Luduena, R. F.; Tusznski, J. A.; Damaraju, S., Designing and
Testing of Novel Taxanes to Probe the Highly Complex Mechanisms by Which Taxanes Bind to
Microtubules and Cause Cytotoxicity to Cancer Cells. Plos One 2015, 10 (6), e0129168.
329. Bramer, T.; Paulson, M.; Edwards, K.; Edsman, K., Catanionic Drug-Surfactant Mixtures:
Phase Behavior and Sustained Release from Gels. Pharmaceutical Research 2003, 20 (10), 1661-
1667.
330. Berlepsch, H. v.; Böttcher, C.; Dähne, L., Structure of J-Aggregates of Pseudoisocyanine
Dye in Aqueous Solution. J. Phys. Chem. B 2000, 104 (37), 8792-8799.
331. Honary, S.; Zahir, F., Effect of Zeta Potential on the Properties of Nano-Drug Delivery
Systems - A Review (Part 1). Tropical Journal of Pharmaceutical Research 2013, 12 (2), 255-
264
332. He, C.; Hu, Y.; Yin, L.; Tang, C.; Yin, C., Effects of particle size and surface charge on
cellular uptake and biodistribution of polymeric nanoparticles. Biomaterials 2010, 31, 3657–3666.
333. Xiao, K.; Li, Y.; Luo, J.; Lee, J. S.; Xiao, W.; Gonik, A. M.; Agarwal, R. G.; Lam, K.
S., The effect of surface charge on in vivo biodistribution of PEG-oligocholic acid based micellar
nanoparticles. Biomaterials 2011, 32, 3435-3446.
334. Finke, J. H.; Richter, C.; Gothsch, T.; Kwade, A.; Buttgenbach, S.; Muller-Govmann,
C. C., Coumarin 6 as a Fluorescent Model Drug: How to Identify Properties of Lipid Colloidal
Drug Delivery Systems via Fluorescence Spectroscopy. European Journal of Lipid Science and
Technology 2014, 116, 1234-1246.
335. Yu, X.; Song, Y.; Di, Y.; He, H.; Fu, D.; Jin, C., Enhanced tumor targeting of cRGD
peptide-conjugated albumin nanoparticles in the BxPC-3 cell line. Scientific Reports 2016, 6,
31539.
336. Jiang, W.; Yang, L.; Qui, L.; Xu, J.; Yang, X.; Wang, J.; Zhoud, H.; Wang, D.,
Multifunctional Hybrid Nanoparticles based on Sodium Carboxymethylcellulose-graft-histidine
and TPGS for Enhanced Effect of Docetaxel. RSC Adv. 2015, 5.
337. Langner, M., The Intracellular Fate of Non-Viral DNA Carriers. Cellular & Molecular
Biology Letters 2000, 5 (3), 295-313.
338. Wasungu, L.; Hoekstra, D., Cationic Lipids, Lipoplexes and Intracellular Delivery of
Genes. J Control Release 2006, 116(2), 255-64.
339. Wiethoff, C. M.; Middaugh, C. R., Barriers to Nonviral Gene Delivery. J Pharm Sci. 2003,
92(2), 203-17.
340. Selbo, P. K.; Weyergang, A.; Hogset, A.; Norum, O. J.; Berstad, M. B.; Vikdal, M.;
Berg, K., Photochemical Internalization Provides Time- and Space- Controlled Endolysosomal
Escape of Therapeutic Molecules. J. Control Release 2010, 148 (1), 2-12.
341. Kaler, E. W.; Herrington, K. L.; Murthy, A. K.; Zasadzinski, J. A. N., Phase Behavior and
Structures of Mixtures of Anionic and Cationic Surfactants. J. Phys. Chem. 1992, 96, 6698-6707.
342. Sakai, H.; Matsumura, A.; Yokoyama, S.; Saji, T.; Abe, M., Photochemical Switching of
Vesicle Formation Using an Azobenzene-Modified Surfactant. J. Phys. Chem B 1999, 103, 10737-
10740.
343. Egres, R. G.; Wagner, N. J., The Rheology and Microstructure of Acicular Precipitated
Calcium Carbonate Colloidal Suspensions Through The Shear Thickening Transition. Journal of
Rheology 2005, 49 (3), 719-746.

197

344. Draper, E. R.; Wallace, M.; Honecker, D.; Adams, D. J., Aligning self-assembled perylene
bisimides in a magnetic field. Chemical Communications 2018, 54, 10977-10980.
345. Avrami, M., Kinetics of Phase Change I - General Theory. J. Chem. Phys. 1939, 7, 1103-
1112
346. Yatcilla, M. T.; Herrington, K. L.; Brasher, L. L.; Kaler, E. W., Phase Behavior of
Aqueous Mixtures of Cetyltrimethylammonium Bromide (CTAB) and Sodium Octyl Sulfate
(SOS) J. Phys. Chem. 1996, 100, 5874-5879
347. Raghavan, S. R.; G., F.; W., K. E., Wormlike Micelles Formed by Synergistic Self-
Assembly in Mixture of Anionic and Cationic Surfactants. Langmuir 2002, 18, 3797-3803.
348. Regev, O.; Khan, A., Alkyl Chain Symmetry Effects in Mixed Cationic-Anionic Surfactant
Systems. Journal of Colloid and Interface Science 1996, 182, 95-109.
349. Jurasin, D.; M., V.; Pustak, A.; Smit, I.; Bujan, M.; Filipovic-Vincekovic, N., Lamellar
to hexagonal columnar liquid crystalline phase transition in a catanionic surfactant mixture:
dodecylammonium chloride–sodium bis(2-ethylhexyl) sulfosuccinate. Soft Matter 2013, 9, 3349-
3360.
350. Maiti, K.; Bhattacharya, S. C.; Moulik, S. P.; Panda, A. K., Physicochemical Studies on
Ion-Pair Amphiphiles: Solution and Interfacial Behavior of Systems Derived from Sodium
Dodecylsulfate and n-Alkyltrimethylammonium Bromide Homologues. J. Chem Sci. 2010, 122:6,
867-879.
351. Dummur, D., Optical Properties of Liquid Crystals. In Handbook of Liquid Crystals,
Wiley-VCH Verlag GmbH & Co. KGaA: 2014; Vol. 2. Physical Properties and Phase Behavior
of Liquid Crystals-Physical Properties.
352. Taniguchi, T.; Fujisawa, J.; Shiro, M.; Koshima, H.; Asahi, T., Mechanical Motion of
Chiral Azobenzene Crystals with Twisting upon Photoirradiation. Chem. Eur. J. 2016, 22, 7950-
7958.
353. Koshima, H.; Ojima, N.; Uchimoto, H., Mechanical Motion of Azobenzene Crystals upon
Photoirradiation. J. Am. Chem. Soc. 2009, 131, 6890-6891.
354. Vincekovic, N. F.; Pucic, I.; Popovic, S.; Tomasic, V.; Tezak, D., Solid-Phase Transitions
of Catanionic Surfactants. Journal of Colloid and Interface Science 1997, 188, 396-403.
355. Chapman, R.; Stenzel, M. H., All Wrapped up: Stabilization of Enzymes within Single
Enzyme Nanoparticles. J. Am. Chem. Soc. 2019, 141, 2754-2769.
356. Taniguchi, T.; Asahi, T.; Koshima, H., Photomechanical Azobenzene Crystals. Crystals
2019, 9 (437).
357. Hao, Y.; Huang, S.; Guo, Y.; Zhou, L.; Hao, H.; Barrett, C. J.; Yu, H., Photoinduced
multi-directional deformation of azobenzene molecular crystals. J. Mater. Chem. C 2019, 7 (503).
358. Schulein, M., Protein Engineering of Cellulases. Biochimica et Biophysica Acta 2000,
1543, 239-252.
359. Khademi, S.; Zhang, D.; Swanson, S. M.; Wartenberg, A.; Witte, K.; Meyer, E. F.,
Determination of the Structure of an Endoglucanase from Aspergillus niger and its Mode of
Inhibition by Palladium Chloride. Acta Cryst. 2002, D58, 660-667.
360. Mukerjee, P.; Mysels, K. J., Critical Micelle Concentrations of Aqueous Surfactant
Systems. Washington, D.C. U.S. Dept. of Commerce, National Bureau of Standards, 1971.
361. Patra, N.; Mal, A.; Dey, A.; Ghosh, S., Influence of Solvent, Electrolytes, b-CD, OTAB
on the Krafft Temperature and Aggregation of Sodium Tetradecyl Sulfate. Journal of Molecular
Liquids 2019, 280, 307-313.

198

362. Wolfrum, S. Long Chain Soaps and Alkyl Sulfates in Aqueous Solutions at Room
Temperature. University of Regensburg 2017.
363. Verma, A.; Purkait, M. K., Synthesis, Characterization and Physical Properties Studies of
an Anionic Surfactant. Indian Journal of Chemical Technology 2010, 17, 233-237.
364. Blanco, E.; Rodriguez-Abreu, C.; Schulz, P.; Ruso, J. M., Effect of alkyl chain asymmetry
on catanionic mixtures of hydrogenated and fluorinated surfactants. Journal of Colloid and
Interface Science 2010, 341, 261-266.
365. Chandrasekaran, A. P.; Song, M.; Kim, K.-S.; Ramakrishna, S., Chapter Six- Different
Methods of Delivering CRISPR/Cas9 Into Cells. In Progress in Molecular Biology and
Translational Science, Teplow, D. B., Ed. Elsevier: Cambridge, MA, 2018; Vol. 159, pp 157-176.
366. Wang, H.-X.; Song, Z.; Lao, Y.-H.; Xu, X.; Gong, J.; Cheng, D.; Chakrabortuy, S.;
Park, J. S.; Li, M.; Huang, D.; Yin, L.; Cheng, J.; Leong, K. W., Nonviral Gene Editing via
CRISPR/Cas9 Delivery by Membrane-disruptive and Endosomolytic Helical Polypeptide. PNAS
2018, 115 (19), 4903-4908.
367. Glass, Z.; Lee, M.; Li, Y.; Xu, Q., Engineering the Delivery System for CRISPR-Based
Genome Editing. Trends in Biotechnology 2018, 36 (2).
368. Li, L.; Hu, S.; Chen, X., Non-viral Delivery Systems for CRISPR/Cas9-based Genomic
Editing: Challenges and Opportunities. Biomaterials 2018, 171, 207-218.
369. Liu, C.; Zhang, L.; Liu, H.; Cheng, K., Delivery Strategies of the CRISPR-Cas9 Gene
Editing System for Therapeutic Applications. Journal of Controlled Release 2017, 266, 17-26.
370. Ryu, N.; Kim, M.-A.; Park, D.; Lee, B.; Kim, Y.-R.; Kim, K.-H.; Baek, J.-I.; Kim, W.
J.; Lee, K.-Y.; Kim, U.-K., Effective PEI-mediated Delivery of CRISPR-Cas9 Complex for
Targeted Gene Therapy. Nanomedicine: Nanotechnology, Biology and Medicine 2018, 14, 2095-
2102.

Abstract (if available)

Abstract Azobenzene trimethyl ammonium bromide (azoTAB), a photoresponsive analog of the common dodecyltrimethylammonium bromide (DTAB) surfactant, has been widely studied in areas ranging from biotechnology, nanotechnology, and pharmacology. AzoTAB undergoes a reversible photoisomerization from the relatively-hydrophilic cis form under UV light (???? = 350 nm) to the relatively-hydrophobic trans form under visible light (???? = 434 nm). This switchable hydrophobicity of the surfactant has been used to photoreversibly control the structure-function relationship of enzymes. Specifically, the more hydrophobic trans photoisomer of azoTAB has a greater binding affinity to proteins relative to the cis form, thereby resulting in a greater degree of protein unfolding. In some cases, this unfolding occurs at the active site, leading to a “switched off” enzyme that can be switched back “on” upon UV illumination. In other cases, this unfolding occurs distal to the active site in regions that impart greater flexibility to the enzyme, allowing the enzyme-substrate complex to more easily traverse the activation energy barrier and leading to a “superactive” enzyme. ? In the present work, photocontrol of the structure and function of cellulase enzymes will be explored. Cellulases are generally multienzyme complexes containing endocellulase, exocellulase, and ?-glucosidase that work in a synergistic manner to convert cellulose, the most abundant organic molecule on Earth, into fermentable sugars (i.e., glucose). Bioethanol obtained in this manner thus represents an abundant and sustainable energy source to minimize the dependence on fossil fuels. The process begins with endocellulase or endoglucanase randomly cleaving internal glycoside bonds to convert crystalline cellulose into individual polymer chains. Exocellulase or cellobiohydrolase then liberates cellobiose (two linked glucose units) from the terminal ends of these chains, while ?-glucosidase hydrolyzes cellobiose into two glucose molecules. Notably, cellobiose is an inhibitor of endo- and exo-cellulase, thus, the last enzymatic step can be rate limiting. Addition of azoTAB to ?-glucosidase leads to a dimer ? monomer transition of the enzyme and a corresponding 50% increase in catalytic activity. Endoglucanase activity increase of 45% with azoTAB addition towards Avicel crystalline substate will be correlated with endoglucanase adsoption increase on the crystalline substrate. Furthermore, 45?50% activity enhancement via azoTAB surfactant preserved for all three cellulase enzyme mixture of endoglucanase, cellohydrolase and ?-glucosidase. This superactivity or activity enhancement of cellulase enzymes could have profound impacts on the economic viability of bioethanol (e.g., about one-third of the costs of every gallon of bioethanol are enzyme-related). ? As another application of azoTAB surfactants, photocontrol of the spontaneous self-assembly of “catanionic” surfactant systems (i.e., aqueous solutions of cationic and anionic surfactants) will be discussed. Catanionic systems can form various microstructures depending on the conditions, including spherical and cylindrical micelles, vesicles, multilamellar assemblies and crystalline structures in addition to the monomeric state. As a result, azoTAB-based catanionic systems allow for photoreversible transitions including vesicle ? monomer, vesicle ? micelle, vesicles ? multilamellar and vesicles ? liquid crystals transitions. Notably, azoTAB-based catanionic vesicles will be utilized to facilitate co-delivery of siRNA (i.e., small interfering RNA) and a hydrophobic chemotherapy drug into breast cancer cells, with the photo-initiated vesicle-to-monomer transition with UV illumination used to release the siRNA and drug from the carrier once entered into the cells. Additionally, light-induced vesicle-to-liquid crystal transition will be optimized for cellulase enzyme encapsulation, controlling the reaction as well as recycling enzymes infinitely. Furthermore, the effects of azoTAB chemical structure and solution conditions on vesicle properties will be examined using a range of techniques as a means to optimize the bilayer thickness of vesicles for membrane protein isolation with vesicle ? micelle phototransition.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

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

PDF

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

PDF

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

PDF

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

PDF

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

PDF

Photo-assisted gene therapy using light-responsive catanionic vesicles

PDF

Controlling membrane protein folding with light illumination and catanionic surfactant systems

PDF

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

PDF

Photocontrol of DNA compaction using light-responsive surfactants

PDF

Controlling membrane protein folding using photoresponsive surfactant

PDF

High throughput computational framework for synthesis and accelerated discovery of dielectric polymer materials using polarizable reactive molecular dynamics and graph neural networks

PDF

Physical principles of membrane mechanics, membrane domain formation, and cellular signal transduction

PDF

Flipping the switch on protein activity activity: elastin-like polypeptides assemble into cell switches and vesicles

Asset Metadata

Creator Seidel, Zumra Peksaglam (author)

Core Title Enhancement of biofuel enzyme activity and kinetics with azoTAB surfactants

School Andrew and Erna Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Chemical Engineering

Degree Conferral Date 2021-08

Publication Date 07/18/2021

Defense Date 04/08/2021

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

Tag azobenzene-based surfactants,light responsive surfactants,light triggered drug and gene delivery,OAI-PMH Harvest,photosensitive surfactants,photosensitive vesicles,small-angle neutron scattering,stimuli-responsive nanoparticles,structure-activity relationship

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Lee, Charles Ted, Jr. (committee chair), Nakano, Aiichiro (committee member), Shing, Katherine (committee member)

Creator Email peksagla@usc.edu,pzumra@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC15609010

Unique identifier UC15609010

Legacy Identifier etd-SeidelZumr-9783

Document Type Dissertation

Format application/pdf (imt)

Rights Seidel, Zumra Peksaglam

Type texts

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

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu