(«£3383

 

 

CV
9097

.2!

CU
ECO-E
42:0,
I“)
{3.93

".C
1.9:

2

-fi-

 

 

This is to certify that the
dissertation entitled

CHARACTERIZATION OF MOLECULAR MOTION IN
SELECTED LIPID BILAYER SYSTEMS

presented by

KELLY GREENOUGH

has been accepted towards fulfillment
of the requirements for the

Doctoral degree in Chemistry

 

 

/é~z74//§/&/—7Q

D Major Professor’s Signature

Z/z/a?

Date

MSU is an affinnative-action, equal-opportunity employer

-.-—.-.-n-.-.-.-.—u--o,-.-.-u--¢—-.-—--

..—.---.-----~.-.n.-.-.-.-u-a--.-.-.-n--n—o—--u--—uu... -._.-.-.-._.

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATEDUE

DATEDUE

DAIEDUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

._ Faw“' ’

6/07 p:/C|RCIDaIeDueindd-pt1

 

CHARACTERIZIN G MOLECULAR MOTION IN SELECTED LIPID BILAYER
SYSTEMS

By

Kelly Greenough

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Chemistry

2007

ABSTRACT

CHARACTERIZIN G MOLECULAR MOTION IN SELECTED LIPID BILAYER
SYSTEMS

By

Kelly Greenough

Plasma membranes are central to the existence and function of cellular systems.
These membranes are known to contain a large number of constituents and as our
understanding of these complex and dynamic systems has evolved, it has become clear
that even simple lipid bilayer membranes are heterogeneous, highly dynamic structures.
The complexity of these structures makes studies regarding the environment inside
membranes difficult. For this reason, experiments are performed using model lipid
bilayers. We have produced model lipid bilayers in the form of unilamellar vesicles and
have chosen a chemical probe, 6-(N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)amino) (NBD) to
incorporate into our bilayer systems. NBD is a well characterized fluorophore with strong
one- and two-photon absorption bands. Two phospholipids were chosen, 1,2-dilauroyl-
sn-glycero-3-phosphocholine (DLPC) and l,2-distearoyl-sn-glycero-3-phosphocholine
(DSPC), as were cholesterol and sphingomyelin in order to promote phase segregation in
the model lipid bilayers. The fluorescence lifetime and anisotropy behavior of NBD in
vesicles were measured using time correlated single photon counting (TCSPC). We
utilized both one- and two-photon excitation to allow determination of the Cartesian

components of the chromophore rotational diffusion coefficient as a function of its local

environment. Fluorescence lifetime data indicated that NBD, regardless of being present
as a free or tethered probe, resides in the polar region of the lipid bilayer. One- and two-
photon anisotropy data have been demonstrated to provide complementary information
regarding molecular motions of the chemical probe in the membrane environment and
indicate changes in bilayer organization upon the addition of cholesterol and

sphingomyelin.

To my husband

iv

ACKNOWLEDGEMENTS

I would like to first thank my advisor Gary Blanchard for all of the support and
encouragement he gave me in my years at MSU. He always made it a point to be
available and approachable to all of his students. I very much appreciated his confidence
in me even when I did not believe in myself. I know that I was extremely lucky to have
joined his group, and also very fortunate to have known him.

I would also like to give a special thanks to my friend and colleague, Monika
Dominska. She gave me great ideas for my research and was always there when I needed
someone to talk to. She is such a talented and knowledgeable individual — she is truly the
“Boss”. I will miss our daily lattes, which was often the best part of my day.

I would like to say thank you to my parents, Chuck and Susie Patching, who have
always been there for me and who continue to go above and beyond to help me out.
Much of my success in life I attribute to the fact that I have such incredible and
supportive parents.

I would like to say thanks to my sister, Amy Winston, who has always believed in
my ability to earn my Ph.D., and encouraged me to pursue this goal from the beginning.

A special thanks goes to my brother Tim Patching who was always my “good
luck charm” - especially when I was taking cumulative exams.

Finally, my greatest thank you goes to my husband Dan Greenough. From the
beginning, he supported my decision to go back to school, even if it meant completely
disrupting our lives, because that was what I wanted. He was the one who was there with

me every day throughout this program, standing beside me no matter what. I cannot count

the number of times I wanted to just give up on myself, but he continued to believe in me
and picked me up when I was at my lowest. I truly could not have gotten my degree
without him. Out of everything I have ever achieved in my life, I am most proud that I am

his wife.

vi

TABLE OF CONTENTS

List of Tables ........................................................................................ viii
List of Figures ......................................................................................... x
Chapter 1 — Introduction ............................................................................. 1

Literature Cited ............................................................................. 15

Chapter 2 - Evaluating the Role of Chromophore Side Group Identity in Mediating
Solution Phase Rotational Motion

Introduction ................................................................................. 1 9
Experimental ................................................................................ 21
Results and Discussion .................................................................... 24
Conclusions ................................................................................. 53
Literature Cited ............................................................................. 55

Chapter 3 - Quantitating the Dynamics of NBD Hexanoic Acid in Homogeneous
Solution and in Solutions Containing Unilamellar Vesicles

Introduction ................................................................................. 59
Experimental ................................................................................ 62
Background ................................................................................. 66
Results and Discussion .................................................................... 73
Conclusions ................................................................................. 88
Literature Cited ............................................................................. 89

Chapter 4 — One- and Two-Photon Excited Fluorescence Dynamics of NBD in lipid
Bilayers. Dependence of Chromophore Dynamics on Bilayer Composition

Introduction ................................................................................. 92
Experimental ................................................................................ 96
Results and Discussion .................................................................. 100
Conclusions ................................................................................ 125
Literature Cited ........................................................................... 126

Chapter 5 — Dynamics of a Tethered Chromophore in Phosphoglycerol Vesicles

Introduction ................................................................................ 129
Experimental .............................................................................. 1 3 3
Results and Discussion .................................................................. 136
Conclusions ................................................................................ 156
Literature Cited ........................................................................... 157
Chapter 6 — Conclusions .......................................................................... 162

vii

LIST OF TABLES

Table 2.1 Fluorescence lifetimes of the chromophores BBD, SNBD and NBDHA in the
solvents used in this work. Solvent polarities according to the ET(3O) scale were
taken from Ref. 86. Uncertainties reported here are 110 for at least six individual
determinations. Lifetimes were the same for both one and two photon excitation
to within the uncertainty of the measurement .......................................... 31

Table 2.2 Experimental reorientation times for BBD using one photon excitation (1) and
two photon excitation with linear polarization (2L) and circular polarization (2C).
Values of Dx and D2 are extracted from the experimental data according to Eqs.
2.3 and 2.13. Uncertainties reported here are $10 for at least six individual
determinations .............................................................................. 41

Table 2.3. Experimental reorientation times for SNBD using one photon excitation (1)
and two photon excitation with linear polarization (2L) and circular polarization
(2C). Values of D)( and D2 are extracted from the experimental data according to
Eqs. 2.3 and 2.13. Uncertainties reported here are $10 for at least six individual
determinations .............................................................................. 42

Table 2.4. Experimental reorientation times for NBDHA using one photon excitation (1)
and two photon excitation with linear polarization (2L) and circular polarization
(2C). Values of Dx and DZ are extracted from the experimental data according to
Eqs. 2.3 and 2.13. Uncertainties reported here are $10 for at least six individual
determinations .............................................................................. 43

Table 2.5. Calculated dielectric friction time constants, DSE reorientation time constants
and their sum, reported at 10R. Solvent properties viscosity, dielectric constant
and Debye dielectric relaxation time constant are from references ”’94‘97. For the
calculation of 1935 (S=O.667) (Eq. 2.14), v = 215 A3 for BBD; v = 297 A3 for
SNBD; v = 245 A3 for NBDHA.48 For the calculation of W, a = 3.7 A and w =
10.6 D for BBD; a = 4.1 A and u* = 12.2 D for SNBD; a = 3.9 A and u* = 9.9 D
for NBDHA ................................................................................. 45

Table 3.1 Solvent-dependence of D2 and DX for NBDHA in neat solvents ................. 79

Table 3.2. Dependence of NBDHA one- and two-photon excited anisotropy decay time
constants on vesicle composition. Abbreviations: DLPC = 1,2-Dilauroyl-sn-
Glycero-3-Phosphocholine; DSPC = 1,2-Distearoyl-sn-Glycero-3-
Phosphocholine; chol = cholesterol; spm = sphingomyelin .......................... 84

Table 4.1 One photon excitation, fluorescence lifetimes and anisotropy decay time
constants in vesicles of different compositions. Chol = cholesterol, SPM =
sphingomyelin ............................................................................. 1 15

Table 4.2. Two photon excitation (linear polarization), fluorescence lifetimes and
anisotropy decay time constants in vesicles of different compositions ............ 117

viii

Table 4.3. Two photon excitation (circular polarization), fluorescence lifetimes and
anisotropy decay time constants in vesicles of different compositions ............ 118

Table 4.4. Cartesian components of the rotational diffusion coefficient for NBD C12-
HPC, and resulting viscosities calculated from these D values ..................... 120

Table 4.5 Calculated cone angles for hindered rotor model based on zero and infinite
time anisotropy data for NBD-Clz-HPC in vesicles of varying composition... . . 121

Table 5.1. One-photon excited fluorescence lifetimes of NBD-PG in vesicles.
Experimental data were fit to the function f(t) = Alexp(-t/rl) + Azexp(-t/'rz).
Uncertainties are i lo for at least six individual determinations .................. 141

Table 5.2. Experimental reorientation time constants and calculated values of rotational

diffusion coefficients for NBD-PG in the vesicles of compositions indicated in the
first column ................................................................................ 152

ix

LIST OF FIGURES

Figure 1.1 - The time correlated single photon counting (TCSPC) system ................... 5
Figure 1.2 - An experimental single exponential anisotropy decay ............................ 8

Figure 1.3 - The initial orientational distributions of excited state molecules for one-
photon, two-photon linear, and two-photon circular polarization ................... 13

Figure 2.1 — Structures of the substituted NBD chromophores used in this work .......... 23

Figure 2.2 — (a) Normalized absorption and emission spectra of BBD in the n-alcohols
methanol through l-pentanol. (b) Normalized absorption and emission spectra of
BBD in the polar aprotic solvents acetonitrile (ACN), N,N-dimethyl forrnamide
(DMF) and dimethyl sulfoxide (DMSO) ................................................ 26

Figure 2.3 - (a) Normalized absorption and emission spectra of SNBD in the n-alcohols
methanol through l-pentanol. (b) Normalized absorption and emission spectra of
SNBD in the polar aprotic solvents acetonitrile (ACN), N,N-dimethyl fonnamide
(DMF) and dimethyl sulfoxide (DMSO) ................................................ 27

Figure 2.4 - (a) Normalized absorption and emission spectra of NBDHA in the n-
alcohols methanol through l-pentanol. (b) Normalized absorption and emission
spectra of NBDHA in the polar aprotic solvents acetonitrile (ACN), N,N-dimethyl
fonnamide (DMF) and dimethyl sulfoxide (DMSO) .................................. 28

Figure 2.5 - Dependence of chromophore fluorescence lifetime on solvent polarity. The
solvent polarity is given in the form of the ET(3O) scale, and the fluorescence
lifetimes are in ps. CI = BBD, O = SNBD and if = NBDHA ....................... 30

Figure 2.6 - (a) Representative I”(t) and I i(t) data for SNBD in l-pentanol with one
photon excitation. (b) Corresponding R(t) function generated from the data
according to Eq. 1. The line through the data is the best fit single exponential
decay function, with the residuals of the fit being distributed around zero
intensity ...................................................................................... 35

Figure 2.7 - (a) Representative I”(t) and I _1_(t) data for SNBD in l-pentanol with linearly
polarized two photon excitation. (b) Corresponding R(t) function generated from
the data according to Eq. 2.6. The line through the data is the best fit single
exponential decay function, with the residuals of the fit being distributed around
zero intensity ................................................................................ 38

Figure 2.8 - (a) Representative I”(t) and I _L(t) data for SNBD in l-pentanol with circularly
polarized two photon excitation. (b) Corresponding R(t) function generated from
the data according to Eq. 2.6. The line through the data is the best fit single
exponential decay function, with the residuals of the fit being distributed around
zero intensity ................................................................................. 39

Figure 2.9 - (a) One photon experimental reorientation times (0) for BBD as a function
of solvent viscosity. The dashed line is calculated from the DSE model (Eq.
2.14), the dotted line is the calculated dielectric friction contribution (Eq. 2.15)
and the solid line is the sum of the two model contributions. (b) One photon
experimental reorientation times (0) for SNBD as a function of solvent viscosity,
with the dashed, dotted and solid lines having the same assignments as for panel
(a). (c) One photon experimental reorientation times (0) for NBDHA as a
function of solvent viscosity, with the dashed, dotted and solid lines having the
same assignments as for panel (a) ........................................................ 46

Figure 3.1 - TEM image of unilamellar vesicles produced by extrusion. The average
vesicle size is 109 :I: 28 nm based on 25 individual determinations ................. 65

Figure 3.2 - Dependence of NBDHA one-photon excited reorientation time constant on
solvent viscosity. The slope of the best-fit line is used to extract the viscosity-
dependence of the rotational diffusion constant DZ .................................... 76

Figure 3.3 - Normalized absorption and emission spectra of NBDHA in neat water
(dashed line) and in a solution containing DLPC vesicles (solid line). Spectra for
all vesicle-containing solutions were identical ......................................... 81

Figure 3.4 - Dependence of NBDHA reorientation on vesicle composition ................ 85

Figure 4.1 - Structures of compounds used in this work: (a) NBD-CleCP, (b) DSPC,
(c) DLPC, (d) sphingomyelin, and (e) cholesterol .................................... 94

Figure 4.2 — TEM micrographs of unilamellar vesicles. Scale bars indicate distances of
200 nm ....................................................................................... 99

Figure 4.3 - Normalized emission spectra for NBD-CIZHPC in unilamellar vesicles.
Emission maxima are centered at ca. 550 nm in all cases ........................... 103

Figure 4.4 - Lifetime decay of NBD-CszPC in unilamellar vesicles composed of DLPC.
The lifetime is a double exponential decay with 11:321 3: 22 ps and 12:1331 i
147 ps. ..................................................................................... 104

Figure 4.5 - Anisotropy decay for one-photon excitation of NBD-CIZHPC in unilamellar
vesicles composed of DLPC and cholesterol. Residuals of the fit to the data are

shown as solid circles. For this particular run, 1.’ = 195 i 27 ps and R(oo) = 0.15 :1:
0.01 .......................................................................................... 124

Figure 5.1 - Normalized absorption and emission spectra of NBD-PG in ethanol (left) and
structure of 16:0-12:O NBD-PG (right) ................................................ 137

Figure 5.2 - Example fluorescence lifetime data (points) for DLPG vesicles containing
NBD-PG, showing a fit to a double exponential decay functionality. (r1=327 ps,

12:2123 ps) The solid line through the data is the fit, with the residuals plotted as

fluctuations about zero intensity ........................................................ 140

xi

Figure 5.3 - Pressure-area isotherms for Langmuir-Blodgett monolayers of (a): DLPG (1 ),
DLPG + Cholesterol (2), DLPG + Cholesterol + sphingomyelin (3), DLPG +
NBD-PG (4), DLPG + NBD-PG + Cholesterol (5), DLPG + NBD-PG +
Cholesterol +sphingomyelin (6). (b): DSPG (1), DSPG + Cholesterol (2), DSPG
+ Cholesterol + sphingomyelin (3), DSPG + NBD-PG (4), DSPG + NBD-PG +
Cholesterol (5), DSPG + NBD-PG + Cholesterol +sphingomyelin (6) ............ 142

Figure 5.4 - Pressure-area isotherms for Langmuir-Blodgett monolayers of DSPC (1),
DSPC + Cholesterol (2), DSPC + NBDPC (3), DSPC + NBDPC + Cholesterol
(4), NBDPC (5) ........................................................................... 144

Figure 5.5 - Two photon excited anisotropy decay for vesicles containing DSPG, NBD-
PG and cholesterol. Points are experimental data, solid line is the fit of the data to
an exponential decay and the residuals of the fit are shown centered about the zero
intensity line ............................................................................... 148

xii

CHAPTER 1

INTRODUCTION

Lipid bilayers have attracted a great deal of attention in recent years due to the
crucial role they play in cellular processes. Plasma membranes are central to the function
of cells as they serve as a barrier to the external environment and support transmembrane
proteins. These proteins play a key role in metabolic reactions, viral and bacterial
infection, nutrient transport to cells, as well as drug delivery. In order to gain an
understanding of these processes, knowledge of the membranes and the environment

within lipid bilayers must first be gained.

Cell membranes were first described as lipid bilayers by Gorter and Grendel in
1925,1’2 and the lipid bilayer structure became the accepted model for cell membranes
after further studies by Davson and Danielli in 1935.3 However, at that time, it was
believed that the lipids constituting a membrane were fixed, with proteins covering the
outside of the bilayer. The fluid mosaic model introduced by Singer and Nicolson in
19723’4 proposed that the lipids constituting the membrane are not fixed, but rather are
fluid. This led to the idea of biomembranes as lipid matrices containing both peripheral
and integral proteins. A key element in this model is that the proteins and lipids
composing the bilayer have motional freedom. The fluid mosaic model is not entirely
sufficient to describe the bilayer and the motions within it, because the lipids exist in
several phases, and other components of the membrane (i.e. cholesterol) affect mobility
within the bilayer.3’4 Dynamics of proteins and lipids within the bilayer membrane

facilitate life processes, and in order to fully understand how such processes affect

cellular functions, it is first necessary to understand the structure and organization of the

membrane.

A major function of membrane lipids is to form amphipathic bilayers that
surround cells. Three common classes of lipids found in mammalian cell membranes
include glycerophospholipids, sphingolipids, and sterols, yet there is great variety in the
identity of the lipid head groups, and both length and unsaturations present in the acyl
chains. A characteristic of lipid bilayers is their transition temperature, Tm, the point at
which the lipid layer transitions from the solid phase, 3, to a liquid disordered phase, Id.
At a temperature above the transition temperature, the lipid monolayer or bilayer is fluid
and disorganized, lateral motion of lipid molecules is rapid, and there is very loose lipid
packing. At a temperature below the transition temperature, the lipids exist in a solid, gel-
like state, and lipid molecules that comprise the monolayer or bilayer are tightly packed

together.

Recent research has shown that plasma membranes are exceedingly complex
heterogeneous structures, and are composed of multiple different constituents.4'9 In
studies utilizing biomimetic membranes, it has been found that, upon mixing, rather than
random blending of lipids resulting in a homogeneous bilayer, the formation of micro-
domains is seen.5 ’8‘10’18 These structures, which are sometimes referred to as “lipid rafts”,

6,7,10,14,19-26

have been found to be enriched in cholesterol and sphingomyelin, yet the

question remains if these lipid raft structures form in actual plasma membranes.

The formation of such structures was first realized when certain membrane

fragments were found to be insoluble in cold non-ionic detergents such as Triton X-l 00.27

These detergent-resistant membranes, or DRMs, were enriched in cholesterol and
sphingolipids. It became apparent that detergents could solubilize membrane domains in
different physical states differentially.27 Insoluble domains were classified as a “liquid-
ordered” (10) phase, because they shared characteristics of both the Id and s phases —— high
lipid packing density, similar to the solid phase, yet like the 14 phase in that relatively
rapid lateral motion of lipid molecules is possible. The 10 state is distinct and coexists

with 1d or solid states.

Although the existence of lipid raft domains has yet to firmly be established in
plasma membranes, several studies have been performed that specifically address the
question as to whether raft structures exist, and under what circumstances they form.
Various methods have been used to explore lipid raft formation including fluorescence
microscopy, fluorescence recovery after photobleaching (FRAP), single particle tracking
(SPT), nuclear magnetic resonance (N MR), atomic force microscopy (AF M) and others.7'
9"3'15’28'33 Time-resolved fluorescence techniques have been used widely to investigate

biomimetic membranes due to the high sensitivity of such experiments.

The long term goal of this project is to quantitate the chemical, physical and
viscoelastic properties of lipid bilayer membranes. We produced model lipid bilayers in
the form of vesicles by the extrusion method.”36 This technique uses an extruder
composed of two syringes and a polycarbonate membrane with pores of specific
diameters. A lipid solution is forced through the membrane several times, and the result is
lipid vesicles with a diameter roughly equal to that of the pore sizes. Extrusion has been
shown to reproducibly produce unilamellar vesicles of a desired diameter. We formed

such structures with fluorescent molecular probes embedded within and tethered to the

3

bilayer, and monitored the lifetime and motional properties of the probe molecules using
time-correlated single photon counting (TCSPC). Specifically, experiments designed to
determine fluorescence lifetime and molecular reorientation have assisted in our
understanding of the polarity, viscosity, and organization within lipid bilayers. The model
bilayers used in this study were composed mainly of phospholipids, similar to plasma
membranes but, in order to study possible lipid raft structures, cholesterol and
sphingomyelin were used as well. Our experiments determine how the bilayer properties

change as a function of bilayer composition.

We used time-resolved fluorescence spectroscopy to probe lipid bilayer
heterogeneity and to examine the environment within a lipid bilayer membrane. In such
experiments, a fluorescent probe is selected, incorporated into the model bilayer, and
fluorescence lifetime and depolarization measurements provide insight into the
chromophore environment. For this work, experiments utilizing both one- and two-
photon excitation were used in order to determine the rotational diffusion coefficient of

the chromophore embedded in the bilayer.

All lifetime and anisotropy data were acquired using a time correlated single
photon counting (TCSPC) system (Figure 1.1). The source laser is a CW mode-locked
Nd:YAG laser (Coherent Antares 76—S) that produces 30 W average power at 1064 nm
with 100 ps pulses at 76 MHz repetition rate. The second (3W average power) or third
harmonic (l W average power) of the output of this laser excites a cavity dumped dye
laser (Coherent 702-2). For our experiments, based on the absorption properties of the
chromophore used, output was produced using a Rhodamine 610 dye (Exciton, 532 nm

pump) which produced light at ~650 nm for two- photon excitation experiments, or with

4

Stilbene 420 dye (Exciton, 355 nm pump), producing ~460 nm light for one photon
excitation experiments. The dye laser output is typically 5 ps pulses at a repetition rate of
3.8 MHz for both excitation wavelengths. The dye output beam is split into a reference

channel and a sample beam.

mode-locked YAG laser

  
 

 

 

cavity dumped dye laser

 

 

Figure 1.1 The time correlated single photon counting (TCSPC)
system

The sample is excited by a vertically polarized pulse and is collected 90° to the
excitation beam. Fluorescence was collected at 0°, 54.7° and 90° with respect to the
vertically polarized excitation pulse. Fluorescence signals are spectrally filtered using a
subtractive double monochromoator (American Holographic DBlO-S) which reduces the
effects of dispersion-induced time broadening of the collected 1i ght,3 7 and are detected
using a Hamamatsu R3 809U microchannel plate photomultiplier tube (MCP-PMT)
detector. The signals are processed using a Tennelec 454 quad constant fraction
discriminator and Tennelec 864 time-to-amplitude converter (TAC) and biased amplifier.

The TAC measures the time between the excitation pulse and the first arriving photon,

and is operated in reverse time mode to avoid dead—time limitations of the electronics.”38

The TAC counts are observed with an oscilloscope and are sent to a Tennelec PCA-II
multichannel analyzer (MCA). Collection wavelength, polarization and data acquisition

were controlled using National Instruments LabVIEW® 7.1 software.

The emission transients collected at 54.7° provide fluorescence lifetime data,
while those collected at 0° and 90° are used for fluorescence depolarization
measurements. Lifetime values provide information on the polarity of the environment in
which the probe molecule localizes. Specifically, in lipid bilayer environments where
drastic differences in polarity exist, the lifetime values can yield detailed information
regarding the chromophore’s surroundings, such as whether the probe molecule is in the
non-polar lipid tail region, near the glycerol backbone region, or near the bilayer-aqueous
layer interface. In addition, chromophores with unique fluorescent properties can be
selected to provide further detail on the specific location of probe molecules, and the
polarity of those surroundings. The decay functionality also provides insight as to the
number of environments sensed by the chromophore; a single exponential anisotrOpy
decay indicates the chromophore population exists in a single environment, or at least a
group of rapidly-exchanging environments, while a double exponential anisotropy decay
indicates the chromophore population is sampling multiple, distinct non-exchanging

environments.

Time-resolved fluorescence anisotropy measurements provide information about
the rotational dynamics of the molecule of interest. In such experiments, randomly
oriented sample molecules in the ground state are excited by the vertically polarized

excitation beam. The molecules in the sample with their absorption dipoles oriented

6

parallel to the incident light will be preferentially excited, leading to a non-random
excited state population. The primary means in which the sample molecules restore their
initial random orientation is through rotational diffusion. The rotational motion of a
molecule in solution or in a lipid bilayer is determined by a number of factors, including
the size and shape of the molecule, the viscosity of its environment, and the nature of the
chromophore’s interactions with neighboring molecules. Depending on these various
factors, the rotational motions of the probe molecules may be inhibited, and it is useful to
follow the rotational relaxation of solute molecules after photoselection. Fluorescence
depolarization measurements can provide direct measurements of the rotational decay
time, tor, which help characterize rotational dynamics by allowing determination of the
rotational diffusion coefficient. Such experiments have been used extensively to

3038-44

investigate the environment in both solvents and model membranes , and utilized

that information to gain insight about the organization in these systems.

Anisotropy measurements have been used extensively to investigate the rotational
dynamics of molecules in a range of environments, and there is a well established
theoretical framework for such measurements.37’38‘40’45 In these experiments, emission
transients are collected at 0° (parallel) and 90° (perpendicular) relative to the vertically
polarized excitation pulse. The raw data are combined to generate the induced
orientational anisotropy function (Eqn.1 .1):

: 1|[(t)—1_L(t)
I”(t)+21i(t)

 

R(t)
(1.1)

Where 1H (t) is the emission intensity collected at 0° relative to the excitation polarization,
and I i (t) is the emission intensity collected at 90° relative to the excitation polarization.

The value of R(t) is then plotted vs. time (Figure 1.2), and can exhibit up to five
exponential decay components, but the typical case is to observe one or two components.
As the molecules re-randomize, the anisotropy decays exponentially in time, until the

sample molecules are once again randomly oriented and the value of R is zero.

r(O) ..........

r(t)

‘l
l~l

. 7...] i,
l. ' ‘I . "Jim: ‘\.. 'll5.!.

i, I I . v ‘ ‘LI.

time (ps)

Figure 1.2 An experimental single exponential anisotropy decay

 

 

The reorientation time, “[0,, reflects rotational motion of the chromophore. For
chromophore molecules that are free to rotate in solution, the anisotropy decay curve
yields the reorientation time, and for a spherical molecule, to, is related to the rotational

diffusion constant, D, by the modified Debye-Stokes-Einstein equation46'48:

T ::7727
OR

where l7 is the viscosity of the surrounding medium, V is the hydrodynamic volume of the
probe molecule, T is the absolute temperature, k3 is the Boltzmann constant, f is a
frictional coefficient with values from 0 to 1, and S is a shape factor that is dependent on

the shape of the probe molecule.

In most cases, the probe molecule is not perfectly spherical, but has an ellipsoid
shape, and the relationship between the decay functionality of R(t) and the rotational
diffusion coefficients and the angle between the absorption and emission transition dipole
moments has been described in detail by Chuang and Eisenthal.40 They treated the form
of R(t) for a general ellipsoid, and related the experimental anisotropy decay to both the
orientation of the absorption and emission transition dipole moments and the rotational
diffusion coefficient by choosing Cartesian axes such that the z-axis was perpendicular to
the molecular plane, and the transition dipole lied along either the x- or y-axis.37’38’40'45
They treated the form of R(t) for a general ellipsoid, using two limiting cases, termed
prolate or oblate rotors, where the rotational motion around the axis that is distorted is no
longer equal to the rotation about the other axes. In the case of a prolate rotor, the
rotational diffusion coefficient about the x-axis is greater than that about either the y- or z-
axes. In the case of an oblate rotor, rotations about the z-axis are slower than rotations

about either the y- or x-axes. These differences are reflected in the values of the rotational

diffusion coefficients.

When the motional freedom of a molecule is restricted in some manner, such as
incorporation into a bilayer structure and/or tethered to an interface, the data are usually
treated using the hindered rotor model.32’37’44'49’S3 This model takes into account the fact

that the initially non-random chromophore population is not free to re-randomize

completely. The inability to re-randomize completely is manifested as a non-zero
anisotropy decay at long times. The decay time constant I, is related to 00, the semi-angle
of the cone which the tethered molecule sweeps out, and the “wobbling” diffusion

coefficient DW which describes the motion about it’s tethering bondz32’37”"4’49'53

79.?
1':

24D“, (13)

While the cone angle is a useful measure of the average disorder in the system, Dw
provides limited detail on chromophore motion. However, the Cartesian components of

DW can be separated to give a more complete picture of chromophore dynamics.

For a hindered rotor, the anisotropy decays according to Equation 1.4. R(O) is the
initial anisotropy value, and R(oo) is related to the size of the cone in which the
chromophore resides. A large cone angle means a large volume is swept out by the
chromophore indicative of an environment where the chromophore has substantial
freedom. Conversely, a small cone angle is reflective of the chromophore residing in a
comparatively restricted environment.

R(t) = R(oo) +[R(0) — R(oo)]expf’flm) (1 4)

For either model used, depending on the orientation of the probe molecule’s
transition dipole moments, there can be ambiguity in resolving the Cartesian components
of the rotational diffusional coefficient. Using one-photon anisotropy experiments in this
work, determination of only one of the Cartesian components of the rotational diffusion

coefficients is possible. Therefore the diffusion coefficients found by single photon

10

excitation do not provide a full description of rotational dynamics. For this reason, we
have also used the induced orientational anisotropy function using two-photon excitation.
In 1931, Maria Goppert-Mayer first predicted the ability of a molecule to
simultaneously absorb two photons.54 It was believed that two photons, of energies hvi
and hvz, when summed would correspond to an excited state of an atom. It was not until
the 1960’s, with the invention of the laser that two photon absorption was experimentally
observed. Kaiser and Garrett first observed the phenomenon in CanzEu2+ crystals in

1961 54.55

Two photon excitation (TPE) is a third order non-linear process where the

intensity of the fluorescence signal is dependent on the square of the incident powerz55

IccP2 (1.5)

Therefore, high peak intensities are necessary. TPE offers several advantages over
one-photon excitation (OPE).56 First, because of the longer wavelength, resonant one
photon absorption does not occur, minimizing sample heating and damage. The laser
source is focused onto a small area to achieve the high intensity necessary for two photon
absorption, resulting in less background interference. Another advantage of two-photon
excitation is the different selection rules for electronic transitions than for one-photon
excitation. Two-photon absorption requires an even-parity transition, while one-photon
absorption requires an odd parity transition between the ground and excited states. Thus
the ground and excited states can have the same parity in two-photon excitation. This
property of two photon excitation was widely used in the 1970’s to assign symmetry

properties to previously inaccessible excited states.”59 In two-photon excitation, two

11

photons are absorbed simultaneously, and so two electric field vectors are needed to
describe the interaction with the process, therefore the amplitude for the process is given

by the double projection of two field vectors onto a molecular tensor, represented by a

3x3 matrix:57’60‘6l
Sxx xy S x:
S = S.” 22' S)”:
SIX 53)} SI: (1 .6)

Reorientation dynamics probed by two-photon excited fluorescence can provide
more detailed information regarding rotational diffusion than by utilizing one-photon
excitation. The Johnson group and others have established a theoretical framework with
which to interpret anisotropy data gathered using two photon excitation.3 0’61'65 An added
benefit in utilizing TPE for study of lipid bilayers is that it uses lower energy photons so
thermal damage of the bilayer structure is less likely. This characteristic of TPE makes it

a particularly valuable technique for studying fragile biological membranes.

Reorientational dynamics probed by two-photon excited fluorescence can provide
complementary information regarding rotational motion of molecules. When used in
combination, data acquired using both means of excitation allow determination of the
Cartesian components of the rotational diffusion coefficient. For both modes of
excitation, the motional dynamics of the molecules is similar, however, the initial
anisotropic distribution of excited state molecules is dependent on different factors for the
two means of excitation. These experiments, in turn, give different experimental

reorientation times, and thus different molecular information.

12

For both one- and two-photon excitation, the value of the initial anisotropy
depends on the extent of photo-selection. If the molecule’s absorption transition moment
is oriented parallel to the excitation, they have the highest probability of absorption. That
probability for molecules to be excited by one-photon excitation is proportional to c0520,
and for two-photon excitation, the probability of absorption is proportional to c0540,
where 0 is the angle between the incident electric field polarization and the axis along
which the transition moment lies. Therefore, the two experiments yield different initial

distributions of excited molecules as displayed in Figure 1.3.

   

one-photon two-photon two-photon

(linear) (circular)

Figure 1.3 The initial orientational distributions of excited state
molecules for one-photon, two-photon linear, and two-photon circular
polarization

For one-photon excitation, the anisotropy depends on the relationship between the
absorption and emission transition dipole moments (uab and pm, respectively), while for
two-photon experiments, R(t) depends on the molecular tensor, S, and the emission

transition dipole moment (Eqns 1.7).

ram; (I) = r1 (0)(#ab (O) ‘ .110", (1))
r (’)=r2(0)(5(0)'#em(1)) (1.7)

13

The molecules comprising the initial excited state distribution will re-randomize
to become randomly oriented again, and by comparing the experimental data for one- and
two-photon excitation, a more complete and less ambiguous picture of the rotational
diffusion coefficients is determined.

For our experiments, we have chosen probes from the 7-nitrobenz-2-oxa-1,3-
diazole (NBD) family. The fluorescence behavior of this chromophore has been shown to
depend sensitively on its environment.3 9’41’66'68 Chapter 2 examines three substituted
NBD chromophores and discusses the role of the side group in mediating solution phase
rotational motion. Initial experiments in which an NBD chromophore has been
incorporated into a bilayer are presented in Chapter 3 where NBD-hexanoic acid was
inserted into a lipid bilayer composed of phospholipids, cholesterol, and a sphingolipid.
Experiments utilizing tethered chromophores and embedding them in lipid bilayers are
outlined in Chapter 4. Chapter 5 discusses experiments dealing with lipids with different
headgroups. Finally, Chapter 6 provides some final conclusions as well as presents ideas

for future research.

14

Literature Cited

(1) Chapman, D. Langmuir 1993, 9, 39.

(2) Gorter, E., Grendel, F. J. Exp. Med. 1925, 41, 439.

(3) Singer, S. J ., Nicholson, G.L. Science 1972, 172, 720.

(4) Subczynski, W. K. e. a. Cell. Mol. Bio Lett. 2003, 8, 147.

(5) Scherfeld, D.; Kahya, N.; Schwille, P. Biophysical Journal 2003, 85, 3758.
(6) Stottrup, B. L., Stevens, D.S., Kellar, S.L. Biophysical Journal 2005, 88, 269.
(7) Veatch, S. L. a. K., S. L. Biophysical Journal 2003, 85, 3074.

(8) Veatch, S. L. a. K., Sarah L. Biochimica et Biophysica Acta 2005, 1746, 172.
(9) Wilson, B. S. e. a. Molecular Biology of the Cell 2004, I5, 2580.

(10) Wisniewska, A.; Draus, J.; Subczynski, W. K. Cellular & Molecular Biology
Letters 2003, 8, 147.

(l 1) Crane, J. M., Tamm, L.K. Biophysical Journal 2004, 86, 2965.

(12) Dietrich, C., Bagatolli, L.A., Volovyk, Z.N., Thompson, N.L., Levi, M.,
Jacobson, K., Gratton, E. Biophysical Journal 2001, 80, 1417.

(13) Hancock, J. F. Molecular Cell Biology 2006, 7, 456.
(14) Hao, M.; Maxfield, F. R. Journal of Fluorescence 2001, 11, 287.

(15) Prenner, E., Honsek, G., Honig, D., Mobius, D., Lohner, K. Chemistry and
Physics of Lipids 2007, 145, 106.

(16) Silvius, J. R. Biochimica et Biophysica Acta 2005, 1746, 193.
(l 7) Simons, K.; van Meer, G. Biochemistry 1988, 27, 6197.

(18) Suzuki, T., Suzuki, Y. Biol. Pharm. Bull. 2006, 29, 1538.
(19) Jacobson, K., Dietrich, C. Trends in Cell Biology 1999, 9, 87.

(20) Bittman, R., Kasireddy, C. R. , Mattjus, P., Slotte, J .P. Biochemistry 1994, 33,
1 1776.

15

(21) Garmy, N., Taieb, N., Nouara, Y., F antini, J. Journal of Lipid Research 2007.
(22) Ramstedt, B., Slotte, J .P. FEBS Letters 2002, 531, 33.

(23) Rog, T., Pasenkiewicz-Gierula, M. Biophysical Journal 2006, 91, 3756.

(24) Smaby, J. M., Brockman, H.L., Brown, R.E. Biochemistry 1994, 33, 9135.

(25) Smaby, J. M., Momsen, M., Kulkarni, V. S., Brown, R.E. Biochemistry 1996, 35,
5696.

(26) Terova, B., Heczko, R., Slotte, J .P. Biophysical Journal 2005, 88, 2661.

(27) Xin-Min, L. M., M.M., Smaby, J. M., Brockman, H. L., Brown, R. E.
Biochemistry 2001, 40, 5954.

(28) London, E., Brown, D.A. Biochimica et Biophysica Acta 2000, 1508, 182.

(29) Birch, D. J. S. Spectrochimica Acta, Part A: Molecular and Biomolecular
Spectroscopy 2001, 5 7A, 2313.

(30) Girard, C.; Martin, O. J. F .; Dereux, A. Physical Review Letters 1995, 75, 3098.
(31) Johnson, J. M.; Ha, T.; Chu, S.; Boxer, S. G. Biophysical Journal 2002, 83, 3371.

(32) Korchowiec, B., Paluch, M., Corvis, Y., Rogalska, E. Chemistry and Physics of
Lipids 2006, 144, 127.

(33) Lipari, G.; Szabo, A. Biophysical Journal 1980, 30, 489.
(34) Polozov, I. V. a. G., Klaus. Biophysical Journal 2006, 90, 2051.
(35) Hunter, D. G.; Frisken, B. J. Biophysical Journal 1998, 74, 2996.

(36) Moscho, A.; Orwar, O.; Chiu, D. T.; Modi, B. P.; Zare, R. N. Proceedings of the
National Academy of Sciences of the United States of America 1996, 93, 1 1443.

(37) Patty Philipus, J .; Frisken Barbara, J. Biophysical Journal 2003, 85, 996.
(38) Lakowicz, J. R. Principles of Fluorescence Spectroscopy, 2nd ed., 1999.
(39) Blanchard, G. J. Applied Spectroscopy 2001, 55, 1 10A.

(40) Chattopadhyay, A.; London, E. Biochimica et Biophysica Acta, Biomembranes
1988, 938, 24.

16

(41)

(42)
9328.

(43)
6351.

(44)

(45)
327.

(46)

Chuang, T. J .; Eisenthal, K. B. Journal of Chemical Physics 1972, 57, 5094.

Dela Cruz, J. L.; Blanchard, G. J. Journal of Physical ChemistryA 2001, 105,

Greenough, K. P.; Blanchard, G. J. Journal of Physical Chemistry B 2006, 110,

Koan, M. M.; Blanchard, G. J. Journal of Physical Chemistry B 2006, 110, 16584.

Mazeres, S.; Schram, V.; Tocanne, J .-F.; Lopez, A. Biophysical Journal 1996, 71,

McPhie, P. Principles of Fluorescence Spectroscopy, Second ed. Joseph R.

Lakowicz, 2000; Vol. 287. pp. 291-333.

(47)
(48)
(49)
(50)
(51)
(52)
(53)
(54)
(55)
(56)
(57)
(58)
(59)
(60)

(61)

Blanchard, G. J. Journal of Chemical Physics 1987, 87, 6802.

Edward, J. T. Journal of Chemical Education 1970, 47, 261.

Zwanzig, R.; Harrison, A. K. Journal of Chemical Physics 1985, 83, 5861.
Perrin, F. Journal de Physique et le Radium 1936, 7, 1.

Karpovich, D. S.; Blanchard, G. J. Langmuir 1996, 12, 5522.

Kinosita, K., Jr.; Ikegami, A.; Kawato, S. Biophysical Journal 1982, 3 7, 461.
Kinosita, K., Jr.; Kawato, S.; Ikegami, A. Biophysical Journal 1977, 20, 289.
Szabo, A. Journal of Chemical Physics 1984, 81, 150.

Tao, T. Biopolymers.l969. 8. 609.

So, P. T., Masters, B. R. Micros. Res. and Tech. 2004, 69, 3.

Garrett, C. G. B., Kaiser, W. . Phys. Rev. Lett., 7, 229.

Jovin, T. M. e. a. Biophysical Journal 2000, 78, 1589.

McClain, W. M., Harris, R. A. Excited States; Vol. 3. pp.l-56.

McClain, W. M. Journal of Chemical Physics 1972, 5 7, 2264.

McClain, W. M. Journal of Chemical Physics 1973, 58, 324.

17

(62)
(63)
(64)
(65)
(66)
(67)

(63)

Blanchard, G. J ., J iang, Y. Journal of Physical Chemistry 98, 98, 6436.
Wan, C.; Johnson, C. K. Journal of Chemical Physics 1994, 101, 10283.
Callis, P. R. Journal of Chemical Physics 1993, 99, 27.

Chen, S. Y.; Van Der Meer, B. W. Biophysical Journal 1993, 64, 1567.
Johnson, C. K.; Wan, C. Topics in Fluorescence Spectroscopy 1997, 5, 43.
Wan, C.; Johnson, C. K. Chemical Physics 1994, 179, 513.

Fery-Forgues, S.; F ayet, J.-P.; Lopez, A. Journal of Photochemistry and

Photobiology, A.‘ Chemistry 1993, 70, 229.

18

CHAPTER 2

EVALUATING THE ROLE OF CHROMOPHORE SIDE GROUP
IDENTITY IN MEDIATING SOLUTION PHASE ROTATIONAL
MOTION

Introduction

The motion of a molecule in solution is controlled by a number of factors, ranging
from its size, shape and polarity to the nature of its interactions with its immediate
environment. It is these interactions that play a critical role in mediating processes such
as chemical reaction kinetics, for example. As a result of a broad effort over the past
decade or more, a series of chromophores has been identified that are capable of
providing information on their local environment with both their time- and frequency-
domain responses. Molecules such as pyrene are known to exhibit solvent-polarity
dependent fluorescence spectral profiles, and the fluorescence lifetime of this molecule
depends sensitively on the amount of oxygen present.l Other chromophores have been
developed that are tethered to specific structures, allowing them to locate in
comparatively well-defined regions of heterogeneous systems. The so-called lock-and-
key approach to the examination of crystallization phenomena makes use of
chromophores that have as pendant side groups the crystallizing entity?“1 This approach
to the optical interrogation of molecular-scale phenomena has proven to be useful to a

broad range of biological and chemical investigations.

One chromophore that has been used widely for the study of biological and

biomimetic systems is 7-nitrobenz-2-oxa-1,3-diazole, NBD.5’12 The reason for the

19

popularity of this chromophore is that its fluorescence spectral profile and lifetime both
depend sensitively on the polarity of its immediate environment. In addition, the
chromophore NBD is amenable to synthetic substitution or modification of its amine side
group, rendering it useful for localization within heterogeneous systems. A variety of
sterols, lipids and other compounds are available with a pendant NBD chromophore. A
recurring question in these studies is the role that the chromophore side group plays in
mediating its dynamics. In an effort to address this issue, we have undertaken a study of
the reorientation and fluorescence lifetime of NBD possessing several different side
groups in a series of polar protic and aprotic solvents. Our data indicate that the
reorientation dynamics of the NBD chromophore depend on the identity of the
chromophore side group to a limited extent and that we can account for the experimental
data in the context of a combination of frictional and dielectric interactions between the

dipolar chromophore and its immediate environment.

20

Experimental

Materials. The fluorescent probes (figure 2.1) succinimidyl 6-(N—(7-nitrobenz—2-
oxa-l,3-diazol-4-yl)amino) hexanoate (SNBD) and 6-(N -(7nitrobenz-2-oxa-1,3-diazol-4-
yl)amino hexanoic acid (N BDHA) were obtained from Molecular Probes Inc. and used
without further purification. The probe 4-benzylamino-7-nitrobenzofurazan (BBD) was
purchased from Sigma-Aldrich and was also used as received. The solvents methanol, 1-
propanol, l-butanol, l—pentanol, dimethyl sulfoxide (DMSO), N,N-dimethylformamide
(DMF) and acetonitrile (ACN) were purchased from Sigma-Aldrich in their highest
purity available and were used as received. Ethanol (95%) was distilled in-house. For
time-resolved fluorescence measurements, the chromophore concentrations were 10'4 M

or less in all cases.

Steady State Measurements. All absorption spectra were recorded on a Cary
model 300 double beam UV-Visible absorption spectrometer, with 1 nm spectral
resolution. All emission spectra were recorded on a Spex Fluorolog 3 spectrometer at a

spectral resolution of 3 nm for both excitation and emission monochromators.

Time Correlated Single Photon Counting Measurements. All fluorescence
lifetime and anisotropy data were collected using a time correlated single photon
counting (TCSPC) system that has been described in detail elsewhere.‘3 We provide a
brief recap of its salient properties here. The source laser is a CW mode-locked Nd:YAG
laser (Coherent Antares 76-S) that produces 100 ps 1064 nm pulses at 76 MHz repetition
rate. The second or third harmonic of the output of this laser is used to excite a cavity

dumped dye laser (Coherent 702-2), operating with Rhodamine 610 dye (Exciton, 532

21

nm pump) for two photon excitation experiments, or with Stilbene 420 dye (Exciton, 355
nm pump) for one photon excitation experiments. The dye laser outputs were 5 ps pulses
at a repetition rate of 4 MHz for both output wavelengths (460 nm and 650 nm).
Fluorescence transients centered between 535 and 545 nm were detected using a
Hamamatsu R3 809U microchannel plate photomultiplier tube detector with a Tennelec
454 quad constant fraction discriminator and Tennelec 864 time-to-amplitude converter
and biased amplifier used for signal processing. Data were collected using a virtual
instrument (VI) written with LabVIEW® 7.1 software. For this system the instrument
response time is typically 35 ps fwhm. Fluorescence transients were collected at

polarizations of 0°, 54.7° and 90° with respect to a vertically polarized excitation pulse.

22

I
O O
HN HN
N N
/ \ / \
/0 IO
N N
N02 N02
BBD SNBD

O—L'L‘

N02

NBDHA

Figure 2.1 Structures of the substituted NBD chromophores used in this work.

23

Results and Discussion

There are several aspects of our experimental data on the chromophores SNBD,
BBD and NBDHA that require discussion. The NBD chromophore is known to exhibit a
steady state spectral response that depends on the polarity of its local environment. We
show in Figs. 2-4 the absorption and emission spectra of the three chromophores in the
solvents used here. We find that the steady state spectra of the chromophores changes
very little with the identity of the alcohol solvent (Figs 2a — 4a), but that there is a marked
variation in absorption and emission maxima in the polar aprotic solvents ACN, DMF
and DMSO (Figs. 2b — 4b). For a given chromophore, the spectra in ACN are most
similar to those in the alcohols, and greater red-shifting of both absorption and emission
bands is seen for DMF and DMSO. Attempting to understand these spectral shifts in the
context of solvent properties or polarity indices does not yield any insight into the
chemical basis for these spectral shifts. Likewise, it appears that the Stokes shifts for
these molecules do not change significantly or in a manner correlated to any solvent
property or polarity index that we have tested. These spectra provide us with the
information needed to perform the time-resolved experiments, but give limited insight
into the chemical and/or physical basis for the observed band positions and Stokes shifts.
For this reason, we have examined the emission bands of these chromophores in the time

domain.

We consider next the relationship between the chromophore fluorescence lifetime
and the polarity of the medium in which the chromophore resides. We note that the
fluorescence lifetimes of BBD, SNBD and NBDHA are quite similar, and this is not

surprising. The NBD chromophore is known to exhibit a solvent polarity-dependent

24

fluorescence lifetime,5'7’9'll but the issue of what comprises “solvent polarity” is not well
understood. We have examined whether or not there is a correlation between the
fluorescence lifetime of the NBD chromophore and a variety of solvent properties such as
dielectric constant, refractive index, Debye dielectric relaxation time and solvent
viscosity. We have found no correlation between fluorescence lifetime and any of these
solvent properties, which is not surprising. The nanosecond time scale over which the
chromophore population decays following excitation differs from the characteristic time
scales which the above parameters sense, in many cases by several orders of magnitude.
The interactions between solvent and solute must derive from dipole-dipole and dipole-
induced dipole coupling, but the typical measures of solvent-solvent coupling listed
above do not access the region(s) of the frequency-dependent dielectric response of the

medium that dominate solvent-solute interactions.”15

The literature is replete with attempts to create solvent polarity scales, with the
most notable being the py scale,l6"7 the 115* scale'g'zl and the ET(30) scalezz’23 The py
scale is based on the emission spectrum of pyrene, which is sensitive to solvent mediation
of the vibronic coupling between its Si and S2 electronic states.1 For the 1r* and E700)
scales, the solvent “polarity” is gauged by the spectral shifts of a polar dye molecule,
with the mechanism(s) of these spectral shifts not being well understood at the molecular
level. Perhaps because of this lack of fundamental understanding, these latter two solvent
polarity scales have been related to various solvent properties through phenomenological

linear free energy relationships, with each factor weighted according to the creator(s) of

25

normalized linear response

normalized linear response

 

 

 

 

1 0 P pentanol methanol a
0.8 —
.. / \
0.6 l— / \
0.4 i
. ‘ \
0.2 / / \\\,
0.0 1 I n I I I n i l I
350 400 450 500 550 600 650
wavelength (nm)
1 0 ' DMF\ DhI/ISO DMF b
. - o» ~ DMSO
_ AC); AC3 /
0.8 —
0.6 — l
0.4 - I)
0.2 I\
0.0 r . I . I )1 1 L 1 l I I
350 400 450 500 550 600 650

wavelength (nm)

Figure 2.2 (a) Normalized absorption and emission spectra of BBD in the n-
alcohols methanol through l-pentanol. (b) Normalized absorption and emission
spectra of BBD in the polar aprotic solvents acetonitrile (ACN), N,N-dimethyl
formamide (DMF) and dimethyl sulfoxide (DMSO).

26

 

 

 

 

1.0: ”“1““
an r methanol
23. 0.8 - , ..
g ' l
,g 0.6 L ’
.6 .
13’ 0.4 -
”:5
g 0-2 \
a

0.0 I 1 I ' A “.2- - J 1 I

350 400 450 500 550 600 650
wavelength (nm)
- DMF

g 1.0 '- .t D7180 " b
g - ACN /
a 0.8 - \
:3 . DMF
.13 0.6 ~ /
3 . . DMSO
E 0'4 ‘ ‘ ACN
"c8 .
E 0.2 ,. if \
C \

0.0 . 1 . 1 1 . 1x “A... _..1 a 1

350 400 450 500 550 600 650

wavelength (nm)

Figure 2.3. (a) Normalized absorption and emission spectra of SNBD in the
n-alcohols methanol through l-pentanol. (b) Normalized absorption and
emission spectra of SNBD in the polar aprotic solvents acetonitrile (ACN),
N,N-dimethyl forrnamide (DMF) and dimethyl sulfoxide (DMSO).

27

 

 

normalized linear response

 

 

 

350 400 450 500 A 550 600 650

wavelength (nm)
DMF DMSO/

 

 

O -
g 1.0 - ACN\ DMSO
c, .
§ ()8 4
§ .
,5 ()6 — D00?
,3 .
,8 0.4R
13
§ 0.2K
C
0350 600 650

wavelength (nm)

Figure 2.4 (a) Normalized absorption and emission spectra of NBDHA in
the n-alcohols methanol through l-pentanol. (b) Normalized absorption and
emission spectra of NBDHA in the polar aprotic solvents acetonitrile (ACN),
N,N-dimethyl formamide (DMF) and dimethyl sulfoxide (DMSO).

28

the polarity scale. It is useful to note that despite the success of these polarity scales, the
system properties that are empirically factored into them are, for the most part, related to
the polarizability of the solvent in some manner. These polarity scales can, in fact prove
useful, as is the case here. We find that the fluorescence lifetime of the NBD
chromophore correlates reasonably well with the ET(30) polarity scale (Fig. 2.5, Table
2.1). This information is useful for predictive or comparative purposes of estimating the
polarity of the chromophore local environment. Because NBD derivatives find wide use
in the interrogation of biological systems such as lipid bilayer structures, the correlation
we report in Fig. 2.5 is of potential utility in understanding the average chemical
environments of complex biological systems. Unfortunately, the analogous correlation
does not exist in the frequency domain, between the steady state spectral maxima and the
ET(30) scale. This is not surprising given the extent to which inhomogeneous broadening

contributes to the spectral features of this complex organic chromophore.

We note that as the solvent polarity increases, the lifetime of the chromophore(s)
decreases. This is expected because of the polar nature of the chromophore. Despite the
poorly defined term of solvent polarity, we expect qualitatively that as the strength of
(polar) interactions increases between solvent and solute, the more opportunity there is
for structural and/or electronic wavefunction distortion of the solute. Such interactions
between solvent and solute serve to increase the opportunity for nonradiative decay from

the solute excited state.

We consider next the rotational diffusion properties of these chromophores.
Because the theoretical foundation of reorientation measurements is well established,

there is a wealth of information available from this type of measurement. We use both

29

10500
10000
9500
9000
8500
8000
7500
7000
6500
6000
5500

fluorescence lifetime (ps)

 

 

 

1?? D BBD
f o SNBD
_ if: NBDHA
l a
.. . O
f. 1:: ii! a
' e
I»
_ El ’ 15
o
_ a 1: air
i = t
:_ a . t
l
.. E
J I _1_ l 1 l 1 l L I 1 I
44 46 48 50 52 54 56

Solvent polarity (ET(30))

Figure 2.5 Dependence of chromophore fluorescence lifetime on solvent polarity.
The solvent polarity is given in the form of the ET(30) scale, and the fluorescence
lifetimes are in ps. CI = BBD, o = SNBD and A = NBDHA.

30

Table 2.1 Fluorescence lifetimes of the chromophores BBD, SNBD and NBDHA in
the solvents used in this work. Solvent polarities according to the ET(30) scale were
taken from Ref. 22. Uncertainties reported here are :hlo for at least six individual
determinations. Lifetimes were the same for both one and two photon excitation to
within the uncertainty of the measurement.

Solvent ET(30) BBD In (ps) SNBD In (ps) NBDHA tn (ps)

polarity
ACN 46.0 8385i22 9031355 10283:t51
MeOH 55.5 5585il3 6129il4 6136i21
DMF 43.8 8319127 8944i16 9208i35
EtOH 51.9 6136i10 6685i24 7078i32
PrOH 50.7 6208i16 6822i46 7129122
DMSO 45.0 7553i60 7764i52 83 54il9
BuOH 50.2 6620i20 7199i17 7515159
PeOH 49.1 6972i12 7622i23 7866i60

one photon and two photon excitation because they provide complementary information.
For both modes of excitation, we monitor emission from the same excited state. We
recap the information available from the two excitation methods and what we can extract

from the experimental data for BBD, SNBD and NBDHA.

One-Photon Excited Fluorescence Anisotropy. In one-photon excited anisotropy
measurements, the sample is excited by a vertically polarized light pulse, and the
resulting emission is collected parallel and perpendicular to the incident beam. The
induced orientational anisotropy is given by

I||(t)—1_L(t)
1H (t)+ 21i(t)

 

R (t) = (2.1)

Where I“(t) is the intensity of emission collected at a polarization angle of 0° with respect
to the incident vertical polarization, and I i(t) is the intensity of emission collected at a
polarization angle of 90° with respect to the incident polarization. Chemically useful
information is contained in the decay time constant(s) of R(t), Eq. 1. R(t) can contain up
to five exponential decays in theory, but only one or two decays are observed
experimentally. The values of R(O) can range from -0.2 to 0.4 for one-photon excited
anisotropy measurements, depending on the angle between the excited and emitting

transition dipole moments.

The relationship between the anisotropy decay obtained experimentally and the
molecular properties of the rotating molecule have been described by Chuang and
Eisenthal.24 These molecular properties include the rotational diffusion constants and the

angle between the excited. and emitting transition dipole moments. The typical

32

assignment of Cartesian axes to the chromophore are to have the z-axis perpendicular to
the molecular 1: system plane, and the transition dipole lying either along the x (long) axis
or the y (short) axis of the molecular plane. We follow this convention for the NBD
chromophores considered here. Two limiting cases can be examined to evaluate the
functionality of the anisotropy decay. By estimating the shape of the volume swept out
by the rotating molecule as oblate (Dz #3 Dx = Dy), or prolate (Dx at Dy =DZ) rotors, we can
simplify Chuang and Eisenthal’s equations significantly. Even with these simplifications,
there remains ambiguity in determining whether the prolate or oblate rotor shape best
represents the molecule in question because of the orientation of the transition dipoles
relative to the rotor axes. For parallel absorbing and emitting transition dipole moments,
oriented along the long (x) axis of the ellipsoid of rotation, the functional forms of R(t)

4
arez

oblate R(t) = 0.1exp(—(2DI + 41): )t) + 0.3 exp(—6Dxt) (2.2)
prolate R(t) = 0.4 exp(—6D:t) (2.3)
Similarly, for parallel absorbing and emitting transitions that are short (y) axis polarized,
oblate R(t) = 0.4 exp(—(4Dx + 2D: )t) (2.4)
prolate R(t) = 0.1exp(—(4DX +ZD__)t)+0.3exp(—6D__t) (2.5)

For the NBD derivatives we have examined, all anisotropy decays are single exponential,
(an example is provided in Fig. 2.6) a condition consistent with either a prolate rotor with

a long-axis polarized transition, or an oblate rotor with a short axis polarized transition.

33

Because of this ambiguity, we have also used two photon excitation to provide
complementary information and thus resolve the Cartesian components of the rotational

diffusion constant.

T wo-Photon Excited Fluorescence Anisotropy. Two photon excitation involves
the photoselection of an anisotropic orientational distribution via the two-photon tensor,
S, in contrast to the transition dipole moment, u, accessed by one-photon excitation.
Which two photon tensor elements are accessed depends on the polarization of the
incident light. Induced orientational anisotropy decay functions are generated for both

linearly (ri(t)) and circularly (r2(t)) polarized excitation and are given by:

linear linear
Ill (0 - I; (t)
lmcur [wear
1” (t)+21i (t)
Circular circular
1” (t ) — II. (0

Circular Circular
21ll (t)+]i (z)

 

r10) =
(2.6)

 

13(1):

For both excitation polarizations, the anisotropy decays are characterized by two time
constants, To and 12, weighted according to the spectroscopic and dynamical properties of

the chromophore.”27

r1 (1) = r1 (0) [c0 exp(—t / to) + c2 exp(—t / 12)]

(2.7)
r, (t) = r, (0) [atO exp(-t / To) + d, exp(—t / r2 )]

The anisotropy decay for either two photon excitation polarization should contain two
exponentials, but this is not seen in all cases. Our data are characterized by a single

exponential decay time constant, and there are two possible explanations for this finding.

34

 

 

 

 

 

6000 —
5000 —
3 4000 —
g .
8 3000 —
2000 —
1000 —
0.1 11Lm111l11+1l1111l1111lL111l
0 2500 5000 7500 10000 12500 15000
time (ps)
_ b
0.4 - f5
‘3‘ 0.2 -
P o.‘}?..°~.
offlllli'ai.'f7°’-a "'l». l , _ .. °:"'.'1..;'
0'0 ' ‘ ’ ‘ """l 1241...}..-
1 I 1 1 1 l 1 1 l 1 1 l 1 1 l 1 1 I
O 1000 2000 3000 4000 5000
time (ps)

Figure 2.6 (a) Representative I“(t) and I 1(t) data for SNBD in l-pentanol
with one photon excitation. (b) Corresponding R(t) function generated from
the data according to Eq. 1. The line through the data is the best fit single
exponential decay function, with the residuals of the fit being distributed

around zero intensity.

35

The first is that the two time constants To and 12 are similar enough that they cannot be
resolved outside of the experimental uncertainty. The other possibility is that some of the

prefactors in Eqs. 2.7 are too small to observe. The prefactors ci and d; have been

8,25-27

described in detail elsewhere, and we present them for the illustrative purposes.

xx ‘ yy

(fla + b)[3(fia + 1;)st + 3(-J§a + (2)5; + 2bS s + 4195;]

 

 

 

 

CO 2 2 . . (2.8)
7N (35... + 35;. + 25,,5” + 45;, )
(a — fib) [3(a — film; + 3(a + £11)sz + 2asnsw + 405;] 9
c = ' " ' , 2.
2 7N2(3S; +3513“, +2538”, +453) ( )
(fia + b)[(\/§a + ms; + (-J§a + 1353 — 4bSuSW + 6125;] 2 10
d: ' ' -
0 14N2(S; +53}. —SUSW +353.) ( )
d (a — \/§b) [(a — Jim; + (a + Jim; — 4asusfl, + @531 (2 1 1)
2 = .
14M2 (5; + 5.39, — sits” + 353),)
a = Jim), — DI)
bzzp~ —Dy —Dx +2A
. , (2.12)
N2 = a2 + b2

A4m+m+w—Qq—qQ—qu

In order to determine more information about the rotational diffusion coefficients, we
return to the limiting cases of prolate (DK 7: Dy =DZ) and oblate (DZ :6 DM = Dy) rotors.
Using these ellipsoidal shapes to simplify Eqs. 2.12 we obtain the prefactors in Eqs.2.8-

2.1] for the terms c and d as follows:

36

Prolate: J30+b=2\/3(D__ ~Dx) a—x/3b20

Oblate: fia+b=4(1): —D_,) a—fib=4J§(D, —D__)

Given the fact that we obtain a single exponential anisotropy decay for both one photon
(Fig. 2.6) and two photon (Figs. 2.7, 2.8) excitation in these experiments, the
experimental data are most consistent with the NBD chromophores reorienting as prolate
rotors in the systems we have examined. For a prolate rotor, the time constant To for two

photon excitation is

3(1),2 + 219213,)
2D, + Dz
8(Df + 212(1),)(65; — 3s; + S S + 253,)

XX )J'

 

S;(7Dx+17D:)—4Sfy(2Dx—Dz)+(S S +2s;,)[ +4Dx—Dz]

 

(2.13)

By setting Syy = 1 and extracting normalized values of Sxx and Sxy from r1(0) and r2(0),
then using the value of D2 obtained from the one photon measurements (Eq. 2.3), we can
extract D)( from the two photon excitation data. We present in Tables 2.2-2.4 the
experimental reorientation times and the resulting values of D2 and Dx that we extract
from our anisotropy data as a function of chromophore and solvent. As expected, the
reorientation times for one photon excitation and two photon excitation of each
chromophore differ, in some cases quite noticeably. This is an expected result, as is clear
from comparing Eq. 2.3 with Eq. 2.13, for example. These data show that while there are
subtle differences between the substituted chromophores, they all produce qualitatively
the same result of Dz/Dx ~ 2. This finding is significant in that it implies that the side

groups exert a limited influence on the motional properties of

37

1200 -

 

 

 

 

 

 

_ a
1000 -
m 800 -
§ 600 ; ~
0 .
400 -
200 —
0.|.41.1..L.lni..|....n....1....|
O 2500 5000 7500 10000 12500 15000
time (ps)
. b
0.4 —
T5“ 0.2 — i
P ’ - 2.2...." g . s ' . - 0 °
00 l w, ..'_. .;.g-‘.._:'|.‘ ’ '.I. '...'..‘ :1IOa;-:” :‘
1 l 1 1 1 l L n n l n n l l 1 .l n n I
O 1000 2000 3000 4000 5000
time(ps)

Figure 2.7 (a) Representative I”(t) and I i(t) data for SNBD in l-pentanol with
linearly polarized two photon excitation. (b) Corresponding R(t) function
generated from the data according to Eq. 2.6. The line through the data is
the best fit single exponential decay function, with the residuals of the fit
being distributed around zero intensity.

38

1400 _
1200

1 000

counts

800 '
600 '
400 '
200 '

 

 

, M.,

I I 1 I 1 I

I 4

 

 

1 l 1 1 1 1 1 1 1 1 1 1 1 1 1 1
5000 7 500 10000 12500 15000
time (ps)

2500

b

.. ... . . O ' ‘ I
D II “Amt"..iIP-l.‘ . ‘ - '2 - I I ”“1. I I. ' O in" I

l
. u 11 11
.I ' 'v' I' t , I‘I ‘ r’ ‘1‘ '1' I ' “ I I! .‘" 'I-ll ' 1" il n" 1| p.

I k I I l l l l l l l

 

0 1000

1 1 1 I 1 1 I
2000 3000 4000 5000

time (ps)

Figure 2.8. (a) Representative I”(t) and I i(t) data for SNBD in l-pentanol with
circularly polarized two photon excitation. (b) Corresponding R(t) function
generated from the data according to Eq. 2.6. The line through the data is the best
fit single exponential decay function, with the residuals of the fit being distributed
around zero intensity.

39

these molecules, which is likely the result of averaging motions over a large number of

side group conformations for all three chromophores.

We note that the assignment of a prolate effective rotor shape for these
chromophores is based on the observation of single exponential decays for both the one
and two photon excitation data. It is, of course, possible that there could be two
exponential decays present in our data that remain unresolved either due to the similarity
of the time constants or to the limited S/N ratio of the experimental data. In the limit that
the time constants are too similar to resolve, the physical picture that emerges from a
prolate rotor would change little, and the conclusions drawn from interpreting the data as
single exponentials would not change. If the latter were the case, we would expect to see
at least hints of deviation from single exponential behavior in the residuals of the fits to
the data. In no cases do we observe any such indication. While the interpretation of our
data are ultimately limited by the S/N ratio of the data, as is the case for many
experimental studies, the consistency and reproducibility of our findings lend confidence

to our interpretation of single exponential anisotropy decays.

We consider next how to relate our findings to solvent-solute interactions. The
relationship between molecular rotational motion and the properties of the surrounding

solvent medium has been described by the modified Debye-Stokes-Einstein model,28

 

V
7057-: = [:7 7(5‘ ,, (2.14)
B

40

Table 2.2 Experimental reorientation times for BBD using one photon excitation (1)
and two photon excitation with linear polarization (2L) and circular polarization
(2C). Values of Dx and D1 are extracted from the experimental data according to

Eqs. 2.3 and 2.13. Uncertainties reported here are :lzlo for at least six individual
determinations.

Solvent Viscosity 10R(1) TQR(2L) TQR(2C) Dz' Dx Dz/Dx

Tl, (GP) (PS) (PS) (133) (MHZ) (MHZ)

ACN 0.37 88d:9 72:1:15 921:9 1894 959 2.0

MeOH 0.54 96il3 79:1:11 793:6 1736 900 1.9

DMF 0.79 102i3 103i6 123:1:14 1634 846 1.9
EtOH 1.07 139i10 148i15 140i6 1199 595 2.0
lPrOH 1.95 199i10 205i7 199i11 838 419 2.0

DMSO 1.99 170:1:27 151:1:4 196i23 980 428 2.3

BuOH 2.54 267i25 249i19 25217 624 304 2.1

PeOH 3.61 3333:19 365:1:32 346:1:24 501 247 2.0

41

Table 2.3. Experimental reorientation times for SNBD using one photon excitation
(l) and two photon excitation with linear polarization (2L) and circular polarization
(2C). Values of D, and D2 are extracted from the experimental data according to
Eqs. 2.3 and 2.13. Uncertainties reported here are :tlo for at least six individual
determinations.

Solvent Viscosity TQR( 1) TQR(2L) TQR(2C) Dz Dx [3sz
'1» (GP) (P5) (P5) (P5) (MHZ) (MHZ)
ACN 0.37 125i3 81:1:22 763:2 1333 826 1.6
MCOH 0.54 182d:l5 113:1:6 112i2 916 497 1.8
DMF 0.79 177i5 1402127 136:1:11 942 535 1.8
EtOH 1.07 23 8:1:25 179:1:2 195:1:118 700 396 1.8
PI'OH 1.95 351i31 292:1:14 328327 474 269 1.8
DMSO 1.99 285i15 239il8 288i18 585 322 1.8
BuOH 2.54 398:1:18 324:1:19 396i5 419 246 1.7
PCOH 3.61 477i19 465i14 499i19 349 187 1.9

42

Table 2.4. Experimental reorientation times for NBDHA using one photon
excitation (1) and two photon excitation with linear polarization (2L) and circular
polarization (2C). Values of D, and D2: are extracted from the experimental data
according to Eqs. 2.3 and 2.13. Uncertainties reported here are £10 for at least six
individual determinations.

Solvent Viscosity TQR( l) TQR(2L) TQR(2C) Dz Dx Dz/Dx
n. (c?) 035) (135) (138) (MHZ) (MHZ)
ACN 0.37 50i3 523:3 49i3 3370 1887 1.8
MeOH 0.54 131i24 1123c10 111:1:10 1270 610 2.1
DMF 0.79 162i11 1343213 109i3 1030 599 1.7
EIOH 1.07 174i6 158:1:13 1 80:1:21 960 566 1.7
PrOH 1.95 302i13 283i12 339i50 550 324 1.7
DMSO 1.99 2733210 204i22 250:1: 1, 6 610 352 1.7
BuOH 2.54 570i26 402:1:20 426256 290 165 1.8
PeOH 3.61 724i15 456:1:4 539:1:35 230 128 1.8

43

Where the time constant for rotational motion, 131335, is determined from the anisotropy
decay data. The term 11 is the solvent bulk viscosity, V is the hydrodynamic volume29 of
the rotating moiety, f is a frictional interaction term, taken to be 1 for polar systems, and
S is the shape factor derived by Perrin3O to account for non-spherical solutes. The
quantity 12955 is equal to 6D], where D = ‘/2.(Dx+Dy+Dz). We can thus relate the

predictions of the DSE model to the experimental data.

We show in Tables 2.2-2.4 the experimentally measured one photon excited
reorientation times, and compare these data to the predictions of the DSE model with S =
0.667, calculated for a prolate rotor with an aspect ratio of two3 1’32 (Table 2.5). In all
cases, the experimental reorientation times are longer than those predicted by Eq. 2.14
(Fig. 2.9), dashed line vs. experimental data). Such a relationship between experiment

and model has been seen for a number of molecules,”35

and the origin(s) of such a
discrepancy have been ascribed to a number of factors. Certainly, some portion of this
difference arises from the DSE model itself. Eq. 2.14 was derived for an ellipsoidal
solute reorienting in a continuum solvent, thereby accounting only for frictional
interactions. To compound matters, the estimation of V and S can give rise to significant
uncertainty because of the intrinsic dynamics and variability in the range of structural
conformations that these chromophores can take on over time. Also, depending on the
solvent system used, strong and persistent solvent-solute interactions can give rise to a
rotating entity that is larger, on average, than the bare chromophore.4’34’35 Such a solvent-

attachment approach to understanding experimental reorientation data has been useful for

solvents capable of hydrogen bonding.36 For solvents that are not capable of

44

Table 2.5. Calculated dielectric friction time constants, DSE reorientation time
constants and their sum, reported at 1011- Solvent properties viscosity, dielectric
constant and Debye dielectric relaxation time constant are from references 34’37’38.
For the calculation of Toss (S=0.667) (Eq. 2.14), V = 215 A3 for BBD; V = 297 A3 for
SNBD; v = 245 A3 for NBDHA.” For the calculation of 1d,, a = 3.7 A and 11* = 10.6
D for BBD; a = 4.1 A and 11* = 12.2 D for SNBD; a = 3.9 A and 11* = 9.9 D for
NBDHA.

BBD SNBD NBDHA

SOIvent VISCOSIty Dielectric TD Tdf TDSE TOR Tdf TDSE TOR Tdf TDSE TOR

m (CH Const, 80 (ps) (ps) (138) (ps) (p8) (ps) (p8) (p5) (195) (p8)
ACN 0.37 38.0 4 0.2 30 30 0.2 40 40 0.2 33 33
MeOH 0.54 33.7 56 3.4 42 45 3.4 58 61 2.5 48 51
DMF 0.79 36.7 12 0.8 61 62 0.8 85 86 0.6 70 71
EtOH 1.07 24.5 337 30.6 82 113 29.3 115 144 24 94 118
PrOH 1.95 20.5 430 46.5 151 198 44.4 210 254 36 174 210

DMSO 1.99 46.5 21 0.8 153 154 0.8 214 215 0.6 177 178

BuOH 2.54 17.5 668 75.4 198 273 72.1 271 343 58 225 283

PeOH 3.61 13.9 927 129 280 409 123 388 511 99 321 420

45

10R (p8)

 

 

 

 

flflNNWbJ-D-ALII
UIOUIOUIOUIOUIO
OOOOOOOOOOO
'I'I'I'IjT'I'I'I'I'Ij

10R (p8)

 

 

 

 

0.

O
O
Lit
h—l
O
p—a
LII
N
O
[\J
kl]
DJ
0
b)
(II

M

O

O
'I'I'I'I'TTI‘T‘I

 

 

 

 

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5

solvent viscosity (cP)

Figure 2.9 (a) One photon experimental reorientation times (0) for BBD as a
function of solvent viscosity. The dashed line is calculated from the DSE model (Eq.
2.14), the dotted line is the calculated dielectric friction contribution (Eq. 2.15) and
the solid line is the sum of the two model contributions. (b) One photon
experimental reorientation times (0) for SNBD as a function of solvent viscosity,
with the dashed, dotted and solid lines having the same assignments as for panel (a).
(c) One photon experimental reorientation times (0) for NBDHA as a function of
solvent viscosity, with the dashed, dotted and solid lines having the same
assignments as for panel (a).

46

forming H-bonds, such an explanation is not appropriate. Another approach to
understanding such data has been to postulate that the reorienting molecule, in certain
circumstances, can possess added effective volume owing to the conformation of its side
groups. As noted above, such an explanation is difficult to support because of the labile
nature of the side groups that would have to be responsible for creating any such
“volume”, and the fact that within the most widely accepted model for the calculation of
hydrodynamic volume, there is no means to account for conformational variations within
a given molecule.29 A third explanation for the difference between experiment and
theory is that the frictional interactions between the solvent and solute are very strong,
giving rise to values of f that exceed unity. Such interactions have been examined
theoretically39 but values of f in excess of unity derived from experimental data are
necessarily suspect. Assigning physical or chemical significance to such an explanation
would hinge of the identification of strong, persistent solvent-solute interactions such as

hydrogen bonding or dipole-dipole interactions.

In addition to the reasons put forth above, there can be other contributions to the
interactions between solvent and solute. The dielectric coupling between the solvent
medium and the solute can also contribute to the solute reorientation times. Dielectric
friction is a mechanism for solvent-solute interactions that has a well established

theoretical foundation4048

and it has proven useful in explaining a host of reorientation
data for sometimes complex systems.37 In this model, there are two contributions to

solvent-solute interactions; viscoelastic frictional forces, which are accounted for through

the DSE model (Eq. 2.14), and dielectric interactions, which can be modeled by:

47

 

*2
7‘” = 1.2311631} (2'15)
Where [4* is the solute excited state dipole moment, a is the zero-frequency dielectric
constant of the solvent medium, I'D is the Debye relaxation time of the solvent and a is
the radius of the sphere approximating the solute cavity in the solvent. In this model, the
contribution of dielectric friction to the overall reorientation time of the chromophore
depends on the solute property u*2/a3, which effectively represents the torque exerted on
the solvent by the solute as it moves, the solvent susceptibility to this force, and the
characteristic time constant for solvent motions, I'D. It is this latter term that plays a
dominant role in determining the contribution from dielectric friction. For hydrogen-
bonding solvents, “CD is relatively long, reflecting the extensive H-bonding network
formed in alcohols, and in aprotic solvents, In is short, mediated primarily by dipole-
dipole or dispersion interactions. This dependence on 19 provides an important signature

that can be compared to experimental data, as we consider below.

A word is in order about the values of 11* we used in the calculation of Tar. We
have calculated the excited state dipole moments for BBD, SNBD and NBDHA at the
semi-empirical level, using Hyperchem® v. 6.0 with the PM3 parameterization. While it
would be preferable to obtain an experimental estimate of the dipole moments of these
molecules, none of them exhibit a sufficient solavatochromaticity nor are there any
reliable experimental data available on the ground state dipole moments of these

molecules. Thus we are left to estimate the values of 11*, and experience has indicated

48

that the PM3 parameterization provides useful results for polar organic chromophores

such as NBD.

We show in Table 2.5 and Figs. 2.9 the comparison of our experimental data to
model predictions where 10R (solid line) is the sum of 13055 (Eq. 2.14, dashed line) and Tdf
(Eq. 2.15, dotted line). There are a number of interesting features to be noted in the
comparison of the experimental data to the model. First, we note that in all cases the
experimental reorientation time is longer than that calculated by the model. As discussed
above, this systematic difference between model and data can be accounted for either in
the context of an underestimation of the frictional terms, a systematic error in the
calculation of the hydrodynamic volume, V, or direct solvent-solute interactions. We
note that the data for NBDHA, the only solute studied here that has a side group capable
of participating strongly in H-bonding interactions, provides the greatest difference
between model and experiment. It is likely that strong solvent-solute interactions, in
some cases approaching “solvent attachment,” are responsible for at least a portion of this
difference. We also consider that for BBD and SNBD, the experimental reorientation
times do not regress to a prediction of 10R ~ 0 for zero viscosity. Such behavior has been
seen before and it has sometimes been explained in the context of the inertial
contributions of molecular motion in the limit of zero viscosity. When inertial
contributions are calculated for dyes similar in size to these NBD derivatives, the inertial
contributions account for only a few ps at most.49 Another possible explanation is that
for low solvent viscosities there is a different boundary condition for frictional solvent-
solute interactions than there is for higher viscosity systems. We work with several

solvents of relatively low viscosity (ACN, MeOH, DMF) and if such a change in

49

boundary condition were to occur with increasing viscosity, it is likely that we would
sense it. The origin of the non-zero intercept for the data in Figs. 2.9 will require further

investigation to understand in detail.

For all of the chromophores we have examined here, there is a significant feature
in the experimental data. The data for 1-propanol (n=1.95 CF) and DMSO (n=1.99 cp)
are consistent in that for all chromophores the reorientation time of DMSO is always
faster than that of 1-propanol (Figs. 2.9). This finding stands in contrast to the DSE
model (Eq. 2.14) but is predicted accurately in the context of dielectric friction
contributions to the data (Eq. 2.15). The apparent anomaly results from the short ID of
DMSO compared to that for 1-propanol, and this feature is modeled well by Eq. 2.15.
This feature is not seen as prominently for the other polar aprotic solvents that are
characterized by short TD values because the ID values for the alcohols in the same
viscosity range are also short. There is insufficient contrast between methanol and
acetonitrile to resolve contributions from dielectric friction, but there is a small
observable effect for DMF and a prominent effect seen for comparing DMSO to 1-
propanol data (Fig. 2.9). We take the comparison of the functional form of the data and

the model to indicate that dielectric friction does, in fact, contribute to our results.

The offset between experiment and model cannot be explained fully in the context
of the DSE model. It is also possible that the contribution from dielectric friction has not
been estimated correctly owing to the limitations of the model. The systematic difference
between model and experiment could result from our use of the calculated solute excited
state dipole moments, or from the limitations associated with the assumption of the

spherical solute shape, as implied by the a3 term. In Table 2.5, where the frictional and

50

dielectric contributions to the overall model reorientation time are indicated, it is clear
that dielectric friction plays a more important role in alcohols than it does in aprotic
solvents, underscoring the importance of persistent solvent organization as expressed
through TD. While we have found literature values for ID, we recognize that this is a
difficult quantity to obtain and that there may be significant uncertainties in the values we
use. We note that for both the frictional and dielectric contributions to the model, when
there is a systematic difference between model and experiment, the implication is that
one of the quantities used in the calculations is consistently under-estimated. It is also
possible that there are other contributions to the experimental data that remain
unaccounted for, such as changes in the frictional boundary condition between solvent
and solute as the ratio of solvent to solute size varies systematically. If there is any
indication of such an effect, it appears to be most prominent for BBD (Fig. 2.9a) and
SNBD (Fig. 2%), where the smallest solvents appear to be modeled least well. Such a
finding would appear to stand in conflict with the DSE model, which was derived under
the assumption of a continuum solute, a condition approximated most closely by small
solvents and large solutes. Regardless of these subtle features in the data, there is overall

good agreement between model and experiment, as shown in Fig. 2.9.

We recognize that invoking dielectric friction as a contributor to our experimental
data is difficult to verify by other means, especially in light of other systems that have
shown varying contributions from this same effect. Several studies have concluded that
dielectric friction does not contribute significantly to their data, even with polar solvents
and solutes.”5 I In those cases, coumarin derivatives were used as the solutes and the

interpretation of the data may have been complicated by the presence of multiple excited

51

electronic states in close energetic proximity.52 In other cases, where systems where
comparatively less polar chromophores such as DPP and DMDPP have been used,
dielectric friction can contribute to the motional behavior of these molecules.53 It would
appear that the contribution of dielectric friction to experimental reorientation data can be
important, but the extent to which dielectric friction contributes is system-dependent.

The fact that the functional form of our experimental data match relatively well with the
predictions of Eq. 2.15 suggest that NBD derivatives do in fact experience dielectric

friction in polar solvents.

52

Conclusions

We have shown in this work that, despite the different side groups present on the
NBD chromophores, the dynamics of the molecules are fundamentally quite similar. In
all cases the fluorescence lifetime of the NBD chromophore exhibits the same solvent
polarity-dependence, indicating that the side groups are neither coupled strongly to the
electronic excited state of the NBD chromophore, nor is there a significant steric
interaction between side group and chromophore that interferes with chromophore-

solvent interactions.

The rotational diffusion data for both one and two photon excitation of the NBD
chromophore reveal remarkably similar motional behavior for the three molecules,
analogous to the lifetime data. The use of both one and two photon excitation provides
complementary information and allows the evaluation of the ratio of the Cartesian
components of the rotational diffusion constant. For the three molecules we have
studied, Dz/Dx ~ 2. When our experimental reorientation data are compared to the well-
established Debye-Stokes-Einstein model, we find that in all cases the experimental
reorientation time is substantially longer than predicted. We account for this discrepancy
to a substantial extent by considering the role that dielectric friction plays. When
dielectric friction is accounted for, the reorientation times predicted by model are much
more similar to the experimental data, including the apparent discrepancy between H-
bonding alcohol solvents and polar aprotic solvents. Taken collectively, our
measurements show that the identity of the side group on the NBD chromophore plays

little measurable role in determining the dynamics of the NBD chromophore, and that

53

both frictional and dielectric effects must be taken into account to understand the

comparatively strong interactions between NBD and polar solvents.

54

Literature Cited

(1) Karpovich, D. S.; Blanchard, G. J. Journal of Physical Chemistry 1995, 99, 3951.

(2) Rasimas, J. P.; Berglund, K. A.; Blanchard, G. J. Journal of Physical Chemistry
1996, 100, 17034.

(3) Rasimas, J. P.; Berglund, K. A.; Blanchard, G. J. Journal of Physical Chemistry
1996, 100, 7220.

(4) Rasimas, J. P.; Blanchard, G. J. Journal of Physical Chemistry 1995, 99, 1 1333.

(5) Chapman, C. F.; Liu, Y.; Sonek, G. J .; Tromberg, B. J. Photochemistry and
Photobiology 1995, 62, 416.

(6) Chattopadhyay, A.; London, E. Biochimica et Biophysica Acta, Biomembranes
1988, 938, 24.

(7) Fery-Forgues, S.; Fayet, J .-P.; Lopez, A. Journal of Photochemistry and
Photobiology, A: Chemistry 1993, 70, 229.

(8) Greenough, K. P.; Blanchard, G. J. Journal of Physical Chemistry B 2006, 110,
6351.

(9) Lin, 3.; Struve, W. S. Photochemistry and Photobiology 1991, 54, 361.

(10) Mazeres, S.; Schram, V.; Tocanne, J .-F.; Lopez, A. Biophysical Journal 1996, 71 ,
327.

(1 1) Mukherjee, S.; Chattopadhyay, A.; Samanta, A.; Soujanya, T. Journal of Physical
Chemistry 1994, 98, 2809.

(12) Tsukanova, V.; Grainger, D. W.; Salesse, C. Langmuir 2002, 18, 5539.
(13) Dewitt, L.; Blanchard, G. J.; Legoff, E.; Benz, M. E.; Liao, J. H.; Kanatzidis, M.

G. Journal of the American Chemical Society 1993, 115, 12158.

55

(14) Garg, S. K., NSmyth, C. P. Journal of Physical Chemistry 1965, 69, 1294.

(15) Nee, T. W.; Zwanzig, R. Journal of Chemical Physics 1970, 52, 6353.

(16) Dong, D. C.; Winnik, M. A. Photochemistry and Photobiology 1982, 35, 17.

(17) Doug, D. C.; Winnik, M. A. Canadian Journal of Chemistry 1984, 62, 2560.

(18) Kamlet, M. J .; Abboud, J. L.; Taft, R. W. Journal of the American Chemical
Society 1977, 99, 6027.

(19) Kamlet, M. J .; Hall, T. N.; Boykin, J .; Taft, R. W. Journal of Organic Chemistry
1979, 44, 2599.

(20) Taft, R. W.; Abboud, J. L.; Kamlet, M. J. Journal of the American Chemical
Society 1981, 103, 1080.

(21) Brady, J. E.; Carr, P. W. Journal of Physical Chemistry 1982, 86, 3053.

(22) Reichardt, C. Angewandte Chemie International Edition in English 1979, 18, 98.

(23) Elias, H.; Dreher, M.; Neitzel, S.; Volz, H. Zeitschriftfuer Naturforschung, Teil
B: Anorganische Chemie, Organische Chemie 1982, 3 7B, 684.

(24) Chuang, T. J .; Eisenthal, K. B. Journal of Chemical Physics 1972, 5 7, 5094.

(25) Johnson, C. K.; Wan, C. Topics in Fluorescence Spectroscopy 1997, 5, 43.

(26) Wan, C.; Johnson, C. K. Journal of Chemical Physics 1994, 101, 10283.

(27) Wan, C.; Johnson, C. K. Chemical Physics 1994, l 79, 513.

(28) Debye, P. Polar Molecules; Chemical Catalog Co.: New York, 1929.

(29) Edward, J. T. Journal of Chemical Education 1970, 47, 261.

56

(30) Perrin, F. Journal de Physique et le Radium 1936, 7, 1.

(31) Blanchard, G. J. Ultrafast Stimulated Emission Spectroscopy. In Topics in
Fluorescence Spectroscopy. Volume 5. Nonlinear and Two-Photon Induced
Fluorescence; Lakowicz, J. R., Ed.; Plenum Press: New York, 1997; Vol. 5; pp 255.

(32) Cantor, C. R.; Schimmel, P. R. Biophysical Chemistry, Vol. 1; Mir: Moscow,
USSR, 1984; Vol. 1.

(33) Dutt, G. B.; Srivatsavoy, V. J. P.; Sapre, A. V. Journal of Chemical Physics 1999,
110, 9623.

(34) Blanchard, G. J. Journal of Physical Chemistry 1988, 92, 6303.

(35) Blanchard, G. J .; Cihal, C. A. Journal of Physical Chemistry 1988, 92, 5950.

(36) Stevenson, S. A.; Blanchard, G. J. Journal of Physical ChemistryA 2006, 110,
3426.

(37) Simon, J. D.; Thompson, P. A. Journal of Chemical Physics 1990, 92, 2891.

(38) Maroncelli, M.; Maclnnis, J .; Fleming, G. R. Science 1989, 243, 1674.

(39) Srinivas, G.; Bhattacharyya, S.; Bagchi, B. Journal of Chemical Physics 1999,
1 10, 4477.

(40) Kivelson, D.; Spears, K. G. Journal of Physical Chemistry 1985, 89, 1999.

(41) Madden, P.; Kivelson, D. Journal of Physical Chemistry 1982, 86, 4244.

(42) Dote, J. L.; Kivelson, D. Journal of Physical Chemistry 1983, 8 7, 3889.

(43) Dote, J. L.; Kivelson, D.; Knobler, C. M. Journal of Chemical Physics 1980, 73,
3519.

57

(44) Dote, J. L.; Kivelson, D.; Schwartz, R. N. Journal of Physical Chemistry 1981,
85, 2169.

(45) Cross, A. J .; Simon, J. D. Journal of Chemical Physics 1987, 86, 7079.

(46) Hubbard, J. B.; Wolynes, P. G. Journal of Chemical Physics 1978, 69, 998.

(47) van der Zwan, G.; Hynes, J. T. Journal of Physical Chemistry 1985, 89, 4181.

(48) Vijayadamodar, G. V.; Chandra, A.; Bagchi, B. Chemical Physics Letters 1989,
161, 413.

(49) Blanchard, G. J .; Wirth, M. J. Journal of Physical Chemistry 1986, 90, 2521.

(50) Dutt, G. 8.; Raman, S. Journal of Chemical Physics 2001, 114, 6702.

(51) Homg, M.-L.; Gardecki, J. A.; Maroncelli, M. Journal of Physical ChemistryA
1997, 101, 1030.

(52) J iang, Y.; McCarthy, P. K.; Blanchard, G. J. Chemical Physics 1994, 183, 249.

(53) Dutt, G. B. ChemPhysChem 2005, 6, 413.

58

CHAPTER 3

QUANTITATING THE DYNAMICS OF NBD HEXANOIC ACID IN
HOMOGENEOUS SOLUTION AND IN SOLUTIONS CONTAINING
UNILAMELLAR VESICLES

Introduction

Lipid bilayers have attracted a great deal of attention because of the central role
these structures play in mediating life processes. As our understanding of these structures
has improved, it has become clear that plasma membranes are exceedingly complex
heterogeneous structures comprised of multiple different constituents, with the relative
amount of each depending on the function of the cell. Early work on understanding
plasma membranes pointed to the existence of some bilayer constituents that were not
detergent-soluble, and more detailed analysis of these insoluble fractions revealed that
they were enriched in cholesterol content. This finding, combined with work on multi-
component model bilayer systems lead to the notion that “lipid rafts” may exist in cellular
membranes.l Such rafts are thought to be regions of locally high cholesterol
concentration, with sphingolipids playing a supportive role in raft formation owing to the
ability of this family of lipids to hydrogen bond. While there is abundant evidence in the

2-4

literature for phase separation in two and three component bilayer model systems, the

existence of lipid rafts in plasma membranes remains to be established firmly.

We are interested in biomimetic lipid bilayer structures because of the possibility

that these structures can be used to support transmembrane proteins in their active forms.

59

A first step in this work is to understand the local environments formed by model bilayer
structures as a function of their composition. In order to gain information on these
heterogeneous systems from a molecular perspective, we are interested in imbedding
selected chromophores in these structures. In this paper, we report on the optical
properties of the chromophore (6-(N-(7nitrobenz-2-oxa-1,3-diazol-4-yl)amino hexanoic
acid), NBDHA. The NBD chromophore is used widely for investigations of bilayer
systems, and we focus here on both the details of chromophore motion in neat solvents
and in solutions containing unilamellar vesicles. We have chosen to study the hexanoic
acid derivative of this chromophore in an effort to understand where in the unilamellar
vesicle structures such an amphiphilic chromophore will reside. Our specific choice of a
chromophore possessing an amphiphilic side group that differs in structure from either
the phospholipid or cholesterol bilayer constituents was deliberate. We are interested in
understanding both polarity-mediated and structurally-mediated interactions within lipid
bilayers, and this work centers on polarity-mediated interactions. We have chosen to use
an NBD derivative because of the extensive knowledge base that already exists for this
chromophore.5 '” Our data in neat solvents reveal that NBDHA reorients as a prolate
rotor with an aspect ratio DZ/Dx ~ 2. For NBDHA in vesicle-containing solutions, we
find that the chromophore experiences the same environment regardless of vesicle
composition or the specific phospholipid used. This finding, in conjunction with steady
state absorbance and emission data, and fluorescence lifetime data, point to the NBDHA
residing in close proximity to the head group region of the vesicles. Because the pKa of

NBDHA is ca. 4.5 and the aqueous vesicle solutions are buffered to pH 8, the

60

chromophore exists in a predominantly deprotonated (anionic) form, facilitating

association with the phosphocholine headgroups of the lipids.

61

Experimental

Materials. The fluorescent probe 6-(N -(7nitrobenz-2-oxa-1,3-diazol-4-yl)amino
hexanoic acid (N BDHA) was obtained from Molecular Probes Inc. and used without
further purification. The solvents methanol, l-propanol, l-butanol, l-pentanol, dimethyl
sulfoxide (DMSO), N,N-dimethylformamide (DMF) and acetonitrile were purchased
from Sigma-Aldrich in their highest purity available. Water and ethanol (95%) were
distilled in-house. For time-resolved fluorescence measurements, the concentration of
NBDHA in solvents and the concentration of NBDHA in the vesicles for one—photon
experiments was 10'5 M, while the concentration of NBDHA in vesicles for two-photon

experiments was 10'4 M.

For the construction of unilamellar vesicles, the lipids 1,2-Di1auroyl-sn-Glycero-
3-Phosphocholine (DLPC, mp = -1°C), l,2-Distearoyl-sn-Glycero-3-Phosphocholine
(DSPC, mp = 55°C), egg sphingomyelin (SPM) and powdered cholesterol were obtained
from Avanti Polar Lipids, Inc. Although DLPC and DSPC are not major components of
plasma membranes, they were chosen based on their significant difference in transition
temperatures. The vesicle constituents were prepared in a 1:121 mole ratio of
phospholipid : cholesterol : sphingomyelin or a 2:1 mole ratio of phospholipid :
cholesterol for the multicomponent systems. For the mixed lipid systems, the ratio of
phospholipids used was 1 :1. A total of nine vesicle systems were produced with varying
composition. The vesicle compositions were as follows: DLPC, DLPC/cholesterol,
DLPC/cholesterol/SPM, DSPC, DSPC/cholesterol, DSPC/cholesterol/SPM,
DLPC/DSPC, DLPC/DSPC/cholesterol, and DLPC/DSPC/cholesterol/SPM. Chlorofonn

was evaporated from the lipid solutions, then the vesicle constituents were dissolved in a

62

4:1 benzenezmethanol solvent system, followed by another evaporation. The resulting
lipid mixture was then dissolved in a 10 mM solution of Tris® buffer (Aldrich) in Milli-Q
water. The Tris® buffer was adjusted for a pH ~ 8. An equal volume of the NBDHA-
containing solution (2x10'4 M) was added next for a final total concentration of the
vesicle constituents of 1 mg/mL, and 10“1 M in NBDHA. This probe concentration
corresponds to ~7000 chromophore molecules per 100 nm diameter vesicle. The average
distance between chromophores is calculated to be ~30 A, a distance that could give rise
to efficient Forster excitation transport. To test for this possibility, we compared our
time-resolved and steady state data for 10'4 M solutions to the same data for 10'5 M

solutions, and found that they were identical to within the experimental uncertainty.

We used the extrusion method'z'l5 to produce small unilamellar vesicles of
approximately 100 nm diameter. The lipid suspensions experienced five freeze-thaw-
vortex cycles to aid in mixing of the constituents. The suspensions were frozen in liquid
nitrogen for 5 minutes, followed by thawing in a hot water bath (5 minutes), then
vortexed for approximately 2 minutes. For vesicle formation, a syringe-based mini-
extruder from Avanti Lipids, Inc. was used. The lipid suspension was extruded eleven
times through a polycarbonate filter with 100 nm diameter pores to produce vesicles.

TEM measurements were used to verify vesicle size following extrusion.

Steady State Measurements. All absorption spectra were recorded on a Cary
model 300 double beam UV-Visible absorption spectrophotometer, with 1 nm spectral
resolution. All emission spectra were recorded on a Spex Fluorolog 3 spectrometer at a

spectral resolution of 3 nm for both excitation and emission monochromators.

63

Time Correlated Single Photon Counting Measurements. All lifetime and
anisotropy measurements of solvent and vesicle solutions were collected using a time
correlated single photon counting (TCSPC) system. This system has been described in
detail elsewhere and we provide only a brief recap of its salient features here. The source
laser is a CW mode-locked Nd:YAG laser (Coherent Antares 76-S) that produces 100 ps
1064 nm pulses at 76 MHz repetition rate. The second or third harmonic of the output of
this laser is used to excite a cavity dumped dye laser (Coherent 702-2), operating with
Rhodamine 610 dye (Exciton, 532 nm pump) for two photon excitation experiments, or
with Stilbene 420 dye (Exciton, 355 nm pump) for one photon excitation experiments.
The dye laser outputs were 5 ps pulses at a repetition rate of 4 MHz for both excitation
wavelengths. Fluorescence signals were detected using a Hamamatsu R3 809U
microchannel plate photomultiplier tube detector with a Tennelec 454 quad constant
fraction discriminator and Tennelec 864 time-to-amplitude converter and biased amplifier
used for signal processing. For this system the instrument response time is ~ 35 ps.
Fluorescence was collected at 0°, 54.7° and 90° with respect to the vertically polarized

excitation pulse.

Transmission Electron Microscopy. Samples were fixed with a 2% osmium
tetroxide solution in 0.1 M cacodylate buffer (pH ~7.4). TEM images (Fig. 3.1) were
acquired using a J EOL 100CX TEM operated at an accelerating voltage of 100 kV. The
TEM data reveal an average vesicle size of 109 i 28 nm based on 25 individual

determinations.

64

 

 

Figure 3.1 TEM image of unilamellar vesicles produced by extrusion. The average vesicle
size is 109 :t 28 nm based on 25 individual determinations.

65

Background

In this section, we consider the theoretical framework within which we interpret
time-resolved fluorescence anisotropy and molecular reorientation data for either one- or
two-photon excitation. For either mode of excitation, we monitor emission from the 81
state of the NBDHA chromophore. While the treatment of fluorescence anisotropy data

'6'19 we believe it is important to provide a brief discussion

has been reported elsewhere,
of this subject to facilitate the interpretation of our experimental data. We start with a

discussion of molecular reorientation data acquired using one photon excitation.

One-photon excited fluorescence anisotropy. For one photon excitation, the
exciting pulse is polarized vertically and fluorescence transients are collected at 90° with
respect to the excitation axis, at polarizations parallel (vertical) and perpendicular
(horizontal) to the excitation polarization. These raw data are combined according to Eq.

3.1 to generate the induced orientational anisotropy function, R(t) (Eq. 3.1)

 

= Ilia—Int)

R(t) 1r,,(t)+21l (t)

(3.1)

The chemically important information contained in R(t) is manifested in the number of
exponential decays recovered, their relative fractional contribution, and the time
constant(s) of the decay(s). While R(t) can contain up to five exponential decays, the
most common case is for a one component decay, with a two component decay being
observed under certain circumstances. The zero-time value(s) of R(t) range from -0.2 to
0.4 for one photon excitation, depending on the angle between the excited and emitting

transition dipole moments.

66

The relationship between the decay functionality of R(t) and molecular properties
including the rotational diffusion constants and the angle between the excited and
emitting transition dipole moments was described by Chuang and Eisenthal.16 That
treatment considers the functional form of R(t) for an arbitrary ellipsoid of rotation, with
limiting cases of either an oblate or prolate ellipsoid being used to simplify the general
expression for R(t). J udicious assignment of Cartesian axes to the rotating molecule can
serve to simplify the functionality of R(t). Generally, the excited transition dipole
moment is taken as the x-axis, and for nominally planar molecules such as NBD the TC-
system defines the x-y plane. For an oblate rotor, DZ at DK = Dy, and for a prolate rotor Dx
at Dy = DZ.16"8"9 For excitation and emission transition dipole moments aligned parallel

to one another and along the long (x) axis, R(t) decays according to Eqs. 3.2

oblate R(t) =(1—10—jexp(—(2Dx + 4D: )t) + [1—36Jexp(—6Drt)
(3.2)

prolate R(t) = ($Jexp(—6th)

If the transition moments are oriented along the z-axis, an oblate rotor will decay as a
single exponential with an exponential term of (4D,,+2DZ).'6"8'20 These various
possibilities have been treated elsewhere in detail, and for the outwardly simple case of a
single exponential decay, there can be ambiguity in resolving the Cartesian components
of D.'6 For this reason, we have also measured the induced orientational anisotropy
function of NBDHA using two-photon excitation. While the essential physics of
molecular motion remain the same regardless of the excitation mechanism, the means by
which the initial anisotropic distribution is excited can have a profound influence on the

functional form of the resulting fluorescence transients. By comparing the experimental

67

data for one- and two-photon excitation, we obtain a more complete and less ambiguous
picture of the rotational diffusion constant for NBDHA in the various environments

which we examine.

Two-photon excited fluorescence anisotropy. Reorientation dynamics probed by
two photon excited fluorescence can, in principle, provide more detailed, or at least
complementary information on the rotational motion that characterizes NBDHA in
solution. We are interested in using two-photon excitation because it is a nonlinear
process and thus provides selective excitation of the chromophore of interest without
unwanted background absorption that can give rise to heating in thermally sensitive
materials or in fragile biological systems. In our discussion of two-photon excited
fluorescence anisotropy data, we follow the treatment put forth by the Johnson group?“23

Our goal is to relate the two-photon excited anisotropy decays of NBDHA in selected

environments to the Cartesian components of its rotational diffusion constant.

Two-photon absorption is a third order nonlinear process, where two photons
from the same incident electric field are absorbed simultaneously. Two-photon
absorption is not an efficient process, and measuring the attenuation of the incident
electric field is not a practical means of detection. Typically, for two-photon resonant
excitation, the population placed in an excited electronic state as a result of two photon
absorption relaxes by internal conversion to the 8; state, where subsequent radiative
decay is detected. The components of the molecular hyperpolarizability that can be

accessed by the incident electric field(s) are represented by a tensor.21'24

68

For most planar chromophores such as NBD, a limited number of tensor elements
contribute to two photon absorption, which are typically written as a 2x2 tensor,
containing only 8..., Sxy(=Syx) and SW2"24 We are interested primarily in the relative

values of the tensor elements Sxx and SW, and thus normalize these values to SW =1.

Two-photon excitation provides information on the Cartesian components of the
rotational diffusion constant of a chromophore that is complementary to that available
from one photon excitation measurements. Because the details of the orientational
anisotropy relaxation depend on both the shape of the initial photo-selected anisotropic
population and on the two photon tensor elements accessed, the functional form of the
anisotropy decay will depend on whether circularly or plane-polarized light was used for
two photon excitation. The anisotropy functions generated from the linearly (rl(t)) and
circularly (r2(t)) polarized light are given by

In..." linear
.m: (:;>;,’;,.,-,<;3,
ll 1

Iiflrcular (t) _ Iiircu/ar (t)
circular Circular
21H (t)+1l (t)

 

(3.3)

 

r20):

Both zero-time anisotropies and anisotropy decay time constants are recovered from the
experimental data. For both excitation polarizations, the transient anisotropy decays are
characterized by the same two time constants, To and 1:2, each weighted according to the

spectroscopic properties of the chromophores. The zero-time anisotropies for these two

69

different excitation polarizations are related to the two photon tensor elements according

to E9- 3-4, for systems where Sxx. Sxy (:Syx) and Syy are the dominant elements. 2"23

 

 

“(O)=lz(su+sw) +(sn—5W)2+4sfy2+9(sfx_sfy)
7 2(SH+SW) +(SU_S ) +45;
2 2 (3.4)
13(s..-s,y) —(su+s,;) +1zsfy+6(s;_5§y)
5(0):; ‘ 2 - 2 2
(S~“+Syy) +3 Six—SW) +12Sxy

Knowing the experimental values of r1(0) and r2(0), we can determine the relative values
of Sxx and SW when Syy is normalized to unity. Because we are concerned with the
relative values of the tensor elements and not their absolute magnitudes, normalization of
the tensor elements to SW does not reduce the relevant information content of our

measurements.

Rotational diffusion will re-randomize the initially anisotropic orientational

distribution of the emitting molecules according to Eqs. 3.5, 2"23

rl (t) = rl (0)[cO exp(—t / to) + c2 exp(—t / 72)]

(3.5)
r2 (t) = r2 (0)[do exp(—t / To) + d2 exp(—t / r2 )]

where To and 12 are two different relaxation time constants. The relative contribution of
the two time constants is given by the prefactors c, and d,, which are in turn related to the

rotational diffusion constant and the two photon tensor elements, as described in Eqs. 3.6

 

— 3.10:2"23
(#3.. + b)[3(J3'a + ms; + “450 + b15122». + ”er5” + 4” S 3'1 (3 6)
CO _ 7N2(3Si + 3S): + ZSxxSyy + 4S:) 7 .

70

_ (a -— 73b) [3(a — has; + 3(a + fins; + 2asnsw + 4.153,]

 

 

 

c _ 3.7
2 7N2 (3s; + 35;, + zsnsyy + 4s; ) ( )
(fia + b)[(«/3a + as; + (—\/30 + ms; — 41.535”, + 61153,]
d = , 3.8
0 14N2(s§, +53, -s,,syy +3sfy) ( )
(a —- 73b) [(a — Ems; + (a + fins; - 4.1535” + 6.153;]
d, = (3.9)

141V2 (53, + sf; — Susy; + 353,)

The relative amplitudes of the decay time constants will provide information about the

values of the Cartesian components of the rotational diffusion constant. The parameters

a, b, N and A are related to the diffusion constant by

a = 3(Dy — Dx)
b=2Dz “D. —D)r +2A
' (3.10)
N 2 = a2 + b2
A = (of + 123+ of — 12,1); — DyD: — 0,13,)"2
We use these relationships'é’lg‘20 and those developed by Johnson and coworkers”'23 to

treat our two photon excited fluorescence anisotropy data on NBDHA in solvents and

unilamellar vesicle structures.

Eqs. 3.5 indicate that the two photon anisotropy decay should exhibit two
exponential components, and our experimental data yield single exponential decays in all
cases (vide infra). This result has been observed previously,” and obtains under the
condition that the experimental time constants are similar, with the single time constant
representing a weighted average of the two decay time constants. As is the case for

Chuang and Eisenthal’s treatment, invoking ellipsoidal rotor shapes for reorientation can

71

serve to simplify the form of the expressions for the decay time constants to and t2. Our
experimental data for one photon excitation suggest that NBD reorients as a prolate rotor,

Dx at Dy = Dz, (vide infra), and we apply this effective rotor shape to the equations for to

21-23

and 12. The equations derived by the Johnson group can be simplified to yield,

 

 

 

2
5,100, +170,)—4sf;(20, —0,) +(s,,s +252) 3(D. + 2134394.40, - 0,
PRO = . yy xy 20x + Dz (311)
" 8(0,2 +20,0,)(6s:, -3sf,, +s,,s,, +253)
2
85:,(20, + 0,) — 453,120, — 0,) —- (2s,,s,, + 353,) 60, — 50, + 9(1): + 200,)
PRO _ I I h 2D): + Dz (3 12)

 

T
r2 24(0,2 + 2D,D,)(25.i - Sf. — 2545.1 + 353)

The various factors that contribute to the two-photon excited anisotropy decay time
constants are not related in an intuitive manner to the rotational motion of the molecule,
in contrast to the time constant(s) recovered for one photon excitation. Treating NBD as
a prolate rotor, we can determine Dz (Eqs. 3.2) and DK (Eq. 3.13).

0 175:0, 4.12530 + 2s,,0, — 4&lefo — 161,530: — 8115,,0, + 241,0,2
" _ 7s; + 43,, — 967,52 0 — 321,530, —16s,,0,rI + 48r,D,

XX 2

(3.13)

 

We use this information to advantage in probing the immediate environment of NBDHA

in a series of solvent and unilamellar vesicle systems of varying composition.

72

Results and Discussion

As noted in the Introduction, we are interested in understanding the motional
dynamics of the chromophore NBDHA in both neat solvents and in the presence of
unilamellar vesicles. The issues of primary concern are obtaining a detailed
understanding of the dynamics of this chromophore, and whether or not it localizes to
specific region(s) within the unilamellar vesicle structures. We consider first the
dynamics of NBDHA in selected neat solvents. Previous literature reports have
demonstrated that the polarity of the environment surrounding the NBD chromophore
plays a significant role in determining its fluorescence lifetime.5 '10 Our data are

6'8'10 showing a decrease in fluorescence lifetime of

consistent with these previous results,
NBDHA with increasing solvent polarity. The excited state population decay of NBDHA
is single exponential in all cases, consistent with the expected first order dynamics of a
single emitting species. Fluorescence lifetimes of NBDHA in neat solvents are between
6 and 10 ns. The lifetime values determined by one-photon and two-photon excitation
are the same to within the experimental uncertainty, which is expected because radiative
emission originates from the same excited electronic state for both modes of excitation.

It is difficult to extract detailed chemical information from fluorescence lifetimes because
of the system-specific nature of such data and the absence of a comprehensive theoretical
framework relating excited state lifetime to either local environment or molecular

properties for systems where the emission bands experience significant inhomogeneous

broadening.

73

For these reasons, we turn to the information available from fluorescence
anisotropy data. For NBDHA in neat solvents, the one-photon excited fluorescence
anisotropy decays are single exponential. Based on Chuang and Eisenthal’s treatment, as
indicated in Eq. 3.2, the single exponential decay of R(t) indicates that NBDHA is a
prolate rotor with t = 1/(6Dz)16’20 The one photon reorientation times we recover are
related directly to the viscosity of the local medium, as described by the modified Debye-

Stokes-Einstein (DSE) equation, 16'18'20

W

T : 3.14
()R kBTS ( )

Where n is the solvent bulk viscosity, V is the solute hydrodynamic volume VNBDHA =
245 A3 ,25 f is the frictional solvent-solute boundary condition, and S is the solute shape
factor. We show in Fig. 3.2 the dependence of TOR on solvent bulk viscosity, and from
those data we determine that Dz/n = 877 i 51 MHz/0P. Despite this useful relationship,
there remains the potential ambiguity in determining the Cartesian components of D
based on the axis along which the excited transition moment is oriented. We are thus in
need of achieving a more thorough understanding of the motional dynamics of NBDHA.
In an attempt to lift this ambiguity, we have performed measurements of the transient

fluorescence anisotropy of NBDHA for two photon excitation.

For the reorientation data acquired using two-photon excitation, the recovered anisotropy
decay times are not related to the viscosity of the local medium in the same manner as the
data for one-photon excitation, but are a permutation of D7, Dx, and the two photon tensor
elements. The one-photon anisotropy data for NBDHA suggest that this chromophore

reorients as a prolate rotor in the neat solvents we have studied. It is also possible, under

74

certain circumstances, that the one-photon data could be consistent with the behavior of
an oblate rotor, depending on the orientation of the transition dipole moments relative to
the dominant rotor axis. In an effort to resolve this potential ambiguity and determine
whether NBDHA reorients as a prolate rotor or an oblate rotor, we have treated the two-
photon data assuming both rotor shapes. Assuming an oblate rotor shape, yields
physically unrealistic, negative values for Dx and D2 that are not consistent between the
one-photon and two-photon data. Based on these findings, we assert that NBDHA
reorients as a prolate rotor in the neat solvents we have studied. We report the values of
D2 and Dx in the solvents used here in Table 3.1. These data show that the ratio of
diffusion coefficients is DZ/Dx ~2 for all solvents, a result that is possible only by
comparing the two independent data sets, analogous to the information extracted from
anisotropy measurements on Rhodamine 640 when excited to two different initial

. 26
electromc states.

With this information in hand, we now have a relevant framework within which to
evaluate the behavior of NBDHA in the presence of unilamellar vesicles. We have
synthesized unilamellar vesicles from solutions containing 10‘4 M NBDHA. A variety of
unilamellar vesicle compositions were used to determine whether or not the NBDHA
chromophore functions as a sensor of vesicle structure and composition. The addition of
cholesterol and sphingomyelin to phospholipid vesicle structures is known to induce
phase segregated regions within the bilayer structures.2'4 Ongoing work in the Blanchard
laboratory shows that varying the composition of bilayers does, in fact, influence the
dynamics of the constituents.27 Those results were obtained using chromophores tethered

to specific bilayer constituents such as cholesterol or a

75

800 -
700 h
600 h
500 '

400

10,08)

300
200

100

 

 

I 1 I 1 I 1 I 1 I # I 1 I 1 I

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

solvent viscosity (cP)

Figure 3.2 Dependence of NBDHA one-photon excited reorientation time constant
on solvent viscosity. The slope of the best-fit line is used to extract the viscosity-
dependence of the rotational diffusion constant Dz.

76

phospholipid, and they thus provided information on the region(s) of interest within the
bilayer structure. A central goal of this work is to determine whether or not the NBDHA
chromophore locates within specific vesicle regions. The underlying question is the
structural homology required of a probe molecule in order for it to locate within specific

regions of the bilayer.

To help understand the location of NBDHA in vesicle systems, we consider first
the fluorescence lifetime of this chromophore. As noted earlier, there is limited
information available from fluorescence lifetime measurements by themselves, but when
such data are compared to NBDHA lifetimes in other systems, it is possible to garner
some useful environmental insight. For each vesicle system, the NBDHA excited state
lifetime decays exhibited a single exponential functionality. Previous literature reports
have shown that the fluorescence lifetime of NBD in an aqueous or polar environment is

d, 6’7’9’”) suggestive of strong interactions between solvent and

typically sub-nanosecon
solute, such as hydrogen bonding and dipolar coupling. We note that there is evidence in
the literature for direct chromophore-chromophore interactions between NBD moieties
under certain experimental conditions.ll Given both the fimctionality and time constant
of the lifetime data we recover for our systems, we do not believe that chromophore
dimers contribute significantly to our results. We recover the interesting result that,
regardless of vesicle composition, the NBDHA lifetimes are all ca. 1 ns, suggesting that
the environment sensed by the NBDHA chromophore is of similar polarity in all cases. If

this explanation is true, the steady state fluorescence spectra of NBDHA should be

similar for all systems, and we consider these data next.

77

The steady state spectra of NBDHA in vesicles and in water are shown in Fig. 3.3.
The NBDHA absorption band maximum and shape is substantially different in water than
it is in vesicle-containing solutions. The absorption maximum for NBDHA in vesicle-
containing solutions is ~475 nm, and is centered at ~484 nm in water. We conclude from
these data that NBDHA is not in the aqueous phase in vesicle-containing solutions. The
chromophore NBD is known to have a fluorescence quantum yield that depends
sensitively on the polarity of its environment. In water, NBD has a low quantum yield, 7’8
and in our measurements of NBDHA we were unable to detect a significant time-

resolved fluorescence signal in water. Based on both the steady state absorption data and

the time-resolved emission data, it is clear that in vesicle-containing solutions, NBDHA
does not reside in the bulk water region of this system, and that for all vesicle-containing

solutions the chromophore senses essentially the same environment.

We consider next the information content of the reorientation data for NBDHA in
vesicle-containing solutions. When considering anisotropy decays for NBDHA in the
presence of vesicles, we have to decide on the most appropriate model to use in
interpreting these data. Depending on where in the solution the chromophore resides,

either the DSE (free rotor, Eq. 3.14) modelzg'3O or the hindered rotor model3 "33 could be

used to interpret the data.

The hindered rotor model has previously been used to describe the motion of a
chromophore embedded in a membrane and must be considered here based on the steady
state absorption data. In this model, the motion of a chromophore is restricted by its

environment, which is modeled as a cone.

78

Table 3.1 Solvent-dependence of D2 and D, for NBDHA in neat solvents.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

solvent viscosity (cP) Dz (MHz) Dx (MHz) Dz/D,
acetonitrile 0.369 3370 i 6 1887 i 1 2.1
methanol 0.544 1270 i 18 610 i 1 1.7
DMF 0.794 1030 i7 599 i 1 1.7
ethanol 1.074 960 i: 3 566 i l 1.8
propanol 1.945 550 :1: 4 324 :t 1 1.8
DMSO 1.987 610i4 35211 1.7
butanol 2.544 290 j: 5 165 :t 1 1.7
pentanol 3.619 230 i 2 128 d: l 1.8

 

79

 

The cone angle that characterizes the chromophore motional freedom is related

to the measured time constant of the anisotropy decay according to Eq. 3.15.

7193
‘Z’ :
240,,

 

[3.15]

The time constant associated with motional relaxation of the chromophore is not
determined unambiguously because of the dependence of this quantity on both the cone

angle in which the chromophore is free to move and on the “wobbling” diffusion constant

of the chromophore as it moves about its bond to its anchoring functionality. We note
that in this model the measured time constant behaves differently than the quantity 10R
(Eq. 3.14), which is measured for a free chromophore in solution. For the hindered rotor,
the anisotropy decay time I scales with the cone angle swept out. For a restrictive
environment (small 00), a rapid time constant is measured because the chromophore
samples a small volume, and for a relatively free environment (large 00), t is long
because of the comparatively greater volume the chromophore has access to. In contrast,
for a free rotor, a fast time constant TOR is reflective of a low viscosity environment and a
slow time constant implies a high viscosity, restrictive environment. Also, for hindered
rotors, the anisotropy does not decay to zero at t = 00 as it does in bulk solutions, because
a restricted range of motional freedom is available to the chromophore. In principle, the
observation of a non-zero R(oo) value would indicate that the hindered rotor is the
appropriate model. Because we are using vesicles, however, and restriction in the

orientational distribution associated with chromophore confinement within the bilayer is

80

1.0 —
A
“O
Q)
.53, 0.8 ..
E
8
v 0.6 —
<1)
U)
:1
O
9;
4,3 0.4
In
8 r.
.5
'— 0.2 - -.

0.0 - " " .............................

L I l I l I I I l J l I l I

 

 

350 400 450 500 550 600 650 700

wavelength (nm)

Figure 3.3. Normalized absorption and emission spectra of NBDHA in neat water
(dashed line) and in a solution containing DLPC vesicles (solid line). Spectra for all
vesicle-containing solutions were identical.

81

orientationally averaged over 41: steradians macroscopically owing to the spherical shape
of the vesicles, R(oo) = 0 for all measurements. The choice of the model most appropriate

for the interpretation of our data must therefore be made based on other factors.

We consider that, if NBDHA were bound to the surface of the unilamellar
vesicles, it would enjoy substantial motional freedom. We make this structural argument
based on the fact that, for the lipids we use, there is a cationic quaternary ammonium
species in closest proximity to the water at the lipid-solution interface. Because we make
our measurements at pH 8, the NBDHA is fully deprotonated (precluding carboxylic acid
tail-to-tail chromophore dimer formation), and if the chromophore were to interact with
the vesicle, as suggested by the steady state absorption data, an ionic interaction would be
possible. In this picture, the nominally tethered chromophore would experience a cone
angle of 200 ~ 180°, which is a limiting case that renders the chromophore equivalent to a
free rotor. Also, the strength of the ionic interaction postulated above would be on the
order of several kcal/mol at most, and the persistence time for such interactions would be
on the order of hundreds of ps. For both of these reasons, we assert that our reorientation
data for NBDHA in the presence of vesicles are best understood in the context of the free
rotor model rather than the hindered rotor model. Interpreting our reorientation data
within the framework of the free rotor model yields the result that NBDHA reorients as a
prolate rotor in the presence of vesicles, qualitatively the same as for its behavior in neat
solvents. We have synthesized vesicles in a range of compositions, including the use of
two different lipids with melting points substantially below (DLPC, mp = -1°C) and
substantially above (DSPC, mp = 55°C) room temperature, and with the addition of

cholesterol and sphingomyelin. These different vesicle compositions are known to

82

produce phase-segregated bilayer structures, and the local environments created by these
different phases vary widely in both their polarity and rigidity.“’34 Despite the wide
range of environments within the vesicles that are available to NBDHA, we recover
essentially the same one-photon excited reorientation times, regardless of vesicle
composition.(Table 3.2, Fig. 3.4) While there may be some slight variation between
individual systems, these differences are barely different from the experimental
uncertainty. For this reason we do not attempt to provide any more detailed, system-
dependent interpretation of these values. We also note that, while the two photon excited
anisotropy decays produce different time constants than those seen for the one photon

data, as expected, the same behavior is seen for all of the anisotropy decay time data sets.

Treatment of the one-photon excited anisotropy decay times of NBDHA in
vesicle-containing solutions in the context of a free rotor allows Ton to be related to the
viscosity of the environment through the DSE equation (Eq. 3.14), and the observation of
single exponential anisotropy decays allows assignment of a prolate rotor shape to
NBDHA in the presence of vesicles, based on our measurements of NBDHA in neat
solvents (vide supra). Comparing the reorientation times of NBDHA in vesicle-
containing solutions to the same data for NBDHA in neat solvents indicates that the local
environment of NBDHA is characterized by a viscosity of ~1 cP, a value fully consistent
with an aqueous environment. These data, taken in conjunction with the steady state data
(Fig. 3.3), suggest that the probe resides in the polar headgroup region, in the immediate

vicinity of the bilayer.

83

Table 3.2. Dependence of NBDHA one- and two-photon excited anisotropy decay
time constants on vesicle composition. Abbreviations: DLPC = 1,2-DilauroyI-sn-
Glycero-3-Phosphocholine; DSPC = l,2-Distearoyl—sn-Glycero-3-Phosphocholine;

chol = cholesterol; spm = sphingomyelin.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Vesicle system Ton (one tum, 1111111111 1" (ps) Dz/D,
photon, (two (two
ps) photon, photon,
98) pS)
DLPC 163i10 121i5 102i14 1173i88 1.8
DLPC/chol. 166 i 4 127 d: 5 95 :1: 4 1168 at 66 1.8
DLPC/chol/spm 194 i 5 132 d: 19 106 :t 16 1177 :b 62 1.7
DSPC 186:1:6 126:1:8 97:9 1144i45 1.8
DSPC/chol. l76i6 109d:6 107i10 1132i23 1.7
DSPC/chol/spm 165 i 4 122 d: 4 97 i 10 1145 :1: 44 1.7
DLPC/DSPC 159:1:6 125:1:5 102:1:4 1157:48 1.9
DLPC/DSPC/chol. 161 :t 8 126 :b 8 89 :L- 9 l 147 i 40 1.7
DLPC/DSPC/chol/spm 186 d: 5 87 :1: 2 75 i 5 1067 :1: 41 2.0

 

84

 

 

 

220 -

200 -

aaflofiommeofio

666me5an
88639
m
emflofiommo mm
m
p
350me m
b
88 .m
V
eamsofibxfio
2385
2.5

Figure 3.4 Dependence of NBDHA reorientation on vesicle composition.

85

The two-photon excited anisotropy decays for NBDHA in solvents were modeled
as both a prolate and an oblate rotor to determine which model provided results that were
consistent with the one photon data. We have treated the two photon data for NBDHA in
vesicle-containing solution the same way, and the results of this treatment are that a self-
consistent result is obtained only if NBDHA is modeled as a prolate rotor. For the two-
photon excited anisotropy data, single exponential decays are seen, and the reorientation
times for all systems are the same for both linearly and circularly polarized excitation,
regardless of vesicle composition. We apply Eq. 3.13 to our data to extract the quantity
Dx. Combining the two-photon and the one-photon data we recover a ratio Dz/Dx~2,
(Table 3.2) the same ratio that we determined for NBDHA in neat solvents. The
consistency of the results between the two different types of environments argues for the
applicability of the free rotor model for the interpretation of NBDHA molecular motion

in vesicle-containing solutions.

We note that it is, in principle, possible for the NBDHA chromophore to be
incorporated to a limited extent (shallow incorporation) into the bilayer structure, and
under this condition it would sense a polar environment that is likely not too different
than that of bulk water.”38 We believe that the reorientation data we have presented
allow us to distinguish between the possibility of shallow incorporation and the
chromophore being at the bilayer interface. If the chromophore were incorporated to any
significant extent within the bilayer structure, the ratio of Dz/Dx would be altered by
virtue of its spatial confinement. We find experimentally that the ratio Dz/Dx for
NBDHA in vesicle-containing solutions is the same as that in neat liquid solvents,

indicating that it resides in an environment that affords substantial orientational freedom

86

that varies very little if at all with bilayer composition. The only environment that fits

these criteria fiilly is the interface between the bilayer and the bulk liquid.

87

Conclusions

We have studied the steady state and time resolved optical properties of the
chromophore NBDHA in a series of neat solvents and in vesicle-containing solutions.
We have used both one photon excitation and two photon excitation for our anisotropy
measurements and analyzing these data to find a self-consistent solution allows us to
determine that the chromophore reorients as a prolate rotor with Dz/Dx ~ 2 in all cases.
For NBDHA in vesicle-containing solutions we observe reorientation dynamics that are
independent of vesicle composition, even though the steady state absorption and
fluorescence lifetime data on these systems indicate that the chromophore does not reside
in a purely aqueous environment. The location of the NBDHA chromophore that is most
consistent with the entire body of steady state and time-resolved emission data is that the
chromophore resides at the bilayer-solution interface, near the zwitterionic
phosphocholine headgroups. In this region, interactions between the headgroups and the
NBDHA are possible. All solutions are at pH ~8, therefore the NBDHA is essentially all
deprotonated, allowing ionic interactions to dominate in this system. The larger
implication of this work is that there needs to be a very close match between
chromophore side group functionality and a bilayer constituent for there to be
incorporation into specific regions of the bilayer structures. We do note that the
chromophore interaction with the vesicles is selective for the phospholipid regions, but in

a way that is not sensitive to any lipid local organization.

88

Literature Cited

(1) Simons, K.; van Meer, G. Biochemistry 1988, 2 7, 6197.

(2) Crane, J. M.; Tamm, L. K. Biophysical Journal 2004, 86, 2965.

(3) Hao, M.; Maxfield, F. R. Journal of Fluorescence 2001, 11, 287.

(4) Scherfeld, D.; Kahya, N.; Schwille, P. Biophysical Journal 2003, 85, 3758.

(5) Chapman, C. F.; Liu, Y.; Sonek, G. J .; Tromberg, B. J. Photochemistry and
Photobiology 1995, 62, 416.

(6) Chattopadhyay, A.; London, E. Biochimica et Biophysica Acta, Biomembranes
1988, 938, 24.

(7) F ery-F orgues, S.; Fayet, J .-P.; Lopez, A. Journal of Photochemistry and
Photobiology, A: Chemistry 1993, 70, 229.

(8) Lin, 8.; Struve, W. S. Photochemistry and Photobiology 1991, 54, 361.

(9) Mazeres, S.; Schram, V.; Tocanne, J .-F.; Lopez, A. Biophysical Journal 1996, 71,
327.

(10) Mukherjee, S.; Chattopadhyay, A.; Samanta, A.; Soujanya, T. Journal of Physical
Chemistry 1994, 98, 2809.

(1 1) Tsukanova, V.; Grainger, D. W.; Salesse, C. Langmuir 2002, 18, 5539.
(12) Hunter, D. G.; Frisken, B. J. Biophysical Journal 1998, 74, 2996.
(13) Johnson, J. M.; Ha, T.; Chu, S.; Boxer, S. G. Biophysical Journal 2002, 83, 3371.

(14) Moscho, A.; Orwar, O.; Chiu, D. T.; Modi, B. P.; Zare, R. N. Proceedings of the
National Academy of Sciences of the United States of America 1996, 93 , 11443.

89

(15)

(16)

(17)

(18)

(19)
287.

(20)

Patty Philipus, J .; Frisken Barbara, J. Biophysical journal 2003, 85, 996.

Chuang, T. J .; Eisenthal, K. B. Journal of Chemical Physics 1972, 5 7, 5094.

Jiang, Y.; Blanchard, G. J. Journal of Physical Chemistry 1994, 98, 6436.

Blanchard, G. J. Journal of Chemical Physics 1987, 87, 6802.

McPhie, P. Principles of Fluorescence Spectroscopy, Second Edition 2000; Vol.

Tao, T. Time-dependent fluorescence depolarization and Brownian rotational

diffusion coefficients of macromolecules. PhD Thesis, 1969.

(21)

(22)

(23)

(24)

(25)

(26)
9328.

(27)
16584.

(28)

(29)

(30)

Johnson, C. K.; Wan, C. Topics in Fluorescence Spectroscopy 1997, 5, 43.

Wan, C.; Johnson, C. K. Chemical Physics 1994, 179, 513.

Wan, C.; Johnson, C. K. Journal of Chemical Physics 1994, 101, 10283.

Callis, P. R. Journal of Chemical Physics 1993, 99, 27.

Edward, J. T. Journal of Chemical Education 1970, 4 7, 261.

Dela Cruz, J . L.; Blanchard, G. J. Journal of Physical Chemistry A 2001, 105,

Koan, M. M., Blanchard, G. J. . Journal of Physical Chemistry B 2006, 110,

Debye, P. Polar Molecules; Chemical Catalog Co.: New York, 1929.

Perrin, F. Journal de Physique et Ie Radium 1936, 7, 1.

Zwanzig, R.; Harrison, A. K. Journal of Chemical Physics 1985, 83, 5861.

90

(31) Kinosita, K., Jr.; Ikegami, A.; Kawato, S. Biophysical Journal 1982, 37, 461.

(32) Kinosita, K., Jr.; Kawato, S.; Ikegami, A. Biophysical Journal 1977, 20, 289.

(33) Szabo, A. Journal of Chemical Physics 1984, 81, 150.

(34) Wisniewska, A.; Draus, J .; Subczynski, W. K. Cellular & Molecular Biology
Letters 2003, 8, 147.

(35) Costa, E. J. X.; Shida, C. S.; Biaggi, M. H.; Ito, A. S.; Lamy-Freund, M. T. FEBS
Letters 1997, 4116, 103.

(3 6) Dhanvantari, S.; Arnaoutova, I.; Snell, C. R.; Steinbach, P. J .; Hammond, K.;
Caputo, G. A.; London, E.; Loh, Y. P. Biochemistry 2002, 41 , 52.

(37) Drummen, G. P. C.; van Liebergen, L. C. M.; Op den Kamp, J. A. F .; Post, J. A.
Free Radical Biology & Medicine 2002, 33, 473.

(3 8) Lee-Gau Chong, P. W. v. d. M., B. Thompson, T.E. . Biochimica et Biophysica
Acta 1985, 813, 253.

91

CHAPTER 4

ONE- AND TWO-PHOTON EXCITED FLUORESCENCE
DYNAMICS OF NBD IN LIPID BILAYERS. DEPENDENCE OF
CHROMOPHORE DYNAMICS ON BILAYER COMPOSITION

Introduction

Plasma membranes are central to the existence and function of cellular systems.
These membranes are known to contain a large number of constituents, and as our
understanding of these complex and dynamic systems has evolved, it has become clear
that even simple lipid bilayer membranes are heterogeneous, highly dynamic structures.

”"2 as a result of exploring the differential solubility of

The discovery of “lipid rafts
plasma membrane constituents has given rise to a sustained effort aimed at resolving
whether or not these structures exist in plasma membranes. Phase separated regions have
been readily observed in model bilayer systems composed of three constituents,“ most
commonly a phospholipid, a sphingolipid, and cholesterol in varying amounts. However,
compelling experimental evidence for the existence of these raft structures in plasma
membranes remains to be presented. As a consequence, there is a great deal of effort
aimed at understanding the chemical and physical basis for phase separation in the

comparatively simpler model systems, so that the structural factors that give rise to phase

separation can be evaluated in more complex, natural plasma membranes.

Our interest in heterogeneous bilayer structures is for both pragmatic and
fundamental reasons. Ultimately, we want to create a biomimetic interface containing

active trans-membrane proteins that can be used for chemical sensing. The first step in

92

that process is gaining control over the bilayer structure(s) we will use to support the
trans-membrane proteins. We are thus interested in understanding the molecular
processes that mediate phase separation in model bilayers, and have studied the
organization and dynamics of model bilayer systems using a tethered fluorescent probe.
We utilize the spectroscopic properties of the tethered chromophore 2-(12-(7-nitrobenz-
2-oxa-1,3-diazol-4-yl)amino)dodecanoyl-l -hexadecanoyl-sn-glycero-3-phosphocholine
(N BD Clz-HPC, Figure 4.1) to interrogate local organization in our model bilayers. This
chromophore, which is connected to an acyl chain terminus of a phospholipid,
incorporates into the bilayer structure of unilamellar vesicles, and we can control the
composition of these vesicles readily. We note that the NBD chromophore is well known
to exhibit a solvent polarity-dependent fluorescence lifetime,6'9 but we find that the
motional properties of this chromophore shed more light on the organization of the
bilayer environment. The transient fluorescence anisotropy decay measurements we use
have a well established theoretical framework for their interpretation, both for free and
tethered chromophores. In this work we excite the sample using two different
spectroscopic processes, one- and two-photon excitation, because the use of two different
modes of excitation allows us to extract detailed information on the motional properties
of the chromophore. Our data show that the NBD chromophore exists within a
significantly restricted environment in phospholipid bilayers, and that both motion about
the tethering bond and longer timescale precession within the bilayer structure play a role
in the dynamics of this system. When multiple bilayer constituents are added, the
dynamics of the NBD chromophore change, with molecular-scale motion in the multi-

component medium appearing more rapid and isotropic than in a pure lipid bilayer. Our

93

 

Figure 4.1 Structures of compounds used in this work: (a) NBD-CuHCP, (b)
DSPC, (c) DLPC, (d) sphingomyelin, and (e) cholesterol.

94

infinite time anisotropy data indicate that the composition of the bilayer mediates the

mesoscopic rigidity of the system substantially, with cone angle of chromophore varying

from ca. 60° for the pure lipid bilayer to ca. 10° for bilayers comprised of lipid,
cholesterol and sphingomyelin. This effect has been seen before10 and our data point
collectively to the creation of molecular scale disorder and a mesoscopic increase in
bilayer rigidity. The only structural motif consistent with these findings is that of a

heterogeneous, phase separated bilayer structure.

95

Experimental

Materials. The fluorescent probe 2-(12-(7-nitrobenz-2-oxa-l,3-diazol-4-
yl)amino)dodecanoyl-l-hexadecanoyl-sn-glycero~3-phosphocholine (N BD C1 z-HPC) was
obtained from Molecular Probes Inc. and used without further purification. The bilayer
constituents 1,2-dilauroyl-sn-glycero-3-phosphocholine (DLPC, mp = -1°C), 1,2-
distearoyl-sn-glycero-3-phosphocholine (DSPC, mp = 55°C), egg sphingomyelin (SPM)
and powdered cholesterol were obtained from Avanti Polar Lipids, Inc. and used as

received.

For the construction of unilamellar vesicles, constituents were prepared in a 1:121
mole ratio of phospholipid : cholesterol : sphingomyelin or a 2:1 mole ratio of
phospholipid : cholesterol for the multicomponent systems. For the mixed lipid systems,
the ratio of phospholipids used was 1:1. A total of six vesicle solutions were produced:
NBD C1 z-HPC/DLPC, NBD Clz-HPC/DLPC/cholesterol, NBD C12-
HPC/DLPC/cholesterol/SPM, NBD Clz-HPC/DSPC, NBD Clz-HPC/DSPC/cholesterol,
and NBD C12-HPC/DSPC/cholesterol/SPM. NBD Clz-HPC. The concentration of NBD
Clz-HPC in each of these bilayer structures was maintained at ~10'4 M. Chloroform was
evaporated from the lipid solution, then it was dissolved in a 4:1 benzene : methanol
solution. The benzene : methanol solvent was then evaporated and the resulting lipid
mixture was dissolved in 10 mM Tris® buffer (Aldrich) adjusted for pH~8 in Milli-Q
water. For further mixing, the solution was placed in liquid nitrogen for 5 minutes, then
in a hot-water bath for 5 minutes, and vortexed for approximately 2 minutes. This
process was repeated four more times. The extrusion method was used to produce

unilamellar vesicles of ~100 nm diameter. A syringe-based mini-extruder (Avanti Polar

96

Lipids) was used to extrude the lipid suspension between a polycarbonate filter eleven

times to produce unilamellar vesicles of ~100 nm diameter (Figure 4.2).

Steady State Measurements. All absorption spectra were recorded with 1 nm
spectral resolution using a Cary model 300 double beam UV-Visible absorption
spectrophotometer. All emission spectra were recorded with a Spex Fluorolog 3
spectrometer at a spectral resolution of 3 nm for both excitation and emission

monochromators.

Time Correlated Single Photon Counting Measurements. Time-resolved
fluorescence lifetime and anisotropy measurements were performed using a time
correlated single photon counting (TCSPC) system. This system has been described in
detail elsewherell and we provide only a brief recap of its features here. The source laser
is a CW mode-locked Nd:YAG laser (Coherent Antares 76-S) that produces 100 ps 1064
nm pulses at a 76 MHz repetition rate. The second or third harmonic of the output of this
laser is used to excite a cavity dumped dye laser (Coherent 702-2), operating at 650 nm
with Rhodamine 610 dye (Exciton, 532 nm pump) for two photon excitation experiments,
or at 460 nm with Stilbene 420 dye (Exciton, 355 nm pump) for one photon excitation
experiments. The dye laser outputs were 5 ps pulses at a repetition rate of 4 MHz for
both excitation wavelengths. Fluorescence signals were detected using a Hamamatsu
R3 809U microchannel plate photomultiplier tube detector with a Tennelec 454 quad
constant fraction discriminator and Tennelec 864 time-to-amplitude converter and biased
amplifier used for signal processing. For this system the instrument response time is ~ 35

ps. Time-resolved fluorescence intensities were collected at 0°, 54.7° and 90° with

97

respect to the vertically polarized excitation pulse, and the response function was not

deconvoluted from the data.

98

 

Figure 4.2 TEM micrographs of unilamellar vesicles. Scale bars indicate distances
of 200 nm.

99

Results and Discussion

The central purpose of this work is to understand the molecular scale environment
that exists within a lipid bilayer structure and to elucidate the role of bilayer constituents
such as cholesterol and sphingomyelin in mediating these dynamics. Information
available from this work bears on issues such as the environment available to trans-
membrane proteins and the local viscosity of these fluid bilayer structures. To evaluate
the local environment(s) formed by lipid bilayer structures, we have used the
chromophore NBD tethered to a phospholipid (NBD Clz-HPC) as a probe in unilamellar
vesicle structures. While we are primarily concerned with the dynamics of the
chromophore, it is important to have some knowledge of the specific region within the
unilamellar vesicle where the tethered chromophore resides. We have addressed both of
these issues using time-resolved fluorescence measurements, where lifetime data provide
information on the polarity of the chromophore local environment and anisotropy decay
data acquired using both one- and two-photon excitation yield detailed information on the
motional properties of the chromophore. Because these two types of measurements are
sensitive to different aspects of the chromophore environment, we consider the results of

each separately.

We first examine the fluorescence lifetime of NBD Clz-HPC imbedded in
vesicles. For NBD, a substantial body of literature has established that the steady state
absorption and emission spectra and the fluorescence lifetime of this compound are
sensitive to the polarity of its immediate environment.6'8"2'l4 The absorption and

emission maxima for NBD shift to the red with increasing polarity, while the Stokes shift

100

for this chromophore remains approximately constant over a range of solvents.6'8’]2"4

Our steady state absorbance data show that for all vesicle solutions, the absorption
maximum of NBD occurs at ~460 nm and the emission maximum is centered at ~550 nm
(Figure 4.3). Time-resolved fluorescence lifetime data show that for both one and two-
photon excitation, the S) population decay of NBD C1 z-HPC contains two exponential
components in all cases, indicating that the chromophore resides in at least two
environments of differing polarity. The long lifetime component for all cases is between
~0.8 and 1.3 ns, while the short lifetime component is between ~100 and 400 ps (Figure
4.4). While these lifetimes are significantly different, the difference in environments for
these two sub-populations is likely not very great. From our prior work, we have found
that the hexanoic acid derivative of NBD selectively partitions in phosphocholine
bilayers into the region near the lipid head group, and for this chromophore the recovered
lifetime is ca. 1.1 ns. Other literature7 reports a short lifetime for NBD derivatives in an
aqueous environment, suggesting that our NBD chromophore is in all cases in the vicinity
of the lipid head groups. We note that the lifetime of NBD derivatives in solvents of
moderate polarity exhibit lifetimes on the order of 8 ns,'5 indicating that for these
experiments the tethered NBD chromophore is not in close proximity to the interior of the
bilayer. It is reasonable to expect that the acyl chain to which the NBD chromophore is
attached has “folded back”, leaving the NBD in relatively close proximity to the polar
headgroup region of the vesicles, rather than in the nonpolar core region.12 The
immediate environment of a chromophore has long been recognized for having an
influence fluorescence lifetime. Molecules that reside in an environment where there is a

dielectric gradient on the length scale of the chromophore can experience a substantial

101

modulation of their fluorescence lifetime, and we believe that to be the situation here.”"'9

Within the bilayer structure, there is a gradient in dielectric response from the acyl region
(8 ~ 4) to the aqueous region (8 ~ 82) over a distance of several Angstroms. Where in
this gradient the NBD chromophore resides, and the orientation of its transition moment
relative to the gradient, will influence its fluorescence lifetime. Thus, the presence of two
distinct fluorescence lifetime components in our data should not be taken as a necessary
indication of two substantially different environments, but rather, these data signify
subtle positional differences for sub-populations of the tethered chromophore. These data
do not provide sufficient information by themselves, however, to draw any firm structural
or organizational conclusions. In order to understand the local environment of the NBD

chromophore in more detail, we turn to fluorescence anisotropy measurements.

For our experimental conditions, the fluorescence anisotropy data for NBD C12-
HPC in vesicles result from molecular reorientation. There is a well established
theoretical framework for the interpretation of fluorescence anisotropy data, under
conditions of both one and two photon excitation. For both means of excitation, the
essential physics of the relaxation process remains the same — an initially anisotropic
population returns to an equilibrium orientational distribution. The difference between
one and two photon excitation measurements lies in the initial orientational distribution
of the anisotropic photo-selected population and the relationship between the incident
electric field polarization(s) and the excited state orientational distribution. For this
reason, the relationship between the time constant(s) of the induced orientational

anisotropy decay and the Cartesian components of the rotational diffusion constant will

102

1.0 —

 

 

 

0.8 — /
A?
U2
:1 /
<1) ,
E 0.6 -
"U
4,3, ' .
'23 0.4 - /
§ ,1
C
0.2 -
/ ::
0.0 1 I 1 I 1 I 1 I 1 I
450 500 550 600 650 700

wavelength (nm)

Figure 4.3. Normalized emission spectra for NBD-CuHPC in unilamellar vesicles.
Emission maxima are centered at ca. 550 nm in all cases.

103

 

 

1000-
3
800-
+2 6001 '
=1 .
8 O
400» 3 -
200— ;
i
O , M1 .
O 1000 2000 3000 4000 5000
time (ps)

Figure 4.4. Lifetime decay of NBD-C12HPC in unilamellar vesicles composed of
DLPC. The lifetime is a double exponential decay with n=321 :l: 22 ps and rz=133l

i 147 ps.

104

differ for the two modes of excitation, and we take advantage of this difference to extract
information which is not available from either measurement individually.For one-photon
excitation, the induced orientational anisotropy function, R(t), is generated from the

polarized emission transients according to Eq.4.1,

 

R(t) _ [MD—[10)

" 1,,(1) + 21i (t) (4'1)

where I”(t) and I 1(t) are the emission intensities polarized parallel and perpendicular to
the vertically polarized excitation pulse. The chemical information extracted from these
measurements is contained in the number of exponential decays and their time constant(s)
that constitute R(t). A note is in order about the experimental data we present here. Our
instrument response time is typically 35 ps fwhm and several of the anisotropy decay
time constants we report here are ca. 100 ps. Our past experience has consistently shown
that the deconvolution of the instrument response function from our raw data changes the
observed anisotropy decay time constant by less than the experimental uncertainty, even
for decays with time constants as fast at 100 ps. Because of the limited S/N ratio that
characterize our data and our desire to not manipulate the experimental data more than

necessary, we have not attempted deconvolution.

While R(t) may be characterized by up to five decays, typically only one or two
decays are resolved experimentally. The relationship between the functionality of R(t)
and the chromophore molecular properties has been described in detail previously.”22

Chuang and Eisenthal’s treatment of R(t) allows for this function to be simplified to two

general cases, where the shape of the rotor describes the time-averaged volume swept out

105

by the rotating species.22 A Type I rotor is characterized by a distortion along the axis
coincident with the transition moment axis and a Type II rotor is characterized by a
distortion from a spherical shape along an axis perpendicular to the transition moment
axis. These rotor designations correspond to the more traditional descriptions of prolate
and oblate rotor as follows: For a Type I rotor with the transition moments along the ith
axis, D, > D,- = D, (i,j,k = x,y,z) are prolate, and D, < D,- = D, are oblate. For a Type II
rotor, D. > D, = D,- is prolate and D, < D, = DJ- is oblate. We prefer to use the Type I and
Type II designations here because the notion of prolate and oblate rotors is not
necessarily physically clear in the context of a restricted or tethered environment. There
are two general solutions to these cases?"22 For the transition moment polarized along

the ith axis, with D, 715 DJ- : Dk,

R(t) = 0.4 exp(-—6D,,t) (4.2)
And for D, 95 D, = D',

R(t) = 0.1exp(—6D,.t)+ 0.3 exp(—(ZD, + 4D,, )t) (4.3)

Despite these simplifications to the time-averaged shape of the reorienting
species, there remains ambiguity in the interpretation of anisotropy decay data, arising
form the orientations of the absorbing and emitting transition dipole moments. In order
to resolve which of these conditions obtains for our results, we must have knowledge of
the absorption and emission transition dipole moment orientations for NBD Clz-HPC.
For the NBD chromophore, the one photon absorption and emission transition dipole
moments are in the rt-system plane with x designating the long molecular axis and y

designating the short axis. Based on this assignment, and the location of the connection

106

of the phospholipid acyl chain to the chromophore (Figure 1), the motional constraints
relevant to NBD Clz-HPC are Dy 76 Dx = Dz. With this axis assignment Eqs. 4.2 and 4.3
can be modified to describe the possible decay functionalities for R(t) in terms of the

Cartesian axes,
x-polarized transition R(t) = 0.3 exp(—(4Dy + 20, )t) + 0.1exp(—6D,t) (4.4)
y-polarized transition R(t) = 0.4 exp(—6D,t) (4.5)

The fact that we recover single exponential decay anisotropies for all one-photon excited
measurements (vide infra) suggests that the S) H So transitions are y-axis polarized for
NBD. While it is possible that our S/N ratio does not allow us to resolve a second
component, analysis of the two photon excited data for either x- or y-polarized emission
yields physically meaningful results (i. e. D, > 0) only for y-polarized emission (vide
infra). The recovered time constant for one photon excitation is therefore related to the
motion about the x-axis. A simplistic interpretation of this treatment implies that the
chromophore is free to rotate in solution, which is not the case with NBD Clz-HPC. The
information obtained on Dx must therefore be reflective of motions within the bilayer
structure that are either precessional or random walk in nature, and we discuss this point
in more detail below, after considering the information content of two-photon excited

anisotropy data.

We borrow from J ohnson’s excellent treatment of two—photon excited anisotropy
decay data to interpret our results.9‘23'26 Induced orientational anisotropy decay functions
are generated for both linearly (r1(t)) and circularly (r2(t)) polarized light and are given
by:

107

Ill/mew (t) _ I‘ll-incur (t)

Illlmear (t) + 2111mm (t)

Circular Circular
r(t)- In (”-11 (I)
2 - Circular Circular
21ll (t) + 1, (t)

 

r.(t) =
(4.6)

 

For both excitation polarizations, the anisotropy decays are characterized by two time

constants, to and r2, weighted according to the properties of the chromophore.9’24’26

rl (t) = r1 (0)[co exp(-t / To) + c2 exp(—t / 12)] (4 7)
r.<r>=r.(O)[d..exp(—t/r.>+d.exp(—t/r.>l '

The relative contributions of the time constants are given by the prefactors c, and (1,.

These prefactors are25

xx ,ty

(Ea + b>[3<fia+ 053. + 3<—fia + 053. + M S + 48931

 

 

 

 

4.8

CO 7N2(3s; + 3s; + 28,8; + 45; ) ( )
(a — Em [3(a — Ems; + 3(a + Ems; + 2as,,s; + 4115;] 4 9

9 7 * 7N2(3S,2, + 3s; + 25,5; + 45;) ( ' )

d (Ea + m [(Ea + ms; + (—Ea + ms; — 4bS,,S;, + 668;] 4 10

0 - 14N2(S§, + S; _ 5.85,... + 35;) * ( . )

(a _ .51.) [(a _ Ems; + (a + Ems; — 4as,,s; + 6.18;]

d, = ' ' ' (4.1 1)

14N2(S_,2_, +Sf. -s s +3s;)

n )y

The relative amplitudes of the decay time constants will provide information about the
values of the Cartesian components of the rotational diffusion constant. The parameters

a, b, N and A are related to the diffusion constant by

108

a=fi<D.—D.>
b:20,-0,,-0,+2A
N2 =a2+b2

A = (0,2 + 0; + 03 — 0,0, — 0,0, — 0,0,)“2

(4.12)

While Eqs. 4.7 indicate that two anisotropy decay time constants should be recovered for
two photon excitation measurements, this is not always the case. For cases where 10 and
12 are similar, the time constants observed experimentally are 1,) (linear polarization) and
1,2 (circular polarization), which are weighted averages of 10 and 12.25 It is also possible,
depending on the spectroscopic and motional properties of the chromophore, that one or
more of the terms c, and d, could assume values of zero, and in this case, the time
constant observed is simply the time constant not associated with the zero-coefficient
term. This situation obtains for our experimental conditions, with Dx = Dz, leading to the
term (\l3a+b) = O. The imposition of this condition causes the prefactors co and do to
become zero, and for our experimental conditions, 1,1=12, and 1,2=12 where 12=(6D+2A)'
1.25 The reorientation times recovered from linearly polarized excitation and circularly

polarized excitation are predicted to be identical, and 1 is related to both Dy and szs

12 = 1
(20, + 40,.)

 

(4.13)

The time constant recovered from two-photon excited decays contains contributions from
both D): and Dx, while the one-photon excited anisotropy decay time constant is related to
D... From these two measurements we can, in principle, determine Dx and Dy uniquely.

We caution, however, that the chromophore motions are restricted to some extent by the

109

fact that they are tethered to the phospholipid, and we consider the consequences of this

structural condition next.

At this point it is useful to note that the R(O) values that we extract from our
experimental data do not approach the theoretical limits of either the one or two photon
treatments of the data. The quantity R(O), for both one and two photon excitation, is
related to the angle between the excited and emitting transition dipole moments. This
quantity is not related to the rotational diffusion coefficient, and there is a measurable
perturbation of the recovered time constants only where R(O) ~ 0.27 This condition
occurs for an angle of 54.7° between the transition moments for one photon excitation
and at 353° for two photon excitation.28 The fact that our R(O) values all lie in the ~O.1-
0.3 range indicates that the determination of the rotational diffusion coefficients are
essentially unaffected by the angle between the transition moments for either type of

excitation.”27

The theoretical constructs used to interpret fluorescence anisotropy data depend
on the constraints imposed on the chromophore by its local environment. When the
motional freedom of a molecule is restricted in some manner, such as incorporation into a
bilayer structure and/or tethering to an interface, the data are typically treated within the
framework of the hindered rotor model.”32 This model takes into account the fact that
the initially non-random chromophore distribution selected by the excitation pulse is not
free to rerandomize completely. The inability of the chromophore to relax into a fully
random orientational distribution is manifested as a non-zero infinite-time anisotropy,

R(oo). The decay time constant for the anisotropy function is the time required for the

110

system to relax from the non-random orientational distribution selected by the excitation

pulse to its steady state orientational distribution, and this time constant is given by

793
TH” = 240

 

(4.14)

Where ()0 is the semi-angle of the cone which the (tethered) molecule sweeps out as it
moves within its confines, and DW is the “wobbling diffusion constant” which describes
the motion of the chromophore about its tethering bond. This description of molecular
motion is not determined sufficiently for our purposes. While the quantity 00 is useful as
a measure of average disorder in the system, the quantity Dw provides very little in the

way of detailed molecular scale insight into chromophore motion.

It is important to consider at this point the role of symmetry and timescale in
extracting R(oo) information from our data. This issue arises because of the nmnber of
phenomena that are operating simultaneously on the chromophore. Specifically, the
translational mobility of the chromophore within the vesicle walls and the rotational
motion of the vesicle as a whole need to be considered. Based on the average size of our
vesicles (ca. 100 nm diameter), we estimate its rotational diffusion time to be ca. 125 ps.
The time scale of vesicle reorientation is thus long relative to chromophore local
rotational motion. Translational motion of the vesicle constituents (DT ~ 50 pmz/s)
allows the chromophore to sample the surface area of the vesicle in ca. 500 ps. Thus, on
the time scale of chromophore reorientation, neither vesicle rotation nor chromophore
translation will contribute to any discernible extent. Our ps time scale measurements

sample exclusively the molecular rotational motion of the chromophore. We can thus

111

extract meaningful R(oo) values from our experimental data, and relate that information to

local organization within the vesicle wall.
For a hindered rotor, the time-course of R(t) behaves according to Eq. 4.15:

R(I) = R(00)+(R(0)-R(00))6XP(-t/THR) (4-15)

Where 1HR is given by Eq. 4.14. The quantity R(oo) is related to the size of the cone in

which the chromophore resides,

1/2

R (0°)

— =0.5 c056 1+cos6l 4.16
Our experimental data for R(O) and R(oo) are tabulated in Table 4.5, and from these data
we calculate cone angles 00 in the range of ~60° to ~10° for these bilayer structures. We
will return to consideration of the significance of these data after dealing with the

motional component of our signals.

If we were to interpret our experimental anisotropy data simply in the context of
the hindered rotor model, we would be left in all cases with the anisotropy decay time
constant being associated with the same physical quantities 00 and Dw, and this situation
would not account for the excitation and system dependencies we see in our data. Our
experimental data (Table 4.1) demonstrate that this is not the case, so that a more detailed
interpretation than is afforded by Eq. 4.15 is required. We are not aware of any
treatments in the literature that attempt to make correspondence between the hindered

12032.33

rotor mode and the more general treatment22 for fluorescence anisotropy decays

which consider explicitly the Cartesian components of the rotational diffusion coefficient.

112

Regardless of the extent of confinement, the relaxation of the excited chromophore
population will scale as P2[u(0)°u’(t)] for one photon excitation, and as P2[S(0)-u’(t)] for

two photon excitation. Both of these expressions can be cast in terms of Dx, Dy and D2.9

As discussed above, we can treat the fluorescence anisotropy data for one and two photon
excitation according to Chuang and Eisenthal22 and Johnson et al.,9’Z3'26 and can extract
values for Dx and Dy for NBD Cl2-HPC from the data. Based on the functional form of
the anisotropy decays for both one and two photon excitation, the NBD chromophore
reorients as a Type I rotor with Dy at DK = Dz, with the absorption and emission transition
dipole moments oriented along the chromophore y-axis. Attempts to locate the transition
dipole moment(s) along the chromophore x (long) axis yield expressions that are
inconsistent with the functional form and time constants of our experimental data. For an
x-polarized transition, our experimental data yield at least one negative diffusion
coefficient, and the results for one- and two-photon excitation are not self-consistent.

The anisotropy decay treated by the various models must represent the same physical
phenomenon; the re-randomization of an initially non-random orientational distribution,

and there must thus be a connection between the two models.

For NBD-C12-HPC, the y-axis corresponds to that axis along which the
chromophore is tethered, x is the long in-plane axis and z is perpendicular to the x-y
plane. In this model, the unique motion for the tethered chromophore will be about the
tethering, y-axis. Apparent rotational motions about the x- and z- axes will actually be

either precessional or random walk motions within the (nominally) conic volume the

113

tethered chromophore sweeps out. We consider next the experimental data, and how the

anisotropy decay times differ due to the different modes of excitation.

We present in Tables 4.1 — 4.3 the fluorescence lifetime and anisotropy decay
time data for excitation of NDB Cl2-HPC by one photon excitation at 460 nm (Table 4.1),
and by two photon excitation with linearly (Table 4.2) and circularly (Table 4.3)
polarized light at 650 nm. For each system we recover a two component fluorescence
lifetime, and for a given vesicle system, the lifetimes we recover are the same to within
the experimental uncertainty for all modes of excitation because the emitting state is the
same in all cases. The anisotropy decay data shown in Tables 4.1 — 4.3 reveal a

dependence on both the bilayer composition and on the mode of excitation.

We have interrogated the dynamics of NBD C12-HPC unilamellar vesicles
comprised solely of the phospholipid as well as vesicles containing cholesterol and/or
sphingomyelin in addition. The intent of these experiments was to understand the
constraints imposed on the chromophore by the phospholipids and to evaluate the role of
the added constituents on bilayer organization over a range of length scales. The one-
photon excited anisotropy decay time constant is related to the rotational diffusion
coefficient by 1 = (6Dx)". For a Type I ellipsoid tethered along its y-axis, Dx = DZ and
the recovered one-photon reorientation time constants represent motion that lies between
precessional and random walk, as constrained by its tether. This motion, while not fully
rotational, reflects an angular displacement with time, and will appear as a nominally
rotational motion to our experiment because of the inherent distribution of environments

that the chromophore occupies. In comparison, for two-photon excited measurements,

114

Table 4.1 One photon excitation, fluorescence lifetimes and anisotropy decay time

constants in vesicles of different compositions. Chol = cholesterol, SPM =

 

 

 

 

 

 

 

sphingomyelin.
Vesicle composition Fluorescence Fluorescence 103 (ps)
Lifetime 11 (ps) Lifetime 12 (ps)
NBD-C12HPC, DLPC 321 i 22 (60:h1%) 1331 :1: 147 572 :1: 55
(402t1%)
NBD-C12HPC, DLPC, 179 :t 18 (83il%) 808 i 64 (17i1%) 215 d: 35
chol
NBD-C12HPC, DLPC, 132 i 20 (67i2%) 581 i 71 (33i2%) 191 :1: 18
chol, SPM
NBD-C12HPC, DSPC 144 i 25 (42:1:3%) 1044 d: 158 789 :1: 17
(58i3%)
NBD-C12HPC, DSPC, 119 i 4 (72d:3%) 793 i: 9 (28i3%) 110 i: 19
chol
NBD-C12HPC, DSPC, 135 d: 8 (97i3%) 750 i 60 (3i3%) 94 :1: 11

chol, SPM

 

 

 

 

 

115

 

where the anisotropy decay time constant has contributions from D)( and Dy (Eq. 4.13),
we anticipate that the time constants will, in general be different, and this is seen
experimentally. We note that, as predicted on theoretical grounds, the anisotropy decay
times acquired for linearly and circularly polarized two photon excitation yield the same

time constants for a given system.

We present in Table 4.4 the calculated diffusion coefficients Dx and Dy for NBD-
C12-HPC in vesicles of varying composition, along with the calculated viscosity and the
Dy/Dx aspect ratio. There are several interesting pieces of information in these data
which provide significant insight into the organization of our bilayers. We have extracted
the values for Dx and Dy from the experimental anisotropy decay data (Tables 4.1-4.3)
and Eqs. 4.5 and 4.13. In order to estimate the local viscosity sensed by our

chromophore, we have used the modified Debye-Stokes-Einstein equation.20'22’34’35

77W 1 1 (4.17)

 

TOR k " — =
,15 6D 40, + 20,

where n is the viscosity of the medium surrounding the reorienting moiety, V is the solute
hydrodynamic volume, which we take to be the volume of the NBD chromophore for this
work, VNBD = 106 A3 ,3 6 f is the frictional solvent-solute boundary condition, and S is the

solute shape factor. We calculate the value of D from our experimental quantities Dx and

Dy and with the assumption that D, = Dz. D is defined as the average of the Cartesian

components, D = (Dx+Dy+Dz)/3 = (2Dx+Dy)/3. The modified DSE equation was derived

to describe the rotational motion of an ellipsoid in a continuum fluid, where there is no

116

Table 4.2. Two photon excitation (linear polarization), fluorescence lifetimes and
anisotropy decay time constants in vesicles of different compositions.

 

 

 

 

 

 

 

Lipid composition Fluorescence Fluorescence Lifetime 103 (ps)
Lifetime 11 (ps) 12 (ps)
NBD-C12HPC, DLPC 247 1 8 (5812%) 1029 1 17 (4212%) 123 1 9
NBD-C12HPC, DLPC,
166 1 1 1 (7618%) 1061 1 82 (2418%) 183 1 9
chol
NBD-CI2HPC, DLPC,
12715 (571l4%) 944113(43114%) 10616
chol, SPM
NBD-CI2HPC, DSPC l 15 1 6 (6812%) 890 1 13 (3212%) 189 1 7
NBD-C12HPC, DSPC,
1181 15 (6313%) 931 137 (3713°/o) 156 17
chol
NBD-PC, DSPC, chol,
106 1 8 (7015%) 883 1 33 (3015%) 105 1 3
SPM

 

 

 

 

 

 

117

Table 4.3. Two photon excitation (circular polarization), fluorescence lifetimes and
anisotropy decay time constants in vesicles of different compositions.

 

Lipid composition Fluorescence Fluorescence 10R(ps)

Lifetime 11 (ps) Lifetime 12 (ps)

 

NBD-C12HPC,DLPC 26619(6011%) 1110115(4011%) 117113

 

NBD-C12HPC, DLPC, 171 1 9 (5111%) 1087 1 24 (4911%) 166 1 18

 

chol
NBD-C12HPC, DLPC, 128 1 13 910 1 38 (4717%) 80 1 11
chol, SPM (5317%)

 

NBD-CnHPC, DSPC 142 1 6 (6611%) 1039 118 (3411%) 176 1 3

 

NBD-C12HPC, DSPC, 202 1 21 1014 144 (5214%) 160 1 8

01101 (4814%)

 

NBD-C12HPC, DSPC, 114 1 9 (6713%) 879 :1 29 (331307.) 101 1 9

chol, SPM

 

 

 

 

 

 

118

tethering attachment. While our application of Eq. 4.17 to a bilayer structure is clearly at
odds with the original intent of the model, we believe that we are justified in making this
approximation because of the fluid nature of the bilayer system. As noted above, it is fair
to consider that there will be motions associated with both the chromophore and with the
vesicle itself. We assert that these motions are well isolated from one another in the time
domain. Specifically, if we apply Eq. 4.17 to the vesicle as a whole, (r ~ 500 A), we
calculate a reorientation time of 126 us; six orders of magnitude longer than the
chromophore reorientation time. Even if this calculation were in error by as much as a
factor of two, such a difference in time scales would render the overall motion of the

vesicle undetectably slow on the timescale of our measurements.

We recognize that the viscosity values we extract from the DSE model are likely
characterized by a systematic offset from the actual viscosities, but our primary purpose
in using this model is that it allows a basic comparison to be made across the range of
experimental conditions we have employed here. We obtain values of ca. 7.8 cP for
NBD Clz-HPC in DLPC (mp = -1°C) and ca. 11.6 CF for NBD Clz-HPC in DSPC (mp =
55°C). These values are entirely consistent with recent data we have acquired for DMPC
above and below its gel-to-fluid transition temperature using a different, non-tethered
probe.37 We note that the aspect ratio, Dy/Dx, is significantly greater than unity for the
NBD chromophore in the pure lipid bilayer, consistent with molecular-scale motional
freedom constrained by the presence of significant lipid acyl chain organization. With
the addition of cholesterol to the bilayer, we see two noteworthy changes. The first is a

decrease in the measured viscosity (much more pronounced for DSPC), indicating a

119

 

Table 4.4. Cartesian components of the rotational diffusion coefficient for NBD C12-
HPC, and resulting viscosities calculated from these D values.

 

 

 

 

 

Lipid composition D, (MHz) Dy (MHz) 11 from D Dy/D,l
NBD-CIZHPC, DLPC 291 1938 7.8 6.7
NBD-ClePC, DLPC, 775 1041 7.5 1.3

chol
NBD—ClePC, DLPC, 873 2252 4.9 2.6
chol, SPM
NBD-ClePC, DSPC 211 1261 11.6 6.0

 

 

 

 

 

 

 

120

Table 4.5 Calculated cone angles for hindered rotor model based on zero and
infinite time anisotropy data for NBD-Clz-HPC in vesicles of varying composition.

 

 

 

 

 

 

 

 

 

choL

SPM

 

 

 

 

Vesicle 1' HR (ps) R(0) R(oo) 00

composition
NBDPC, DLPC 572 1 55 0.15 1 0.01 0.02 1 0.01 61 1 7
NBDPC, DLPC, 215 1 35 0.19 1 0.02 0.10 1 0.04 37 1 12
chol
NBDPC, DLPC, 191 1 18 0.37 1 0.07 0.35 1 0.04 11 1 16
chol, SPM
NBDPC, DSPC 789 1 17 0.15 1 0.03 0.04 1 0.01 51 1 8
NBDPC, DSPC, 110 1 19 0.20 1 0.07 0.09 1 0.05 41 1 18
chol

BDPC, DSPC, 94 1 11 0.19 10.02 0.14 1 0.03 26 1 11

 

121

 

decrease in the organization of the bilayer on a molecular length scale. Along with this
change we see that Dy/Dx changes from 6.7 to ca. 1.3 for DLPC, and from 6.0 to 0.5 for
DSPC. This remarkable change is a direct indication of the local disorder introduced by
cholesterol in the region sensed by NBD. Examination of Dy and DK values reveals that
Dy decreases upon addition of cholesterol, while Dx increases slightly. The decrease in
Dy implies that the cholesterol added to the system is either interacting with the NBD
chromophore providing some type of impediment to motion,10 but the implications of the
changes in Dx are less clear. The addition of sphingomyelin continues these trends,
consistent with the further disruption of local organization in the portion of the bilayer
structure sensed by the NBD chromophore. Taken collectively, these data point to
significant interactions between the chromophore and all the bilayer constituents,
indicating that the addition of multiple constituents to the bilayer gives rise to a loss of

short-range organization.

To this point we have considered the transient anisotropy decay and the
information contained within it relative to short-range bilayer organization. The infinite
time anisotropy information contained in the one-photon excited anisotropy data shed
light on the longer-range structural heterogeneity of the bilayers. We consider these
results in the context of the hindered rotor model, where the measured infinite time
anisotropy (Figure 4.5, Eq. 4.16) is related to the volume swept out by the probe. We
observe (Table 4.5) that with the addition of cholesterol and sphingomyelin to the
bilayers, there is a monotonic trend to smaller cone semi-angles, 00, which is seen for
both lipids. On its face, such a trend would appear to be in conflict with the reorientation

data. We believe that there is no conflict. It is well known that the addition of

122

cholesterol to bilayers makes the bilayers more rigid.lo The increase in rigidity is
consistent with a heterogeneous system, where the packing of mesoscopic domains give
rise to greater rigidity, and less translational mobility in the overall system. The fact that
we sense local loss of organization with the reorientation transients and mesoscopic
increase of organization with the infinite time anisotropy data suggests that the NBD

chromophore resides relatively near the domain boundaries in these systems.

With this information on the reorientation dynamics of NBD C1 2-HPC, we return
to the issue of the two component fluorescence lifetime decay. The time constants
recovered from the lifetime data (ca. 200 ps, 1.1 ns) are both indicative of polar
environments, and it is likely that they represent two different configurations where the
chromophore is “folded back” to approach more closely the polar interface between the
bilayer head groups and the aqueous environment. Both of these populations apparently
sense relatively similar local environments and are thus sensitive to the same loss of short
range organization associated with the addition of cholesterol and sphingomyelin to the
bilayer. The remarkable changes we note in the anisotropy dynamics are likely averaged
over some range of “depths” within the bilayer structure, and owing to the polarity
sensitivity of this chromophore, the lifetime data reflect this particular aspect of the local

environment with greater sensitivity than the reorientation data.

123

0.2 -

R(t)
O

       
 

 

. 0 . .0
‘ g .0.
b . I. . o . . . . . o
o o. 0 00 I .0 v. 0' 0. .Q o ; . . o. . .
..' “'1". ""3 ~~""': 4'. ' .' . " .’2 '° ~° ~'-.'-.-.°.-:-' ~-'--: .
0 O "'5‘“. ““ " ' '1 ' ’°.' .:.~'.’-.- I.“ -. .9: c - - ns. t .' " ~"‘
. i .0. o 9 0.. . . .' 1 0 ’
"3‘5.“ ..' . 10"." 0...... .a ‘ ’0'.‘ 0. ' ' ’ ' .00 ~ .0
° .- '1'. °: ' :. ’”0 7‘ 33‘ °
. O .. o O. ‘ 0.. . O. .‘ 9.
o G' o
_O 1 l l 1 l 1 l 1 l l n I 1 l ' l I

 

-500 0 500 1000 1500 2000 2500 3000 3500 4000
time (ps)

Figure 4.5. Anisotropy decay for one-photon excitation of NBD-C12HPC in
unilamellar vesicles composed of DLPC and cholesterol. Residuals of the fit to the

data are shown as solid circles. For this particular run, 1' = 195 :1: 27 ps and R(oo) =
0.15 1 0.01.

124

Conclusions

We have studied the steady state and time resolved optical properties of the
chromophore NBD-CnHPC in vesicles containing aqueous solutions. We used both one-
and two-photon excitation to gather the information required to determine both Dy and
D... From these measurements, we estimate the viscosity of the local medium of the
chromophore, as well as Dy/Dx. For vesicles comprised of pure phospholipid, we
determine a viscosity of 11.6 cP for gel phase DSPC and 7.8 cP for fluid phase DLPC,
consistent with other recent reports on bilayer viscosity.37 Upon the addition of
cholesterol, the bilayer viscosity decreases due to changes in the local organization within
the bilayer structure(s), and the addition of sphingomyelin causes a further decrease in
bilayer viscosity. For pure phospholipid vesicles, Dy/Dx ~ 1, and for vesicles comprised
of two or more components, Dy/Dx ~ 1. Such changes underscore the extent of the local
disorder being imposed on the bilayer. Infinite time anisotropy data reveal that longer
range organization in the bilayers is enhanced with the addition cholesterol and
sphingomyelin, consistent with the formation of a heterogeneous bilayer structure. Taken
collectively, our reorientation and cone angle data on NBD-Clz-HPC in a range of bilayer
systems point to the chromophore residing in relatively close spatial proximity to the
boundaries between heterogeneous regions within these bilayers. For these complex
bilayer systems, local and mesoscopic organization change in different ways with the

addition of cholesterol and sphingomyelin.

125

Literature Cited

(1) Simons, K., Ehehalt, R. Journal of Clinical Investigation 2002, 110, 597.

(2) Simons, K.; van Meer, G. Biochemistry 1988, 27, 6197.

(3) Crane, J. M.; Tamm, L. K. Biophysical Journal 2004, 86, 2965.

(4) Hao, M.; Maxfield, F. R. Journal of Fluorescence 2001, 11, 287.

(5) Scherfeld, D.; Kahya, N.; Schwille, P. Biophysical Journal 2003, 85, 3758.

(6) Chattopadhyay, A.; London, E. Biochimica et Biophysica Acta, Biomembranes
1988, 938, 24.

(7) Fery-Forgues, S.; Fayet, J .-P.; Lopez, A. Journal of Photochemistry and
Photobiology, A: Chemistry 1993, 70, 229.

(8) Mazeres, S.; Schram, V.; Tocanne, J .-F.; Lopez, A. Biophysical Journal 1996, 71,
327.

(9) Wan, C.; Johnson, C. K. Chemical Physics 1994, 179, 513.

(10) Cao, H.; Tokutake, N.; Regen, S. L. Journal of the American Chemical Society
2003, 125, 16182.

(1 1) DeWitt, L.; Blanchard, G. J .; LeGoff, E.; Benz, M. E.; Liao, J. H.; Kanatzidis, M.
G. Journal of the American Chemical Society 1993, 115, 12158.

(12) Huster, D.; Muller, P.; Arnold, K.; Herrmann, A. Biophysical Journal 2001, 80,
822.

(13) Lin, 8.; Struve, W. S. Photochemistry and Photobiology 1991, 54, 361.

(14) Mukherjee, S.; Chattopadhyay, A.; Samanta, A.; Soujanya, T. Journal of Physical
Chemistry 1994, 98, 2809.

126

(15) Greenough, K. P.; Blanchard, G. J. Journal of Physical Chemistry A 2007, 111,
558.

(16) Drexhage, K. H. Journal of Luminescence 1970, 1,2, 693.

(17) Girard, C.; Martin, O. J. F.; Dereux, A. Physical Review Letters 1995, 75, 3098.

(18) Lukosz, W.; Kunz, R. E. Optics Communications 1977, 20, 195.

(19) Lukosz, W.; Kunz, R. B. Journal of the Optical Society of America 1978, 67,
1607.

(20) Tao, T. Time-dependent fluorescence depolarization and Brownian rotational
diffusion coefficients of macromolecules. PhD Thesis, 1969.

(21) Blanchard, G. J. Journal of Chemical Physics 1987, 8 7, 6802.

(22) Chuang, T. J .; Eisenthal, K. B. Journal of Chemical Physics 1972, 5 7, 5094.

(23) Johnson, C. K.; Wan, C. Topics in Fluorescence Spectroscopy 1997, 5, 43.

(24) Lakowicz, J. R.; Gryczynski, 1.; Gryczynski, Z.; Danielsen, E.; Wirth, M. J.
Journal of Physical Chemistry 1992, 96, 3000.

(25) Pauls, S. W.; Hedstrom, J. P.; Johnson, C. K. Chemical Physics 1998, 23 7, 205.

(26) Wan, C.; Johnson, C. K. Journal of Chemical Physics 1994, 101, 10283.

(27) Blanchard, G. J .; Wirth, M. J. Journal of Chemical Physics 1985, 82, 39.

(28) Cho, M. Journal of Chemical Physics 2003, 119, 7003.

(29) Karpovich, D. S.; Blanchard, G. J. Langmuir 1996, 12, 5522.

(30) Kinosita, K., Jr.; Ikegami, A.; Kawato, S. Biophysical Journal 1982, 37, 461.

127

(31) Kinosita, K., Jr.; Kawato, S.; Ikegami, A. Biophysical Journal 1977, 20, 289.

(32) Szabo, A. Journal of Chemical Physics 1984, 81, 150.

(33) Lipari, G.; Szabo, A. Biophysical Journal 1980, 30, 489.

(34) Zwanzig, R.; Harrison, A. K. Journal of Chemical Physics 1985, 83, 5861.

(35) McPhie, P. Principles of Fluorescence Spectroscopy, Second Edition 2000; Vol.
287.

(36) Edward, J. T. Journal of Chemical Education 1970, 4 7, 261.

(37) Koan, M. M., Blanchard, G. J. . Journal of Physical Chemistry B 2006, 110,
16584.

128

CHAPTER 5

DYNAMICS OF A TETHERED CHROMOPHORE IN
PHOSPHOGLYCEROL VESICLES

Introduction

Plasma membranes are central to the function of mammalian cells. Despite the
fact that a single mammalian plasma membrane may contain as many as 500 different
lipid and sterol compounds, it has emerged that their distribution within the membrane is
not homogeneous. "7 Lipid bilayer membranes are heterogeneous, complex structures,
exhibiting phase separation even in comparatively simple model systems. The existence
of such phase separation has given rise to the notion of “lipid rafts”,2'5’8'20 and their
presence and characterization in both model bilayers and plasma membranes is an area of
active research. Although there is ample evidence of phase separation in two- and three—
component systems, the existence of lipid raft structures in plasma membranes remains to
be proven. We are interested in these heterogeneous bilayer systems to determine
whether or not such “rafts” in simple model systems can be used to support
transmembrane proteins in their active forms. To address this issue, we have investigated
organization and dynamics within one, two and three component model bilayer structures
and have examined how organization and dynamics varies with the chemical
functionalities present, in both the polar and nonpolar regions of the bilayers.“22 In this
manuscript, we consider the role of the lipid head group identity in mediating the

dynamics of chromophores placed in close spatial proximity to the bilayer head groups.

129

In the vicinity of the lipid bilayer head group region, the environment is
dominated by the presence of water, and optical spectroscopic probes indicate that this
environment is slightly more viscous than bulk water.23 Despite the distinct structural
differences between the polar and nonpolar regions of lipid bilayers, it is clear that each
has a structural influence on the other, both in terms of the local motional freedom
available to individual molecules, and in terms of longer range translational freedom. A
key issue in understanding the organization of lipid bilayer structures is knowledge of
how the polar phospholipid head groups interact with the aqueous environment in which
they are in contact. The most widely used phospholipid family for such studies is the
phosphocholines (PCs). The headgroups of these lipids are terminated by a
trimethylamino moiety that carries a formal charge which is strongly solvated by water,
but does not possess lone pairs which hydrogen bond extensively with surrounding water.
The phosphoester functionality, however, does possess the lone pairs with which to
hydrogen-bond with water. Phosphoglycerols (PGs), in addition to having the
phosphoester functionality, also contain two hydroxyl moieties, enabling more extensive
hydrogen bonding with water. The issue we are concerned with in this paper is whether
or not the structural differences between the PC and PG lipid head groups gives rise to
significant differences in the dynamics of bilayers made from these lipids, and we

interrogate the bilayer dynamics by means of a tethered chromophore.

We use the chromophore 7-nitrobenz-2-oxa-1,3-diazole, NBD, a molecule that
has been used extensively for interrogating biological and biomimetic systems?“32 This
chromophore is characterized by a fluorescence spectral profile and lifetime that both

depend sensitively on the polarity of its immediate environment?“32 In addition, the

130

chromophore NBD is available with a variety of sidegroups, or attached to lipid head
groups or acyl chains, making it a particularly attractive choice for the study of lipid

bilayers.

We have reported previously on the dynamics of NBD tethered to a PC lipid.21 In
that work we examined the fluorescence lifetime and induced orientational anisotropy
decay of the NBD chromophore as a function of bilayer composition. We used NBD
tethered to a phospholipid acyl chain, and found that the environment experienced by the
chromophore depends sensitively on the composition of the bilayer. The addition of
cholesterol and sphingomyelin to the phospholipid bilayer reduced the organization of the
chromophore environment significantly but did not influence the (conic) voltune of the
restricted region in which the chromophore can orientationally relax. The functional
form of the anisotropy and lifetime data pointed to the chromophore residing in the
proximity of the bilayer polar head group region with the local viscosity of the bilayer
region sensed by the probe being ca. 2 CF. The most plausible explanation for this
finding is that the chromophore tethered to the lipid acyl chain is able to “loop back” and
reside in proximity to the phospholipid head group.28 In this work, we examine the
lifetime and anisotropy decay dynamics of the molecule (l-Palmitoyl-2-[l2-[(7-nitro-2-

1 ,3-benzoxadiazol-4-yl)amino]dodecanoyl]-sn-Glycero-3 -[Phospho-rac-(1-glycerol)]

(N BD-PG). Use of this chromophore allows us to gauge the effect of the lipid headgroup
on interactions with their local environment for a system that is expected to hydrogen-
bond extensively both among lipid headgroups and with surrounding water. We use one-

and two-photon excitation to determine fluorescence lifetime and anisotropy decay

131

dynamics. The data we present here show that the identity of the lipid headgroup plays a

significant role in chromophore behavior.

132

Experimental

Materials. The fluorescent probe 1-Palmitoyl-2-[12-[(7-nitro-2-1,3-
benzoxadiazol-4-yl)amino]dodecanoyl]-sn-Glycero-3-[Phospho-rac-(1-glycerol)] (N BD-
PG) was obtained from Avanti Polar Lipids, Inc. and was used without further
purification. For the construction of unilamellar vesicles, the lipids 1,2-Dilauroyl-sn-
Glycero-3—[Phospho-rac-(1-glycerol)] (DLPG, mp = -3°C), 1,2-Distearoyl-sn-G1ycero-3-
[Phospho-rac-(l-glycerol)] (DSPG, mp = 55°C), egg sphingomyelin (SPM) and
powdered cholesterol were obtained from Avanti Polar Lipids, Inc. The vesicle
constituents were prepared in a 1:1:1 mole ratio of phospholipid : cholesterol :
sphingomyelin or a 2:1 mole ratio of phospholipid : cholesterol for the multicomponent
systems. The tagged phospholipids in these systems were present at 10’4 M. A total of six
vesicle solutions were produced: (1) NBD-PG/DLPG, (2) NBD-PG/DLPG/cholesterol,
(3) NBD-PG/DLPG/cholesterol/SPM, (4) NBD-PG/DSPG, (5) NBD-
PG/DSPG/cholesterol, and (6) NBD-PG/DSPG/cholesterol/SPM. Chloroform was
evaporated from the lipid solution, and the remaining solid was dissolved in a 4:1
benzene : methanol solvent system. The benzene : methanol solvent was then
evaporated, and the resulting lipid mixture was dissolved in 10 mM Tris® buffer
(Aldrich) adjusted to pH~8 in MilliQ water. The resulting concentration of NBD-PG in
each vesicle solution was ~10“4 M. For further mixing, the solution was put through five
freeze-thaw-vortex cycles, where for each cycle the sample is placed in liquid nitrogen
for 5 minutes, then in a hot water bath for 5 minutes, then vortexed for 2 minutes. A

syringe-based mini-extruder from Avanti Lipids was used to extrude the lipid suspension

133

through a polycarbonate filter eleven times to produce unilamellar vesicles of ~100 nm

diameter. l "23

Steady State Measurements. Absorption spectra were recorded on a Cary model
300 double beam UV-Visible absorption spectrophotometer, with 1 nm spectral
resolution. Emission spectra were recorded on a Spex Fluorolog 3 spectrometer at a

spectral resolution of 3 nm for both excitation and collection monochromators.

Time Correlated Single Photon Counting Measurements. All lifetime and
anisotropy data were acquired using a time correlated single photon counting (TCSPC)

21‘23'33’34 and we recap only

system. This system has been described in detail elsewhere,
its salient features here. The source laser is a CW mode-locked Nd:YAG laser (Coherent
Antares 76-S) that produces 30 W average power at 1064 nm with 100 ps pulses at 76
MHz repetition rate. The second harmonic (532 nm, 3W average power) or third
harmonic (355 nm, 1 W average power) of the output of this laser excites a cavity
dumped dye laser (Coherent 702-2), operating with Rhodamine 610 dye (Exciton, 532
nm pump) for two photon excitation experiments, or with Stilbene 420 dye (Exciton, 355
nm pump) for one photon excitation experiments. The dye laser output is typically 5 ps
pulses at a repetition rate of 3.8 MHz for both excitation wavelengths. Fluorescence
signals are spectrally filtered using a subtractive double monochromator (American
Holographic) and are detected using a Hamamatsu R3 809U-51 microchannel plate
photomultiplier tube (MCP-PMT) detector. The signals are processed using a Tennelec
454 quad constant fraction discriminator and Tennelec 864 time-to-amplitude converter

and biased amplifier. For this system the instrument response time is ca. 35 ps fwhm.

Fluorescence was collected at 0°, 54.7° and 90° with respect to the vertically polarized

134

excitation pulse. Collection wavelength, polarization and data acquisition were

controlled using National Instruments LabVIEW® 7 .1 software.

Pressure-Area Isotherms. Pressure-area isotherms were generated using a
Langmuir-Blodgett Trough (N ima Technology Ltd. Model 612D). The lipid solutions
were prepared at a concentration of 1 mg/mL in chloroform, with a 2:1 molar ratio of
phospholipidzcholesterol and 1:1:1 phospholipid, cholesterol, SPM. For these
experiments, the fluorescent lipid and the phospholipid were present in equimolar

amounts.

135

Results and Discussion

We are interested in understanding the motional dynamics of 1-Palmitoyl-2-[l2-
[(7-nitro-2- l ,3-benzoxadiazol-4-yl)amino]dodecanoyl]-sn-Glycero-3-[Phospho-rac-( l -
glycerol)] (N BD-PG) (Figure 5.1) in unilamellar vesicles comprised of phosphoglycerol
lipids, sphingomyelin and cholesterol. We are concerned not only with the dynamics of
the chromophore, but also the specific region within the unilamellar vesicle where the
tethered chromophore localizes. We have obtained pressure-area isotherm data to gauge
the organization of our bilayers and the extent to which the presence of the chromophore
perturbs the bilayer. We have measured both the fluorescence lifetime and anisotropy
decay using one- and two-photon excitation of the chromophore incorporated into the
vesicles. Using two different excitation methods provides access to mutually
complementary information regarding the motion of the chromophore. With these data,
we can evaluate the Cartesian components of the rotational diffusion coefficient, a task

which is not possible using either of these excitation methods individually.

Previous literature reports have demonstrated that the polarity of the NBD
chromophore environment plays a significant role in determining its fluorescence
lifetime.23‘25'32 NBD is a polar chromophore, and it is reasonable to expect this molecule
to reside preferentially in a polar environment.”’25‘27’28 The absorption maximum for

NBD shifts to the red in polar solvents,25'27’3O

and our steady state absorbance data show
that for all vesicle-containing solutions, the NBD absorption maximum is ~460 nm and

its emission maximum is ~5 50 nm (Figure 5.1), consistent with a polar environment.

This finding is consistent with what is known about tethered NBD is bilayer systems.

136

(l6

Ox1~

normalized intensity

(12

 

0.0
350

I l
400

l

450

l m

J
500 550

wavelength (nm)

L

l

600

CH1

CH1

VPII)

650 700

NH
/N\
o
\ /
N
N02

Figure 5.1 Normalized absorption and emission spectra of NBD-PG in ethanol (left)

and structure of l6:0-12:0 NBD-PG (right).

137

When NBD is attached to the acyl chain of a lipid and incorporated into a lipid bilayer, it
is known that the NBD moiety will “loop back” into the polar region of the bilayer, in

close proximity to the lipid head group.2"25'28

The NBD fluorescence lifetime data for both one- and two-photon excitation yield
the same result, which is an expected result because the emitting state for both means of
excitation is the S1 (71*) state. Time-resolved fluorescence lifetime data show that the
excited state population decay of NBD-PG in vesicles contains at least two exponential
components in all cases (Figure 5.2), indicating the presence of the probe in multiple
environments characterized by a range of polarities. The long lifetime component for all
cases ranges from ~1300 ps to 2200 ps, while the short lifetime component is between
300 ps and 400 ps. These two lifetimes differ significantly in each vesicle system, and
their fractional contribution to the overall decay also varies with the composition of the
vesicle (Table 5.1). These environmentally-dependent differences in lifetime can be
attributed to the gradient in the dielectric response of the bilayer across the headgroup
region, ranging from e~2 in the nonpolar portion of the bilayer to a~80 in the aqueous
region above the head group region, a distance of ca. 10 A. It is this significant dielectric
gradient across the bilayer head group region that modulates the NBD lifetime data?”8
rather than significant differences in chromophore position within the bilayer. Thus, the
two-component lifetime should not be taken to indicate two radically different

environments; rather, the chromophore senses a large change in dielectric response as a

result of subtle positional differences within the bilayer head group region.

138

To evaluate the organization of the phospholipids and the extent to which the
presence of the NBD chromophore affects organization, pressure—area isotherms of

Langmuir-Blodgett monolayers were obtained for each of the six vesicle-forming

solutions (after mixing but not extruded), both with and without the tethered
chromophore (Figures 5.3a and 5.3b). For ordered systems characterized by a high Tm,
such as DSPG, the pressure-area isotherm shows a slight increase in lateral packing
density as well as an apparent slight increase in organization with the addition of
cholesterol (Figure 5.3b), and we note this same qualitative trend for DLPG, although the
isotherms appear to be more liquid-like under all conditions for this lipid (Figure .
5.3b).5‘6‘9’39 Upon addition of SPM to DSPG films, there is notable expansion of the
bilayer; the opposite is seen for the disorganized monolayer of DLPG when SPM is

4,39-48

added (Figure 5.3b). Such results have been observed previously, and this behavior

most often attributed to the cooperative effect of SPM and cholesterol when both are

39’40‘4345 The surface monolayers that did not contain the NBD-

present in the bilayer.
tagged lipid showed that DSPG films exist as solids, presumably with ordered acyl
chains, and the DLPG films behave as liquids. The addition of the chromophore-
containing lipid to the monolayer caused the monolayers to become more liquid-like in
all cases, owing to the disorder induced by the presence of this species. Similar results
were seen for the isotherms obtained for phosphocholine lipid monolayers (Figure 5.4).

These data underscore the significant role that the lipid acyl chains as well as non-lipid

constituents play in determining overall organization within the layers.

139

3500

3000 1-

2500

2000

1

counts

1500

1000

I

500

l

 

 

l L l l l l J_ J l l l l

0 1000 2000 3000

I l 1 1 I

4000 5000
time (ps)

Figure 5.2 Example fluorescence lifetime data (points) for DLPG vesicles containing
NBD-PG, showing a fit to a double exponential decay functionality. (1'1=327

ps, 12:2123 ps) The solid line through the data is the fit, with the residuals

plotted as fluctuations about zero intensity.

140

Table 5.1. One-photon excited fluorescence lifetimes of NBD-PG in vesicles.
Experimental data were fit to the function f(t) = Alexp(-t/‘n) + Azexp(-t/‘tz).
Uncertainties are :1: 10' for at least six individual determinations.

 

 

 

 

 

 

 

 

 

Vesicle system 11 (ps) 12 (ps) Al/(A1+A2)
NBD-PG, DLPG 347 1 20 2211 1 31 0.27 1 0.01
NBD-PG, DLPG, chol 425 1 17 1394 1 38 0.40 1 0.02
NBD'PG’ DLPG’ cm" 332 1 5 2232 1 50 0.81 1 0.003
SPM
NBD-PG, DSPG 415 1 20 1305 1 16 0.34 1 0.01
NBD-PG, DSPG, chol 374 1 13 1800 1 70 0.64 1 0.01
Egam’ DSPG’ cm" 390 1 8 1340 1 19 0.62 1 0.02

 

 

 

 

141

 

   
   

 

 

 

 

 

 

50 —
x
. "\.\ 13 —DLPG(1) a
_ ‘- --------------- DLPG+Chol(2)
g 40 ------- ~---DLPG+Chol+SPM(3)
Z - 21‘ ------- DLPG+NBD (4)
E 30 - - ----------- --DLPG+NBD+Chol(5)
2 _ DLPG+NBD+ChoI+SPM (6)
:3 \ "1
a 20 — 6
10 r \‘f-‘Jr‘
r
O r : L L L .1 1 '1 1 . :
20 40 60 80 100 120 140
area/molecule (A2)
60 - b
‘ —DSPG (1)
a 50 - ------- DSPG+Chol (2)
2 - ----------- DSPG+Chol+SPM (3)
E 40 - ----------- DSPG+NBD (4)
v - -. - ------------- DSPG+NBD+ChoI (5)
“g 30 — '~-..‘-~~-~-~~ DSPG+NBD+Chol+SPM (6)
m _ .
8 20 - E .
0,, _ “.\\\ \x
10 _ \ ..
0 - 1:§\“- : \i ““31 r a
20 80 100 120 140

 

area/molecule (A2)

Figure 5.3 Pressure-area isotherms for Langmuir-Blodgett monolayers of (a):
DLPG (l), DLPG + Cholesterol (2), DLPG + Cholesterol + sphingomyelin
(3), DLPG + N BD-PG (4), DLPG + NBD-PG + Cholesterol (5), DLPG +
NBD-PG + Cholesterol +sphingomyelin (6). (b): DSPG (1), DSPG +
Cholesterol (2), DSPG + Cholesterol + sphingomyelin (3), DSPG + NBD-PG
(4), DSPG + NBD-PG + Cholesterol (5), DSPG + NBD-PG + Cholesterol
+sphingomyelin (6).

142

When NBD-PG is incorporated into the LB monolayers, a significant amount of disorder
results. With the addition of tethered NBD, DSPG/PC lipid monolayers exhibit pressure-
area isotherms characteristic of liquids at room temperature (Figures 5.3b and 5.4). This
is not a surprising result, because the NBD-containing lipid is comparatively large and
bulky, a condition which will interfere with tight lipid packing. Following cholesterol
incorporation into the chromophore-containing monolayers, we do not observe an
increase in fluidity. For LB films that do not contain the chromophore, the addition of
cholesterol gives rise to a reduction in organization, producing an isotherm with a more
fluid—like transition than is seen for the lipid-only monolayer (Figures 5.3a and 5.3b).
The chromophore concentrations used in these isotherm measurements are higher than
those used for the spectroscopic measurements to make clear the role of the chromophore
in layer organization. It is reasonable to expect that the presence of the chromophores
perturbs local organization for any chromophore concentration, and it is this environment
that is sensed by the probe spectroscopic response. Despite the limitations to this type of
measurement, we find that the local organization of our bilayer vesicles is affected
measurably by the addition of cholesterol and sphingolipid, and we consider these issues

HBXI.

Anisotropy decay data for NBD-PG in vesicle systems are capable of revealing
significant detail about the chromophore local environment. Interpretation of anisotropy
decay data obtained using one- and two-photon excitation differs because of the different
initial orientational distribution selected by these two methods and the different
contributions of spectroscopic and diffusional terms to the anisotropy decays. The

information obtained from the one-photon excitation is complementary to that extracted

143

 

 

 

50 — .
A 17- ——DSPC (1)
E 40 r . 1 ------ DSPC + Chol (2)
2 I .......... DSPC +NBDPC (3)
E ‘1. '3 .......... Dspc+NBDPC+Chol (4)
2 30 _ .1 - ----------- NBDPC (5)
o. 20 - 5
10 - “1;;
0 I“ 1 . . . r: iiiiiiii i """" :-
20 60 80 100 120 140

 

area/molecule (A2)

Figure 5.4. Pressure-area isotherms for Langmuir-Blodgett monolayers of DSPC
(l), DSPC + Cholesterol (2), DSPC + N BDPC (3), DSPC + NBDPC +
Cholesterol (4), NBDPC (5).

144

from the two-photon excitation measurements. These data, taken collectively, provide a
detailed description of tethered NBD motion in lipid bilayers. We treat the information
contained in each body of data separately before considering that available through a

comparison of the two data sets.

For one photon excitation, there is a well-established framework for the
interpretation of anisotropy decay data.34’49'55 The induced orientational anisotropy
function, R(t), is generated from the polarized fluorescence signal intensities (Eq. 5 .1),

= 1110—110)
1,,(z)+ 21,0)

 

R(t) (5.1)

where the excitation is vertically polarized and the quantities I“(t) and I 1(t) are the
polarized emission transients. The chemical information extracted from these data is
contained in the functional form of R(t). While up to five exponential decays may be
present in R(t),50 the most common case is to observe one or two exponential decays.
The relationship between the decay functionality of R(t) and the chromophore transition
moment(s) and rotational diffusion constant has been described in detail.34"m'55 Chuang
and Eisenthal’s treatment50 described R(t) as a function of rotor shape and transition
moment orientation, with the limiting cases being either a Type I or Type II rotor, which
simplifies the general expression for R(t). A Type I rotor is characterized by a distortion
along the axis containing the transition moment, and a Type II rotor is characterized by a
distortion along an axis perpendicular to the transition moment.2| For a transition
moment along the ith axis (one of the Cartesian axes), there are two general solutions to

the anisotropy decay:

145

TypeI R(t) = 0.4 exp(—6Dkt) (5.2)
Type II R(t) = 0.1exp(—6D,t)+ 0.3 exp(—(2D, + 4D,, )t) (5.3)

The information that can be extracted from R(t) thus depends on the orientation of the
absorption and emission transition dipole moments of the chromophore. The NBD
transition moment lies in the 7: plane, where we take the x-axis to be the long molecular
axis, and the y-axis is coincident with the short in-plane axis. The bond tethering the
NBD moiety to the lipid molecule therefore is oriented along the y-axis, and under this
condition, NBD is characterized by Dy 9E Dx = Dz. For this system, the anisotropy can
decay as either a single or double exponential, depending on the orientation of the

transition (x or y axis):

x-polarized transition R(t) = 0.3 exp(— (4D), + 2Dx )t) + 0.1exp(—6Dxt) (5.4)
y-polarized transition R (t) = 0 .4 exp(—6Dxt) (5.5)

Our experimental data are, to within our ability to determine, single exponential
anisotropy decays, suggesting that the transition is y-axis polarized, and the recovered
reorientation time is reflective of motions about the chromophore x- and z-axes.
Specifically, the recovered anisotropy decay time constant for one photon excitation is
given by (6D,.)". We recognize that, owing to the presence of the tether, NBD is not
allowed to rotate freely about the x-axis, and the recovered time constant is representative
of wobbling and angular tilting about the tethering bond. It is clear, however, that the
motional relaxation behavior of NBD in our vesicle structures is not well characterized

when one photon excited fluorescence anisotropy is the only method used. This

146

information becomes more useful when compared to the results from two photon

excitation, and we consider these data next.

A brief word is in order regarding the assignment of rotor shape for these data.
The interpretation of our results in the context of the one-photon excited anisotropy decay
being a single exponential ultimately relies on the S/N ratio of our data, which is limited
(Figure 5.5). We note that in the limit of an isotropic rotor (Dx~D,/~Dz), Eqs. 5.4 and 5.5
reduce to the same functional form. Comparing the one- and two-photon excitation data,
which we discuss next, we find that, except for one system, we in fact recover isotropic
rotor behavior for the tethered NBD chromophore (vide infra), rendering the issue of

distinguishing between Eq. 5.4 and Eq. 5.5 largely moot for this system.

To understand the anisotropy decay data acquired using two—photon excitation,
we borrow from the treatment presented by the Johnson group.”62 Induced orientational
anisotropy decay functions are generated for both linearly (r1(t)) and circularly (r2(t))

polarized excitation (Eqs. 5.6),

linear linear
r.(t) = I," (”—13 (t)
Ill/mar (t) + 211mm (t)

Ilc'm‘ular (t) _ [in-calm (I)

1
r2 (1) = Circular circular
21H (t)+]i (t)

 

(5.6)

 

For both excitation polarizations, the anisotropy decays are characterized by two time

constants, to and 12, weighted according to the spectroscopic and motional properties of

147

0.2

 

 

f
0.1 ~ .
E P . .ooa .. .... a. .. :. -.u..... .:............ ...'
:21": . ”[20, " .:.1 '1 “1.1.: .: Ill . '1I .1 1‘ 1" [1'1
00 . f it 1. l 1 ,HEW:.llltllMIHMIIliyfill5111111111111151511!111 15.111111115111111llllll‘l'l
1 11111 1 11111111111111: ' 111.311.1111 ”l
l 1
_0- 1 1 1 1 1 1 l . . . 1 l i . l . L
0 500 1000 1500
time (ps)

Figure 5.5 Two photon excited anisotropy decay for vesicles containing DSPG,
NBD-PG and cholesterol. Points are experimental data, solid line is the fit of
the data to an exponential decay and the residuals of the fit are shown
centered about the zero intensity line.

148

the chromophore. The anisotropy decays are given by,58’59’6"62

rl (t) = r1 (O)l:coe—'/r0 + cze_’/t2:l

 

 

 

 

 

_ r - . 5.7
r2(t):r2(0)[d0e ” . +d2e ” ] ( )
Where the exponential prefactors are60

(13a + b) 736/3}: +b)s; + 3(—\/3a 1 ms; + 26835,, + 46531
C = - - ' — 5.8
0 7N2(3s:, + 3s; + zsnsw +4sfy) ( )
(a - J36)[3(a — has; + 3(a + has; + 2asns”, +4as;,]
c = 5.9
2 7N2(3s; +3sfy +2515” +4s;) ( )
d (fia+b)[(fia+b)s§, +(—\/3a+b)Sfy —4bSnSyy +6bey]
0 14N2(sj, +83, —s,,s”, +3sfy) ( )
(a—xl3b)[(a—«/3b)SZ +(a+\/3b)SZ —4aS s +6as2 ]
d2 = * W U ,, x. (5.11)

 

141V2 (53, + sf, - snsx, + 3sz

and the quantities a, b, N and A are related to the rotational diffusion constant D by

influx—1),)
b=2D —D —D +2A

‘ y x (5.12)
N2=a2+b2

Aqm+m+m—Qq—qQ-QQW

While Eqs. 5.7 predict two exponential decay components for each excitation
polarization, we recover only one decay time constant in each case, and this is a

commonly observed situation. Further, for each system studied here, the anisotropy

149

decay time constant for linear and circular excitation polarization is the same, 1,1 (linear)
= 1,2 (circular). The most general interpretation of single exponential decays for two
photon excitation is that the recovered time constants are simply weighted averages of To

0
and 12,6

_ (corO + c212)

 

’1

c0 +c2
r _ (doz'0 +d27'2)
r2 .-
d0 +d2

(5.13)

 

This condition does not, in general lead to the same decay time constant for both linear
and circular excitation polarization. Because Dx ~ Dz for our systems (from one photon
excitation data), the term (V3a+b) in Eqs 5.8 and 5.10 is zero, making the prefactors co
and do equal to zero for DK 2 Dy, and this condition is met to within the experimental
uncertainty, save for one case (vide infra). The time constants we recover are thus Tr|=T2,
and r,2=12 where ”cz‘-‘(6D+2A)'l.60’6l This situation is consistent with the fact that we
recover the same time constants, to within the experimental uncertainty for both linear
and circular excitation polarization, and for these conditions, 1' has contributions from
both Dy and Dx:

1'2 = 1
(40. + 213..)

 

(5.14)

We determine D)( from one-photon excitation, and use that information, in
combination with two-photon data to determine Dy. We recognize that the values of Dx
and Dy we recover are not free rotational motions about each axis because of the tethering

attachment to the phospholipid. These quantities represent random walk (Dx and D2) and

150

predominantly rotational (Dy) motions of the chromophore. With this information in
hand, we are now in a position to consider the changes in chromophore local environment

associated with variations in bilayer composition.

The anisotropy decay time constants, calculated values of Dy and Dx, and Dy/Dx
are summarized in Table 5.2. Vesicles comprised of DLPG vesicles are characterized by
Dy/Dx = 5.7, suggesting a substantially anisotropic environment for the chromophore,
where there is relatively little lateral (Dz, Dx) freedom and the dominant motion is about
the y-axis. With the addition of cholesterol and/or sphingomyelin to the bilayers, the
ratio Dy/Dx decreases to 1.3 1 0.1 (cholesterol) and 1.1 1 0.2 (cholesterol and
sphingomyelin). The approach to an isotropic environment suggests that there is a
decrease in organization of the bilayer structure with the addition of cholesterol and/or
sphingomyelin, in agreement with our earlier findings for phosphocholine bilayers
formed with DLPC and DSPC.2| The reduction of the ratio Dy/Dx in these bilayers
suggests that these structures are becoming less organized and perhaps more fluid,
consistent with the functional form of the pressure area isotherms for the corresponding

Langmuir Blodgett monolayers (Figures 5.3a and 5.3b).

The anisotropy data on Dy/Dx shows a consistent trend for DLPC, DSPC, and DLPG,
with comparative confinement in the phospholipid vesicles, and an increase in the
motional freedom when cholesterol and/or sphingomyelin are added. However, DSPG
vesicles show a different bilayer composition dependence through Dy/Dx, even though the

isotherms change in a manner similar to the other monolayers when the NBD-containing

151

Table 5.2. Experimental reorientation time constants and calculated values of
rotational diffusion coefficients for NBD-PG in the vesicles of compositions
indicated in the first column.

 

 

 

 

 

 

 

 

Vesicle 1: 1 10 (one tOR i 10
on (two photon, D (MHz) D (MHz) D /D
system photon, ps) x Y Y "
ps)
DLPG 678 1 34 167 1 21 246 1 13 1400 1 7 5.7 1 0.3
DLPG, Chol 93 1 5 77 1 13 1797 1 97 2272 1 49 1.3 1 0.1
311;;6’ Cho" 81110 79113 20811249 21991125 1.1102
DSPG 138111 153127 1212196 1059148 08710.1
DSPG,chol 117121 121129 14901291 14271146 09610.2
25,5? cm" 82 1 14 88 1 22 2084 1 380 1841 1 190 0.88 1 0.2

 

 

 

 

 

 

152

 

lipid is added. The DSPG vesicles that contain the tethered NBD are characterized Dy/Dx
= 0.87 1 0.10, and that value remains constant to within the experimental uncertainty
with the addition of cholesterol (Dy/Dx = 0.96 1 0.20) and sphingomyelin (Dy/Dx = 0.88 1
0.20). This finding indicates the environment probed by the NBD-PG for DSPG-based
systems does not change its effective shape with the addition of cholesterol and/or
sphingomyelin. The ratio of Dy/Dx does not provide the complete picture, however.
When examining trends in D = l/3(Dx+Dy+Dz), we see that D increases for both DSPC
and DSPG lipids with the addition of cholesterol, and increases slightly more with the
addition of sphingomyelin in addition (Table 5.2). Thus the qualitative organization of
the local environment does not appear to change for DSPG vesicles with the addition of
cholesterol and/or sphingomyelin, but the addition of these species does introduce more
fluidity into the portion of the bilayer sensed by the tethered chromophore. There are a
few characteristics of the DSPG bilayers that could account for these observations.
Comparisons between DSPC and DSPG show differences in organization of pure
phospholipid bilayers. Although the transition temperatures are comparable, the
difference between the two lipids lies in the identity of the headgroup. The glycerol
headgroup will participate in a network of hydrogen bonding with the neighboring
headgroups while choline does not. The extensive hydrogen bonding results in stronger
interactions between lipids and chromophore, and thus more ordered bilayers. Although
this factor could account for differences between DSPG and DSPC, a comparison must
be made between DLPG and DSPG systems. Both DLPG and DSPG contain the glycerol
headgroup that will participate extensively in hydrogen bonding between headgroups, the

aqueous layer outside the bilayer, and the NBD group as it folds back to be in close

153

proximity to the hydrophilic head group region of the bilayer. The difference in behavior
for NBD in the DLPG and DSPG vesicles suggests that for DSPG the local environment
of NBD is more isotropic than it is for DLPG. Given the commonality in head group
structure for DLPG and DSPG, and our data indicating the lifetimes of tethered NBD are
the same in DLPG and DSPG vesicles, the only factor that can be responsible of the
observed behavior is the length of the acyl chains, and differences in bilayer organization
that result from this difference. Because the chromophore has a C16 acyl chain, if this
chain were all-trans, it would span the DLPG monolayer and penetrate to some extent
into the opposing leaflet, whereas for DSPG this is not the case. The consequences of
this difference may be that the chromophore-containing lipid is more translationally
constrained in DLPG. The functional form of this finding is somewhat surprising
because DSPG has a higher melting point than DLPG, and intuition would suggest
behavior analogous to that seen for DSPC and DLPC. Further investigation is required to

understand this behavior fully.

For chromophores imbedded in an intrinsically anisotropic medium, such as the
lipid bilayers we examine here, the data are often treated using the hindered rotor model.
The chromophore we use in this work, NBD-PG, consists of an NBD moiety attached to
the aliphatic chain of a phosphoglycerol lipid. Due to attachment of the chromophore to
a lipid, the hindered rotor model is an appropriate model for describing the behavior of
NBD-PG. This model has been used previously to describe the one-photon excited
anisotropy decay behavior of other chromophores embedded in a bilayer.57’63’64 In this
model, the motion of the chromophore is restricted by its environment as well as its

attachment to a membrane constituent; we anticipate the motion of the NBD

154

chromophore to sweep out a conical volume. The cone angle (200) that characterizes
chromophore motion is related to the infinite time anisotropy, R(oo), and the zero time

anisotropy, R(0).

Roo V2
[Pi—0}] =0.5(cosBO(1+cosBO)) (5.15)

Where R(OO) is related to the degree of confinement imposed by the chromophore
environment and R(0) is related to the spectroscopic properties of the probe. From our
experimental data we extract R(0) and R(OO), and calculate the cone semi-angle 00
according to Eq. 5.15. We present these results for NBD—PG in our vesicles in Table 5.2.
Notably, the cone angle values are the same to within the experimental uncertainty for all
systems studied, and are relatively large, indicating the local environment of the NBD
chromophore is reasonably constant and disordered for the PG lipids. This finding is in
contrast to previous results we reported for phosphocholine lipids,21 where the addition of
cholesterol and sphingomyelin produced a marked decrease in cone angle, reflective of
the system becoming more organized (Table 5.2). The comparative constancy of the
cone angle for the PG lipids, with or without cholesterol and/or sphingomyelin, is
consistent with the hydrogen-bonding amongst the PG head groups mediating layer
organization as sensed by the NBD chromophore, in contrast to PC lipids, where acyl

chain organization is thought to dominate.

155

Conclusions

We have investigated the dynamics of the tethered NBD-PG chromophore
contained within unilamellar vesicles comprised of phosphoglycerols DLPG and DSPG,
with or without cholesterol and sphingomyelin added. Our data point to some similarities
between vesicles comprised of phosphoglycerol and phosphocholine lipids, but the ability
of the phosphoglycerol head groups to form hydrogen bonded networks provides an
additional means of mediating bilayer organization in the proximity of the polar head
groups. These findings point to comparatively subtle differences between the
phosphocholine and phosphoglycerol bilayers, and underscore the ability of spectroscopic
probes to interrogate these differences. Using both one- and two-photon excitation for
the acquisition of time-domain fluorescence data, we are able to extract information on
the Cartesian components of the rotational diffusion constant, revealing variations in the
confinement of the chromophore by the bilayer that depend on the bilayer constituents.
The apparently constant cone of confinement experienced by the chromophore suggests a
significant role of head group hydrogen bonding in mediating organization within the

polar regions of phosphoglycerol lipid bilayers.

156

Literature Cited

(1) Araiso, T., Koyama, T. Japanese Journal of Physiology 1995, 45, 187.

(2) Brown, D. A., London, E. Biochemical and Biophysical Research
Communications 1997, 240, 1.

(3) Brown, D. A., London, B. Journal of Membrane Biology 1998, 164, 103.

(4) McConnell, H. M., Radhakrishnan, A., Xin-Min, L., Brown, R.E. Biochimica et
Biophysica Acta 2001, 1511, 1.

(5) Scherfeld, D.; Kahya, N.; Schwille, P. Biophysical Journal 2003, 85, 3758.
(6) Simons, K.; van Meer, G. Biochemistry 1988, 27, 6197.

(7) Wisniewska, A.; Draus, J .; Subczynski, W. K. Cellular & Molecular Biology
Letters 2003, 8, 147.

(8) Brown, R. E. Journal of Cell Science 1998, 111, 1.

(9) I Crane, J. M.; Tamm, L. K. Biophysical Journal 2004, 86, 2965.
(10) Hao, M.; Maxfield, F. R. Journal of Fluorescence 2001, 11, 287.
(l 1) Hunter, D. G.; Frisken, B. J. Biophysical Journal 1998, 74, 2996.

(12) Lee-Gau Chong, P. W. v. d. M., B. Thompson, T.E. . Biochimica et Biophysica
Acta 1985, 813, 253.

(l 3) McConnell, H. M., Radhakrishnan, A. Biochimica et Biophysica Acta 2003, I610,
159.

(14) McMullen, T. P. W., Lewis, R.N.A.H, McElhaney, R.N. . Current Opinion in
Colloid and Interface Science 2004, 8, 459.

157

(15) Mitchell, D. C., Litman, B.J. Biophysical Journal 1998, 75, 896.

(16) Samsonow, A. V., Mihalyov, 1., Cohen, PS. Biophysical Journal 2001, 81.

(I7) Sankaram, M. B., Thompson, T.E. Biochemistry 1990, 29, 10670.

(18) Silvius, J. R. Biochimica et Biophysica Acta 2003, 1610, 174.

(19) Slotte, J. P. Chemistry and Physics of Lipids 1999, 102, 13.

(20) Webb, S. J ., Greenaway, K., Bayati, M., Trembleau, L. Organic and
Biomolecular Chemistry 2006, 4, 2399.

(21) Greenough, K. P. a. B., G. J. Chemical Physics 2007.

(22) Koan, M. M., Blanchard, G. J. . 2006.

(23) Greenough, K. P. a. B., G. J. The Journal of Physical Chemistry B 2006, 110,
6351.

(24) Chapman, C. F.; Liu, Y.; Sonek, G. J .; Tromberg, B. J. Photochemistry and
Photobiology 1995, 62, 416.

(25) Chattopadhyay, A.; London, E. Biochimica et Biophysica Acta, Biomembranes
1988, 938, 24.

(26) Fery-Forgues, S.; Fayet, J .-P.; Lopez, A. Journal of Photochemistry and
Photobiology, A: Chemistry 1993, 70, 229.

(27) Greenough, K. P. a. B., G. J. The Journal of Physical ChemistryA 2007, 111, 558.

(28) Huster, D.; Muller, P.; Arnold, K.; Herrmann, A. Biophysical Journal 2001, 80,
822.

(29) Lin, 8.; Struve, W. S. Photochemistry and Photobiology 1991, 54, 361.

158

(30) Mazeres, S.; Schram, V.; Tocanne, J .-F.; Lopez, A. Biophysical Journal 1996, 71,
327.

(31) Mukherjee, S.; Chattopadhyay, A.; Samanta, A.; Souj anya, T. Journal of Physical
Chemistry 1994, 98, 2809.

(32) Tsukanova, V.; Grainger, D. W.; Salesse, C. Langmuir 2002, 18, 5539.

(33) Goldie, S. N., Blanchard, G.J. Journal of Physical Chemistry A 1999, 103, 999.

(34) J iang, Y.; Blanchard, G. J. Journal of Physical Chemistry 1994, 98, 6436.

(35) Drexhage, K. H. Journal of Luminescence 1970, 1,2, 693.

(36) Girard, C.; Martin, O. J. F.; Dereux, A. Physical Review Letters 1995, 75, 3098.

(37) Lukosz, W.; Kunz, R. E. Optics Communications 1977, 20, 195.

(3 8) Lukosz, W.; Kunz, R. E. Journal of the Optical Society of America 1978, 6 7,
1607.

(39) Bittman, R., Kasireddy, C. R. , Mattjus, P., Slotte, J .P. Biochemistry 1994, 33,
11776.

(40) Koopman, E. Sphingomyelin-cholesterol superlattices as detected with Langmuir
isotherms: their potential role in myelin and demyelination; Husted, C. A., Ed.;
University of California, Santa Barbara, 2003; Vol. I.

(41) Korchowiec, B., Paluch, M., Corvis, Y., Rogalska, E. Chemistry and Physics of
Lipids 2006, 144, 127.

(42) Prenner, E., Honsek, G., Honig, D., Mobius, D., Lohner, K. Chemistry and
Physics of Lipids 2007, 145, 106.

(43) Rog, T., Pasenkiewicz-Gierula, M. Biophysical Journal 2006, 91 , 3756.

159

(44)

(45)
5696.

(46)

(47)

(48)

(49)

(50)

(51)

(52)
9328.

(53)

Smaby, J. M., Brockman, H.L., Brown, R.E. Biochemistry 1994, 33, 9135.

Smaby, J. M., Momsen, M., Kulkami, V. S., Brown, R.E. Biochemistry 1996, 35,

Stottrup, B. L., Stevens, D.S., Kellar, S.L. Biophysical Journal 2005, 88, 269.

Terova, B., Heczko, R., Slotte, J .P. Biophysical Journal 2005, 88, 2661.

Yuan, C., Johnston, L.J. Journal of Microscopy 2002, 205, 136.

Blanchard, G. J. Journal of Chemical Physics 1987, 8 7, 6802.

Chuang, T. J .; Eisenthal, K. B. Journal of Chemical Physics 1972, 5 7, 5094.

Debye, P. Polar Molecules. In Chemical Catalog Co. New York, 1929.

Dela Cruz, J. L.; Blanchard, G. J. Journal of Physical Chemistry A 2001, 105,

Lakowicz, J. R.; Gryczynski, 1.; Gryczynski, Z.; Danielsen, E.; Wirth, M. J.

Journal of Physical Chemistry 1992, 96, 3000.

(54)

McPhie, P. Principles of Fluorescence Spectroscopy, Second ed. Joseph R.

Lakowicz, 2000; Vol. 287.

(55)

(56)

(57)

(58)

(59)

Zwanzig, R.; Harrison, A. K. Journal of Chemical Physics 1985, 83, 5861.

Callis, P. R. Journal of Chemical Physics 1993, 99, 27.

Chen, S. Y.; Van Der Meer, B. W. Biophysical Journal 1993, 64, 1567.

Johnson, C. K.; Wan, C. Topics in Fluorescence Spectroscopy 1997, 5, 43.

Johnson, J. M.; Ha, T.; Chu, S.; Boxer, S. G. Biophysical Journal 2002, 83, 3371.

160

(60)

(61)

(62)

(63)

(64)

Pauls, S. W.; Hedstrom, J. F.; Johnson, C. K. Chemical Physics 1998, 23 7, 205.

Wan, C.; Johnson, C. K. Chemical Physics 1994, 179, 513.

Wan, C.; Johnson, C. K. Journal of Chemical Physics 1994, 101, 10283.

Lipari, G.; Szabo, A. Biophysical Journal 1980, 30, 489.

Szabo, A. Journal of Chemical Physics 1984, 81, 150.

I61

CHAPTER 6

CONCLUSIONS

In this work, we have studied the fluorescence lifetime and anisotropy decay of
chromophores from the NBD family, both as free probe molecules and those tethered to
phospholipid molecules. These fluorescent behavior of these chromophores were
examined using one- and two-photon excitation as a means to determine the Cartesian
components of the rotational diffusion coefficient in both solvents and in model lipid
bilayers. The information gained from this work will provide an initial understanding of

the phase segregation within lipid bilayers.

Prior to incorporation into a bilayer, the fluorescence behavior of three substituted
NBD chromophores was studied in solvents. Chapter 2 showed that, despite different side
group attachments on the NBD chromophores, the dynamics of these molecules are
similar. The fluorescence lifetimes were dependent on the solvent polarity, while the
rotational diffusion coefficients found by both one- and two-photon excitation yielded a
ratio of Dz/Dx ~ 2 regardless of the identity of the solvent. Comparing the experimental
data with the Debye-Stokes-Einstein model it was found that the reorientation time was
much longer that the predicted time constant. We attributed this to a contribution from
dielectric friction, which, when accounted for the discrepancies between the predicted
and experimental time constants. Overall, the identity of the sidegroup on the
chromophore played little role in determining the rotational dynamics of the

chromophore in solution.

162

An NBD chromophore, NBD-hexanoic acid, was incorporated into bilayers
composed of phospholipids, cholesterol, and sphingomyelin. Chapter 3 reports on the
fluorescence behavior of the chromophore in vesicle systems. The fluorescence lifetime
values suggested that, regardless of vesicle composition, NBD-hexanoic acid was located
at the bilayer-aqueous interface. Similar results obtained for the anisotropy data, with the
free chromophore reorienting in the same manner in each vesicle system. These data
taken together suggest that in order to gain information about phase segregation and
overall organization of the bilayer, a chromophore tethered to a lipid molecule must be

used in these studies.

In Chapter 4, an NBD moiety tethered to a phospholipid molecule was used to
probe lipid bilayer heterogeneity. Again, both one- and two-photon excitation were
utilized to determine the fluorescence lifetime and anisotropy in lipid vesicles. Using the
tethered chromophore, double exponential lifetime decays were obtained, indicating the
chromophore was sensing two environments. The experimental anisotropy decay was
measured, and from that, the viscosity of the probe’s surroundings and the Cartesian
components of the rotational diffusion coefficient were calculated. Significant changes in
viscosity were measured upon the addition of both cholesterol and sphingomyelin to the
phospholipid. These changes in viscosity were attributed to changes in local organization
within the bilayer structure. The ratio Dy/Dx was used as a measure of bilayer
organization, and substantial decreases of Dy/Dx were the result of cholesterol and
sphingomyelin being added to the system. Furthermore, changes in the cone angle

showed increased organization within the bilayer upon addition of components. Taken

163

collectively, the data point to the NBD group residing near boundaries between

heterogeneous regions within the model lipid bilayers.

Vesicles comprised of lipids with a glycerol headgroup, as opposed to a choline
headgroup were used for studies outlined in chapter 5. NBD tethered to a
phosphoglycerol lipid was the chromophore incorporated into bilayers. The experimental
data show some similarities in chromophore behavior as compared to the phosphocholine
vesicles, yet the ability of the phosphoglycerol headgroups to participate in hydrogen
bonding provide an additional means of mediating bilayer organization in the proximity

of the lipid headgroup region of the bilayer.

Overall, this work shows the benefits of utilizing both one- and two-photon
excitation to examine the fluorescent behavior and rotational dynamics of a chromophore.
When incorporated into model bilayer systems, these experiments probe the organization
of lipid bilayers and show that even in simple systems, formation of phase segregation
occurs within the bilayer providing further evidence of heterogeneity of lipid bilayer

systems.

164

IIIIIIIIIIIIIIIIIIIIIIIIIIIIIII

1|111111lllllllllllllll11Hllllllllllllllllll