ufinflyhmmu

~
0 I "I 0
.‘abnxwz

MN”!

1%
an..."
n?
.664
1....
find»?
2h:
a...
v

3..
92...“...5
2!... .

.XI».
I .l...
~

{1.3.9.
.h4|| .Dn

$511.1»?
1. 3

«0.2134 arm , ‘

 

 

mats

This is to certify that the
dissertation entitled

UNDERSTANDING LIPID BLAYER DYNAMICS USING
FLUORESCENCE SPECTROSCOPY

presented by

 

MONIQUE MARIE LAPINSKI

LIBRARY
Michigan State
University

has been accepted towards fulfillment
of the requirements for the

 

 

 

I

PhD. degree in Chemistry

 

 

Major Professor’s Signature
7~/ 96%)?

Date

 

MSU is an affinnative-action, equal-opportunity employer

_A-._.—-.-u--.--u---—~— —

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.

 

DAIEDUE

DAIEDUE

DAIEDUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/07 p:/CIRC/DateDue.indd-p.1

 

UNDERSTANDING LIPID BILAYER DYNAMICS USING
FLUORESCENCE SPECTROSCOPY

By

Monique Marie Lapinski

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

UNDERSTANDING LIPID BILAYER DYNAMICS USING
FLUORESCENCE SPECTROSCOPY

By

Monique Marie Lapinski

Lipid vesicles are important because of their use as models for biological
membranes. We were particularly interested in using the rotational and translational
dynamics of chromophores sequestered within the bilayers comprising vesicles so that we
may understand more about the factors that govern interactions between bilayer
constituents. We find that two common methods for vesicle preparation yield the same
rotational and translational diffusion dynamics for chromophores imbedded in them
indicating that the molecular scale organization of bilayers is determined by interactions
between bilayer constituents, which themselves do not depend on the manner in which
the bilayer is formed. We report that small concentrations of certain bilayer species can
have substantial effects on the overall properties of the bilayer systems. In particular, it is
possible to have interactions between chromophores which perturb the system in such a
way that the bilayer itself is not being interrogated properly. This finding emphasizes the
necessity in using small concentrations of fluorescent probes when performing bilayer
studies. We were successful in using the reorientation dynamics of a nonpolar
chromophore to determine the phase transition temperature of a lipid comprising the
bilayers in vesicles. We found that the introduction of an unsaturated lipid into the same
system, in small concentrations, caused a substantial and measureable change in the

phase transition temperature of the system owing to the order was disrupted in the

system. We studied the effects that the radius of curvature of vesicles has on the
measured phase transition temperature of lipids in the bilayers and found that we could
detect the phase change in vesicles up to 800 nm in diameter. At sizes larger than this,
the data suggested that the chromophore changes location in the bilayer with the decrease
in the radius of curvature, underscoring the effect that the size of the vesicle has in

mediating the dynamics of bilayer constituents.

To Dr. Robert Stach,
the professor that inspired me to pursue chemistry in college

iv

ACKNOWLEDGEMENTS

There are several people that I need to thank that have helped me along the way
during my graduate career at Michigan State University. First and foremost, I must thank
my parents, who instilled in me, from a very young age, the importance and value of
education. My father once told me that “education is the fig thing, that once acquired,
can never be taken away.” Now, as my formal education is drawing to a close, I realize
how true that statement is, and also how truly blessed I am to have parents that value
education so clearly.

My husband, Steve, has been just as encouraging and equally supportive of each
of my endeavors. It must be a true challenge to be married to a scientist, especially when
the day’s experiments prove to not go as planned. The patience that he has shown for me
through the years is definitely a blessing, and I am lucky to have someone in my life that
is so devoted to me and my love for chemistry.

Dr. Robert Stach, my undergraduate mentor and advisor, is the person responsible
for developing my real love for biochemistry. His enthusiasm for teaching and his
encouragement of learning during my undergraduate education challenged me in every
way possible. It was at his insistence that I became an undergraduate chemistry tutor and
developed my own passion for teaching. And now, I will continue to follow in his
footsteps.

Dr. Gary Blanchard, my graduate advisor, provided guidance and support from

the first day that I joined his research group. His willingness to help and his overall

V

dedication to research made him an excellent advisor. Over the years, with true patience,
he never once doubted my abilities. He continued to challenge me on all levels and has
now provided me with the tools that I need to be successful.

Finally, I must thank my committee members, Dr. David Weliky, Dr. Greg Swain
and Dr. Gavin Reid for serving on my committee. In addition, I must also thank Dr.
Robert Ofoli and his research group in the Chemical Engineering and Material Science
department on campus, as they were instrumental in conducting the experiments using

supported lipid bilayers in Chapter 3.

vi

TABLE OF CONTENTS

List of Tables ........................................................................................ ix
List of Figures ........................................................................................ xi
List of Abbreviations .............................................................................. xv
Chapter 1 -- INTRODUCTION ....................................................................... 1

Literature Cited ................................................................................. 9

Chapter 2 — A COMPARISON OF LIPOSOMES FORMED BY SONICATION AND
EXTRUSION: ROTATIONAL AND TRANSLATIONAL DIFFUSION OF AN
IMBEDDED CHROMOPHORE

Introduction ................................................................................... l 1
Experimental .................................................................................. 14
Results and Discussion ...................................................................... 20
Conclusions .................................................................................... 33
Literature Cited ............................................................................... 34

Chapter 3 — UNDERSTANDING INTERMOLECULAR INTERACTIONS IN
UNILAMELLAR VESICLES COMPRISED OF MULTIPLE CONSTITUENTS

Introduction ................................................................................... 38
Experimental .................................................................................. 42
Results and Discussion ...................................................................... 46
Conclusions ................................................................................... 64
Literature Cited ............................................................................... 65

Chapter 4 — THE ROLE OF PHOSPHOLIPID HEADGROUPS IN MEDIATIN G
BILAYER ORGANIZATION. PERTURBATIONS INDUCED BY THE PRESENCE
OF A TETHERED CHROMOPHORE

Introduction ................................................................................... 69
Experimental .................................................................................. 72
Results and Discussion ...................................................................... 77
Conclusions .................................................................................... 91
Literature Cited ............................................................................... 93

vii

Chapter 5 — GAUGIN G THE EFFECT OF IMPURITIES ON LIPID BILAYER PHASE
TRANSITION TEMPERATURE

Introduction ................................................................................... 96
Experimental ................................................................................. 100
Results and Discussion ..................................................................... 104
Conclusions .................................................................................. 120
Literature Cited .............................................................................. 121

Chapter 6 — INTERROGATING THE ROLE OF LIPOSOME SIZE IN MEDIATING
THE DYNAMICS OF A CHROMOPHORE IN THE ACYL CHAIN REGION OF A

PHOSPHOLIPID BILAYER
Introduction ................................................................................. 125
Experimental ................................................................................ 128
Results and Discussion .................................................................... 134
Conclusions ................................................................................. 148
Literature Cited ............................................................................. 149
Chapter 7 — CONCLUSIONS AND FUTURE WORK .................................... 153

viii

LIST OF TABLES

Table 2.1 - Dynamic Light Scattering (DLS) results for the extruded and
sonicated liposomes. Uncertainties in the data are estimated to be :h 5% of
the average values ............................................................................... 23

Table 2.2 - Results from Time Correlated Single Photon Counting (TCSPC)
studies using extruded and sonicated liposomes with DOPC and NBD-PC ............. 27

Table 2.3 — Results from Time Correlated Single Photon Counting (TCSPC)
studies using extruded and sonicated liposomes with DOPC, cholesterol and
NBD-PC .......................................................................................... 27

Table 2.4 — Results from Fluorescence Recovery After Pattern Photobleaching
(FRAPP) using supported bilayer membranes from extruded and sonicated
vesicles. The samples contained DOPC and NBD-PC .................................... 32

Table 2.5 — Results from Fluorescence Recovery After Pattern Photobleaching
(FRAPP) using supported bilayer membranes from extruded and sonicated
vesicles. The samples contained DOPC, cholesterol and NBD-PC ...................... 32

Table 3.1 — Calculated viscosity dependence of the rotational diffusion

coefficient and anisotropy decay time constant for NBD-cholesterol. The

diffusion coefficients and reorientation times were calculated based on a
hydrodynamic volume of 528A3 for the NBD-cholesterol molecule, f =1 ,

S=1 and T=298 K ................................................................................ 63

Table 5.1 — Viscosities of DMPC unilamellar vesicles as a fimction of 14:1 PC
concentration ................................................................................... 1 1 1

Table 5.2 — Calculated translational diffusion coefficients for perylene in DMPC
unilamellar vesicles as a function of 14:1 PC concentration. Viscosities were
determined from 10R with use of Eq. 5.2, and DT values were calculated

according to Eq. 5.3 . .. ................................................................................................. 118

ix

Table 6.1 —- Comparison of extrusion membrane pore diameter to the

experimental diameter of the extruded vesicles. Experimental diameters

were determined by DLS measurements. Uncertainties are $10 for three

individual determinations ...................................................................... 132

Table 6.2 — Calculated viscosity at 19°C for the vesicles with varying
diameters ........................................................................................ 145

LIST OF FIGURES

Figure 1.1 — Schematic of the Time Correlated Single Photon Counting (TCSPC)
System .............................................................................................. 5

Figure 2.1 — Solution phase absorption and emission spectra of NBD-PC and
DOPC vesicles. Spectra have been normalized for clarity of presentation ............... 16

Figure 2.2 — TEM images of vesicles produced by extrusion and sonication.
(a) Extruded vesicles without cholesterol. (b) Sonicated vesicles without
cholesterol. The scale bar is 500 nm in both images ....................................... 21

Figure 2.3 — Representative plot of a F RAPP recovery curve for a DOPC/NBD-PC
supported bilayer membrane ..................................................................... 30

Figure 3.1 — Transmission electron microscopy (TEM) image of a unilamellar

vesicle. Top pane: Wide area view of aggregated vesicles. Scale bar = 1 pm.

Bottom pane: Images of individual vesicles under higher magnification

conditions. Scale bar = 200 nm ................................................................................... 44

Figure 3.2 — Top: Absorption and emission spectra of NBD-cholesterol and the
absorption spectrum of Rhodamine-PE in Tris® buffer. The spectra have been
corrected for the extinction coefficients of the two compounds and the

fluorescence quantum yield of NBD-cholesterol is assumed to be 0.5 for the

sake of presentation. Bottom: Structures of NBD-cholesterol and
Rhodamme-PE 47

Figure 3.3 — Fluorescence lifetime decay of NBD-cholesterol in a unilamellar
vesicle containing 33 mol% cholesterol, 11 mol% sphingomyelin and 56 mol%
DMPC. The recovered time constant of this decay is 3150 i 30 ps ...................... 50

xi

Figure 3.4 — (a) Fluorescence lifetime(s) of NBD-cholesterol as a function

of amount of Rhodamine-PE added to unilamellar vesicles. The appearance

of two decay components is consistent with excitation transport in a

heterogeneous system. (b) The fractional contribution of the fast fluorescent

lifetime component of NBD-cholesterol as a function of Rhodamine-PE

concentration .................................................................................... 55

Figure 3.5 — (a) Polarized emission transients for NBD-cholesterol in a

unilamellar vesicle containing 33 mol% cholesterol, 11 mol% sphingomyelin

and 56 mol% DMPC. The acquisition polarization of the transients is indicated

in the figure relative to the vertical excitation pulse polarization. (b) Decay of

the induced orientational anisotropy function, R(t), produced from the data

shown in panel (a). The time constant of this anisotropy decay is 1362 i 61 ps ...... 57

Figure 3.6 — Dependence of Rhodamine-PE anisotropy decay time on
Rhodamine-PE concentration ................................................................. 58

Figure 3.7 — Dependence of NED-cholesterol anisotropy decay time on
Rhodamine-PE concentration ................................................................. 60

Figure 4.1 — Left: Structures of the phospholipid DMPC and the chromophore
containing phospholipid Rhodamine-PE. Right: Solution phase absorption

and emission spectra of Rhodamine-PE in ethanol. Spectra have been

normalized for clarity of presentation. . . . . . . . . . . . .. ........................................................ 73

Figure 4.2 — TEM images of vesicles produced by extrusion and used in this

work. Top pane: Image of an isolated unilamellar vesicle of ca. 150 nm in

diameter. Bottom pane: Image of multiple vesicles. The average diameter

of these vesicles was determined to be 84 :l: 22 nm based on measurements

made from this and other TEM images ...................................................... 74

Figure 4.3 — Steady state emission spectra of Rhodamine-PE in DMPC vesicles.

The chromophore concentrations are 0.0025 mol%, 0.025 mol% and 0.25 mol%,

as indicated. Temperature-dependence of the emission band profiles is also

indicated .......................................................................................... 79

xii

Figure 4.4 — (a) Experimental polarized fluorescence transients for
Rhodamine-PE in 100 nm diameter DMPC vesicles. (b) Induced orientational
anisotropy function, R(t), generated from the data shown in panel (a) ................... 84

Figure 4.5 - (a) Time constants (Eq. 4.2) extracted from anisotropy data for
Rhodamine-PE in DMPC vesicles as a function of chromophore concentration

and system temperature. (b) Confining cone semi-angle measurements from

the experimental data (Eq. 4.4) ................................................................ 86

Figure 4.6 — Dependence of Dw on chromophore concentration and system
temperature. The quantity Dw is calculated from Eq. 4.3 using the
experimentally derived quantities 10R and 00 87

Figure 5.1 — Transmission electron microscope images of a unilamellar vesicle
used in this work. The scale bar for both images is shown in the lower right
corner ........................................................................................... 102

Figure 5.2 — Structures ofDMPC, 14:1 PC and perylene................................... 105

Figure 5.3 — Reorientation time of perylene in unilamellar vesicles comprised

of DMPC as a function of temperature. There is a discontinuous change in rOR

at the gel-to-fluid phase transition temperature. Error bars are :lzlo for at least

four individual determinations at each temperature ...................................... 106

Figure 5.4 — Reorientation time of perylene in DMPC unilamellar vesicles

containing (a) 0.3 mol% 14:1 PC and (b) 0.7 mol% 14:1 PC as a function of
temperature. Error bars are :1: lo for at least four individual determinations at

each temperature ............................................................................... 1 14

Figure 5.5 — Reorientation time of perylene in DMPC unilamellar vesicles

containing (a) 1.0 mol% 14:1 PC and (b) 1.5 mol% 14:1 PC as a function of
temperature. Error bars are i 10 for at least four individual determinations at

each temperature ............................................................................... 1 15

Figure 5.6 — Variation of measured T.m in DMPC unilamellar vesicles as a
function of 14:1 PC concentration ........................................................... 116

xiii

Figure 6.1 — Structures of the phospholipid, DMPC and the chromophore, perylene.
Bottom right: Steady state absorption and emission spectra of perylene in DMPC
vesicles. Spectra have been normalized for clarity of presentation .................... 130

Figure 6.2 — Plot showing the linear relationship between the nominal pore
diameter in the extruder membrane and the measured diameter of
DMPC/perylene vesicles from DLS ......................................................... 133

Figure 6.3 — Anisotropy data for perylene in 100 nm diameter DMPC vesicles

at 25°C, showing a decay that is best fit by two exponential components. The

fitted line is indicated as the solid line and the residuals of the fit are shown

centered around the zero line. The fitted function is

R(t) = Alexp(-t/tl) + Azexp(-t/'cz), where A1 = 0.07 d: 0.01, r] = 271 :l: 6 ps,

A2 = 0.20 i 0.01 and 1:; = 2238 d: 56 ps ..................................................... 137

Figure 6.4 - Reorientation times obtained for 100 nm diameter vesicles

containing DMPC and perylene. The data from the earlier work were single

time constants, while the data from the current work represent weighted

averages of two time constants ............................................................... 138

Figure 6.5 — Reorientation times obtained for perylene in DMPC vesicles as a
function of temperature and vesicle diameter. (a) 100-800 nm vesicle
diameters. (b) 1000-5000 nm vesicle diameters ......................................... 139

Figure 6.6 — Cartesian components of the diffusion coefficient, Dx and D2,

obtained for perylene in DMPC vesicles as a function of temperature and vesicle
diameter. (a) 100-800 nm vesicle diameters. (b) 1000-5000 nm vesicle

diameters ........................................................................................ 141

Figure 6.7 — Reorientation times obtained for perylene in DMPC vesicles as a
filnction of vesicle diameter at 19°C ......................................................... 142

Figure 6.8 — Cartesian components of the diffusion coefficient, Dx and D2,

obtained for perylene in DMPC vesicles as a function of vesicle diameter
at 19°C ............................................................................................ 143

xiv

LIST OF ABBREVIATIONS

TCSPC ............................................. Time correlated single photon counting
CFD .......................................................... Constant fraction discriminator
TAC ............................................................. Time to amplitude converter
MCP ........................................................................ Microchannel plate
PMT ...................................................................... Photomultiplier tube
MCA ................................................................... Multichannel analyzer
TEM ...................................................... Transmission electron microscopy
DLS .................................................................. Dynamic light scattering
FRAPP .............................. Fluorescence recovery after pattern photobleaching
sBLM ........................................................... Supported bilayer membrane
SUV ................................................................. Small unilamellar vesicle
DOPC .......................................... 1 ,2-dioleoyl-sn-glycero-3-phosphocholine
NBD-PC ................................. 1 -oleoyl-2-[1 2-[(7-nitro-2- l ,3-benzoxadiaz 1-4-

yl)amino]dodecanoyl]-sn-glycero-3 -phosphocholine

HEPES ........................... N-2-Hydroxyethylpiperazine-N’-2-ethanesulfonic acid
NBD ...................................................... 7-nitrobenz-2-oxa-1,3-diazol-4-yl
SPM ............................................................................. Sphingomyelin

XV

DMPC ................................................ 1 ,2-dimyristoyl-sn-phosphatidylcholine

NBD-cholesterol .............................. 25-(N-(7-nitrobenz-2-oxa- 1 ,3-diazol-4-yl-
methyl)amino)-27-norcholesterol

Rhodamine-PE .................. l ,2-Dimyri stoyl-sn-Glycero-3-Phosphoethanolamine-
N- (Lissamine Rhodamine B Sulfonyl Ammonium salt)

DMPE .............................. 1 ,2-Dimyristoyl-sn-Glycero-3-Phosphoethanolamine
NMR ............................................................ Nuclear magnetic resonance
Tris® ........................................................ Tris hydroxymethylaminoethane
14: 1 PC ................................. 1,2-dimyristoleoyl-sn-glycero-3-phosphocholine
HEPES® ......................... N-2-Hydroxyethylpiperazine-N’-2-ethanesulfonic acid

xvi

Chapter 1

INTRODUCTION

Lipids and lipid bilayers have recently attracted a great deal of attention from
researchers because of their use in a variety of scientific, medical and pharmacological
applications. The lipid bilayers that are used are often in the form of liposomes; small,
spherical structures that can form when a lipid comes into contact with an aqueous phase.
The driving force for this self assembly process is the tendency of the hydrophobic
portion of the lipid molecule to minimize its contact with water. Therefore, the
hydrophilic polar heads of the lipid molecule will face the aqueous surroundings and the
hydrophobic tail portions of the lipid will constitute the interior of the bilayers. If this
structure closes on itself, a lipid vesicle (liposome) is formed. The utility of this vesicular
structure lies in the ability to trap various constituents within the small confines of the
bilayer. In doing so, researchers are able to further study the physical and molecular
factors that affect lipids and lipid bilayers.

The first researchers to study lipids found in cell membranes were Evert Gorter
and his assistant, F. Grendel. In their classic experiment with red blood cells,1 the
researchers were able to demonstrate that lipid molecules could from a double layer, as
well as a monolayer. In addition, by using a modified device similar to a Langmuir
trough, they were able to show that the surface area of the lipids extracted from the red
blood cells was approximately twice the area of the cells themselves. Gorter and Grendel

repeated their initial studies with red blood cells using erythrocytes from several animals

and humans and concluded that “[red blood cells] are covered by a layer of fatty
substances that is two molecules thickl.”

Later, researchers discovered that the work of Gorter and Grendel was limited by
the measurement techniques available at the time. They had not extracted the lipids from
the red blood cells quantitatively, and had also underestimated the surface area of the
erythrocytes. These two errors essentially cancelled and their conclusions were basically
correct.2 This finding ultimately led to the idea that cell membranes may, in fact, consist
of lipid bilayers.

In 1957, J .D. Robertson proposed one version of the membrane model, which was
based primarily on electron microscopy studies that he had conducted.3 Under the high
magnification of the transmission electron microscope (TEM), membranes displayed a
characteristic outer shaded region that was darker in color than the inner shaded region.
Robertson proposed that the outer layers were protein layers and the inner region was the
lipid bilayer. However, Robertson’s description was not entirely correct.

Once a more detailed understanding of proteins and their interactions with
membranes was developed, the current model for the cell membrane was proposed. This
model, known as the fluid mosaic model, was proposed by Singer and Nicolson in 1972,4
and continues as the membrane model in use today. The fluid mosaic model includes the
basic lipid structure first proposed by Gorter and Grendel, but also describes globular
proteins (both integral and peripheral) that are able to reside within the lipid bilayer,

rather than just on the membrane surface, as Robertson had once proposed.

Using modern molecular dynamics methods and a host of spectroscopic
techniques, much work is in progress to understand more about membranes and lipid
bilayers. It is known that the lipid bilayer that forms cell membranes is a two-
dimensional liquid,5 but the organization of this liquid continues to be an ongoing study
among researchers. There have been repeated attempts to introduce lateral
heterogeneities and lipid microdomains into systems to order to provide further insight
into the structures and dynamics of the lipids that govern membrane bilayers and the
ordering of these systems.("8 Because a large fraction of all biological membranes are
believed to exist in a fluid state under physiological conditions, it is of particular
importance to establish the factors and mechanisms that govern the ordering of lipid
bilayers.9

A drastic change in the order of a bilayer system occurs at the so-called main
phase transition temperature (Tm). At this temperature, the hydrocarbon tails undergo a
change from being predominantly trans in configuration to a structure that is dominated
by a significant number of trans-gauche conformers.lo There are several factors that can
affect phase changes in a lipid system. These factors include, but are not limited to
pressure, temperature, the amount of hydration and the basic structural properties of the
lipid(s), including the length of the hydrocarbon chain and the composition of the
headgroup. We were interested in studying various regions of lipid bilayers to
understand more about the factors that affect the ordering and measured transition

temperatures of various systems.

For most of our studies, we prepared lipid vesicles using the extrusion method,l 1'
'6 where a lipid solution is forced via syringe through polycarbonate membranes with
defined pore sizes. The advantage in using the extrusion method for vesicle preparation
is that we are able to control the sizes of the vesicles and can be assured the samples are
unilamellar in structure. With the lipid vesicles, we then chose to explore some factors
that would affect the phase transition temperature including the radius of curvature of the
vesicles and the addition of an unsaturated lipid into the systems.

For these studies, we employed Time Correlated Single Photon Counting
(TCSPC) as a means of determining the time-domain fluorescent properties of
chromophores either embedded within or tethered to the lipid bilayers. In doing so, we
were able to gain insight into the lipid bilayer environment in which the fluorescent probe
was residing, including the polarity and viscosity of the surrounding area. In addition, we
were able to use the data that we obtained from TCSPC to evaluate the effects that the
addition of an unsaturated lipid and a fluorescent probe has on the measured transition
temperature of the system. We were also able to determine the effects that the size of the
size of the liposome has on the transition temperature.

For the time-resolved fluorescence spectroscopy measurements that we include in this
dissertation, the light source was a continuous wave (CW) mode-locked Nd:YAG laser
(Coherent Antares 76-s) that produced 100 ps pulses at a 76 MHz repetition rate. A
cavity-dumped dye laser (Coherent 702-2) was operated with either Stilbene 420 dye
(Exciton) and excited by the third harmonic of the source laser, or was operated with

Pyromethene 567 dye (Exciton) that was excited by the second harmonic of the source

laser. Sample fluorescence was detected using a microchannel plate — photomultiplier
tube (MCP-PMT, Hamamatsu R3 809U), and the electronics used to temporally resolve
the fluorescence transients were a constant fraction discriminator (CFD, Tennelec 454)
and a time-to-amplitude converter/biased amplifier (TAC, Tennelec 864), (Figure 1.1).
The fluorescence lifetime data were collected at 547° with respect to the vertical
excitation polarization, while the reorientation data were collected at polarizations

parallel (0°) and perpendicular (90°) to the vertically-polarized incident light.

    
 

pulse
diagnostics

  
  

2a), 3a)

   

cavity dumped dye laser

  
 

1‘”
fiber optic
switchable

  
 
 
  
  
 

420 nm — 470 nm (30))
550 nm - 590 nm (20))

  

polarization
control
scrambler
I

.

diode CFD delay

    
 
 

v.1."-

 
 

/

x’ I,
G-T monochromator MCP PMT

prism

Figure 1.1. Schematic of the Time Correlated Single Photon Counting
(TCSPC) system.

One of the parameters that we can determine using a chosen chromophore and the
lipid bilayer structures is the fluorescence lifetime of the probe. The fluorescence
lifetime can be described as the average amount of time that the chromophore molecule
remains in the excited state before the emission of a photon takes place. The data
contained in the fluorescence lifetime measurements therefore provide valuable insight
into the environment in which the chromophore resides. When the exponential lifetime
decay is analyzed, the resulting number of fluorescence lifetimes corresponds to the
number of different environments that the chromophore senses. A single exponential
lifetime decay indicates that the chromophore is sensing a single environment, and the
lifetime value (measured in picoseconds) can, for certain chromophores, provide a gauge
for the polarity of that environment. Multiple exponential lifetime decays are possible,
especially for tethered chromophores that exhibit the ability to “loop back” into an
environment with differing polarity.”20

Another parameter that we can measure using TCSPC is the rotational
reorientation time(s) of the chromophore. The reorientation time, usually measured in
nanoseconds or picoseconds, is described as the amount of time that it takes for a
chromophore to reorient itself to a random distribution following excitation. The
reorientation time is derived by incorporating the measured intensities collected parallel

(I “(t)) and perpendicular (I.L(t)) to the polarized incident light to form the induced

anisotropy function shown in Eqn. 1.1.

Ill(t)-I.L(t)

1" (t)+ 21i (t)

 

R(t): (1.1)

The generated exponential decay curve is then fitted to determine the
reorientation time(s) of the chromophore. The number of exponential decays contained
in R(t) is determined by the effective rotor shape of the probe, and in principle, can
contain up to five exponential decays. However, it is often difficult to resolve more than
two of these decays simply based on symmetry conditions and the limitations in the
signal-to-noise of the measurements.

The recovered reorientation time(s) can be used in the modified Debye-Stokes-

21-23

Einstein Equation shown in Eqn. 1.2, to determine the viscosity of the environment in

which the chromophore resides.

T = W
0" kBTS

(1.2)
In this equation, 11 represents the viscosity, V is the hydrodynamic voltune of the
chromophore, k3 is the Boltzmann constant and T is the temperature. The terms, f and S,
are added to this equation to account for deviations from ideal behavior. The fi'ictional
coefficient, f, is equal to unity in the “stick” limit, which applies to very strong solute-
solvent interactions. For weaker interactions, the “slip limit” is applied.23’24 The shape
factor term, S, is derived from Perrin’s equations” and added to the Debye-Stokes-
Einstein equation to account for deviations of the chromophores from perfectly spherical
shapes. This term can vary from close to zero to 1, where unity represents a spherical
reorienting chromophore.

The utility of these studies resides in the fact that we were able to use several

chromophores with varying structural properties to probe different areas of the lipid

vesicles. In Chapter 2, we used a chromophore tethered to a lipid, 1-oleoyl-2-[12-[(7-
7

nitro-2- 1 ,3-benzoxadiazol-4-yl)amino]dodecanoyl]-sn-glycero-3-phosphocholine (1 8: 1-
12:0 NBD-PC), to determine whether two widely used methods of lipid vesicle
preparation yield the same results, both in the rotational and translational dynamics of the
chromophore. In Chapter 3, two tethered chromophores that are localized in different
regions of the vesicles, NBD-cholesterol and Rhodamine-PE, to determine how
compositional changes in one region give rise to structural changes in other bilayer
regions. Chapter 4 addresses the use of a tethered polar chromophore, Rhodamine-PE,
and the effects that the concentration of this probe has on the recorded rotational
dynamics of the chromophore in the polar region of the lipid bilayer systems. In Chapter
5, a nonpolar chromophore, perylene, is used in varying lipid systems to determine
whether the measured reorientation time of perylene can be used as a suitable means for
detecting the phase transition temperature of lipid bilayer systems. Also, in this chapter,
the effect that an unsaturated lipid has on the transition temperature of the system is
evaluated. Chapter 6 uses the same probe and lipid system as the one tested in Chapter 3,
but the diameter of the vesicles is varied from 100 nm to 5 pm to determine how the
radius of curvature affects the measured transition temperature of the systems. Chapter 7

contains some final conclusions and possible representative future studies.

Literature Cited

(1) Gorter, E.; Grendel, F. Journal of Experimental Medicine 1925, 41, 439.

(2) Sadava, D. E. Cell Biology: Organelle Structure and Function; Jones and
Bartlett: Boston, 1993.

(3) Robertson, J. D. J Biophys Biochem Cytol FIELD Full Journal Title: The
Journal of biophysical and biochemical cytology 1957, 3, 1043.

(4) Singer, S. J .; Nicolson, G. L. Science FIELD Full Journal Title'Science
(New York, NY.) 1972, 175, 720.

(5) Simons, K.; Vaz, W. L. C. Annual Review of Biophysics and Biomolecular
Structure 2004, 33, 269.

(6) Glaser, M. Current Opinion in Structural Biology 1993, 3, 475.
(7) Jain, M. K.; White, H. B., III. Advances in Lipid Research 1977, 15, l.

(8) Vaz, W. L.; Almeida, F. F. Current Opinion in Structural Biology 1993, 3,
482.

(9) Metso, A. J .; Jutila, A.; Mattila, J .-P.; Holopainen, J. M.; Kinnunen, P. K.
J. Journal of Physical Chemistry B 2003, 107, 1251.

(10) Pabst, G.; Amenitsch, H.; Kharakoz, D. P.; Laggner, P.; Rappolt, M.
Physical Review E: Statistical, Nonlinear, and Soft Matter Physics 2004, 70, 021908/ 1.

(11) Driessen, A. J .; van den Hooven, H. W.; Kuiper, W.; van de Kamp, M.;
Sahl, H. G.; Konings, R. N.; Konings, W. N. Biochemistry 1995, 34, 1606.

(12) Hunter, D. G.; Frisken, B. J. Biophysical Journal 1998, 74, 2996.

(13) MacDonald, R. C.; MacDonald, R. 1.; Menco, B. P.; Takeshita, K.;
Subbarao, N. K.; Hu, L. R. Biochimica et Biophysica Acta 1991, 1061, 297.

(14) Mayer, L. D.; Hope, M. J .; Cullis, P. R. Biochimica et Biophysica Acta
1986, 858, 161.

(15) Subbarao, N. K.; MacDonald, R. 1.; Takeshita, K.; MacDonald, R. C.
Biochimica et Biophysica Acta 1991, 1063, 147.

(16) Unger, E. C. ; MacDougall, P.; Cullis, P.; Tilcock, C. Magnetic Resonance
Imaging 1989, 7, 417.

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

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

(19) F eng, Z. V.; Spurlin, T. A.; Gewirth, A. A. Biophysical Journal 2005, 88,
2154.

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

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

(22) Debye, P. Polar Molecules. In Chemical Catalog Co. New York, 1929.

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

(24) Youngren, G. K.; Acrivos, A. Journal of Chemical Physics 1975, 63,
3846.

10

Chapter 2

A COMPARISON OF LIPOSOMES FORMED BY SONICATION AND
EXTRUSION: ROTATIONAL AND TRANSLATIONAL DIFFUSION
OF AN IMBEDDED CHROMOPHORE

Introduction

Lipid bilayers are fundamental structures for all cellular systems, and a great deal
of research effort has been expended on understanding how to create and interrogate
native and model bilayer structures. Phospholipids are known to form liposome
structures in an aqueous environment, and these structures can be created by any of a
number of methods."4 The resulting, nominally spherical structure is comprised of a
phospholipid bilayer surrounding an aqueous core, with the size of the vesicle ranging
from tens of nm to um, depending on the manner in which the liposomes are formed.
Two common protocols used in the preparation of liposomes are extrusions'lo and bath
sonication,ll with each method yielding liposome structures with different average sizes
and size distributions.

Extrusion is currently one of the most common methods for controlling vesicle
size and for producing unilamellar liposomes.5'10 With this method, a lipid suspension is
forced through a polycarbonate membrane with a well defined pore size to produce
vesicles with a characteristic diameter near the pore size of the membrane used in its
preparation. This technique avoids problems associated with the removal of organic
solvents or detergents from the final preparations used in other protocols, and the

technique can be applied to a wide variety of lipid species and mixtures.12 Perhaps of

11

most significance is that the resulting vesicle average size and size distribution are
reproducible from batch-to-batch due to the physical processing method.

Sonication is also used widely for the preparation of liposomes. With this
method, lipid suspensions are mixed using acoustic energy from either a bath or a probe
tip sonicator. The induced pressure breaks up the larger, multilamellar vesicles present in
the sample and forms smaller vesicles that may be either unilamellar or multilamellar in
composition.11 The time over which lipid solutions are sonicated affects the size of the
vesicles, with the smallest radius for phospholipid vesicles being in the range of 10.25 d:
0.55 nm, independent of the phospholipid hydrocarbon chain length.13 The primary
advantage of sonication over extrusion is that sonication is less time-consuming.
However, the resulting mean liposome batch-to-batch diameter and size distribution are
not as reproducible as those made by extrusion.

The goal of this study was to evaluate the behavior of small unilamellar vesicles
(SUVs) prepared by bath sonication and by extrusion, to determine if the method of
preparation has any influence on selected bilayer properties. This is an important issue
because it bears on whether or not experimental data acquired for bilayer lipid
membranes formed by these two different methodologies can be compared directly. We
are interested in the mean diameter and size distribution of liposomes formed by the two
techniques, and examine this issue using dynamic light scattering (DLS) and transmission
electron microscopy (TEM). We are also interested in the molecular scale and
mesoscopic properties of the lipid bilayers formed by the two methods. We address this

issue through the measurement of the rotational motion of the chromophore incorporated

12

in the liposomes using time correlated single photon counting (TCSPC), and its
translational diffusion in planar supported bilayer lipid membranes (sBLMs) using
fluorescence recovery after pattern photobleaching (F RAPP). We find that, while the
physical morphology (mean diameter and size distribution) of the liposomes formed by
the two methods is outwardly different, the molecular scale diffusional behavior of
species within the bilayers is essentially the same. The addition of cholesterol to the
liposomes alters the observed dynamics of the tethered chromophore, with similar

dynamical results for liposomes formed using both methods.

13

Experimental

Vesicle preparation by extrusion. Phospholipids 1,2-dioleoyl-sn-glycero-3-
phospho-choline (DOPC), 1 -oleoyl-2-[12-[(7-nitro-2-1 ,3-benzoxadiazol-4-
yl)amino]dodecanoyl]-sn-glycero-3-phosphocholine (18: 1-12:0 NBD-PC) and cholesterol
were purchased from Avanti Polar Lipids Inc. (Alabaster, AL) and used as received. For
each sample, 1 mM lipid solutions containing DOPC/NBD-PC (98:2 mol/mol) or
DOPC/cholesterol/NBD-PC (68:30:2 mol/mol) were prepared. The chloroform solvent
was evaporated from the lipid mixtures and the dry film was hydrated using HEPES®
buffer (Sigma-Aldrich, St. Louis, MO) containing NaCl (Fisher Scientific, Fair Lawn,
NJ). The goal was to obtain a final lipid concentration of 1 mg/mL. The buffer (10 mM
HEPES®, 150 mM NaCl, pH 7.4) was prepared using purified water from a Milli-Q Plus
water purification system (Millipore, Bedford, MA). The mixtures were processed five
times through a freeze-thaw-vortex cycle to ensure complete mixing of the sample
constituents. Each cycle consisted of immersion in liquid nitrogen for five minutes,
followed by immersion in a 60°C water bath for five minutes, then vortexing the thawed
sample for ca. two minutes. After the freeze-thaw-vortex cycling, the solutions were
extruded through two 400 nm pore diameter polycarbonate membrane filters using a
mini-extruder (Avanti Polar Lipids Inc., Alabaster AL). The pre-extruded vesicle
suspension was then extruded eleven times through two polycarbonate membranes
(Avanti Polar Lipids Inc., Alabaster, AL) with a nominal pore diameter of 100 nm. All

extrusions were performed at room temperature.

14

Vesicle preparation by sonication. Lipid mixtures with the same compositions as
those used for the extrusion experiments were used to form liposomes by sonication. The
chloroform-solvated lipid mixtures were dried under nitrogen and placed under vacuum
at -45°C for 3 hours. The dried lipid mixtures were then hydrated in buffer (10 mM
HEPES®, 150 mM NaCl, pH 7.4), and sonicated for 20 minutes using a bath ultrasonic
cleaner (Branson 1510, Branson Ultrasonic Corporation, Danbury, CT). Fresh liposome
solutions were prepared prior to the beginning of each experiment.

Transmission Electron Microscopy (T Ell/0 Imaging. A 10 uL aliquot of each
sample was fixed on a Formvar® nickel-coated grid using a solution containing 2% uranyl
acetate stain in water. Images were acquired using a transmission electron microscope
(TEM, JEOL 100CX) operated at an accelerating voltage of 100 kV.

Dynamic Light Scattering (DLS). DLS was performed with a Protein Solutions
DynaPro-MS/X system (Wyatt Technology Corporation, Santa Barbara, CA) on
DOPC/NBD-PC liposome solutions, with and without cholesterol, to determine mean
vesicle sizes and size distributions. Liposome solutions were diluted by a factor of ten
with buffer and placed in a polyethylene cuvette for measurement.

Steady state spectroscopy. Steady state excitation and emission spectra (Figure
2.1) were acquired for vesicle samples for the purpose of characterizing the NBD
chromophore band positions. We used an emission spectrometer (Spex F luorolog 3,
Edison, NJ) for all measurements, set to a spectral bandpass of 3 nm for both excitation

and emission monochromators.

15

1.0 -
.
0.8 -
>3
3‘:
m ‘
{:1
flag: 0.6 a
"U
0)
.91
g 0.4 4
g .
0.2 A
0.0~

absorbance

 

 

emission
fl ' l ' l ' I ‘ I
400 450 500 550 600

wavelength (nm)

Figure 2.1. Solution phase absorption and emission spectra of NBD-PC
and DOPC vesicles. Spectra have been normalized for clarity of
presentation.

16

Time-resolved fluorescence measurements. Time-domain polarized fluorescence
intensity decays were acquired using a time-correlated single-photon counting (TCSPC)

”’15 and we present only a

system. This system has been described in detail elsewhere,
brief recap of the essential features here. The light source 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 third harmonic of the Nd:YAG laser output is used to excite a cavity-
dumped dye laser (Coherent 702-2) operated with Stilbene 420 dye (Exciton) at 460 nm.
The average output power of this laser is approximately 25 mW, with 5 ps pulses at a 4
MHz repetition rate. The sample fluorescence emission is measured with a microchannel
plate photomultiplier tube (MCP-PMT, Hamamatsu R3809U-51). The electronics used
to temporally resolve the fluorescence transients included a constant fraction
discriminator (CFD, Tennelec 454) and a time-to-arnplitude converter/biased amplifier
(TAC, Tennelec 864). The collection wavelength (545 nm, 10 nm detection bandwidth)
and polarization were computer controlled using LabVIEW® 7.1 code. The fluorescence
lifetime data were collected at 547° with respect to the vertical excitation polarization,
while the reorientation data were collected at polarizations parallel (0°) and perpendicular
(90°) to the vertically-polarized incident light.

Fluorescence Recovery After Pattern Photobleaching (FRAPP). FRAPP was
used to measure lateral diffusion coefficients of fluorescently-tagged lipids in supported
bilayer lipid membranes (sBLMs) on fused silica microscope slides. The experimental

setup consists of a double syringe pump system (Harvard Apparatus, Model 551382)

used to simultaneously infuse and withdraw solutions from a custom made 1 mL flow

17

cell, an inverted microscope (Zeiss Axiovert 135M, Carl Zeiss, Thomwood, NY) with a
32X objective lens (Carl Zeiss, Thomwood, NY), an argon ion laser (95 Lexel Laser,
Fremont, CA), a side-on photomultiplier tube (Hamamatsu R4632, Bridgewater, NJ), a
photon counter (SR400 Stanford Research Systems, Sunnyvale, CA), and a fast
preamplifier (SR445 Stanford Research Systems, Sunnyvale, CA). The 488 nm laser line
was directed through a 5X beam expander (Edmund Optics, Inc., Barrington, NJ), a
system of optical flats16 used to toggle between a low intensity (20uW) monitoring beam
for continuous fluorescence detection and a high intensity (500mW) beam for
photobleaching of fluorophores. A Ronchi ruling (50 lines per inch, Edmund Optics,
Inc., Barrington, NJ) was placed in the back image plane of the microscope to create a
fringe pattern on the silica slide. The beam was then directed through a filter cube (Ex:
450-490/DM: 510/ Em: 515-565). An aperture placed in the image plane in front of the
PMT was used to restrict the observation area. Thus, the illuminated area was
approximately 200 um, while the observed area was 75 pm. This was done to prevent
unbleached fluorophores from outside of the patterned area from reaching the observation
zone during fluorescence recovery measurements. Stripe periodicity in the sample plane
was 25 um.

Deposition of Supported Lipid Bilayers (sBLMs). Fused silica microscope slides
(75x25x1 mm) were purchased from Technical Glass Products, Inc. (Paineville, OH).
The substrates were cleaned by bath sonication (Branson 1510, Branson Ultrasonic
Corporation, Danbury, CT) in detergent solution for 20 minutes, rinsed with DI water,

baked at 160°C for 4 hours and plasma treated (Harrick Plasma, Ithaca, NY) with oxygen

18

under vacuum (200 mTorr) for 10 minutes immediately before bilayer deposition.
Supported bilayer lipid membranes (sBLMs) were deposited by vesicle fusion in a
custom made flow cell. The flow cell was initially washed with buffer, followed by one
hour incubation with liposome solution and a final buffer wash to remove any

unadsorbed liposomes. All experiments were performed at room temperature.

19

Results and Discussion

The purpose of this work was to compare two methods of liposome preparation,
extrusion and bath sonication, to determine if the method of preparation affects the
physical and dynamical properties of the resulting liposomes. DOPC was chosen as the
primary lipid in our studies because of its low transition temperature (-20°C),17 ensuring
that the resulting lipid bilayers are fluid at room temperature. The specific choice of
chromophore (NBD-PC, 18:1 — 12:0) was made to ensure structural similarity to the
phospholipid to which it is tethered (DOPC, 18: 1). In addition, there is a significant body

”3'20 making it an attractive

of knowledge extant on the prOperties of this chromophore,
choice for probing our structures.

We were primarily concerned with three issues in this work: the physical
attributes (mean diameter and size distribution) of liposomes formed by the two methods,
the molecular scale environment of the lipids comprising the liposomes, and the fluidity
of bilayer structures resulting from the two liposome-forming methods. We are also
concerned with how incorporating cholesterol in the liposome affects each of these
issues. We expected that there would be a marked change in the organization of the
liposomes and bilayer lipid membranes on a molecular scale with the addition of this
component. We consider each of these issues individually.

Liposome mean diameter and size distribution. We used TEM and DLS to

evaluate the mean diameter and size distribution of liposomes formed by extrusion and

sonication. TEM images (Figures 2.2a and 2.2b) indicated liposomes formed by the two

20

 

Figure 2.2. TEM images of vesicles produced by extrusion and sonication.
(a) Extruded vesicles without cholesterol. (b) Sonicated vesicles without
cholesterol. The scale bar is 500 nm in both images.

21

methods are both unilamellar in structure. However, while these images are useful in
comparing relative sizes of liposomes, they provide little utility in evaluating absolute
sizes because the liposomes may be distorted by the vacuum and the high energy electron
beam used in imaging could, in principle, damage liposome structures. Such distortions
may appear in the form of non-spherical structures (Figures 2.2a and 2.2b). An
additional issue is that, while appealing from a visual perspective, TEM data do not lend
themselves readily to the evaluation of the size distribution of the liposomes. To address
these issues, we also characterized our liposomes by dynamic light scattering (DLS), a
technique that measures the diameter of particles suspended in solution. Because DLS
collects scattering information on an ensemble of particles, it is a relatively simple matter
to obtain information on the particle size distribution.”24
The mean diameters of vesicles formed by extrusion were 112 :t 6 nm with
cholesterol and 112 d: 6 nm without cholesterol (Table 2.1). Thus cholesterol has no
apparent influence on the sizes of liposomes formed by extrusion. This is not surprising
since we expected that the extruded vesicles would be close to 100 nm in diameter with
or without cholesterol, based on the sizes of pores in the polycarbonate membranes used
to prepare them.5'10 On the other hand, the results for liposomes formed by sonication
gave values of 28 :l: 2 nm without cholesterol and 64 d: 3 nm with cholesterol, showing
significant cholesterol influence on size. While the ultimate sizes of sonicated vesicles
were smaller than those from extrusion, this can controlled by the sonication time.ll We

also note that liposomes formed by sonication had a bimodal size distribution with or

without cholesterol, and liposomes formed by extrusion also had a bimodal size

22

Table 2.1. Dynamic Light Scattering (DLS) results for the extruded and sonicated
liposomes. Uncertainties in the data are estimated to be :I: 5% of the average values.

 

Average diameters of Average diameters of

 

extruded liposomes sonicated liposomes
(um) (um)
DOPC/NBD-PC 112i6 28i2and196i10

 

 

DOPC / cholesterol / NBD-PC 112 :t 6 and 600 i 30 64 i 3 and 259 i 13

 

 

 

 

23

distribution if they contained cholesterol (Table 2.1). It is not clear if there is a
compositional difference between vesicles in the two distribution zones, an issue that is
currently under investigation. The physical size and distribution information is not
necessarily correlated with the molecular scale properties of the bilayer structures.

Molecular-scale motion in the bilayers. To gain information on the molecular
scale dynamical properties of the liposomes, we have measured the fluorescence lifetime
and induced orientational anisotropy of the tethered NBD chromophores contained within
the bilayer structures. The data provided by lifetime and anisotropy measurements are
different but complementary. For all of the liposome systems examined here (sonicated
and extruded, with and without cholesterol), we observed a fluorescence population
decay that can be fitted with a two-component exponential model. For a chromophore
such as NBD, the observation of multiple lifetime components is indicative of the
existence of at least two environments that do not exchange population on the timescale
of the excited state relaxation. This behavior has been reported for NBD tethered to

25'3" and we interpret it in the following

phospholipids in a variety of environments,
context. The fluorescence lifetime of NBD is sensitive to the polarity of its immediate
environment, where the shorter lifetime corresponds to the NBD chromophore existing in
a more polar environment.31 For NBD chromophores tethered to the acyl chain of a
phospholipid, it has been established that the chromophore can “loop back” to position
itself in close proximity to the polar headgroup region of the bilayer in which it resides.“

20’32 The longer lifetime component in this model corresponds to the NBD chromophore

residing closer to the nonpolar region of the lipid bilayer. The fact that we observe two

24

lifetime values that differ by more than 4500 ps while the proposed environments for
NBD are only separated by a few Angstrdms is consistent with the chromophore existing
in a dielectric gradient that changes substantially over the length scale of the
chromophore itself.”34 Thus, our lifetime results are consistent with previous reports
and this information, combined with the recovery of a single anisotropy decay time
constant for each system (vide infra) points to the chromophore residing in a range of
locations, all in close spatial proximity to the bilayer polar head groups.

For the anisotropy decay measurements reported here, the chemical information

contained in these experiments is in the form of the decay functionality of R(t), (Eq. 2.1).

= Iu(t)—I.L(t)

R(t) [||(t)+211_(t)

 

(2.1)

where I“(t) and I _L(t), are the emission intensities polarized parallel and perpendicular to
the vertically polarized incident excitation pulse, respectively. In our experiments, the
NBD chromophore was tethered to a membrane bilayer constituent, and the anisotropy
decay for this system was treated in the context of the hindered rotor model.”38 The
tethered NBD chromophore does not have the structural freedom to orientationally re-
randomize fully within its environment following excitation, which gives rise to an
infinite time anisotropy, R(oo). For such systems, the function R(t) exhibits is governed
by Eq. 2.2,

R(t) = R(OO) + (R(O) - R(°°))€XP(-’/THR) (22)
where R(oo) is the infinite time anisotropy and R(O) is the zero time anisotropy. The

value of R(0) is determined by the angle between the excited and emitting transition

25

dipole moments of the chromophore and R(OO) serves as a gauge for the orientational
confinement that the chromophore experiences. The term THR is related to the

orientational relaxation of the chromophore within its confining volume, and is

aPprOXimated by Eq. 23,38:39
792
THR : _24D0 (23)

W

where 00 is the semi-angle of the confining cone. The semi-angle of the confining conic
volume of the chromophore is related to the zero- and infinite-time anisotropies by Eq.
2.4,39

Mac) "5
[m] = 0.5(cosc90 (1+ c0360 )) (2.4)
where the values of R(oo) and R(0) are determined from the experimental data. We have
determined the values for 90 for the extruded and sonicated vesicles, both with and
without cholesterol (Tables 2.2 and 2.3). For each of the systems tested, the cone angles
are the same within experimental uncertainty, regardless of the method of vesicle
preparation. We note that, for the systems containing cholesterol, the recovered cone
angles are measurably smaller (ca. 60° vs. ca. 40°) for both the extruded and sonicated
liposomes in comparison to the systems containing only the phospholipid and the
chromophore. This finding is consistent with literature reports indicating that the
addition of cholesterol to a bilayer serves to make the bilayer less fluid.”44 Our data

point to not only a slightly longer reorientation time for the cholesterol-containing

system, but also a more restricted volume (00). From these data we find Dw ~ 100 MHz

26

Table 2.2. Results from Time Correlated Single Photon Counting (T CSPC) studies

using extruded and sonicated liposomes with DOPC and NBD-PC.

 

DOPC / NBD-PC

 

 

 

 

131 (PS) 172 (PS) THR (PS) 90(0) Dw
(MHz)
Extruded 15473: 143 601421:57 30251215 623:5 113i27
Sonicated 1931 :t 261 6098 :1: 73 2944 :l: 94 57 :l: 5 98 i 21

 

 

 

 

 

 

Table 2.3. Results from Time Correlated Single Photon Counting (T CSPC) studies

using extruded and sonicated liposomes with DOPC, cholesterol and NBD-PC.

 

DOPC / cholesterol / NBD-PC

 

 

 

 

T I (PS) T2 (PS) THR (PS) 90(°) Dw
(MHz)
Extruded 1671 3:92 6041i81 3591 i277 421:3 44:1:10
Sonicated 2299 :t 182 5909 i 104 3002 :h 220 37 d: 2 41 :l: 8

 

 

 

 

 

 

27

 

 

for the phospholipid bilayer with no cholesterol, and DW ~ 45 MHz with the addition of
cholesterol. It is interesting to note that there is an apparent difference in the
reorientation times for the cholesterol-containing liposomes. This difference in time
constant, while suggestive of a different environment formed by the two methods, cannot
be taken simply at face value. The observed time constant is related to both the confining
cone angle and to the “wobbling” diffusion coefficient, Dw. The cone angles recovered
for these two liposomes are the same to within experimental uncertainty, and when the
fundamental quantity Dw is extracted from the 00 and Ink data, we find that there is no
statistically discernible difference for the liposomes formed by extrusion vs. those by
sonication.

It is important to emphasize that, in all of the anisotropy data we do not find any
measurable difference between the molecular environments of the liposomes formed by
extrusion and sonication. While there is a slight difference in the fast component of the
NBD fluorescence lifetime for sonicated vs. extruded liposomes containing cholesterol,
this finding is not reflected in the anisotropy decay dynamics, which are arguably a more
sensitive probe of the local environment of the chromophore. With this information, we
turn next to the translational diffusion behavior of the tethered chromophore in planar
lipid bilayers formed from each type of liposome.

Translational diflusion in the bilayers. Measurement of translational diffusion in
a sBLM provides information on the mobile fraction and diffusion coefficient of lipids

within the membrane. It is not possible to directly quantitate the translational motion of

28

chromophores while embedded in liposomes. Therefore, to measure the translational
diffusion of chromophores, it is common to reconstitute the liposome in the form of
sBLMs on a planar substrate such as fused silica. The transformation of liposomes to
planar bilayer lipid membranes is accomplished by a series of steps that include vesicle
adsorption, vesicle surface and vesicle-vesicle interactions and vesicle fusion.”47 Once
the bilayer is formed on a planar substrate, fluorescence recovery after pattern
photobleaching (FRAPP) measurements can be used to characterize the translational
motion of the chromophores. In such measurements, a selected region of the planar
bilayer is exposed to intense electromagnetic radiation, causing photodegradation of
chromophores in the illuminated region. Following photobleaching, the recovery of
fluorescence intensity in the bleached region occurs at a rate determined by the
translational diffusion coefficient of the chromophore.

In our FRAPP protocol, a Ronchi ruling placed in the rear image plane of the
microscope projects a pattern of alternating dark and bright fringes on the sample plane.
The recovery of fluorescence emission intensity in the sample following photobleaching
provides information about the diffusion of the fluorescent probe.48 Figure 2.3 shows a
fluorescence recovery curve obtained for a supported lipid bilayer of DOPC containing
NBD-PC as a fluorescent probe. The equation describing pattern photobleaching

recovery in this system is49

f(t)=f(0)+-’2£[1—f(0)][1—[%){exp[- 47:;Dt]+%exp(—36::DI]H (2.5)

 

 

29

 

0.70
0.65 P
0.60 -
0.55 -
0.50 -
0.45 -

0.40

l

 

0.35,

normalized fluorescence intensity (a.u.)

U \k I

0.00 'lML'Illljfllu‘nrwh"Wlmlltjtlllljm MIAMI II

-0.05 ' ‘ I I .
0 10 20 30 40

time (s)

 

 

 

Figure 2.3. Representative plot of a FRAPP recovery curve for a DOPC/NBD-PC
supported bilayer membrane.

30

where fit) is the post-bleach fluorescence (t>0) normalized with respect to the constant
pre-bleach fluorescence (t<0), with t=0 being the time of the bleach pulse, D is the
average diffusion coefficient of the fluorophores, and m is the mobile fraction. The
parameter a is the stripe periodicity of the Ronchi ruling in the sample plane, 25 pm in
our experiments. Equation 2.5 was fitted to the FRAPP data for each of the systems, and
the average best-fit values of D and m were obtained (Tables 2.4 and 2.5). The physical
significance of these quantities is that D is the translational diffusion coefficient, a direct
measure of molecular mobility on mesoscopic length scales, and m is the mobile fraction
of the chromophore. These data provide two important pieces of information (Tables 2.4
and 2.5). The first is that, for a given liposome composition, the fitted values of D and m
are the same for liposomes made by extrusion and sonication. Thus, differences in size
between liposomes prepared by extrusion and sonication do not seem to affect the fluidity
of the resulting bilayer membranes. This corroborates evidence from a quartz crystal
microbalance (QCM) study showing that liposomes ranging in size from 25 nm to 200
nm can form viable bilayer lipid membranes.50 The second important finding is that, as
expected, the presence of cholesterol in the bilayer serves to make the chromophore less
mobile (i.e. less fluid). This result is fully consistent with the anisotropy decay data
presented above, as well as previous findings that the addition of cholesterol to a

phospholipid bilayer decreases the translational diffusion of the lipid.“

31

Table 2.4. Results from Fluorescence Recovery After Pattern Photobleaching

(FRAPP) using supported bilayer membranes from extruded and sonicated vesicles.

The samples contained DOPC and NBD-PC.

 

 

 

 

Diffusion coefficient Mobile fraction of NBD-
(10'8 cm2 s") PC (%)
Extruded 2.24 i 0.25 78 i 18
Sonicated 2.02 :l: 0.15 81 i 11

 

 

 

Table 2.5. Results from Fluorescence Recovery After Pattern Photobleaching

(FRAPP) using supported bilayer membranes from extruded and sonicated vesicles.

The samples contained DOPC, cholesterol and NBD-PC.

 

 

 

 

Diffusion coefficient Mobile fraction of NBD-
(10'8 cm2 3") PC (%)
Extruded 1.65 :l: 0.22 73 :t 10
Sonicated 1.72 :t 0.04 69 :l: 6

 

 

 

32

 

 

Conclusions

The goal of this study was to compare the properties of liposomes prepared by
sonication to those prepared by extrusion on several length scales. Our findings
demonstrate that, while the physical appearance, mean diameter and size distributions of
the liposomes do indeed depend on the manner by which they are prepared, the molecular
and nanoscale organization of the bilayers, as sensed through the dynamics of a tethered
chromophore, does not depend on the manner of liposome formation. Liposomes formed
by the two schemes were characterized in solution and in planar supported lipid
membranes formed via liposome fusion on a fused silica substrate. The effect of
cholesterol on the properties of the liposomes and supported lipid bilayers was also
evaluated. Our data clearly show that whether the liposomes are formed by sonication or
extrusion does not affect the molecular scale organization of these systems. Rather, the
molecular scale organization of the bilayers is determined by interactions between the
constituent species, which themselves do not depend on the manner in which the bilayer
is formed. This finding indicates that, all other factors being equal, data on bilayer
systems formed by sonication can in fact be compared directly to data on bilayer systems

formed by extrusion.

33

Literature Cited

(1) Manconi, M.; Isola, R.; Falchi, A. M.; Sinico, C.; Fadda, A. M. Colloids
and Surfaces, B: Biointerfaces 2007, 5 7, 143.

(2) Wang, S.; Huang, J .; Song, Q.; Fu, H. Journal of Colloid and Interface
Science 2007, 311, 296.

(3) Kim, S. H.; Haimovich-Caspi, L.; Omer, L.; Talmon, Y.; Franses, E. 1.
Journal of Colloid and Interface Science 2007, 31 I , 217.

(4) Chen, K.; Luo, W.; Rahman, Z. U.; Guo, Y.; Schulte, A. Journal of
Dispersion Science and Technology 2007, 28, 471.

(5) Driessen, A. J.; van den Hooven, H. W.; Kuiper, W.; van de Kamp, M.;
Sahl, H. G.; Konings, R. N.; Konings, W. N. Biochemistry 1995, 34, 1606.

(6) Hunter, D. G.; Frisken, B. J. Biophysical Journal 1998, 74, 2996.

(7) MacDonald, R. C.; MacDonald, R. 1.; Menco, B. P.; Takeshita, K.;
Subbarao, N. K. ; Hu, L. R. Biochimica et Biophysica Acta 1991, 1061, 297.

(8) Mayer, L. D.; Hope, M. J .; Cullis, P. R. Biochimica et Biophysica Acta
1986, 858, 161.

(9) Subbarao, N. K.; MacDonald, R. I.; Takeshita, K.; MacDonald, R. C.
Biochimica et Biophysica Acta 1991, 1063, 147.

(10) Unger, E. C.; MacDougall, P.; Cullis, P.; Tilcock, C. Magnetic Resonance
Imaging 1989, 7, 417.

(11) Maulucci, G.; De Spirito, M.; Arcovito, G.; Boffi, F.; Castellano, A. C.;
Briganti, G. Biophysical Journal 2005, 88, 3545.

(12) F eitosa, E.; Barreleiro, P. C. A.; Olofsson, G. Chemistry and Physics of
Lipids 2000, 105, 201.

34

(l3) Cornell, B. A.; Fletcher, G. C.; Middlehurst, J .; Separovic, F. Biochimica
et BiophysicaActa, Biomembranes 1982, 690, 15.

(14) DelaCruz, J. L.; Blanchard, G. J. Journal of Physical Chemistry B 2003,
107, 7102.

(15) 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.

(16) Gajraj, A.; Ofoli, R. Y. Langmuir 2000, 16, 4279.

(17) Schroeder, R.; London, E.; Brown, D. Proceedings of the National
Academy of Sciences of the United States of America 1994, 91, 12130.

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

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

(20) Mazeres, S.; Schram, V.; Tocane, J .-F .; Lopez, A. Biophysical Journal
1996, 91, 327.

(21) Berne, B. J .; Pecora, R. Dynamic Light Scattering - with Application to
Chemistry, Biology and Physics; Wiley: New York, 1976.

(22) Aragon, S. R.; Pecora, R. Journal of Chemical Physics 1976, 64, 2395.

(23) Hallett, F. R.; Craig, T.; Marsh, J .; Nickel, B. Canadian Journal of
Spectroscopy 1989, 3 4, 63.

(24) Hupfeld, S.; Holsaeter, A. M.; Skar, M.; Frantzen, C. B.; Brandl, M.
Journal of Nanoscience and Nanotechnology 2006, 6, 3025.

35

(25) Chattopadhyay, A.; London, E. Biochimica et Biophysica Acta,
Biomembranes 1988, 93 8, 24.

(26) Chattopadhyay, A.; London, E. Biochemistry 1987, 26, 29.

(27) Huster, D.; Mueller, P.; Arnold, K.; Herrmann, A. European Biophysics
Journal 2003, 32, 47.

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

(29) Mukherjee, S.; Raghuraman, H.; Dasgupta, S.; Chattopadhyay, A.
Chemistry and Physics of Lipids 2004, 127, 91.

(30) Raghuraman, H.; Shrivastava, S.; Chattopadhyay, A. Biochimica et
Biophysica Acta, Biomembranes 2007, I 768, 1258.

(31) Greenough, K. P.; Blanchard, G. J. Journal of Physical ChemistryA 2007,
11 1, 558.

(32) Feng, Z. V.; Spurlin, T. A.; Gewirth, A. A. Biophysical Journal 2005, 88,
2154.

(33) Kreiter, M.; Prummer, M.; Hecht, B.; Wild, U. P. Journal of Chemical
Physics 2002, 11 7, 9430.

(34) Drexhage, K. H. Proc. Int. Conf Lumin. 1970, 693.

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

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

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

36

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

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

(40) F ilippov, A.; Oradd, G.; Lindblom, G. Biophysical Journal 2003, 84,
3079.

(41) Kahya, N.; Schwille, P. Journal of Fluorescence 2006, 16, 671.

(42) Liu, J .; Yang, L. Bioorganic & Medicinal Chemistry 2006, I4, 2225.

(43) Nishimura, S. Y.; Vrljic, M.; Klein, L. 0.; McConnell, H. M.; Moemer,
W. E. Biophysical Journal 2006, 90, 927.

(44) Cullis, P. R. FEBS Letters 1976, 70, 223.

(45) Zhdanov, V. P.; Kasemo, B. Langmuir 2001, 17, 3518.

(46) Seifert, U.; Lipowsky, R. Physical Review A: Atomic, Molecular, and
Optical Physics 1990, 42, 4768.

(47) Nollert, P.; Kiefer, H.; Jaehnig, F. Biophysical Journal 1995, 69, 1447.

(48) Smith, B. A.; McConnell, H. M. Proceedings of the National Academy of
Sciences of the United States of America 1978, 75, 2759.

(49) Starr, T. E.; Thompson, N. L. Biophysical Chemistry 2002, 97, 29.

(50) Reimhult, E.; Hook, F.; Kasemo, B. Journal of Chemical Physics 2002,
11 7, 7401.

(51) Oradd, G.; Lindblom, G.; Westerman, P. W. Biophysical Journal 2002,
83, 2702.

37

Chapter 3

UNDERSTANDING INTERMOLECULAR INTERACTIONS IN
UNILAMELLAR VESICLES COMPRISED OF MULTIPLE
CONSTITUENTS

Introduction

Lipid bilayer structures play a critical role, either directly or indirectly, in
essentially all cellular processes. A central issue in understanding the function of lipid
bilayers is the structural relationship between different, phase-segregated regions of these
complex structures. Plasma membranes are thought to contain one hundred or more
different constituents, and at this point in time, understanding structure and dynamics in
such a complex system is not a tractable problem. A substantial body of work has been
performed on comparatively simple three-component model bilayer structures. Model
bilayers containing a phospholipid, a sphingolipid and cholesterol have been shown to
demonstrate phase segregation?7 depending on the amount of each constituent, and this
property has caused some to speculate on the relationship of such heterogeneous model

8'15 that are thought to exist in plasma membranes. In

systems to “lipid raft” structures
this work, we report on our initial efforts to understand how perturbations to one phase of
these bilayer structures influence the organization and dynamics of other phases that
coexist in the same system. Our data on the fluorescence lifetime, rotational motion and
excitation transport behavior of chromophores tethered to specific bilayer constituents,

point to the presence of structural perturbations in one region having an easily

measurable effect on the other region(s). Our data on NBD-cholesterol also shed some
38

light on the orientation of this molecule within the bilayer structure, an issue that is

unresolved at the present time despite significant, ongoing attention in the literature.”"6'

13
For any chemical system, it is the interactions between molecules that determine
both the micrOSCOpic and bulk properties. Lipid bilayers are interesting because they are
both fluid and heterogeneous, and understanding the intermolecular interactions that are
responsible for the existence of such structures remains to be achieved. For three-
component systems containing phospholipids, sphingolipids and cholesterol, it is thought
by some that the sphingolipids associate most closely with cholesterol owing to the
ability of this family of lipids to form hydrogen bonds with the cholesterol hydroxyl
fimctionality.5 It is also important to consider the acyl chains of the phospholipids and
sphingolipids because, depending on the extent of chain saturation, this region of the
bilayer can exhibit significant organization, which can be perturbed by the introduction of
unsaturations.‘9 Indeed, it is the lipid acyl chain length and degree of unsaturation that is
thought to determine the melting point of the resulting bilayer structure. Lipid headgroup
structure can also influence lipid packing. Phosphatidylethanolamines, for example, have
much higher phase transition temperatures and do not exhibit the same steric hindrance
that the corresponding length phosphatidylcholines do, because of the small headgroup
size and hydrogen-bonding capability of the former. The structural complexity of even
relatively simple bilayer model systems requires that experiments be planned carefully if
they are to distinguish between the different intermolecular interactions that collectively

produce the mesoscopic behavior seen experimentally.

39

One of the first issues to be considered in the examination of a heterogeneous
system is the extent to which the organization of one phase or region influences that of
other phases in the overall system. To address this issue we use structurally well
characterized chromophores that are covalently bound to bilayer constituents. This
strategy is the same as that we have employed for “lock and key” interrogation of
crystallization processes,20 and is well established as a means of providing chromophore
localization. We have studied the fluorescence lifetime, fluorescence anisotropy decay
dynamics and excitation transport behavior of these localized chromophores to
understand local organization and longer range order within lipid bilayer structures. We
have incorporated selected chromophores into the bilayer regions of a series of
unilamellar vesicles of varying composition. Phase separation in multi-component
bilayer systems has been noted in a variety of bilayer structural motifs, including
unilamellar vesicles with diameters comparable to those we use in this work.ll

A key issue that we address in this work is in determining how to probe for the
presence of such phase segregation in structures that are ca. 100 nm in diameter. Our
fluorescence lifetime and anisotropy data indicate that the role of sphingomyelin (SPM)
in mediating bilayer organization is not as significant as expected, and that the existence
of chromophores in these bilayers can influence the dynamical properties of the bilayers
themselves. While essentially any bilayer constituent can cause structural perturbations,
the identity and location of the perturbing constituent play important roles in determining
the consequences of the perturbation. Specifically, there is a clear correlation between

the behavior of the chromophores imbedded in the cholesterol phase and the amount of

40

fluorophore-tagged lipid present in the phospholipid phase. Our data reveal interactions
between tagged lipids in the phospholipid phase and provide insight into the structural

details of the cholesterol phase.

41

Experimental

Vesicle Preparation. Powdered cholesterol, the phospholipid 1,2-dimyristoyl-sn-
phosphatidylcholine (DMPC) and egg sphingomyelin (SPM) were purchased from
Avanti Polar Lipids Inc. (Alabaster, AL) and used as received. The chromophores 25-(N-
(7-nitrobenz-2-oxa— l ,3—diazol-4-yl-methyl)amino)-27-norcholesterol (N BD-cholesterol)
and 1,2-Dimyristoyl-sn-Glycero-3 -Phosphoethanolamine-N-(Lissamine Rhodamine B
Sulfonyl Ammonium salt), (Rhodamine-PE) were purchased from Avanti Polar Lipids
Inc., and used as received. For each sample, the lipids, cholesterol, and optical probes
were mixed in selected ratios and the chloroform solvent was evaporated. The samples
all contained a total concentration of 33 mol% cholesterol (the sum of 32.5 mol%
powdered cholesterol and 0.5 mol% NBD-cholesterol). Several different concentrations
of SPM (0, 11, 22 and 33 mol%) were used, and for all of the resulting systems, the
dynamical behavior of the NBD chromophore remained the same to within the
experimental uncertainty for a given concentration of Rhodamine-PE. The rhodamine-
PE concentration was varied between 0 and 1.0 mol%, and NBD cholesterol dynamics
were found to depend sensitively on Rhodamine-PE concentration. The balance of the
vesicle composition was comprised of the C14 phospholipid, DMPC. An appropriate
volume of Tris® buffer (Sigma-Aldrich, Milwaukee, WI) was added to each mixture so
that the final concentration of lipid in the mixture was 1 mg/mL. The buffer (10 mM, pH
7.8) was prepared using purified water from a Milli-Q Plus water pm‘ification system
(Millipore, Bedford, MA). The mixtures were processed five times through a freeze-

thaw-vortex cycle to ensure thorough mixing of the constituents. Each cycle involves

42

freezing the solution by immersion in liquid nitrogen for five minutes, followed by
thawing via immersion in 60°C water (five minutes), and concluding with vortexing (2
minutes). After the freeze-thaw-vortex processing, the solutions were extruded once
through two polycarbonate membrane filters with 400 nm diameter pores using a mini-
extruder (Avanti Polar Lipids Inc.). The resulting extruded vesicle suspension was then
extruded eleven times through two polycarbonate membranes (Avanti Polar Lipids Inc.)
with a nominal pore diameter of 100 nm. The resulting solution contained vesicles with a
size distribution centered around 100 nm as estimated from TEM data (Figure 3.1).2l'26

Steady state spectroscopy. Absorption spectra were obtained using a Cary 300
UV-visible Spectrophotometer. Quartz cuvettes with a 1 cm pathlength were used for all
measurements. Background samples without the fluorophore were used for baseline
correction because of scattering from the vesicle-containing solutions. NBD exhibits
steady state absorption and emission features that depend sensitively on the chromophore
local environment. The 480 nm band position we observe experimentally is fully
consistent with the chromophore residing in a relatively nonpolar environment.27

Steady state fluorescence measurements were performed using a SPEX
Fluorolog—3 spectrophotometer using 1 cm pathlength quartz cuvettes. Excitation and
emission slits with a nominal bandpass of 3 nm were used for all experiments. For
recording excitation spectra, the emission wavelength was kept fixed at 535 nm and
experiments were performed at 298 K. Background intensities of samples in which

NBD-cholesterol was not added were subtracted from each sample spectrum to cancel out

any scattering contributions.

43

 

Figure 3.1. Transmission electron microscopy (TEM) image of a
unilamellar vesicle. Top pane: Wide area view of aggregated vesicles.
Scale bar = 1 pm. Bottom pane: Images of individual vesicles under
higher magnification conditions. Scale bar = 200 nm.

T ime~resolved fluorescence measurements. Time-domain fluorescence intensity
decays were measured at a sample temperature of 27°C (above the melting transition
temperature of DMPC, 24°C) using a time-correlated single-photon counting (TCSPC)

28’” and we briefly recap its

system. This system has been described in detail elsewhere,
salient features here. The light source is a CW mode-locked Nd:YAG laser (Coherent
Antares 76-s) that produces 100 ps pulses at a 76 MHz repetition rate. A cavity-dumped
dye laser (Coherent 702-2) was operated with either Stilbene 420 dye (Exciton) and
excited by the third harmonic of the source laser, or was operated with Pyromethene 567
dye (Exciton) and was excited by the second harmonic of the source laser. Sample
fluorescence was detected with a microchannel plate — photomultiplier tube (MCP-PMT,
Hamamatsu R3 809U) and the electronics used to temporally resolve the fluorescence
transients were a constant fraction discriminator (CFD, Tennelec 454) and a time-to-
amplitude converter/biased amplifier (TAC, Tennelec 864). The fluorescence lifetime
data were collected at 54.7° with respect to the vertical excitation polarization, while the
reorientation data were collected at polarizations parallel and perpendicular to the
vertically-polarized incident light. The instrument response fimction for this system is

typically 35 ps fwhm and we do not attempt to deconvolute the response function from

the recorded fluorescence transients.

45

Results and Discussion

The primary purpose of this work is to examine the dynamics of chromophores
imbedded in specific bilayer regions and the interactions between different, spectrally
overlapped chromophores where each is sequestered in a different bilayer phase. We are
interested in establishing ways to probe local organization in structures on the order of
100 nm in diameter, with the goal being to elucidate structural heterogeneity on length
scales that are well beneath the far field optical diffraction limit. We have used NBD-
tagged cholesterol and a rhodamine-tagged C14 phospholipid (Figure 3.2), and find that
the transient response of these chromophores individually, as well as the excitation
transport dynamics between them, provides insight into the extent of organization within
different regions of the bilayer structure. We have also examined the effect of

8"°"2"3’3° in the bilayers on the dynamics of these chromophores, and find

sphingomyelin
that the chromophore behavior is essentially sphingomyelin-independent.

The first chromophore we have used is 7-nitrobenz-2-oxa—1,3-diazol-4-yl (NBD).
NBD is very weakly fluorescent in water, but in a hydrophobic medium such as a lipid
bilayer, this chromophore exhibits a high fluorescence quantum yield;31 the fluorescence
lifetime of the NBD group is sensitive to the polarity of its environment.3‘°"33 In this
work, we have used the NBD chromophore bound covalently to the aliphatic terminus of
cholesterol, NBD-cholesterol (Figure 3.2). This probe has been used in a variety of
bilayer studies, and there is an ongoing discussion in the literature regarding the

orientation of this molecule in cholesterol-rich regions of the bilayers studied. Some

reports indicate that the NBD chromophore is located closest to the polar phospholipid

46

' 1

- Rhodmnine-PE abscrpu'cn

linear response/103 (L/mol-cm)

  

 

 

Figure 3.2. Top: Absorption and emission spectra of NBD-cholesterol, and the
absorption spectrum of Rhodamine-PE, Tris® buffer. The spectra have been
corrected for the extinction coefficients of the two compounds, and the
fluorescence quantum yield of NBD-cholesterol is assumed to be 0.5 for the sake
of presentation. Bottom: Structures of NBD-cholesterol and Rhodamine-PE.

47

head group region, with the cholesterol moiety oriented in opposition to that of native
cholesterol.”l7 There are also reports that indicate NBD is oriented in some systems
such that the cholesterol moiety lies in correspondence with the other, non-substituted
cholesterol molecules, positioning the chromophore within the nonpolar interior region of
the bilayer.l6’18

Our experimental data shed some light on this issue. Based on the fluorescence
lifetime data (In ~ 2.8 ns) and the functional form of the motional relaxation data (vide
infra), we believe that the NBD-labeled cholesterol molecule is oriented such that the
NBD moiety is sequestered within the nonpolar region of the membrane.5 Specifically,
our NBD-cholesterol fluorescence lifetimes in the absence of Rhodamine-PE are ca. 2.8
ns, and the anisotropy decay times are ca. 1.3 ns. In aqueous environments, NBD has a
fluorescence lifetime that is typically hundreds of ps at most, and the reorientation times
for this chromOphore in an aqueous environment are on the order of 100 ps. Our lifetime
data thus indicate a nonpolar environment and the reorientation data reveal a viscosity of
ca. 10 — 15 cP, consistent with the interior of a bilayer membrane.”’34 This orientation is
consistent with the cholesterol —OH functionality being in contact with the polar solvent
medium, where it is capable of participating in hydrogen bonding with phospholipids and
sphingolipids.

In contrast to the NBD-cholesterol chromophore, there is less uncertainty with

respect to the location of the rhodamine chromophore bound to DMPE because it is

attached to the polar head group of the phospholipid (Figure 3.2). We thus expect it to

48

interact most strongly with the aqueous solvent in the immediate proximity of the bilayer
head groups.

For fluorescence lifetime measurements of complex organic chromophores, there
is not a single clear theoretical framework within which to interpret the data.”39 The
fluorescence lifetime of NBD is sensitive to the polarity of its immediate environment,
but there is little other than an empirical correlation to apply to the interpretation of such
data. For systems where only one of the chromophores is present, we recover a single
exponential fluorescence intensity decay, indicative of a single environment for each
chromophore (Figure 3.3). For systems where both chromophores are present, we
recover multiple exponential decay components under certain circumstances, and we
understand this behavior in the context of F 6rster energy transfer between chromophores.
The lifetime and reorientation behavior of NBD-cholesterol depend very sensitively on
the amount of Rhodamine-PE present. The first issue to consider is excitation transport
between the chromophores, as expected from their spectral overlap (Figure 3.2). We
show in Figure 3.4 the fluorescence lifetime(s) and their fractional contributions as a
function of Rhodamine-PE concentration in the bilayers. We understand these data as
follows. Based on a significant body of data that exhibit phospholipid-cholesterol phase
separation, we believe that our vesicles exhibit a non-homogeneous distribution of
cholesterol within the bilayer.”7 It is known that even for small amounts of cholesterol
and NBD-cholesterol in unilamellar vesicles that phase separation and organization
occurs,3 1 and our experimental results (vide infra) are fully consistent with that structural

motif. While it would be ideal to have a direct measure of the morphology of our

49

 

1200 -
l-
‘r
1000 F r- "=-
:' ."s
45 o -
r 1*
800 - g ..":.
. ~5th
I- t m. .
g 600 — ' "‘1‘ ’
b ’ r t ‘2'“. -3 ..
L- . f. { “Ev 0. '
4a) ‘ 0'... "1.): 0’.
° ‘0 ”l.
O
200 - -‘
O J "M —
1 l l

 

 

Figure 3.3. Fluorescence lifetime decay of NBD-cholesterol in a unilamellar vesicle
containing 33 mol% cholesterol, 11 mol% sphingomyelin and 56 mol% DMPC. The

tirrr: (ps)

recovered time constant of this decay is 3150 :l: 30 ps.

50

vesicles, providing information such as the domain size of the phase separated regions,
we do not have the capability of producing such data. Owing to the small size of the
vesicles, far field optical microscopic techniques cannot provide adequate resolution.
The creation of larger vesicles or the deposition of these bilayers onto a planar substrate
may yield some information, but translational diffusion within

such structures could serve to cloud the correspondence across bilayer structural formats.
A significant goal of this work is to explore the utility of spectroscopic probes in
providing structural information on vesicle structures of such small dimensions.

With the ansatz that NBD-cholesterol is located in the cholesterol-rich regions of
the bilayers, and the Rhodamine-PE is located in the phospholipid regions, for low
Rhodamine-PE concentrations there is only one NBD-cholesterol lifetime, which
decreases with increasing Rhodamine-PE concentration. At 0.25 mol% Rhodamine-PE,
we observe a change in decay functionality from one to two exponential components.
We can interpret these data either in terms of induced structural heterogeneity in the
system where the NBD chromophore finds itself distributed between two non-exchanging
environments, or in terms of excitation transport. Because of the modest amounts of
Rhodamine-PE used and the fact that the rhodamine chromophore resides beyond the
polar head group region of the bilayer, we believe that structural perturbations to the acyl
chain region of the bilayer will be minimal. Due to the extensive spectral overlap of the
chromophores as well as the monotonic decreases in lifetime(s) with increasing optical

acceptor concentration, we assert that these data are reflective of excitation transport in a

51

structurally heterogeneous system, where the short lifetime reflects efficient excitation
transport when a Rhodamine-PE molecule is in close proximity to a cholesterol-rich
region containing an NBD-cholesterol chromophore and the long lifetime data are related
simply to the average distance between non-associated donor and acceptor species.
Because the NBD-cholesterol fluorescence lifetime varies in a regular manner with the
Rhodamine-PE concentration, we believe that the fast lifetime component contains
information on the size of the cholesterol domains in which the NBD-cholesterol resides.
For this system, the dominant mechanism for energy transfer to occur will be by dipolar

coupling, as described by Forster, 4O

 

DA 4 '

2 o - - 2 D 6
k 4‘2; 8.8x10”I£A(V)_{D(v)di7=£_f;.[%] (3.1)
n V n

where k0,, is the rate constant for the transfer of energy from the donor to the acceptor.
The K term is a function of the geometric alignment of the donor and acceptor transition
dipole moments, and for our experimental conditions, we approximate <r<2> = 2/3. While
K2 can vary between 0 and 4, for our experimental geometries, K2 will lie in the region of
1. Lacking specific knowledge of the relative orientations of the donor and acceptor
transition moments, we use the orientationally averaged value of K2 = 2/3. Any
physically reasonable variations in K2 do not affect the interpretation of our data. The
term k,D is the donor radiative decay rate constant in the absence of optical acceptor, n is
the refractive index (n = 1.4, an approximate average of the organic (n~1.45), and the
aqueous (n~1.33) regions of the bilayer) and R is the intermolecular D-A distance. The

integral term is a measure of the spectral overlap between the emission spectrum of the

52

optical donor and the absorption spectrum of the acceptor. The spectral overlap term,
combined with the several constants on the right hand side of Eq. 3.1 is typically referred
to as the critical radius, R0, a quantity which serves as an effective gauge of the range of
this dipolar coupling process. We calculate R0 for our donor-acceptor pair to be 51.4 A, a
value in good agreement with that of most organic chromophore pairs, (R0 ~ 50 A
typically).4043 Based on this value of R0, k,D = 3.6x108 Hz, and kDA between 1.8 and
2.2x109 Hz (the fast fluorescence lifetime range), we calculate R to be between 28.4 A
and 29.4 A. If we had used a value of K2 = 1, the resulting R would be in the range of
30.4 to 31.4 A, a difference that does not affect the physical picture of this system
significantly. There are many geometric approximations that could be used to estimate
the cholesterol domain size from this value of R, but if we presume these domains to be
circular, R would approximate the radius of the circle. The area of the circle would thus
be ca. 2500 A2. If the cholesterol molecules are stacked more or less end-on with respect
to the plane of the circle, this area corresponds to ca. 100 molecules. For a circular
domain of ca. 100 molecules, something on the order of 30 of these molecules will reside
at the phase boundary, a structural situation that could influence the lifetime and
dynamics of the NBD chromophore. Because of the size of the domains and translational
motion within them, it is likely that the reorientation times we observe for this
chromophore represent an average of interior and edge site dynamics. Given the time
scales of the phenomena we report here, there will certainly be significant positional

averaging of the NBD local environment due to diffusion within the cholesterol domains.

53

In the limit of acceptor molecules being distributed unifomly in the lipid phase,
the F 6rster model predicts an increase in acceptor concentration would be related to the
amount of Rhodamine-PE in closest proximity to the cholesterol-rich regions of the
bilayers, yielding a monotonic relationship between acceptor concentration and fractional
contribution of the fast decay to the overall NBD-cholesterol lifetime decay. The
acceptor concentration-dependence of the lifetime pre-exponential factors, however, do
not display this trend and are thus not consistent with a uniform distribution of acceptors
in the lipid phase (Figure 3.4b). There could be several possible explanations for the
decrease in the contribution from the fast lifetime component with increasing acceptor
concentration, with the most likely being that chromophore-chromophore interactions of
the Rhodamine-PE acceptor molecules is occurring in the phospholipid regions of the
bilayer. This is a testable hypothesis.

To measure fluorescence anisotropy decays, we excite the samples with vertically
polarized light and collect fluorescence transients polarized parallel (11(0) and
perpendicular (Ii(t)) to the excitation polarization (Figure 3.5a). We construct the
induced optical anisotropy function, R(t) from polarized emission intensity decays

according to Eq. 3.2.

R( )_ In(’)"lr(t)

7W5 “'2’

The chemical information contained in these data resides in the decay functionality of

R(t) (Figure 3.5b). For the experiments we report here, the chromophores are tethered to

54

 

fluorescence lifetime
0\
O
o

 

1.0

E

0

.52

a

8‘

g 0.8
93 0.6
E

a; 0.4
2 .
O

‘3,» 0.2
B 00.
z .

; i i

0.0 0.2 0.4 0.6 0.8 1.0
mol% Rhodamine-PE
F § § b

slow component

I l
I—O—I l—O—l

l

{ fast component

i

l J l r l J l m l r l

0.0 0.2 0.4 0.6 0.8 1.0
mol% Rhodamine—PE

 

 

Figure 3.4. (a) Fluorescence lifetime(s) of NBD-cholesterol as a function
of amount of Rhodamine-PE added to unilamellar vesicles. The
appearance of two decay components is consistent with excitation

transport

in a heterogeneous system. (b) The fractional contribution of

the fast fluorescent lifetime component of NBD-cholesterol as a function
of Rhodamine-PE concentration.

55

membrane bilayer constituents, restricting degrees of freedom into which the initially
anisotropic distribution can relax.

If Rhodamine-PE molecules are distributed uniformly in the phospholipid phase
of the bilayer, over the concentration ranges we access, we should see no Rhodamine-PE
concentration-dependence to its anisotropy decay. With interactions between acceptor
chromophores, we would see the time constant for rhodamine chromophore motion
increase with increasing Rhodamine-PE concentration, as is seen experimentally (Figure
3.6). Because the rhodamine chromophore is tethered to the polar head group of the
phospholipid, and the dipolar and/or hydrogen bonding interactions between individual
head groups dominate this region of the bilayer structure, a slowing of chromophore
motion is associated with confinement. The NBD-cholesterol fluorescence lifetime data
(Figure 3.4) and the Rhodamine-PE anisotropy data (Figure 3.6) are both consistent with
the intermolecular interactions of the Rhodamine-PE chromophores in the phospholipid
phase of the bilayer, where such interactions simultaneously slow the acceptor
chromophore motion and increase the average donor-acceptor chromophore distance,
thereby reducing the fractional contribution of the fast lifetime component.

As noted above, the rhodamine chromophore reorientation times increase with
confinement because of their proximity to the polar head group region, where the lipid
head groups interact strongly owing to dipolar and hydrogen-bonding interactions. The
motional relaxation behavior of NBD-cholesterol appears to be different. As the

concentration of Rhodamine-PE is increased to 1.0 mol% in the phospholipid phase, the

56

    

  
  
 

 

 

 

 

 

 

140 f a
120 r
100 g
‘3 800 .
8 .
0 600'" I (t)
4001— \*
ZOO'D response function
0 r . : . 'fi 1 i . a
0 2000 4000 6000 8000
time (ps)
0.20 r 0;, b
0.15 *' '3'.
' 'r’vizo: . .
A0.10-3'~’: -.;,. .
a b °' ’3, "0.; dv '2'“. . . .‘ .
m 0.05 ”a .'.m-:-m%.. - - «3 . -

. V' E}; «XML-5 1:- ' .
- :0. 9' o .J:..‘. - d r. «at . .3! ; L... ’Q;%
0.00 .0 ‘ . . 1 Mt“ f ‘13:.

 

 

. "V.
’ ' afflcg ”.335... {”9qu 1,.
-0.05— - °-‘: ...-; “in: ~:- ”514?:
-0.10 1 . 1 l 1 r 'r . .. m - j I
0 2000 4000 6000 8000
time (ps)

Figure 3.5. (a) Polarized emission transients for NBD-cholesterol in a
unilamellar vesicle containing 33 mol% cholesterol, 11 mol% sphingomyelin
and 56 mol% DMPC. The acquisition polarization of the transients is
indicated in the figure relative to the vertical excitation pulse polarization.
(b) Decay of the induced orientational anisotropy function, R(t), produced
from the data shown in panel (a). The time constant of this anisotropy decay
is 1362 :1: 61 ps.

57

800

l

700

600

 

l
r—O—r

Rhodamine-PE anisotropy decay time (ps)

 

 

500.“ I

400—

300:-

200; i

100 0:0 L 0f2 1 0:4 4 0.L6 I 08

mol % Rhodamine-PE

Figure 3.6. Dependence of Rhodamine-PE anisotropy decay time on
Rhodamine-PE concentration.

58

reorientation dynamics of NBD-cholesterol become faster, exhibiting a feature analogous
to a phase transition within the bilayer structure (Figure 3.7).

When confined within a membrane, a molecule is not free to sample all possible
orientations with equal probability, and the hindered rotor model is appropriate for the
treatment of such data.“45 For a planar membrane, R(t) does not decay to zero at long
times; the residual steady state anisotropy results from the orientational constraints the
membrane imposes on the chromophore. The hindered rotor model describes the motion
of the restricted chromophore in the context of confinement within a conic volume of
semiangle 200, where the angle 0 is defined as the average tilt angle of the chromophore
within the cone (0° 5 0 S 90°).44
R(t) = R(OO) + (R(O) - R(00)) eXPU/T) (3-3)
The quantity 1 contains information on the confining cone semi-angle 00 and the motion

of the chromophore within the cone, Dw.

_ 76:

— 3.4
T 241),, ( )

 

The quantity 1 in Eq. 3.4 contains contributions from both 00 and Dw. Other work in our
group has shown that both the structural freedom (00) and effective rotor shape (Cartesian
components of D...) can vary with the composition of the bilayer.34

Previous work on the temperature-dependent reorientation of perylene
sequestered in a DMPC bilayer showed a change in reorientation time by a factor of ca.
1.5 at the phase transition temperature for this lipid. The addition of 0.3 mol% of a

DMPC analog with an unsaturation at the 9-position of the acyl chains yielded a 15

59

—| H H I—5
N DJ -h U!
C O O O
O O O O
' l ' l
r—o—r

l

1100

l

1000
900 r

800 ~

NBD-cholesterol anisotropy decay time (ps)

 

700 L 1 I 1 L r l 1 l 1 1
0.0 0.2 0.4 0.6 0.8 1.0

mol% Rhodamine-PE

Figure 3.7. Dependence of NBD-cholesterol anisotropy decay time on
Rhodamine-PE concentration.

60

degree reduction in phase transition temperature.19 The presence of Rhodamine-PE
affects the local environment of the NBD-cholesterol, as can be seen in the fluorescence
anisotropy data (Figure 3 .4). The functional form of the Rhodamine-PE concentration-
dependence of NBD-cholesterol reorientation (at the fixed temperature of 27°C) resemble
the phase transition data seen for perylene reorientation in DMPC vesicles. The phase
transition for DMPC is known to be 24°C for the pure phospholipid,46 and the addition of
impurities lowers this phase transition temperature dramatically. If our data on NBD—
cholesterol are indeed sensing a phase transition, it must be for a domain other than the
phospholipids. We believe that we are sensing changes in the organization of the
cholesterol-rich domains in these experiments. We understand this finding in the context
of interacting Rhodamine-PE molecules producing “defect” sites in the lipid phase of the
bilayer. The associated disorder of the lipid phase results in a corresponding change
(loss) of order in the cholesterol phase.

It is tempting to estimate the viscosity of the bilayer region sensed by the NBD-
cholesterol. Because we have not acquired fluorescence anisotropy data using both one-
and two-photon excitation, it is not possible to determine the local Viscosities
unambiguously. If we assume that the NBD-cholesterol reorientation is characterized as
a relatively spherical rotor, we can place bounds on the viscosity of the cholesterol-rich
regions by comparing the experimental data to theoretical predictions. We have shown in
earlier work that for the chromophore NBD tethered to a phospholipid acyl chain, the
motional dynamics of this chromophore change significantly, from being a highly

anisotropic rotor, with the dominant motion about its tethering bond in a pure

61

phospholipid bilayer, to being much closer to a spherical rotor in bilayers containing
cholesterol.34 Those data were obtained by direct comparison of one- and two-photon
excited fluorescence anisotropy measurements, and we have not acquired the analogous
information for the system under examination here. Nonetheless, this information
suggests that the NBD chromophore reorients approximately as a spherical rotor in our
systems. With this ansatz, we can estimate the rotational diffusion of NBD-cholesterol

using the modified Debye-Stokes-Einstein equation.“49

_ Vf
= 60 ‘ = "—— 3.5
TOR ( ) kBTS ( )
This model has been discussed in detail before,'9’20’28’34 with the quantities n being

solution viscosity, V representing solute hydrodynamic volume, f being the solvent-
solute frictional interaction coefficient, and S representing the solute shape factor.

We estimate that the viscosity sensed by NBD-cholesterol varies from ca. 10 cP
for low Rhodamine-PE concentrations to ca. 7 cP at higher Rhodamine-PE concentrations
based on a comparison of the experimental data to values calculated from Eq. 3.5 (Table
3.1). This range of Viscosities is reminiscent of our findings for DMPC bilayers in the
region of their gel-to-fluid phase transition,19 suggesting that both the lipid and

cholesterol domains of model bilayer structures are similar in viscosity.

62

Table 3.1. Calculated viscosity dependence of the rotational diffusion coefficient
and anisotropy decay time constant for NBD-cholesterol. The diffusion coefficients
and reorientation times were calculated based on a hydrodynamic volume of 528 A3
I for the NBD-cholesterol molecule, f = l, S = 1 and T = 298 K.

 

 

 

 

 

 

 

Viscosity (cP) Rotational diffusion Reorientation time, to]:
coefficient, D (Hz) (ps)
5 2.60x108 640
7.5 1.73x108 960
10 1.30x108 1280
12.5 1.04x108 1600
15 8.68x107 1920

 

 

 

63

 

Conclusions

The data we have reported underscore the influence that comparatively small
amounts of certain bilayer constituents can have on the overall properties of the system.
These data are consistent with strong coupling between lipid- and cholesterol-rich phases
determining the overall dynamics and fluidity of bilayer structures. Our fluorescence
lifetime and anisotropy data indicate that the NBD chromophore tethered to cholesterol
resides within the bilayer structure and that there is a significant dependence of the
motional freedom of this chromophore on the composition of the phospholipid-rich
regions of the bilayer. The rhodamine chromophore exists at the interface between the
bilayer polar head groups and water, and our fluorescence anisotropy data for this
chromophore indicates interactions between rhodamine chromophores even at relatively
low concentrations. Such intermolecular interactions give rise to a step-wise change in
the reorientation dynamics of the NBD-cholesterol chromophore that is reminiscent of a

phase transition.

64

Literature Cited

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

(2) Keller, S. L. Journal of Physics: Condensed Matter 2002, 14, 4763.

(3) Keller, S. L. Langmuir 2003, 19, 1451.

(4) Keller, S. L.; McConnell, H. M. Physical Review Letters 1999, 82, 1602.

(5) Keller, S. L.; Radhakrishnan, A.; McConnell, H. M. Journal of Physical
Chemistry B 2000, 104, 7522.

(6) Hinderliter, A.; Biltonen, R. L.; Almeida, P. F. F. Biochemistry 2004, 43,
7102.

(7) Sankaram, M. B.; Thompson, T. E. Proceedings of the National Academy
of Sciences of the USA 1991, 88, 8686.

(8) Brown, D. A.; London, B. Journal of Biological Chemistry 2000, 275,
17221.

(9) de Almeida, R. F. M.; Loura, L. M. S.; F edorov, A.; Prieto, M. Journal of
Molecular Biology 2005, 346, 1109.

(10) de Almeida, R. F. M.; Fedorov, A.; Prieto, M. Biophysical Journal 2003,
85, 2406.

(11) De Almeida, R. F. M.; Loura, L. M. S.; Fedorov, A.; Prieto, M.
Biophysical Journal 2002, 82, 823.

(12) Samsonov, A. V.; Mihalyov, 1.; Cohen, F. S. Biophysical Journal 2001,
81, 1486.

65

(13) Wolf, C.; Koumanov, K.; Tenchov, 8.; Quinn, P. J. Biophysical Chemistry
2001, 89, 163.

(14) Loura, L. M.; Fedorov, A.; Prieto, M. Biochimica et Biophysica Acta
2001, 1511, 236.

(15) McMullen, T. P. W.; Lewis, R. N. A. H.; McElhaney, R. N. Current
Opinion in Colloid & Interface Science 2004, 8, 459.

(16) Chattopadhyay, A.; London, E. Biochemistry 1987, 26, 39.

(l 7) Chattopadhyay, A.; London, E. Biochimica et Biophysica Acta 1988, 938,
24.

(18) Scheidt, H. A.; Muller, P.; Herrmann, A.; Huster, D. The Journal of
Biological Chemistry 2003, 278, 45563.

(19) Koan, M. M.; Blanchard, G. J. Journal of Physical Chemistry B 2006, in
press.

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

(21) Driessen, A. J .; van den Hooven, H. W.; Kuiper, W.; van de Kamp, M.;
Sahl, H. G.; Konings, R. N.; Konings, W. N. Biochemistry 1995, 34, 1606.

(22) Mayer, L. D.; Hope, M. J .; Cullis, P. R. Biochimica et Biophysica Acta
1986, 858, 161.

(23) Unger, E. C.; MacDougall, F.; Cullis, P.; Tilcock, C. Magnetic Resonance
Imaging 1989, 7, 417.

(24) Hunter, D. G.; F risken, B. J. Biophysical Journal 1998, 74, 2996.

66

(25) MacDonald, R. C.; MacDonald, R. I.; Menco, B. P.; Takeshita, K.;
Subbarao, N. K.; Hu, L. R. Biochimica et Biophysica Acta 1991, 1061, 297.

(26) Subbarao, N. K.; MacDonald, R. 1.; Takeshita, K.; MacDonald, R. C.
Biochimica et Biophysica Acta 1991, 1063, 147.

(27) Ladokhin, A. S. Fluorescence Spectroscopy in Peptide and Protein
Analysis; John Wiley & Sons: Chichester, UK, 2000.

(28) DelaCruz, J. L.; Blanchard, G. J. Journal of Physical Chemistry B 2003,
107, 7102.

(29) 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.

(30) Wang, T. Y.; Silvius, J. R. Biophysical Journal 2000, 79, 1478.

(31) Mukherjee, S.; Chattopadhyay, A. Biochemistry 1996, 35, 1311.

(32) Chattopadhyay, A.; Mukherjee, S. Journal of Physical Chemistry B 1999,
103, 8180.

(33) Rukmini, R.; Rawat, S. S.; Biswas, S. C.; Chattopadhyay, A. Biophysical
Journal 2001, 81, 2122.

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

in review.

(35) Kawato, S.; Kinosita, K., Jr.; Ikegami, A. Biochemistry 1977, 16, 2319.

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

(37) Loura, L. M.; Fedorov, A.; Prieto, M. Biophysical Journal 2001, 80, 776.

67

(3 8) Loura, L. M. S.; Fedorov, A.; Prieto, M. Journal of Physical Chemistry B
2000, 104, 6920.

(39) Loura, L. M. S.; F edorov, A.; Prieto, M. Biochimica et Biophysica Acta,
Biomembranes 2000, 1 4 6 7, 101.

(40) Forster, T. Ann. Physik 1948, 2, 55.

(41) Birch, D. J. S.; Suhling, K.; Holmes, A. S.; Salthammer, T.; Imhof, R. E.
Pure and Applied Chemistry 1993, 65, 1687.

(42) Ediger, M. D.; Fayer, M. D. Journal of Chemical Physics 1983, 78, 2518.

(43) Jones, G. M.; Wofsy, C.; Aurell, C.; Sklar, L. A. Biophysical Journal
1999, 76, 517.

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

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

(46) Marsh, D. Handbook of Lipid Bilayers; CRC Press: Boca Raton, 1990.

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

(48) Debye, P. Polar Molecules. In Chemical Catalog Co. New York, 1929.

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

68

Chapter 4

THE ROLE OF PHOSPHOLIPID HEADGROUPS IN NIEDIATING
BILAYER ORGANIZATION. PERTURBATIONS INDUCED BY THE
PRESENCE OF A TETHERED CHROMOPHORE

Introduction

Phospholipids are an essential component of plasma membranes,1 and there is a
significant, multidisciplinary research effort underway to understand the structural and
compositional complexity of lipid bilayer structures. The use of model bilayer systems
has attracted a great deal of attention because of their potential utility in the construction
of chemical and biological sensor devices based on biomolecular recognition.
Phospholipids self-assemble spontaneously in an aqueous environment to form a bilayer
structure with the hydrophobic tails directed inward, toward one another and the
headgroups exposed to the aqueous phase. This characteristic organization is responsible
for the ability of biomembranes to act as barriers between two aqueous environments. It
is clear that for such complex systems, both the acyl chain region and the polar
headgroup region of the bilayer contribute to the overall organization.

As noted above, phase transitions have been seen in a variety of lipid bilayer
structures. For a pure phospholipid the gel-to-fluid phase transition is well defined, with
the transition temperature, Tm, depending sensitively on the length of the acyl chains and
the presence and location of any unsaturations. The physical change that occurs at the
gel-to-fluid phase transition is related to the organization of the aliphatic chains, with an
increase in the number of trans-gauche conformers with increasing temperature.2 The

gel phase is a comparatively well-ordered state in which the translational mobility of the
69

bilayer constituents is thought to be limited. Above the phase transition temperature, in
the fluid phase, the phospholipid aliphatic tails are more disordered,3 leading to a
decrease in the thickness of the bilayer,4 and an increase in the rotational and lateral
freedom of the phospholipid molecules. Lipid bilayers may exist in the gel phase or the
fluid phase depending on the temperature, pressure, extent of hydration, and the structural
identity of the phospholipids.

Many studies of bilayer systems have used fluorescent probes to interrogate the
bilayer structure and dynamics. For example, we have established that the gel-to-fluid
phase transition temperature is sensed by perylene sequestered in the nonpolar region of
the lipid bilayer,5 and the results of this work are in excellent agreement with the phase
transition information recorded for the same lipid system using NMR measurements.6 A
more general question that we are interested in is whether or not the presence of the
chromophore in a bilayer system can have a perturbative effect, yielding results that are
not representative of the native bilayer system. We address this issue in this work, using
a fluorescent probe localized at the polar headgroup region of the lipid bilayer. We have
chosen to use a polar chromophore bound to the headgroup region of a phospholipid as
the probe. Our choice is deliberate. For polar chromophores bonded to the phospholipid
acyl chains, there is a well-established tendency for the chromophore to “loop back” so
that it will be in relatively close proximity to the polar headgroup region,7'12 a condition
which is certainly perturbative to the bilayer structure. For a nonpolar chromophore,
such as perylene, we have shown that the phase transition data we recover is the same as

that for lipid bilayers that do not contain the chromophore. With the system chosen, we

70

are interrogating the portion of the bilayer above the polar headgroups, and the presence
of the probe does not alter the organization of the acyl chains. To understand the role that
this tethered chromophore plays in mediating bilayer organization, we have studied its
concentration-dependent fluorescence behavior in unilamellar vesicle structures.

We use the lipid-tethered chromophore 1,2-dioleoyl-sn-glycero-3-
phosphoethanolamine-N-lissamine rhodamine B sulfonyl ammonium salt (Rhodamine-
PE). The rhodamine dyes are characteristically strong absorbers in the visible spectral
region, and their relatively high fluorescence quantum yields and several nanosecond
fluorescence lifetimes are well suited for use at the sub-monolayer level. The polar
nature of the rhodamine chromophore, together with the characteristics listed above and

13’” render it suitable for the present

previous work with rhodamines in our research group
study.

We focus on the rotational motion of Rhodamine-PE in the polar headgroup
region as a function of its concentration and system temperature. We find that there is a
significant dependence of the rotational relaxation dynamics and disorder in the bilayer
structure with increasing chromophore concentration. These findings are augmented by
steady state fluorescence data showing that at the highest chromophore concentrations
there are some chromophore-chromophore interactions. Taken collectively, the data
point to the need for caution in the use of fluorescent probes in bilayer structures, and that

low concentrations of chromophore are required if the experimental data are to be useful

in understanding the properties of the native bilayer.

71

Experimental

Vesicle Preparation. Phospholipids 1,2-dimyristoyl-sn-phosphatidylcholine
(DMPC) and 1,2-dioleoyl-sn—glycero-3 -phosphoethanolamine—N-lissamine rhodamine B
sulfonyl ammonium salt (Rhodamine-PE) were purchased from Avanti Polar Lipids Inc.
(Alabaster, AL) and used as received (Figure 4.1). For each sample, the lipids and the
rhodamine probe were mixed in selected ratios so that the final concentrations of the
samples were 1.5 x 10'3 M lipid, with the Rhodamine-PE concentration being 0.0025
mol%, 0.025 mol% or 0.25 mol%. The chloroform solvent was evaporated from the
lipids and an appropriate volume of Tris® buffer (Sigma-Aldrich) was added to each
mixture so that the lipid concentration was 1 mg/mL, consistent with the 1.5 mM
concentration listed above. The buffer (10 mM, pH 7.8) was prepared using water from a
Milli-Q Plus water purification system GVlillipore, Bedford, MA). The mixtures were
processed five times through a freeze-thaw-vortex cycle to ensure complete mixing of the
constituents. Each cycle consisted of freezing the solution by immersion in liquid
nitrogen for five minutes, followed by thawing via immersion in 60°C water (five
minutes) and concluded with vortexing (approximately two minutes). After freeze-thaw-
vortex processing, the solutions were immediately extruded once through two
polycarbonate membrane filters with 400 nm pore diameter using a mini-extruder (Avanti
Polar Lipids Inc.). The initially extruded vesicle suspension was subsequently extruded
eleven times through two polycarbonate membranes (Avanti Polar Lipids Inc.), each with

a nominal pore diameter of 100 nm. The resulting solution contained vesicles with a size

72

absu'ln'rce ariasim

o
O

0 ° 0.6-
0.4-
0.2~
1.

0.01114JLIAIRLAAI [LJJJ

450 5(1) 550 6(1) 650
wavelength(nm)

DMPC Rhodamine-PE

 

.é‘
m
c:
8
E
“U
“:9
E“
O
:

 

 

Figure 4.1. Left: Structures of the phospholipid DMPC and the chromophore-
containing phospholipid Rhodamine-PE. Right: Solution phase absorption and
emission spectra of Rhodamine-PE in ethanol. Spectra have been normalized for
clarity of presentation.

73

 

Figure 4.2. TEM images of vesicles produced by extrusion and used in this work.
Top pane: lmage of an isolated unilamellar vesicles of ca. 150 nm in diameter.
Bottom pane: Image of multiple vesicles. The average diameter of vesicles was
determined to be 84 d: 22 nm based on measurements made from this and other
TEM images.

74

distribution centered around 85 nm diameter15 '20 with a standard deviation of 20 nm
(Figure 4.2).

T ime-resolved fluorescence measurements. Time-domain polarized fluorescence
intensity decays were acquired using a time-correlated single-photon counting (TCSPC)
system. This system has been described in detail elsewhere, and we briefly recap its
salient features here.21’22 The light source 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 harmonic of the Nd:YAG laser output is used to excite a cavity-dumped dye laser
(Coherent 702-2) operated with Pyromethene 567 dye (Exciton) at ca. 550 nm. The
output of this laser is ca. 75 mW average power, with 5 ps pulses at 4 MHz repetition
rate. Sample fluorescence is detected with a microchannel plate - photomultiplier tube
(MCP-PMT, Hamamatsu R3 809U). The electronics used to temporally resolve the
fluorescence transients included constant fraction discriminators for the signal and
reference channels (CFD, Tennelec 454), and a time-to-amplitude converter/biased
amplifier (TAC, Tennelec 864). The collection wavelength and polarization were
computer controlled using LabVIEW® 7.1 code. The fluorescence lifetime data were
collected at 54.7° with respect to the vertical excitation polarization, while the
reorientation data were collected at polarizations parallel (0°) and perpendicular (90°) to
the vertically-polarized incident light. The temperature of the sample was regulated to i
0.1°C using a water-circulating bath (N eslab RTE-110) connected to a cooling jacket that
holds the sample cuvette. All samples were allowed to equilibrate thermally for ten

minutes prior to data acquisition.

75

Steady state spectroscopy. Steady state excitation and emission spectra were
acquired for our vesicle samples for the purpose of characterizing the rhodamine band
positions. We used a Spex F luorolog 3 emission spectrometer for all measurements, set

to a spectral bandpass of 3 nm for both excitation and emission monochromators.

76

Results and Discussion

The primary goal of this work is to understand the extent to which the presence of
a tethered chromophore in a bilayer structure perturbs the organization of the bilayer.
Clearly, this is an issue that can be addressed to a limited extent using spectroscopic
methods, owing to the molecular length scale over which a fluorophore senses its local
environment. In other words, regardless of chromophore concentration, the local
enviromnent sensed by the chromophore will be perturbed by the presence of the probe.
This is, of course, a generic concern for any optical probe measurement and there is a
significant body of knowledge that has been gained on bilayer and other structures using
optical spectroscopic methods despite this intrinsic limitation. The concern we address in
this work is not related to the local perturbation induced by the probe, but the longer
length-scale disorder introduced to the bilayer by the presence of the chromophore. To
address this issue we focus on the concentration-dependence of the tethered chromophore
Rhodamine-PE in DMPC bilayers. We have measured the steady state and time-resolved
fluorescence behavior of this chromophore incorporated into unilamellar vesicles of ca.
100 nm diameter, at three different chromophore concentrations; 0.0025 mol%, 0.025
mol% and 0.25 mol%. Our data reveal a change in the organization of the bilayer
structure and a change in the local dynamics experienced by the Rhodamine-PE
chromophore as a function of concentration.

As we present our data on the concentration-dependence of the Rhodamine-PE
spectral and dynamical behavior, we also report these data as a function of temperature.

The reason for reporting the temperature-dependence of these data is that DMPC is

77

known to exhibit a gel-to-fluid phase transition at 24°C, and we are interested in
determining whether or not this structural transition in the acyl chain region of the
bilayers can be sensed by a chromophore that is tethered to the headgroup region. While
we clearly do not expect an effect as dramatic as was reported for perylene in the bilayer
acyl chains, the organization of a bilayer represents a balance of interactions between
headgroup and acyl chain regions, and it is of fundamental interest to understand if such
interactions can be sensed by a headgroup-bound chromophore.

Steady state spectroscopy. We have measured the steady state emission behavior
of Rhodamine-PE in unilamellar vesicles as a function of chromophore concentration and
system temperature, and we show these results in Figure 4.3. There are two trends to
note in these data. The first is that for the lower two chromophore concentrations, 0.0025
mol% and 0.025 mol%, the emission spectra appear to be identical and are characterized
by a maximum at 587 nm. For the 0.25 mol% concentration vesicles, the emission
spectrum is shifted to the red by 5 nm and the band profile is different than for the lower
concentrations. For all Rhodamine-PE concentrations, there is no discernible change in
the steady state emission spectra with temperature, indicating that the optical response of
this chromophore is not sensitive to structural changes in the acyl chain region of the
bilayer to which they are tethered.

The spectral shift seen for the higher concentration Rhodamine-PE indicates that
this chromophore is interacting with neighboring chromophores. This is not a surprising
result based on geometric considerations alone. For vesicles containing 0.0025 mol%

Rhodamine-PE, the average separation between chromophores is ca. 160 nm, assuming a

78

 

 

1.0—

0.8—
3*
E
.E 0.6-
1:
G)
.51
Te
E
s 0.4—

0.2—

“ 288K
I 1 l r l r l r l r I r I

 

0.0 '
560 570 580 590 6(1) 610 620 630
wavelength (nm)

Figure 4.3. Steady state emission spectra of Rhodamine-PE in DMPC vesicles.
The chromophore concentrations are 0.0025 mol%, 0.025 mol% and 0.25 mol%,
as indicated. Temperature-dependence of the emission band profiles is also
indicated.

79

' measurements.

homogeneous distribution of chromophores and a phospholipid size of 64 Az/molecule23.
For vesicles containing 0.025 mol% Rhodamine-PE, the average separation between
chromophores is 50 nm, and for 0.25 mol% Rhodamine-PE, the average spacing is 16
nm. For the lower concentrations, it is reasonable to expect little or no discernible
chromophore-chromophore interaction, but for the 0.25 mol% system, the average
distance is sufficiently short that there will be significant chromophore-chromophore
interactions. By way of comparison, a distance of 16 nm is about twice what one would
expect for the F firster critical transfer radius, R0. We will return to a discussion of this
point because of the effect that donor-donor excitation transport can have on reorientation
24-26

In addition to the steady state spectroscopic data, we have performed time-
resolved lifetime and anisotropy decay measurements on Rhodamine-PE in DMPC
vesicles. The purpose of these measurements is to gain insight into the molecular scale
environment(s) of the chromophores. The interpretation of the experimental reorientation
data is somewhat more complex than the standard, solution phase treatment because the
rhodamine chromophore is attached to the polar lipid headgroups, a structural condition
that gives rise to an infinite time anisotropy. The magnitude of the infinite time
anisotropy depends on the extent to which the chromophore is constrained to move
within a fixed conic volume.”28 For such a system, there are two relevant temporal
regions in the data, and the information of interest is derived from the induced

orientational anisotropy decay function, R(t), in Eq. 4.1.

= [MU—11(1)
I"(t)+2Ii(t)

 

R(t) (4.1)

80

The experimental fluorescence transients, I”(t) and I L(t) (Figure 4.4a), are the emission
intensities polarized parallel and perpendicular with respect to the vertically polarized
incident excitation pulse.

The theoretical basis used to interpret fluorescence anisotropy data depends on the
local environment in which the chromophore is present. The hindered rotor model is
used when the motional freedom of a molecule is restricted, as is the case here?!"3 I This
restriction may be the result of incorporation of the chromophore within a confined area
such as a bilayer structure, or may occur when the chromophore is tethered to an
interface. The fact that the chromophore distribution is unable to re-randomize
completely following excitation results in a non-zero infinite-time anisotropy, R(oo). For
such systems, R(t) exhibits the functional form shown in Eq. 4.2,

R(t) = R(OO) + (R(O) - R(w))eXP(-t/THR) (42)
where R(oo) is the infinite time anisotropy and R(0) is the zero time anisotropy (Figure
4.4b). The term R(O) is determined by the angle between the excited and emitting
transition dipole moments of the chromophore, and R(OO) is a gauge of the orientational
confinement that the chromophore experiences. The term rm: is related to the
orientational relaxation of the chromophore within its confining volume, and is

approximated by Eq. 4.3,”28

_. 76:

1H,, — 240 (4.3)

 

where 00 is the semi-angle of the confining cone. The fimdamental quantity of interest is

Dw, the so-called wobbling diffusion constant. This quantity reflects the ability of the

81

tethered chromophore to move within its restricted environment. To evaluate Dw, we
need both 00 and cm. The time constant Ink is available from the experimental R(t) data,
and we show these data as a fimction of chromophore concentration and system
temperature in Figure 4.5a. The cone angle 00 is related to the limiting values R(O) and
R(oo), and what information can be extracted from these data. The semi-angle of the
conic confining volume of the chromophore is related to the zero- and infinite-time

anisotropies by Eq. 4.4,27

M y, —0 5(cose (1+coso )) (4 4)
R(O) "‘ ' o 0 °

We have calculated values for 00 (Eq. 4.4) as a function of chromophore concentration
and temperature (Figure 4.5b). We consider the temperature and concentration-
dependencies of these data individually.

The Ink data shown in Figure 4.5a do not appear to exhibit a temperature
dependence that is related to the known gel-to-fluid phase transition in DMPC at 24°C.
This is not a surprising result because the chromophore is not located within the acyl
chain region of the bilayer, where the structural phase transition is known to occur. It is
also possible that, at least for the higher Rhodamine-PE concentration, that the presence
of the chromophore is itself influencing the phase transition temperature. One interesting
feature of these data is the dependence of the recovered relaxation time In]: on the
chromophore concentration. There are several possible explanations for these data, but
the fact that the concentration-dependence is non—monotonic suggests that the explanation

is more complicated than a single effect, such as the effect of donor-donor excitation

82

transport, for example. It is also noteworthy that the time constants for orientational
relaxation also trend toward longer times with increasing temperature. This finding
outwardly appears counter-intuitive because the local viscosity of the bilayer
environment is expected to decrease with increasing temperature, an effect which could
give rise to shorter rHR values with increasing temperature, if the acyl chains played a
role in mediating the chromophore local environment. The other possible explanation for
such data is that the volume in which the chromophore can move is increasing in size
with increasing temperature, and for this situation the chromophore would require longer
to re-randomize its orientation within the larger conic volume. This is a testable
hypothesis, through the measurement of the restricting cone angle.

We show in Figure 4.5b the dependence of the cone semi-angle on chromophore
concentration and system temperature. There are several interesting trends in these data.
It is clear that the 0.0025 mol% Rhodamine-PE vesicles exhibit a smaller cone angle than
that of either the 0.025 mol% or 0.25 mol% Rhodamine-PE vesicles. These data, by
themselves, point to the role that the chromophore plays in mediating the organization of
the bilayer structures. For the higher probe concentrations, there is clearly less order in
the bilayer structure, but it is not clear from these data that the concentration of the probe,
above a certain point, changes the bilayer organization. As expected, the cone angles
increase with temperature in all cases, consistent with increasing thermal disorder in

these bilayers.

83

0.4

0.0

‘fi

1,0)

 

 

 

2000 8000

l
4000 6000

tirre (ps)

Figure 4.4. (a) Experimental polarized fluorescence transients for
Rhodamine-PE in 100 nm diameter DMPC vesicles. (b) Induced
orientational anisotropy function, R(t), generated from the data shown in

panel (a).

84

Neither the concentration nor the temperature dependence of 131-1R (Figure 4.5a) or
00 (Figure 4.5b) by themselves provide a clear picture of how the bilayer is changing as a
result of these two experimental variables. As noted above, the fundamental quantity of
interest is DW (Eq. 4.3), and with the Tag and 00 information extracted from the
experimental data, we can determine Dw as a firnction of chromophore concentration and
temperature (Figure 4.6). These data contain several interesting trends, and we consider
the concentration-dependence first.

It is clear from Figure 4.6 that Dw increases measurably with increasing
Rhodamine-PE concentration. As noted above, the quantity Dw reflects the ability of the
chromophore to move within its spatially restricted volume, and we ascribe this change to
the disorganization introduced to the bilayer structure by the presence of the
chromophore itself. Because the chromophore is connected to the lipid headgroup, the
notion of conical confinement is less clear, perhaps, than if the chromophore were
connected to the acyl chain. Because of the location of the chromophore, the species
providing the most confinement will be the phosphocholine headgroups adjacent to the
chromophore. As disorder increases in the acyl chain region of the bilayer, the
corresponding organization of the headgroup region will also change, giving rise to the
effect we detect through the chromophore concentration-dependence of Dw. It is clear
from these data that the presence of the chromophore does indeed perturb the

organization of the bilayer.

85

 

 

 

 

 

 

4000 a
' O 0.25 rml%
3500 o (Mfl5nd%s §
- a 0.0025 rml% § §
3000 § §
3‘ 2500 o o o
5% ' §
.4 2000 § §
1500
1W) 3 g I l r l
15 20 25 30 35
temperature (’0
40_ b
35 f

00 (degrees)
13

305

 

H}l FOHF—O—fi

1—0—1
mm

§§
41

 

 

 

 

 

I E
20: i Q
E o 0.25rml%
15_ 6 o 0.025mo1%
: :1 0.0025nDl%
10 ' 1 l l . I r 1
15 20 25 3O 35

temperature (’0

Figure 4.5. (a) Time constants (Eq. 4.2) extracted from anisotropy data for
Rhodamine-PE in DMPC vesicles as a function of chromophore
concentration and system temperature. (b) Confining cone semi-angle
measurements from the experimental data (Eq. 4.4).

86

 

 

 

 

 

 

 

 

80- t
. o O.25mol%
70.. O 0.025mol%
. , v a 0.0025mo|%
60- i
E 50'
E . r t t
o3 40-
‘ i
30- ._
» % i i i f
20..
10-
0 l l L L L
15 20 25 30 35

temperature (’0

Figure 4.6. Dependence of DW on chromophore concentration and system
temperature. The quantity Dw is calculated from Eq. 4.3 using the experimentally
derived quantities Ton and 00.

87

The second noteworthy finding is that Dw remains essentially constant with
temperature for the lower chromophore concentrations, but it decreases with increasing
temperature for 0.25 mol% Rhodamine-PE. This finding can be understood in terms of
the cooperative effects of acyl chain dynamics and headgroup dynamics in a bilayer
system. We have reported previously that perylene reorientation within the nonpolar
region of DMPC bilayers manifests an increase in D with increasing temperature.5 It is
clear from the perylene data that increasing the temperature of the bilayer structure makes
the acyl regions of the bilayers less viscous, thus increasing the mobility of chemical
species in the nonpolar region. Because the headgroups are covalently bound to the
aliphatic tails, we assert that the mobility of the headgroups must, of necessity,
experience a change in (translational) mobility as well. The fact that we observe
experimentally a constant or decreasing chromophore DW with increasing temperature is
consistent with a limited number of physical conditions for our bilayers.

For the low chromophore concentrations, where no temperature dependence is
seen in Dw, we understand this behavior in the context of our orientational anisotropy
measurements not being sensitive to translational diffusion phenomena. For the lower
chromophore concentrations, there is a sufficiently low chromophore density that under
all conditions the chromophores respond as individual entities. The reason that DW
exhibits little or no change with temperature for these systems is that the chromophore is
in an environment dominated by the lipid headgroup region and the water overlayer.
This region of the bilayer system is not expected to change significantly over the

temperature range we examined, and. this expectation is supported by the functional form

88

 

of our data. For the highest chromophore concentration, where the chromophores are, on
average, 16 nm apart, it is possible that interactions between chromophores would give
rise to slower motion, consistent with our findings. In addition to the temperature-
dependent change in dynamics, such chromophore-chromophore interactions would
produce changes in spectroscopic band shape, and we observe this effect clearly (Figure
4.3). While it may be tempting, based on the translational diffusion coefficients derived
for our perylene data,5 to estimate the extent or frequency of chromophore-chromophore
interactions, we refrain from making this calculation because we do not have any
independent means of estimating the energetics of this process.

It is also important to recognize that, for these measurements, donor-donor
excitation transport can contribute to the functional form of the anisotropy decay for high
chromophore concentrations.24'26 The mechanism of such excitation transport is dipolar
coupling, with the length scale over which this process operates being similar to that for
the more widely considered Forster donor-acceptor energy transfer. The chromophores
are nominally separated by ca. two Fdrster critical radii at the highest concentration, and
the signature of this process is the loss of polarization with time as the hopping
excitation(s) sample the orientational distribution present in the system. The separation
of motional and excitation transport processes is often difficult, relying on measuring the
concentration-dependence of the anisotropy decay over the appropriate concentration
range. We have not mapped out the concentration-dependence of R(t) for this system
with sufficient resolution to cleanly separate these two processes. The fact that our data

for the highest chromophore concentration does manifest a temperature-dependence

89

where the lower concentrations do not suggests that there is at least some contribution
from donor-donor transport in this system, and this is a matter which is currently under

investigation.

90

Conclusions

We have found that the steady state and time-resolved fluorescence behavior of
the chromophore Rhodamine-PE in DMPC bilayers can be used to interrogate
organization within the bilayer system. A long-standing issue with any fluorescent probe
method is the extent of the perturbation introduced to the system by the probe itself. We
find direct evidence for such perturbations with the chromophore concentration and
system temperature dependence of the steady state and time-resolved fluorescence of
Rhodamine-PE in DMPC. Specifically, we find that the emission spectral profiles of this
chromophore depend on the chromophore concentration, with a red-shifted emission
spectrum seen for the highest chromophore concentration. Such a spectral shift is
consistent with chromophore-chromophore interactions. The findings of the steady state
data are supported by the time-resolved experimental results. We recover concentration-
and temperature-dependent anisotropy decay dynamics for this chromophore, with no
apparent simple relationship between the measured decay time constant and the
properties of the bilayer system. As has been seen previously for solution phase systems,
the anisotropy decay time constant can vary for reasons not related to changes in the
rotational diffusion constant.32 In this case, where we use the hindered rotor model to

2738 we find that it is also important to account for changes

interpret our dynamical data,
in the extent to which the chromophore is restricted, and this is manifested in the
confining cone semi-angle, 90. This system property also does not, by itself provide full

insight into the bilayer system. When both of these experimentally-derived quantities are

considered, and the fundamental quantity DW is extracted, an intuitive physical picture

91

emerges. As the chromophore concentration is increased, DW increases as a result of
disorganization within both the head group and acyl chain region of the bilayer structure.
For 0.25 mol% Rhodamine-PE we observe a decrease in Dw with increasing temperature,
a phenomenon consistent with chromophore-chromophore interactions which serve to
slow the motion of the individual chromophores. This interpretation of the data is
consistent with the steady state emission spectra, which show significant inter-
chromophore interactions. Extracting the relative contribution to DW from excitation
transport for this system is a difficult task, and current work is under way to elucidate the
role of this process more clearly. The larger significance of this work, however, is the
understanding that low chromophore concentrations must be used in fluorescence probe

measurements if the properties of the bilayer itself are to be interrogated.

92

Literature Cited

(1) Bagatolli, L. A.; Gratton, E. Biophysical Journal 2000, 78, 290.

(2) Pabst, G.; Amenitsch, H.; Kharakoz, D. P.; Laggner, P.; Rappolt, M.
Physical Review E: Statistical, Nonlinear, and Soft Matter Physics 2004, 70, 021908/1.

(3) Stevens, B. C.; Ha, T. Journal of Chemical Physics 2004, 120, 3030.

(4) Nagle, J. F. Annual Review of Physical Chemistry 1980, 31, 157.

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

(6) Oraedd, G.; Lindblom, G. Biophysical Journal 2004, 87, 980.

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

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

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

(10) Feng, Z. V.; Spurlin, T. A. ; Gewirth, A. A. Biophysical Journal 2005, 88,
2154.

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

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

93

(13) Dela Cruz, J. L.; Blanchard, G. J. Journal of Physical Chemistry A 2001,
105, 9328.

(14) Dela Cruz, J. L.; Blanchard, G. J. Journal of Physical Chemistry A 2002,
106, 10718.

(15) Mayer, L. D.; Hope, M. J .; Cullis, P. R. Biochimica et Biophysica Acta
1986, 858, 161.

(16) Hunter, D. G.; Frisken, B. J. Biophysical Journal 1998, 74, 2996.

(17) MacDonald, R. C.; MacDonald, R. 1.; Menco, B. P.; Takeshita, K.;
Subbarao, N. K.; Hu, L. R. Biochimica et Biophysica Acta 1991, 1061, 297.

(18) Driessen, A. J .; van den Hooven, H. W.; Kuiper, W.; van de Kamp, M.;
Sahl, H. G.; Konings, R. N.; Konings, W. N. Biochemistry 1995, 34, 1606.

(19) Unger, E. C.; MacDougall, P.; Cullis, P.; Tilcock, C. Magnetic Resonance
Imaging 1989, 7, 417.

(20) Subbarao, N. K.; MacDonald, R. 1.; Takeshita, K.; MacDonald, R. C.
Biochimica et Biophysica Acta 1991, 1063, 147.

(21) DelaCruz, J. L.; Blanchard, G. J. Journal of Physical Chemistry B 2003,
107, 7102.

(22) 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.

(23) Lague, P.; Zuckermann, M. J .; Roux, B. Biophysical Journal 2001, 81,
276.

(24) Gochanour, C. R.; Fayer, M. D. Journal of Physical Chemistry 1981, 85,
1989.

94

 

461.

289.

(25)

(26)

(27)

(28)

(29)

(30)

(31)

(32)

Ediger, M. D.; Fayer, M. D. Journal of Chemical Physics 1983, 78, 2518.

Baumann, J .; F ayer, M. D. Journal of Chemical Physics 1986, 85, 4087.

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

Szabo, A. Journal of Chemical Physics 1984, 81, 150.

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

Kinosita, K., Jr.; Ikegami, A.; Kawato, S. Biophysical Journal 1982, 37,

Kinosita, K., Jr.; Kawato, S.; Ikegami, A. Biophysical Journal 1977, 20,

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

95

Chapter 5

GAUGING THE EFFECT OF IMPURITIES ON LIPID BILAYER PHASE
TRANSITION TEMPERATURE

Introduction

Phospholipid bilayers have attracted a great deal of interest because of their
central role in cellular function and the potential for the use of model bilayer systems in
chemical and biological sensing applications. As our understanding of phospholipid
bilayers has developed, the structural and compositional complexity of this family of
molecular assemblies has become more apparent. It is now known that natural bilayer
structures contain 100 or more constituents,l and for simple model systems comprised of
as few as two or three different species, phase separations have been observed.“5 While
phase segregation remains to be seen for biologically-derived bilayers, it is likely that
these compositionally heterogeneous structures are characterized by small, fluid domains
that are beneath our ability to resolve. Our interest in phospholipid bilayers lies in their
potential utility as supported bilayer membranes, a structural motif useful in the creation
of selective chemical sensing systems based on biological molecules such as enzymes
and trans-membrane proteins. We are investigating the fluidity and phase segregation
behavior of model bilayer structures in an effort to optimize these films for incorporation
of selected biomolecules in their active form. We focus in this work on the composition-
dependence of bilayer fluidity.

Phospholipids self-assemble spontaneously to form a bilayer structure with the

96

 

hydrophobic tails directed toward the center, and the head groups exposed to the aqueous
phase. Phase transitions have been seen in a variety of bilayers, and it is thought that at
the so-called gel-to-fluid melting transition temperature, Tm, the hydrocarbon tails
undergo a change from being predominantly trans to a structural configuration
characterized by a significant number of trans-gauche conformers.7 The characteristic
phase transition temperature (Tm) for a phospholipid depends sensitively on the aliphatic _
chain length and the presence and location of any unsaturations in the aliphatic chain(s), l
and such phase transitions have been detected by a number of techniques, ranging from

calorimetry8 to fluorescence spectroscopy9 and spin labeling.lo For temperatures below 1

 

Tm, the lipid aliphatic tails are said to exist in the gel phase, a comparatively well-ordered
state where translational mobility of the bilayer constituents is thought to be limited.
Above the phase transition temperature, the phospholipid aliphatic tails exist in a more
disordered fluid phase,11 characterized by more motional freedom and consequently,
more lateral freedom. Lipid bilayers may exist in the gel phase or the fluid phase
depending on the temperature, pressure, extent of hydration, and the structural identity of
the phospholipids.12 This phase transition has received a great deal of research attention,
both experimental and theoretical, because the phase transition temperature is related to
the energetics of lipid-lipid interactions within the bilayer, and phospholipid phase
transitions are considered to be important in regulating the activities of membrane-
associated proteins. 13 ’ '4

There is a significant body of literature focused on measuring Tm for selected

phospholipid bilayer systems.”25 We have undertaken the present study for two reasons.

97

The first is to evaluate the utility of fluorescent probe molecules sequestered within the
nonpolar regions of lipid bilayers to detect phase transitions. The second purpose of this
work is to evaluate the effect of impurities within the bilayers on the characteristic Tm for
the resulting system. Both of these issues are of great potential importance because of the
information content of the fluorescence experiments in addition to the phase transition
temperature. While there has been a great deal of work focused on understanding
structural phase transitions in bilayer systems, there remains the issue of the characteristic
length scale over which the change in organization at the phase transition is seen. By
working with systems containing controlled amounts of impurities, and comparing
microscopic viscosity data to phase transition temperatures, we can assess whether or not
the change in organization associated with a phase transition is reflected accurately by
molecular-scale changes in the nonpolar region of the bilayer on the molecular scale.

For these experiments we use the chromophore perylene. Perylene is a planar
polycyclic aromatic hydrocarbon with a well characterized linear optical response. Its
nonpolar nature makes it well suited to incorporation within the acyl chain region of lipid
bilayers. In this work, we focus on the rotational motion of perylene within the bilayer
nonpolar region as a function of temperature. Not only is rotational diffusion more
sensitive to local environment than fluorescence lifetime, but the details of the rotational
motion provide a gauge of the bilayer viscosity. While perylene has been used
previously in the study of bilayer systems, our direct picosecond time-resolved
measurements afford more detail than has been available from either steady state or

phase—resolved fluorescence measurments.26’27

98

 

The phase transition behavior of phospholipids bilayers from the gel phase to the
fluid phase has been studied in great detail previously.15'25’28 The phase transition
temperatures of many phospholipids are well-docmnented, especially for those
phospholipids containing no unsaturations. For some systems, even more specific
information is known, such as changes in partial specific molar volume, for example.28
However, there are comparatively fewer reports on the effect of “impurities” on Tm for a .
phospholipid bilayer. The work we report here, using 1,2-dimyristoyl-sn- F
phosphatidylcholine (DMPC) and 1,2-dimyristoleoyl-sn-glycero-3-phosphocholine (14:1

PC) as the phospholipid and “impurity”, respectively, demonstrate that even well ordered

 

phospholipid bilayers are characterized by a viscosity of 14.5 d: 2.5 cP in the gel phase
and 8.5 :1: 1.5 cP in the fluid phase. When 14:1 PC is introduced to a DMPC bilayer, we
observe a dramatic decrease in the gel-to-fluid transition temperature from 24°C for the
pure DMPC bilayer, with a gradual approach to a new phase transition temperature of ca.
4°C for concentrations of 14:1 PC on the order of 1 mol%. These findings point to the
relatively small amount of impurity required to disrupt the phase transition temperature of
a bilayer, even though the local viscosity of the bilayers did not sense a corresponding

change in order.

99

Experimental

Vesicle Preparation. Phospholipids l,2-dimyristoyl-sn-phosphatidylcholine
(DMPC) and l,2-dimyristoleoyl-sn-glycero-3-phosphocholine (14:1 PC) were purchased
from Avanti Polar Lipids Inc. (Alabaster, AL) and used as received. Both lipids are
>99% pure, with the details of the analyses being provided on Avanti’s web site
(www.avantilipids.com). Perylene was purchased from Sigma-Aldrich (Milwaukee, WI)
and used without further purification. For each sample, the lipids and the perylene probe
were mixed in selected ratios so that the final concentrations of the samples were 2.9 x
10‘4 M lipid and l x 10'5 M perylene. The chloroform solvent was evaporated and an
appropriate volume of Tris® buffer (Sigma-Aldrich) was added to each mixture so that
the lipid concentration was 1 mg/mL. The buffer (10 mM, pH 7.8) was prepared using
purified water from a Milli-Q Plus water purification system (Millipore, Bedford, MA).
The mixtures were processed five times through a freeze-thaw-vortex cycle to ensure
complete mixing of the constituents. Each cycle was comprised of freezing the solution
by immersion in liquid nitrogen for five minutes, followed by thawing via immersion in
60°C water (five minutes) and concluded with vortexing (approximately two minutes).
After freeze-thaw-vortex processing, the solutions were extruded once through two
polycarbonate membrane filters with 400 nm pore diameter using a mini-extruder (Avanti
Polar Lipids Inc.). The initially extruded vesicle suspension was then extruded eleven
times through two polycarbonate membranes (Avanti Polar Lipids Inc.) with a nominal

pore diameter of 100 nm. Extrusions were performed at room temperature. The resulting

100

solution contained vesicles with a size distribution centered around 100 nm diameter
(Figure 5.1).29'34

Time-resolved fluorescence measurements. Time-domain polarized fluorescence
intensity decays were acquired over a range of temperatures using a time-correlated
single-photon counting (TCSPC) system. This system has been described in detail
elsewhere, and we briefly recap its salient features here.”’36 The light source 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 third harmonic of the Nd:YAG laser output is used
to excite a cavity-dumped dye laser (Coherent 702-2) operated with Stilbene 420 dye
(Exciton) at ca. 430 nm. The output of this laser is ca. 25 mW average power, with 5 ps
pulses at a 4 MHz repetition rate. Sample fluorescence is detected with a microchannel
plate — photomultiplier tube (MCP-PMT, Hamamatsu R3 809U) and the electronics used
to temporally resolve the fluorescence transients were a constant fraction discriminator
(CFD, Tennelec 454) and a time-to-amplitude converter/biased amplifier (TAC, Tennelec
864). The collection wavelength and polarization were computer controlled using
LabVIEW® 7.0 code. The fluorescence lifetime data were collected at 54.7° with respect
to the vertical excitation polarization, while the reorientation data were collected at
polarizations parallel (0°) and perpendicular (90°) to the vertically-polarized incident
light. The temperature of the sample was regulated to :1: 0.1°C using a water-circulating
bath (N eslab RTE-110) connected to a cooling jacket that holds the sample cuvette. All

samples were allowed to thermally equilibrate for ten minutes prior to data acquisition.

101

 

Figure 5.1. Transmission electron microscope images of a unilamellar
vesicle used in this work. The scale bar for both images is shown in the
lower right corner.

102

Steady state spectroscopy. Steady state excitation and emission spectra (not
shown) were acquired for our vesicle samples for the purpose of characterizing the
perylene band positions. We used a Spex Fluorolog 3 emission spectrometer for all
measurements, set to a spectral bandpass of 2 nm for both excitation and emission

monochromators.

103

Results and Discussion

As noted above, there are two major goals of this work. The first is to
demonstrate the feasibility of using a chromophore imbedded in the non-polar region of a
lipid bilayer structure to sense the gel-to-fluid phase transition temperature. The second
purpose of this work is to achieve a more quantitative understanding of the effect that an
impurity has on the phase transition temperature of vesicles comprised primarily of the
phospholipid DMPC. The chosen “impurity” for this series of experiments is 1,2-
dimyristoleoyl-sn-glycero-3-phosphocholine, a l4-carbon phospholipid that closely
resembles DMPC, but contains one cis-unsaturation at the 9- position in each acyl chain
(14:1 PC, Figure 5.2). This choice of “impurity” was intended to introduce irregularities
in the bilayer structures in a relatively well controlled manner. We consider the utility of
perylene as an optical probe of bilayer phase transitions first.

The gel-to-fluid transition temperature of pure DMPC has been established to be
24°C}7 and we find that perylene exhibits a discontinuous change in its rotational
diffusion dynamics in DMPC unilamellar vesicles at this temperature (Figure 5.3).
Phenomenologically, perylene functions as a phase transition probe likely because the
chromophore is sequestered within the portion of the bilayer structure that is thought to
undergo the greatest structural change at the point of phase transition from the gel to the
fluid phases. It is fair to consider that the perylene itself may be responsible for some
perturbation of the bilayer structure. While this possibility must be considered for any

fluorescent probe experiment, we believe that the perturbation imposed on the bilayer by

104

 

\ |+
2 N
O\ O\
O H O H
K ‘0
o 0 0 0
O O
l perylene
I
DMPC 14:1PC

Figure 5.2. Structures of DMPC, 14:1 PC and perylene.

105

1100f

T0R (ps)

700- i

 

5m 1 I n l l I n l 1 I n l
18 20 22 24 26 28 30

temperature(OC)

 

Figure 5.3. Reorientation time of perylene in unilamellar vesicles comprised of
DMPC as a function of temperature. There is a discontinuous change in fan at
the gel-to-fluid phase transition temperature. Error bars are £10 for at least
four individual determinations at each temperature.

106

perylene is modest. Our reasons for this assertion are that, with perylene, we observe the
same phase transition temperature for DMPC that has been measured by other
methods.37 With the addition of relatively small amounts of 14:1 PC, we observe
changes in the phase transition temperature for the bilayer system. Thus, even if the
perylene does perturb the bilayer structure, it is constant for all of our measurements and
the changes in phase transition temperature associated with the introduction of DMPC are
manifested by large changes in the phase transition temperature. Apparently, the planar
structure of the perylene molecule can be accommodated within the nonpolar bilayer
region with less disruption than the presence of cis defects in the chains themselves.

In addition to the ability of perylene to sense the phase transition temperature of
DMPC, the measurements we perform provide other valuable information on the
nonpolar interior region of these bilayers. We determine transient fluorescence

anisotropy dynamics through the induced optical anisotropy function, R(t) (Eq. 5.1)

R(t)-r Mil-11(1)

" I" (t)+ 21,0) (5'1)

 

where 1H (t) and I i (t) are the emission intensities polarized both parallel and

perpendicular to the incident vertically polarized excitation pulse. The chemical
information contained in these experiments resides in the decay functionality of R(t), and
this information differs for chromophores that exist in free solution versus those confined
within a membrane structure. In general, a chromophore can produce an anisotropy
decay with up to five exponential components, but owing to symmetry considerations and

experimental signal-to-noise limitations, it is highly unusual to observe more than two

107

decay components in R(t). These decay components can be related to the Cartesian
components of the rotational diffusion constant, D, and to the orientation(s) of the excited
and emitting transition dipole moments relative to the molecular structure.

The reorientation dynamics of perylene are well understood in a variety of
environments. The excited and emitting transition dipole moments of perylene are
known to reside along its long in—plane axis, which is typically assigned as the x-axis. In
this scheme, the short in-plane axis is designated y and the z-axis is perpendicular to the
chromophore “rt-system plane. For this molecule, a single exponential anisotropy decay is
taken to indicate perylene reorienting as a prolate rotor (Dx > Dy = D2) and a two-
component anisotropy decay is indicative of reorientation as an oblate rotor (Dz > D)( =
Dy). We find experimentally that perylene reorients as a prolate rotor, rendering the
recovered time constant TOR = 6Dz".38 It is significant for several reasons that we recover
a single exponential anisotropy decay for all measurements. First, our earlier work on
perylene in n-alkane bulk solvents revealed a change in the effective rotor shape of the
chromophore as the length of the alkane exceeded the length of the perylene long axis.38
The fact that we do not observe this same change suggests that the perylene may reside in
a region that is not spanned continuously by discrete aliphatic chains. Such a finding is
consistent with perylene residing near the center of the nonpolar region of the bilayer.
The second reason that this finding is significant is that the reorientation time constants
both above and below the phase transition are directly comparable to one another,
allowing the straightforward extraction of local viscosity information (vide infra). While

the use of perylene as a probe of bilayer microviscosity has been questioned before,27 we

108

believe that perylene is reporting accurately on microscale changes in analogous
molecular environments across the gel-to-fluid phase transitions in these bilayers, and
that on both sides of the phase transition, the probe sweeps out the same ellipsoidal
volume of rotation. At the very least, the relative changes in local environment across the
bilayer phase transition are reflected accurately in the perylene reorientation data. We
consider next the relationship between the recovered reorientation time constant and the
local environment of the chromophore.

The modified Debye-Stokes-Einstein equation relates the measured reorientation

time to the quantities of interest.”41

V
r... = g; (5.2)
B

This model has been discussed extensively before, and we will not attempt a detailed
review here. The quantities of interest here are n, the viscosity of the medium
surrounding the reorienting chromophore, V is the solute hydrodynamic volume, f is a
solvent-solute frictional interaction coefficient, and S is a shape factor to account for the
non-spherical shape of the reorienting chromophore. This model has been shown to be in
reasonable agreement with experimental data for a wide variety of systems. While this
model assumes a continuum medium surrounding the reorienting moiety, which is clearly
not an accurate description because the molecular motion we sense is averaged over
many configurations of the local environment, the approximation of a continuum is
phenomenologically reasonable. Using the calculated hydrodynamic volume of perylene
(225 A3) and the experimental reorientation data, we can extract information on the

viscosity of the lipid bilayers. These data lie in the range of ca. 14.5 cP for the gel phase
109

and ca. 8.5 cP for the fluid phase (Table 5.1). These values are broad averages over a
range of experimental conditions, but they are nonetheless important because they
provide significant insight into the fluidity of these bilayer structures, and the influence
that structural disorder has on this property.

To place these values in context, there has been a good deal of interest in the
viscosity of bilayers because this property is a gauge of membrane fluidity and thus the
mobility of membrane constituents. Our understanding of bilayer structures has evolved
substantially over time, and with it has come a range of estimates for the viscosity of
bilayers. Initially it was thought that bilayers were characterized by a viscosity of ca. 100
cP,l with subsequent estimates being revised downward. A recent estimate of plasma
membrane viscosity was that it was similar to “crocodile fat on a warm summer’s day.”1
Absent accurate viscosity data on warm crocodile fat, we chose to quantitate the viscosity
of the bilayer structures directly. The range of Viscosities we recover from our
measurements are in a range one would expect for either a long chain alkanol such as
nonanol or decanol, (8.7 — 10.5 CF)42 or for an extensively hydrogen-bonded liquid such
as ethylene glycol (ca. 16 CF).43 Our findings, based on the fluorescence anisotropy data
we report here, are consonant with other literature reports. X-ray diffraction experiments
“‘45 and simulation studiesll have shown that phospholipids in the gel phase pack less
tightly than a crystal, but still exhibit short-range order. In the gel phase, the lipids are

situated in a nominally hexagonal lattice pattern and the fatty acid chains are extended

fully.ll Our use of a nonpolar chromophore places the probe in a location ideally suited

110

Table 5.1. Viscosities of DMPC unilamellar vesicles as a function of 14:1 PC

 

 

concentration.
Gel Phase Fluid Phase
Mol % TOR (ps) T (K) 11 (cP) TOR (ps) T (K) n (cP)
14:1 PC
0.0 825 i 18 296 15.0 d: 0.3 638 i 23 298 9.0 i: 0.1
0.3 646 i 38 282 11.9 i 0.7 408 :t 67 283 7.1 i 1.2
0.7 850 d: 43 277 14.5 i 0.7 453 i 37 278 7.7 i 0.6
1.0 767 i 73 277 13.0 i 1.2 588 :1: 60 278 10.0 i 1.0
1.5 1061 i 26 271 17.6 dc 0.4 539 :1: 5 273 9.0 :t 0.1

111

to the detection and characterization of changes in bilayer structure associated with the
phase transition.

With the ability to detect phase transitions in bilayer structures using perylene as
an optical “sensor,” we turn our attention to the effect of impurities on the phase
transition temperature and local organization of DMPC bilayers. To evaluate this issue,
we have made samples containing comparatively small amounts of 14:1 PC as an
impurity (0.3 mol%, 0.7 mol%, 1.0 mol% and 1.5 mol%), and have measured the
transition temperatures of the vesicle solutions using perylene reorientation. We show
the temperature-dependent reorientation data in graphical form in Figures 5.4 and 5.5, for
DMPC vesicles containing varying amounts of 14:1 PC. We find that there is a regular
progression in Tm, decreasing with increasing amounts of 14:1 PC (Figure 5.6). For 0.3
mol% 14:1 PC, we observe a 15°C decrease in Tm, indicating that even relatively small
amounts of a structural impurity can give rise to marked changes in the organization of
bilayers. We note that experiments on bilayers comprised of 5 mol% 14:1 PC did not
yield a phase transition temperature in the range we can access (as low as -15°C). Given
that as few as three molecules per thousand can disrupt the organization of a bilayer
substantially, it is fair to question whether or not the viscosity of the bilayers changes
with the addition of 14:1 PC. As shown in Table 5.1, we recover Viscosities that are
possibly slightly different than for vesicles of pure DMPC, but the fact that we are
measuring the reorientation times at substantially different temperatures makes a direct
comparison of viscosity values for the different vesicles difficult. That said, for all of the

viscosity data we report, the fluid phase Viscosities lie in the range of 8.5 :I: 1.5 cP and the

112

gel phase Viscosities are in the range of 14.5 i 2.5 cP. These are measurably different
values for the two phases, but there appears to be no discernible trend in either phase with
respect to the amount of 14:1 PC present. Thus, even though the longer range
organization that is characterized by Tm is compromised by the presence of 14:1 PC, the
local environment(s) sensed by perylene in either the gel or fluid phases is not influenced

measurably by the presence of the structural impurity.

113

1000

I... (p8)
r—O—t
r—O—t
r—O——I
I—O—i

 

 

Temperature (’0

lOOOr

10R (p8)

[fir
l—O—‘l

 

 

Temperature (’0

Figure 5.4. Reorientation time of perylene in DMPC unilamellar vesicles
containing (a) 0.3 mol% 14:1 PC and (b) 0.7 mol% 14:1 PC as a function of
temperature. Error bars are i 16 for at least four individual determinations
at each temperature.

114

1100- a

101)

l
t—O——1

TOR (ps)
. , . .
r——o—4

 

 

 

 

4“) I l I 1 l l
-2 O 2 4 6 8 10
Temperature (’0
11(1) -
. § i b
lCXX) -
9(1) -
1;; 8(1) -
3 .
+9 700 -
6G) -
' o
500 —
4m 1 n l 1 J g l l l l l
-6 -4 -2 0 2 4
Temperature (’0

Figure 5.5. Reorientation time of perylene in DMPC unilamellar vesicles containing
(a) 1.0 mol% 14:] PC and (b) 1.5 mol% 14:1 PC as a function of temperature.
Error bars are :1: 10 for at least four individual determinations at each temperature.

115

 

25-

.3 i

0)

"‘ 20-

8

8.

515—

8

.5 L

3:310-

% . i

I:

:2 5-

e l i

é 0

a j i
_51.L%1.1Lu.11.1¥44

 

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
mol%14zlPC

Figure 5.6. Variation of measured Tm in DMPC unilamellar vesicles as a function of
14:1 PC concentration.

116

The perylene reorientation data provide information on the effective local
viscosity of the DMPC bilayer. This information can be used to estimate the translational

diffusion coefficient, DT, of perylene in these systems according to the Stokes-Einstein

 

equation,46
Dr = 6’93; (5.3)
It 77

Where R is the radius of the diffusing species and n is the viscosity. To estimate R, we
use the hydrodynamic volume of perylene (225 A3) and presume a spherical shape,
yielding R = 3.77 A. We present the calculated values of DT for our systems in Table
5.2. The data show that for all of these systems, there is an increase of a factor of ca. 1.5
in DT across the phase transition, and that, even with the addition of 14:1 PC, DT changes
only modestly. It is, of course, important to consider how our values of DT (~40 umZ/s
below Tm, ~65 umz/s above Tm) compare to other reports of DT. Our values for Dr are in
excellent qualitative agreement with those reported for tethered fluorophores in
phospholipid monolayers, where DT values between 15 and 110 umZ/s were reported,
depending on the compression applied to the phospholipid LB monolayer.47

We provide the calculated DT values in Table 5.2 with a couple of caveats. The
first is that our bilayer structures could be a heterogeneous system, because we do not
know the extent to which the 14:1 PC tends to aggregate within the bilayers. In this limit,
the translational diffusion coefficient we report would not be an accurate representation
of the frustrated translational diffusion that would occur in a heterogeneous system. The

other caution is that both Eqs. 5.2 and 5.3 were derived for a solute contained within a

117

Table 5.2. Calculated translational diffusion coefficients for perylene in DMPC
unilamellar vesicles as a function of 14:1 PC concentration. Viscosities were
determined from TOR data with use of Eq. 5.2, and DT values were calculated

according to Eq. 5.3.

 

 

Gel Phase Fluid Phase
Mol % T (K) 1‘1 (cP) D1 T (K) '1 (cP) Dr
14:1 PC
(”ml/S) (umzls)
0.0 296 15.0 :t 0.3 38 d: 1 298 9.0 :1: 0.1 64 :1: 1
0.3 282 11.9i0.7 46:1:3 283 7.1 :1: 1.2 77:1:11
0.7 277 14.5 d: 0.7 37 d: 2 278 7.7 :t 0.6 70 i 5
1.0 277 13.0i1.2 41i4 278 10.0i1.0 54i5
1.5 271 17.6 :t 0.4 30 d: 1 273 9.0 d: 0.1 59 :1: 1

118

continuum solvent, and for rotational diffusion measurements, there can be significant
deviation from model predictions for cases where the solute and solvent molecules
become similar in size. This deviation reflects a complex solvent-solute frictional
interaction that must involve molecular-scale steric and dipolar processes. The
microscale Viscosities we infer from the reorientation data could, in principle, be off by a
scaling constant, rendering our DT values uniformly high or low, but the scaling constant
should be essentially the same for all measurements. The change in DT across the phase
transition is thus reflected accurately in our reported results. With this frame of
reference, we believe that the relatively modest changes in DT across the phase transition
for our systems is more consistent with a change in the extent of organization within the

bilayer than with any substantial changes in free volume within the bilayer.

119

Conclusions

We have evaluated the utility of fluorescence anisotropy as a means of
characterizing gel-to-fluid phase transitions in phospholipid bilayer structures. We find
that perylene reorientation is a sensitive measure of the bilayer phase transition and that
the incorporation of impurities has a substantial and measurable effect on this phase
transition temperature. We have incorporated perylene, a nonpolar polycyclic aromatic
hydrocarbon, into lipid bilayer structures consisting of pure DMPC and in bilayers
containing DMPC with 0.3 mol%, 0.7 mol%, 1.0 mol% and 1.5 mol% of 14:1 PC over a
range of temperatures surrounding the phase transition for these systems. Despite this
mesoscopic structural effect, the local environment of the perylene probe is not affected
substantially by the addition of 14:1 PC. Our viscosity data above and below the phase
transition temperature provide quantitative information on the fluidity of this family of

bilayer structures.

120

Literature Cited

(1) Edidin, M. Nature Reviews. Molecular Cell Biology 2003, 4, 414.

(2) Discher, B. M.; Maloney, K. M.; Grainger, D. W.; Sousa, C. A.; Hall, S.
B. Biochemistry 1999, 38, 474.

(3) Keller, S. L.; McConnell, H. M. Physical Review Letters 1999, 82, 1602.

(4) Samsonov, A. V.; Mihalyov, 1.; Cohen, F. S. Biophysical Journal 2001,
81 , 1486.

(5) Subramaniam, S.; McConnell, H. M. Journal of Physical Chemistry 1987,
91, 1715.

(6) Yechiel, E.; Edidin, M. Journal of Cell Biology 1987, 105, 755.

(7) Pabst, G.; Amenitsch, H.; Kharakoz, D. P.; Laggner, P.; Rappolt, M.
Physical Review E: Statistical, Nonlinear, and Soft Matter Physics 2004, 70, 021908/1.

(8) Suurkuusk, J .; Lentz, B. R.; Barenholz, Y.; Biltonen, R. L.; Thompson, T.
E. Biochemistry 1976, 15, 1393.

(9) Lentz, B. R.; Barenholz, Y.; Thompson, T. E. Biochemistry 1976, 15,

4521.
(10) Shimshick, E. J .; McConnell, H. M. Biochemistry 1973, 12, 2351.
(11) Stevens, B. C.; Ha, T. Journal of Chemical Physics 2004, 120, 3030.
6 3 (12) Kranenburg, M.; Smit, B. Journal of Physical Chemistry B 2005, 109,
55 .

(13) Epand, R. M. Biochemical Society Transactions 1997, 25, 1073.

121

(14) Tocanne, J. F.; Cezanne, L.; Lopez, A.; Piknova, B.; Schram, V.;
Tournier, J. F.; Welby, M. Chemistry and Physics of Lipids 1994, 73, 139.

(15) Korreman, S. S.; Posselt, D. European Physical Journal E: Soft Matter
2000, 1, 87.

(16) Lewis, R. N.; McElhaney, R. N. Biochemistry 1990, 29, 7946.

(17) Marsh, D.; Watts, A.; Knowles, P. F. Biochimica et Biophysica Acta,
Biomembranes 1977, 465, 500.

(18) Morrow, M. R.; Davis, J. H. Biochemistry 1988, 27, 2024.

(19) Prenner, E. J .; Lewis, R. N.; Kondejewski, L. H.; Hodges, R. S.;
McElhaney, R. N. Biochimica et Biophysica Acta 1999, 141 7, 211.

(20) Mason, J. T. Methods in Enzymology 1998, 295, 468.

(21) Mabrey, S.; Sturtevant, J. M. Proceedings of the National Academy of
Sciences 1976, 73, 3862.

(22) Mantsch, H. H.; McElhaney, R. N. Chemistry and Physics of Lipids 1991,
5 7, 213.

(23) McElhaney, R. N. Chemistry and Physics of Lipids 1982, 30, 229.

(24) Parasassi, T.; Gratton, B; Yu, W. M.; Wilson, P.; Levi, M. Biophysical
Journal 1997, 72, 2413.

(25) Watts, A.; Spooner, P. J. Chemistry and Physics of Lipids 1991, 57, 195.

(26) Khan, T. K.; Chang, P. L. G. Biophysical Journal 2000, 78, 1390.

(27) Lakowicz, J. R.; Knutson, J. R. Biochemistry 1980, 19, 905.

122

(28) Watts, A.; Marsh, D.; Knowles, P. F. Biochemistry 1978, I 7, 1792.

(29) Mayer, L. D.; Hope, M. J .; Cullis, P. R. Biochimica et Biophysica Acta
1986, 858, 161.

(30) Hunter, D. G.; Frisken, B. J. Biophysical Journal 1998, 74, 2996.

(31) MacDonald, R. C.; MacDonald, R. I.; Menco, B. P.; Takeshita, K.;
Subbarao, N. K.; Hu, L. R. Biochimica et Biophysica Acta 1991, 1061, 297.

(32) Driessen, A. J .; van den Hooven, H. W.; Kuiper, W.; van de Kamp, M.;
Sahl, H. G.; Konings, R. N.; Konings, W. N. Biochemistry 1995, 34, 1606.

(33) Unger, E. C.; MacDougall, P.; Cullis, P.; Tilcock, C. Magnetic Resonance
Imaging 1989, 7, 417.

(34) Subbarao, N. K.; MacDonald, R. 1.; Takeshita, K.; MacDonald, R. C.
Biochimica et Biophysica Acta 1991, 1063, 147.

(3 5) DelaCruz, J. L.; Blanchard, G. J. Journal of Physical Chemistry B 2003,
107, 7102.

(36) 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.

(3 7) Marsh, D. Handbook of Lipid Bilayers; CRC Press: Boca Raton, 1990.

(3 8) J iang, Y.; Blanchard, G. J. Journal of Physical Chemistry 1994, 98, 6436.

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

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

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

123

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

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

(44) Janiak, M. J .; Small, D. M.; Shipley, G. G. Biochemistry 1976, 15, 4575.

(45) Tristram-Nagle, S.; Zhang, R.; Suter, R. M.; Worthington, C. R.; Sun, W.
J .; Nagle, J. F. Biophysical Journal 1993, 64, 1097.

(46) Eimer, W.; Pecora, R. Journal of Chemical Physics 1991, 94, 2324.

(47) Peters, R.; Beck, K. Proceedings of the National Academy of Sciences
1983, 80, 7183.

124

Chapter 6

INTERROGATING THE ROLE OF LIPOSOME SIZE IN MEDIATING
THE DYNAMICS OF A CHROMOPHORE IN THE ACYL CHAIN
REGION OF A PHOSPHOLIPID BILAYER

Introduction

Lipid bilayer structures are essential to life processes, and as such have attracted a
great deal of attention. It is well known that phospholipids will self-assemble
spontaneously to form a bilayer structure in an aqueous environment, and under some
conditions the most stable form is that which closes on itself, leaving no exposed
“edges.” Such spherical bilayer structures are commonly referred to as liposomes or
vesicles, and they can be formed in a range of sizes, from ca. 100 nm or less to ca. 10 um
in diameter. This structural format has found widespread use because such structures can
be suspended in an aqueous environment and can be manipulated with relative ease.

These outwardly simple structures are characterized by a range of subtle
molecular interactions that collectively account for why the bilayer structure closes on.
itself. For example, it is known that there are phase transitions in lipid bilayer structures
that are related to the organization of the phospholipid acyl chains. At a characteristic
temperature, Tm, the hydrocarbon chains undergo a structural change from being
predominantly trans in configuration to a structure characterized by a one or more trans-
gauche conformers per chain.l For temperatures below the Tm, the lipid aliphatic tails are
said to exist in a gel phase, a well-ordered state in which the translational mobility and

diffusion of the bilayer constituents is thought to be comparatively limited. Above the

125

phase transition temperature, the phospholipid aliphatic tails exist in a disordered, fluid-
like phase,2 characterized by more motional and lateral freedom of bilayer constituents.3'S
Some of the techniques that have been used in the past to measure this phase transition
include differential scanning calorimetry,6 fluorescence spectroscopy7 and spin labeling.8

Because the phase transition temperature is related to the energetics of lipid-lipid
‘ interactions within the bilayer, and phospholipid phase transitions are considered to be

9’10 there is a

important in regulating the activities of membrane-associated proteins,
significant body of literature focused on measuring Tm for selected phospholipid bilayer
systems.2’9'22 The phase transitions seen for the acyl regions of phospholipid bilayers do
not, however, provide a complete picture of bilayer organization. Vesicle geometry is
also thought to contribute to the properties of the bilayer structure, with the dominant
issue for this case being the extent to which surface curvature influences the organization
of the acyl chains. This is a topic that has been considered extensively in the literature,”
26 and there is some experimental evidence that bilayer surface curvature can influence
the dynamics of the bilayer constituents. Indeed, it is the purpose of this paper to
examine the influence of bilayer surface curvature on the dynamics of a chromophore
imbedded in the acyl chain region. We find that the perylene reorientation depends
sensitively on the size of the liposome in which it is incorporated.

To understand the role of lipid bilayer curvature in mediating the dynamics of an
imbedded chromophore, we have constructed unilamellar vesicles of the phospholipid

DMPC containing 10'6 M perylene as the probing chromophore. The diameter of

prepared liposomes is controlled using the extrusion method,”33 where polycarbonate

126

membranes with defined pore sizes are used as the medium for extrusion. As the pore
sizes in the polycarbonate membranes increase, the radius of curvature of the vesicles
decreases, and we have established the relationship between membrane pore size and the
resulting liposome diameter based on dynamic light scattering (DLS) measurements. We
are interested in two issues. The first is to understand whether or not there is a
relationship between vesicle size and phase transition temperature, and the second is to
understand whether or not the size of the vesicle influences the chromophore dynamics.
We find that for the smaller vesicles (S 800 nm diameter) the phospholipid gel to fluid
phase transition can be detected through the reorientation dynamics of perylene, and this
finding is in agreement with our earlier report.34 We also find that the dynamics of the
perylene chromophore change markedly and abruptly with increasing liposome size.
This finding suggests that as curvature in the bilayer becomes less pronounced, the
perylene chromophore relocates within the bilayer structure. We discuss possible
chromophore locations within the bilayer that are consistent with this finding. Our data
imply that, for smaller vesicles, the chromophore locates preferentially in one of the
bilayer leaflets, although which leaflet is not able to be discerned from our

measurements.

127

Experimental

Vesicle Preparation. The phospholipid l,2-dimyristoyl-sn-phosphatidylcholine
(DMPC) was purchased from Avanti Polar Lipids Inc. (Alabaster, AL) and used as
received. Perylene was purchased from Sigma—Aldrich (Milwaukee, WI) and used
without further purification. The lipid and the perylene probe were mixed so that the
final concentration was 2.9 x 10'4 M DMPC and 1 x 10'6 M perylene in the initial
chloroform solution. The chloroform solvent was evaporated and an appropriate volume
of Tris® buffer (Sigma-Aldrich, Milwaukee, WI) was added to the sample so that the
resulting lipid concentration was 1 mg/mL. The buffer (10 mM, pH 7.8) was prepared
using purified water from a Milli-Q Plus water purification system (Millipore, Bedford,
MA). The mixture was processed five times through a freeze-thaw-vortex cycle to ensure
complete mixing of the lipid and fluorescent probe. Each cycle consisted of freezing the
solution by immersion in liquid nitrogen for five minutes, followed by drawing in a 60°C
water bath (five minutes) and vortexing (approximately two minutes). After freeze-thaw-
vortex processing, the solution was extruded eleven times through a polycarbonate
membrane (Avanti Polar Lipids Inc., Alabaster, AL) with pores of a given diameter

27'” were performed at room temperature.

(ranging from 100 nm to 5 pm). Extrusions
Time-resolved fluorescence measurements. Time-domain polarized fluorescence

intensity decays were acquired over a range of temperatures using a time-correlated

single-photon counting (TCSPC) system. This system has been described in detail

elsewhere, and we briefly recap some of its features here.35’36 The light source is a CW

mode-locked Nd:YAG laser (Coherent Antares 76-s) that produces 100 ps 1064 nm

128

pulses at 76 MHz repetition rate. The third harmonic of the Nd:YAG laser output is used
to excite a cavity-dumped dye laser (Coherent 702-2) operated with Stilbene 420 dye
(Exciton) at ca. 430 nm. The output of this laser is ca. 25 mW average power, with 5 ps
pulses at a 3.8 MHz repetition rate. Sample fluorescence is detected with a microchannel
plate — photomultiplier tube (MCP-PMT, Hamamatsu R3809U) and the electronics used
to temporally resolve the fluorescence transients are constant fraction discriminators
(CFD, Tennelec 454) and a time-to-amplitude converter/biased amplifier (TAC, Tennelec
864). The collection wavelength and polarization are computer-controlled using
LabVIEW® 7.1 code. The fluorescence lifetime data were collected at 54.7° with respect
to the vertical excitation polarization, and the reorientation data were collected at
polarizations parallel (0°) and perpendicular (90°) to the vertically-polarized incident
light. The temperature of the sample was regulated to :l: O.1°C using a water-circulating
bath (Neslab RTE-110) connected to a cooling jacket that surrounds the sample cuvette.
All samples were allowed to thermally equilibrate for ten minutes prior to data
acquisition.

Steady state spectroscopy. Steady state excitation and emission spectra (Figure
6.1) were acquired for our vesicle samples for the pm'pose of characterizing the perylene
band positions. We used a Spex Fluorolog 3 emission spectrometer for all
measurements, set to a spectral bandpass of 2 nm for both excitation and emission

monochromators.

129

 

perylene

excitation

emission

   

 

0.0 -

I n I 4 I n I n I L l n I n I A I A I

3804m4ZO4404604805m520540560
wavelengdflnn)

DMPC

Figure 6.1. Structures of the phospholipid, DMPC and the chromophore,
perylene. Bottom right: Steady state absorption and emission spectra of
perylene in DMPC vesicles. Spectra have been normalized for clarity of
presentation.

130

Dynamic light scattering. Dynamic light scattering (DLS) was performed with a
Protein Solutions DynaPro-MS/X system (Wyatt Technology Corporation, Santa
Barbara, CA) on DMPC/perylene liposome solutions to determine mean vesicle sizes and
size distribution. Liposome solutions were diluted by a factor of fifty with buffer and
placed in a polyethylene cuvette for measurement. We find that there is a direct
correspondence between DLS-measured vesicle size and the nominal pore diameter of

the extrusion membrane (Table 6.1 and Figure 6.2).

131

Table 6.1. Comparison of extrusion membrane pore diameter to the experimental
diameter of the extruded vesicles. Experimental diameters were determined by DLS

measurements. Uncertainties are :1: 10 for three individual determinations.

 

 

 

 

 

 

 

 

 

 

Extrusion nominal Measured mean vesicle
pore diameter (nm) diameter (nm)
100 106 i 3
200 193 d: 14
400 417 :1: 8
600 610 d: 16
800 817 i 23
1000 1050 :t 69
2000 2108 :1: 21
3000 3028 :1: 83
5000 (not measurable)

 

 

 

 

132

3500 -

 

 

 

 

E y = 0.575 + 1.04x

c3 3°°°' R2=0.999

o

3 2500-

'U

8

E 2000-

3

8 1500-

(D

‘“ 1

b

9. 1000-

C)

E .

.9

U 5004

C

m -1

Q)

E 04
I T T ' I ' I f I ' I T I '
o 500 1000 1500 2000 2500 3000

nominal pore diameter from extruder (nm)

Figure 6.2. Plot showing the linear relationship between the nominal pore diameter
in the extruder membrane and the measured diameter of DMPC/perylene vesicles
from DLS.

133

Results and Discussion

The purpose of this work is to understand the dynamics of perylene in DMPC
vesicles and how those dynamics depend on the temperature of the system and the
diameter of the vesicles. As noted above, phospholipids are known to undergo phase
transitions that are related to the organization of their acyl chains, and in a previous study
we reported that the reorientation dynamics of perylene imbedded within DMPC bilayers
are sensitive to this phase transition.34 The interrogation of a phase transition in a lipid
bilayer system requires some consideration of where in the system the probe resides as
well as the characteristic length scale of the structural changes in the bilayer relative to
the length scale sensed by chromophore reorientation. In an attempt to address these
issues, we have studied perylene imbedded in a series of DMPC unilamellar vesicles of
controlled diameter. Our findings relating to the temperature- and vesicle-size
dependence of the recovered perylene dynamics reveal several interesting structural
features. We consider the significance of each of these bodies of data after initial
consideration of the perylene dynamics we recover.

In our previous report we found that perylene imbedded in 100 nm diameter
unilamellar DMPC vesicles was sensitive to the phase transition of the bilayer from the
gel to fluid phase.34 Further, with the addition of 14:1 PC, even in small amounts, there
was a marked change in the observed phase transition temperature. To characterize the
reorientation dynamics of perylene, we combined the experimental fluorescence
transients to form the induced orientational anisotropy function, R(t),

134

R(t) = IMO—11.0)
I"(t)+21l(t)

 

(6.1)

and from the functional form of R(t) we inferred the motional dynamics of the perylene
chromophore. The theory for the interpretation of R(t) is well established, and the
number of exponential decays contained in this function is related to the shape of the
molecule and its motion relative to the orientations of the excited and emitting transition

38"” and can exhibit either a

moments.37 Perylene has been studied extensively before,
single exponential decay or a two-component exponential decay of R(t), depending on
whether it reorients as a Type I or a Type II rotor. We use Type I and Type II
designations rather than the more widely used prolate and oblate terms because of the
potential ambiguity with the latter. For a Type I rotor, the unique rotational axis is
coincident with the transition moment, and for a Type II rotor, the unique rotational axis
is perpendicular to the transition moment. For the purposes of this discussion, we assign
the perylene long in-plane axis as x (coincident with the S] 4— So transition dipole
moment), the short in-plane axis as y and z is the axis perpendicular to the molecular 1:
system plane. For a Type I rotor (DM 3:5 Dy = DZ) under these axis assignments, we expect
the functional form of R(t) to be,

R(t) = 0.4exp(6D:t) (6.2)
and for a Type II rotor (Dz 76 DK = Dy),

R(t) = 0.1exp(—(2DJr + 4Dz )t) + 0.3 exp(—6Dxt) (6.3)

In our earlier work we recovered anisotropy decay functions consistent with a

single exponential decay, and in this work we recover anisotropy functions that can be

135

resolved as two-component decays. The results are consistent with one another (vide
infia), and indicate that perylene reorients as a Type II rotor in the DMPC vesicles we
have formed. The data reported previously34 were characterized by a limited S/N ratio,
thereby allowing the resolution of two anisotropy decay components (Figure 6.3). When
the experimental time constants we report here are reduced to a single weighted average
time constant, the results are the same as the time constants reported in the earlier work
(Figure 6.4). Because we can resolve two anisotropy decay components for perylene, it
is possible to determine D2 and Dx, and substantial information is contained in these
rotational diffusion constants. We have examined how these quantities vary with vesicle
size and system temperature. We consider first the temperature dependence of these
rotational diffusion constants.

We show in Figures 6.5 the temperature-dependent reorientation dynamics of
perylene in unilamellar vesicles from 100 nm to 800 nm diameter (Figure 6.5a) and 1000
nm to 5000 nm diameter (Figure 6.5b). We present these data in the form of weighted
average time constants to make direct correspondence with data we have reported
previously. The temperature-dependent reorientation data show a subtle change in
dynamical behavior in the vicinity of the transition temperature for vesicles up to 800 nm
diameter (Figure 6.5a), but for vesicles 1000 nm and larger (Figure 6.5b), there is no
discernible change in probe dynamics near the phase transition temperature. This finding
suggests that either there is a change in the nature of the phospholipid phase transition or
there is a change in the location of the probe within the bilayer structure. We will return

to a discussion of this point following consideration of the vesicle-size dependence of the

136

0.4 -

 

 

G 0.2 — .
E 3..
to '-
t '3,
3212.2??? 1 . .
'E‘-.~ ': “1.1;,3‘5‘ ‘1‘ ’ .T‘ggz‘.:....;,";°.;,j;" :‘:1 am?
0'0 ' ‘1'_ l ‘ 1V3” ‘1 1M”: 5551"” “twin,- I‘ll" [hill-”11.6111" 1~ anummdr "NI
1" 1 1 ‘ 1 1‘ ll 1‘5 '+ ‘ 1:1..1. .‘lli
1 l 1 . 1 . 1 1 1 ‘ . ‘1
0 ICU) 2000 3%) 4000 50X)

tinx:(ps)

Figure 6.3. Anisotropy data for perylene in 100 nm diameter DMPC vesicles at
25°C, showing a decay that is best fit by two exponential components. The fitted line
is indicated as the solid line and the residuals of the fit are shown centered around
the zero line. The fitted function is R(t) = Alexp(-t/1:1) + Azexp(-t/1:2), where A]

0.07 :b 0.01, 1:; = 271 :1: 6 PS, A; = 0.20 :1: 0.01 and 1'; = 2238 :1: 56 ps.

137

1200~

 

 

 

11001;- . fiunRe£34
: 0 mm
1000-
A 9..- 1
E13
1 1
700_
600'-
sm- 1 I 1 I 1 I 1 I 1 I 1 I
18 20 22 24 26 28 3o

temperature (°C)

Figure 6.4. Reorientation times obtained for 100 nm diameter vesicles containing
DMPC and perylene. The data from the earlier work were single time constants,
while the data from the current work represent weighted averages of two time
constants.

138

 

1200

I

1100-

 

1(XJO

I
r—m-b—Hr—l—t
1—b—l—H—o—1

 

temperature (°C)

 

 

1100-

>00-

l(XX)r

 

10R (ps)
§

1"}!

700-

 

18 20 22 24 26

temperature (°C)

139

H

IHDC-H

Figure 6.5. Reorientation times obtained for perylene in DMPC vesicles as a
function of temperature and vesicle diameter. (a) 100-800 nm vesicle diameters.
(b) 1000-5000 nm vesicle diameters.
data. We note that there is the possibility that the vesicles could change in size during the
phase transition, which we cannot readily account for. However, based on past
research,42’43 there was a minimal change in the average diameter of phospholipid
vesicles over a temperature range that included the gel-to-fluid phase transition.

We were interested in using the measured reorientation times of perylene within
the lipid vesicles to calculate the rotational diffusion coefficient, D, as a function of
temperature. As a Type II rotor, the shorter recovered time constant, 11,0R, is related to

one of the Cartesian components of the rotational diffusion constant by Eq. 6.4.37

1
6D: < >
The longer recovered time constant 12,011, is related to two Cartesian components of the

rotational diffusion constant by Eq. 6.5.37

1

72,011 = m (65)

When both anisotropy decay time constants are used, and the quantities Dx and D, are
extracted (Eq. 6.5), it is clear that D2 is larger than D.x in all cases, and that the most
pronounced change in dynamics occurs for D2 (Figure 6.6).

The temperature-dependent data for a given vesicle size reveals a subtle change in
anisotropy decay dynamics in the region of the DMPC phase transition, but the most
marked functional dependence of the data is seen when the vesicle size—dependence is

plotted for a given temperature. We show these data for 19°C in Figures 6.7-6.8, noting

140

 

' I

‘ I

 

I

 

D (MHz)

"fiiijT'

f I T I I

 

m FO-lt—D-H-D—m 1—0—1

 

 

 

 

 

 

 

D
j
8 28 30
201)-
» b
18(Xl-
1600~ § D2
1400- if
12(X)- é
”1:1”
E lCDO-
; 800- I 1000
’ D XXX)
4m_ 0 5(IX)
L
D
2005 5 E 111: G 0 5
O I l I I I 1 I
18 20 22 24 26 28 30
W60

Figure 6.6. Cartesian components of the diffusion coefficient, D, and D2, obtained
for perylene in DMPC vesicles as a function of temperature and vesicle diameter.
(a) 100-800 nm vesicle diameters. (b) 1000-5000 nm vesicle diameters.

141

1600-

1400- t
' §

1200-1

1000 j

10R (ps)
§

 

I I 1 I 1 I
2000 3(X)0 4000 5000
vesicle diameter (nm)

+ 1

1 I
O 1000

Figure 6.7. Reorientation times obtained for perylene in DMPC vesicles as a
function of vesicle diameter at 19°C.

142

D (MHz)

8

809

I I

©

‘ I

D

.- § 5 ' X a

1G '.. .4 . 1 1 1 1 1 . 1 .

O 1(IX) 2(XIO 3(II) 4(X)0 5(XX)
vesicle diameter (nm)

 

2000
1800
1600
1400

1200 1 D
1000
800
600
400
200
0

Figure 6.8. Cartesian components of the diffusion coefficient, D, and Dz, obtained
for perylene in DMPC vesicles as a function of vesicle diameter at 19°C.

143

that the data at other temperatures exhibit the same size dependence. The anisotropy
decay data change in a step-wise manner; for vesicles smaller than 800 nm diameter we
observe Dz ~ 600 MHz and Dx ~ 60 MHz (Dz/Dx ~ 10). For vesicles 1000 nm diameter
and larger, we recover DZ ~ 1500 MHz and DK ~ 150 MHz (Dz/Dx ~ 10). While theaspect
ratio of the perylene motion remains substantially the same, the reorientation rate of the
chromophore increases by a factor of ca. 2.5 for a small change in vesicle diameter. As
noted above, we observe the same behavior at temperatures above and below the
DMPC phase transition temperature. Such a change in dynamics indicates that there is a
marked change in the local environment of the chromophore as the vesicle diameter
changes from 800 nm to 1000 nm.

Using the calculated Cartesian components (Dx, Dy and Dz) of the rotational
diffusion constant, D, we can calculate D based on Eq. 6.6.

D, + Dy + Dz
D = 3 (6.6)

 

The rotational diffusion constant can then be related to the local environment in which

the chromophore resides based on the modified Debye-Stokes-Einstein equation,44'46

6D kBTS

 

1 = "Vf (6.7)

where 11 is the viscosity of the medium surrounding the reorienting chromophore, V is the
hydrodynamic volume of the chromophore, f is a frictional term to account for solvent-
solute interactions, k3 is the Boltzmann constant and T is the temperature in Kelvin. The
term S is a shape factor determined from Perrin’s equations44 to account for the

nonspherical shape of the chromophore, and we calculate from Perrin’s equations that S =

144

Table 6.2. Calculated viscosity at 19°C for the vesicles with varying diameters.

 

 

 

 

 

 

 

 

 

 

Vesicle diameter (nm) Calculated viscosity
(0?)
100 8.0 :1: 0.8
200 12.4 :1: 2.7
400 11.4 i 1.6
600 11.5 i 1.2
800 8.7 d: 0.3
1000 3.5 i 0.3
2000 4.2 i 1.4
3000 3.5 i 0.3
5000 3.5 d: 0.2

 

 

 

 

145

0.70.47 For each of the systems, using the calculated rotational diffusion constant from
Eq. 6.6 and the hydrodynamic volume for perylene (225 A3),48 the viscosity, n, was
calculated. While the average viscosity for each sample remained approximately
constant over the temperature range that we studied (l9-29°C), we observed a decrease in
the average viscosity in vesicles with diameters greater than 800 nm. Table 6.2 shows
the calculated viscosity as a function of vesicle diameter at 19°C.

The experimental data (at ca. 1000 nm diameter vesicle size) indicate that there is
a change in the chromophore local environment. This change could be the result of either
an alteration of the bilayer structure of DMPC, or it could result from a change in the
location of the chromophore in the bilayer structure. It is known that for vesicles,
curvature in the system gives rise to asymmetric bilayers?"26 This bilayer asymmetry, in
the case of the bilayers we use in this work, must come in the form of the number of
lipids contained in the inner and outer leaflets, and likely also the packing of the acyl
chains. Experimental data on some small vesicles (< 200 nm in diameter) indicates that,
in terms of x-ray measurements, any difference between phosphocholine inner and outer
leaflets cannot be resolved.49 Our dynamical data on perylene rotational motion within
the bilayers indicate, however, that chromophore access to different environments is
modulated by the size of the vesicles, implicating curvature as a factor in this finding.
There is also a large literature, mostly in the form of molecular dynamics simulations,
that indicates that curvature in bilayers does occur and gives rise to asymmetric

bilayers.”26

146

Our data do not point to a specific change in chromophore location with
increasing vesicle size, but there are several aspects of these findings that provide
inferential insight into the organization of the system. The first point is that the
chromophore dynamics become faster with increasing vesicle size. While the most
noticeable change occurs for D2, the ratio of Dz/Dx remains close to constant for all of the
vesicle sizes we have measured. The fact that the diffusion coefficients increase by a P:-
factor of ca. 2.5 over the small change in vesicle size points to the perylene moving to a
region of lower viscosity. Given the nonpolar nature of the chromophore, we suggest that

the perylene is moving from being predominantly within the acyl chains for small

 

I!

vesicles to residing primarily in the region between the leaflets for the more planar
structures. The second, and related point of interest is that the change in dynamics is seen
at the same vesicle size where the motion of perylene appears to stop sensing the acyl
chain gel-to-fluid phase transition. Such a change in what the chr0m0phore senses
implies that the chromophore is not in intimate contact with the portion of the bilayer that
is affected most by the phase transition for larger vesicles. Again, a change of location
from within the acyl chains to between the leaflets is consistent with this observation.

If the assignment of the perylene moving from the acyl chain region to the inter-
leaflet region with increasing vesicle size is correct, an implication of this assignment is
that, because there is necessarily an asymmetry between the inner and outer leaflets for
small vesicles, the perylene chromophore will preferentially reside in one of the leaflets
for small vesicles, owing to more favorable packing. It is possible that the organizational

difference between the inner and outer leaflets is negligible,49 but unfortunately, it is not

147

possible at this point, using simple DMPC vesicles, to evaluate the role of bilayer

asymmetry in mediating perylene dynamics in small vesicles.

Conclusions

We have used fluorescence anisotropy measurements as a means of interrogating
the local environment in lipid vesicle systems of varying diameter. We found that we
could detect the DMPC gel-to-fluid phase transition temperature in vesicles with
diameters of 800 nm or less and could resolve differences in D2, over the temperature
range encompassing the phase transition. With vesicles 1000 nm to 5000 nm in diameter,
the transition temperature could not be resolved and there was no discemable trend in the
rotational dynamics of perylene sequestered within the lipid bilayer. Examining how the
reorientation data varied with vesicle size, however, revealed a marked change in
chromophore dynamics with increasing vesicle size. The data suggest that the perylene
changes location within the bilayer structure as curvature is reduced. This finding
underscores the critical role that bilayer curvature can play in mediating dynamics and
local environment. Further work with more complex bilayer structures will provide
insight into cooperative roles that bilayer constituents may play in altering the influence

of curvature at the molecular level.

148

Literature Cited

(1) Pabst, G.; Amenitsch, H.; Kharakoz, D. P.; Laggner, P.; Rappolt, M.
Physical Review E: Statistical, Nonlinear, and Soft Matter Physics 2004, 70, 021908/ 1.

(2) Stevens, B. C.; Ha, T. Journal of Chemical Physics 2004, 120, 3030.

(3) Van Zoelen, E. J. J .; De Kruijff, B.; Van Deenen, L. L. M. Biochimica et
Biophysica Acta, Biomembranes 1978, 508, 97.

(4) De Kruijff, B.; Van Zoelen, E. J. J.; Van Deenen, L. L. M. Biochimica et
Biophysica Acta, Biomembranes 1978, 509, 537.

(5) Gerritsen, W. J .; Henricks, P. A. J.; De Kruijff, B.; Van Deenen, L. L. M.
Biochimica et Biophysica Acta, Biomembranes 1980, 600, 607.

(6) Suurkuusk, J.; Lentz, B. R.; Barenholz, Y.; Biltonen, R. L.; Thompson, T.
E. Biochemistry 1976, 15, 1393.

(7) Lentz, B. R.; Barenholz, Y.; Thompson, T. E. Biochemistry 1976, I5,
4521.

(8) Shimshick, E. J.; McConnell, H. M. Biochemistry 1973, 12, 2351.

(9) Epand, R. M. Biochemical Society Transactions 1997, 25, 1073.

(10) Tocanne, J. F .; Cezanne, L.; Lopez, A.; Piknova, B.; Schram, V.;
Toumier, J. F .; Welby, M. Chemistry and Physics of Lipids 1994, 73, 139.

(1 l) Kranenburg, M.; Smit, B. Journal of Physical Chemistry B 2005, 109,
6553.

(12) Korreman, S. S.; Posselt, D. European Physical Journal E: Soft Matter
2000, 1, 87.

149

(13) Lewis, R. N.; McElhaney, R. N. Biochemistry 1990, 29, 7946.

(14) Marsh, D.; Watts, A.; Knowles, P. F. Biochimica et Biophysica Acta,
Biomembranes 1977, 465, 500.

(15) Morrow, M. R.; Davis, J. H. Biochemistry 1988, 27, 2024.

(16) Prenner, E. J.; Lewis, R. N.; Kondejewski, L. H.; Hodges, R. S.;
McElhaney, R. N. Biochimica et Biophysica Acta 1999, 141 7, 211.

(17) Mason, J. T. Methods in Enzymology 1998, 295, 468.

(18) Mabrey, S.; Sturtevant, J. M. Proceedings of the National Academy of
Sciences of the United States of America 1976, 73, 3862.

(19) Mantsch, H. H.; McElhaney, R. N. Chemistry and Physics of Lipids 1991,
5 7, 213.

(20) McElhaney, R. N. Chemistry and Physics of Lipids 1982, 30, 229.

(21) Parasassi, T.; Gratton, B; Yu, W. M.; Wilson, P.; Levi, M. Biophysical
Journal 1997, 72, 2413.

(22) Watts, A.; Spooner, P. J. Chemistry and Physics of Lipids 1991, 5 7, 195.

(23) Parthasarathy, R.; Yu, C.-h.; Groves, J. T. Langmuir 2006, 22, 5095.

(24) Cooke, I. R.; Desemo, M. Biophysical Journal 2006, 91, 487.

(25) Lubensky, T. C.; MacKintosh, F. C. Physical Review Letters 1993, 71,
1565.

(26) Baumgaertner, A. Journal of Chemical Physics 1994, 101, 9060.

150

(27) Driessen, A. J.; van den Hooven, H. W.; Kuiper, W.; van de Kamp, M.;
Sahl, H. G.; Konings, R. N.; Konings, W. N. Biochemistry 1995, 34, 1606.

(28) Hope, M. J .; Bally, M. B.; Webb, G.; Cullis, P. R. Biochimica et
Biophysica Acta, Biomembranes 1985, 812, 55.

(29) Hunter, D. G.; Frisken, B. J. Biophysical Journal 1998, 74, 2996.

(30) MacDonald, R. C.; MacDonald, R. 1.; Menco, B. P.; Takeshita, K.;
Subbarao, N. K.; Hu, L. R. Biochimica et Biophysica Acta 1991, 1061, 297.

(31) Mayer, L. D.; Hope, M. J .; Cullis, P. R. Biochimica et Biophysica Acta
1986, 858, 161.

(32) Subbarao, N. K.; MacDonald, R. I.; Takeshita, K.; MacDonald, R. C.
Biochimica et Biophysica Acta 1991, 1063, 147.

(33) Unger, E. C.; MacDougall, P.; Cullis, P.; Tilcock, C. Magnetic Resonance
Imaging 1989, 7, 417.

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

(35) 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.

(36) Dela Cruz, J. L.; Blanchard, G. J. Journal of Physical Chemistry A 2001,
105, 9328.

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

(38) Labhart, H.; Pantke, E. R. Chemical Physics Letters 1973, 23, 482.

(39) Zinsli, P. E. Chemical Physics 1977,20, 299.
151

(40) Klein, U. K. A.; Haar, H. P. Chemical Physics Letters 1979, 63, 40.

(41) Christensen, R. L.; Drake, R. C.; Phillips, D. Journal of Physical
Chemistry 1986, 90, 5960.

(42) Cornell, B. A.; Fletcher, G. C.; Middlehurst, J .; Separovic, F. Biochimica
et Biophysica Acta, Biomembranes 1981, 642, 375.

(43) Ceuterick, F. ; Heremans, K.; De Smedt, H.; Nieuwenhuysen, P.;
Clauwaert, J. Chemical Physics Letters 1979, 62, 341.

(44) Perrin, F. Journal de Physique et le Radium 1936, 7, l.

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

(46) Debye, P. Polar Molecules. In Chemical Catalog Co. New York, 1929.

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

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

(49) Kucerka, N.; Pencer, J .; Sachs, J. N.; Nagle, J. F .; Katsaras, J. Langmuir
2007, 23, 1292.

152

Chapter 7

CONCLUSIONS AND FUTURE WORK

In this dissertation, we used Time Correlated Single Photon Counting (TCSPC)
and fluorescence recovery after pattern photobleaching (F RAPP) to study the
fluorescence lifetime and rotational and translational dynamics of various chromophores
located in different regions of lipid bilayers, with most emphasis on unilamellar vesicles.
In doing so, we were able to study the dynamics of lipid bilayer structures and determine
which factors affect the organizational and structural properties of the systems.

In Chapter 2, we found that two methods of lipid vesicle preparation, extrusion
and sonication, yielded the same translational and rotational dynamic results for the
tethered NBD chromophore that was used, despite any differences that were observed in
the physical appearance, size and size distribution of the samples. While the addition of
cholesterol into these systems made the lipid bilayers more fluid, the results were the
same, regardless of the method of preparation.

In Chapter 3, we found that the lipid bilayer constituents were highly interactive
in our systems, and any changes in one region gave rise to corresponding structural
changes in other bilayer regions. The dynamics of the tethered rhodamine chromophore,
located near the head group region of the lipid bilayer, showed the existence of
interactions between these chromophores, which were mediated by the NBD
chromophore dynamics in cholesterol rich domains.

The effect of increasing chromophore concentration in the lipid vesicles was

studied in Chapter 4, where it was found that both the steady state and dynamical
153

properties of tethered rhodamine change with concentration, in a manner that is
consistent with the introduction of disorganization to the lipid bilayers. This finding
underscored the necessity in using low chromophore concentrations for our studies.

We found in Chapter 5 that we could use a nonpolar chromophore, perylene,
together with its reorientation times to determine the phase transition temperature in
selected lipid bilayer systems. We were able to use the data that we obtained to
approximate the viscosity, as sensed by this probe, in the nonpolar regions of the lipid
bilayers. We also found that the incorporation of an unsaturated lipid into the system, in
varying concentrations, caused substantial changes in the measured transition temperature
of the system.

In Chapter 6, we varied the diameters of the lipid vesicles (from 100 nm to 5 um)
and used the rotational reorientational dynamics of perylene to sense the phase transition.
We found that we could resolve the phase transition in the vesicles with diameters less
than 1 um. In the larger vesicles, the rotational dynamics of perylene were not as fully
resolved.

Future work with this project would involve tethering lipid vesicles onto planar
substrates in order to study additional properties of bilayer systems including the
movement of transmembrane proteins. The incorporation of proteins and additional
bilayer constituents into the systems, including cholesterol and sphingomyelin, would
serve to make the lipid bilayers systems more realistic and aid in the design of future

model membranes.

154

We now realize the necessity in using lower concentrations of chromophores in
spectroscopic studies, and we are further interested in comparing the information
obtained from these studies to electrochemical experiments. The fusion of vesicles onto a
conductive substrate would allow us the ability to measure the electrochemical signal
across the membrane and use the membranes as biosensors. Additionally, it would be
constructive to be able to visualize and size the various domains within lipid bilayers
using fluorescence microscopy to further verify the exact location of the fluorescent

probes.

155

IIIIIIIIIIIIIIIIIIIIIIIIIIIII

1111111111111111111111llllllllllllllll