<%BANNER%>

Bis(monoacylglycerol)phosphate Effects on Model Membrane Morphology

Permanent Link: http://ufdc.ufl.edu/UFE0021335/00001

Material Information

Title: Bis(monoacylglycerol)phosphate Effects on Model Membrane Morphology A Magnetic Resonance Investigation
Physical Description: 1 online resource (122 p.)
Language: english
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: bmp, epr, lbpa, lipid, luv, membrane, mlv, nmr
Chemistry -- Dissertations, Academic -- UF
Genre: Chemistry thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Bis(monoacylglycerol)phosphate (BMP) is a phospholipid found primarily in late endosomes, and has a unique structure due to single acyl chains located at the 3 and 1 prime positions on the glycerol components. BMP is known to play an important role in late-endosome sorting functions, and is also thought to be involved in glycosphingolipid catobilism. When BMP is present in liposomes containing ganglioside GM1, the enzymatic hydrolysis of GM1 to GM2, stimulated by activator proteins, is dramatically enhanced. This work is focused on determining the effect BMP has upon model membrane lipid morphology, and acyl chain dynamics, using magnetic resonance spectroscopy, in order to further understand BMP?s role in lipid catabolism and lysosomal storage disease.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Fanucci, Gail E.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0021335:00001

Permanent Link: http://ufdc.ufl.edu/UFE0021335/00001

Material Information

Title: Bis(monoacylglycerol)phosphate Effects on Model Membrane Morphology A Magnetic Resonance Investigation
Physical Description: 1 online resource (122 p.)
Language: english
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: bmp, epr, lbpa, lipid, luv, membrane, mlv, nmr
Chemistry -- Dissertations, Academic -- UF
Genre: Chemistry thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Bis(monoacylglycerol)phosphate (BMP) is a phospholipid found primarily in late endosomes, and has a unique structure due to single acyl chains located at the 3 and 1 prime positions on the glycerol components. BMP is known to play an important role in late-endosome sorting functions, and is also thought to be involved in glycosphingolipid catobilism. When BMP is present in liposomes containing ganglioside GM1, the enzymatic hydrolysis of GM1 to GM2, stimulated by activator proteins, is dramatically enhanced. This work is focused on determining the effect BMP has upon model membrane lipid morphology, and acyl chain dynamics, using magnetic resonance spectroscopy, in order to further understand BMP?s role in lipid catabolism and lysosomal storage disease.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Fanucci, Gail E.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0021335:00001


This item has the following downloads:


Full Text





BIS(MONOACYLGLYCEROL)PHOSPHATE EFFECT ON MODEL MEMBRANE
MORPHOLOGY: A MAGNETIC RESONANCE INVESTIGATION




















By

CHAD E. MAIR


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2008


































2008 CHAD E. MAIR




























To my mother, and father
Cyra B. and Robert V. Mair









ACKNOWLEDGMENTS

I first thank my advisor Dr. Gail E. Fanucci for allowing me to complete my graduate work

in her research group and for her patience as a mentor. I consider myself fortunate for the two

years of scientific interaction with, in my opinion, one of the top scientific minds in our

department.

I also thank my supervisory committee Professors John Eyler, Clifford Bowers, Joanna

Long, and Tom Lyons. Special thanks are extended to Dr. Long for her help with the setup of

the NMR instrumentation, insightful discussions, and use of her laboratory facilities and to Dr.

Bowers for use of his laboratory facilities. I thank the University of Florida machine shop,

especially Mr. Todd Prox for his patience and help in design and machining of custom parts.

Many fellow graduate students have played an important role in my graduate education.

Evrim Atas and Daniel Kuroda were my sounding boards while I was a member of the Kleiman

group. Luis Galliano has been my scientific confidant and became a close friend after I joined

the Fanucci group. However, I am most grateful to Dr. Lindsay Hardison she has become my

best friend and support group since her arrival at UF in 2002. I would not have finished the PhD.

without her!

Finally, I thank my parents for their encouragement and support, not only during my

academic career but throughout my life.









TABLE OF CONTENTS

page

A CK N O W LED G M EN T S ................................................................. ........... ............. .....

L IST O F T A B L E S ..................................................................................................... . 7

LIST OF FIGURES .................................. .. .. .... ..... ................. .8

A B S T R A C T ............ ................... ............................................................ 1 1

Chapter


1 INTRODUCTION ............... .......................................................... 12

Lipid D description and Classification ......................................................... .............. 12
Typical M em brane Lipid Structure ............................................... ............................. 13
Lipid Self-A assembly and Organization ........................................... ........................... 13
Characterization of Model Membrane Structural Properties............................................... 16
Bis(monoacylglycerol)phosphate ................ .................. .. .......... ............... 18
Bis(monoacylglycerol)phosphate May Be Important to Ganglioside Catabolism.................21
Biological M em branes ........................................................ .......... .. ............ 23
D issertation Outline ..................................................................... .. ............ 24

2 M A TER IA L S A N D M ETH O D S ........................................ .............................................32

M a te ria ls ..................... ...............................................................................................3 2
M ultilam ellar V esicle Preparation ......... ..................................................... ............... 32
P h o sp h ate A ssay ............................................................................3 5
T hin L ayer C hrom atography ......................................................................... ...................35
M magnetic Resonance ............... ................. ........... ............................ 36
D ata P processing ..............................................................................37

3 MAGNETIC RESONANCE APPLICATIONS IN MEMBRANE BIOPHYSICS ...............38

H ydrated L ipid M options and O order ............................................................. .....................38
N itroxide Spin-Probes ...................................... .. ...................... .. ... ....... .............. 40
Description of Nitroxide Spin-Label Order Parameter in Hydrated Lipid Bilayer
Assemblies Obtained by Electron Paramagnetic Resonance............................................40
Orientation of the Nitroxide Spin-Label in Hydrated Lipid Bilayer Assemblies and
Expected EPR Lineshapes ...................... ....... ...................................................... 43
Solid State 31P and 2H NMR of Hydrated Lipid Aggregates................................................45
Magnetic Resonance Line Shapes and Order in Hydrated Lipids............... ..................46











4 MONITORING MODEL BILAYER SOLUBILIZATION BY DETERGENT
M OLECULES USING EPR SPECTROSCOPY ........................................ ..................... 55

M odel M em brane Solubilization .............. .... ..................... ....... ......... .............................55
Characterization of the EPR Line Shapes of Spin-probes Located in Bilayer Aggregates
in the Presence of an Anionic Detergent....................................................................... 58
Hydrated Bis(monoacylglycerol)phosphate Assemblies Solubilized by Sodium Dodecyl
S u lfa te ............................................................................................. 6 1

5 PERTURBATIONS OF LAMELLAR LIQUID CRYSTALLINE ORDER BY
BIS(MONOACYLGLYCEROL)PHOSPHATE............................................... ..................73

Solid State Phosphorus-31 NMR Investigation of Hydrated
Bis(monoacylglycerol)phosphate Aggregation State.................................. ... ..................73
Acyl Chain Order of 1-Palmitoyl-2-Oleoyl-sn-Glycero-3-Phosphocholine
/Bis(monoacylglycerol)phosphate Mixed Vesicles Determined by Electron
Param magnetic Spectroscopy ............................... ...... ................................................. 74
31P NMR of 1-Palmitoyl-2-Oleoyl-sn-Glycero-3-Phosphocholine
/Bis(monoacylglycerol)phosphate mixed M LVs..................................... .................75

6 PERTURBATIONS OF THERMOTROPIC PHASE TRANSITIONS DUE TO
BIS(MONOACYLGLYCEROL)PHOSPHATE............................................... ..................88

Thermotropic Phase Behavior of 1,2-Dipalmitoyl-sn-Glycero-3-Phosphocholine
/Bis(monoacylglycerol)phosphate MLVs Investigated by 2H NMR.............................. 88
Thermotropic Phase Behavior of 1,2-Dipalmitoyl-sn-Glycero-3-Phosphocholine
/Bis(monoacylglycerol)phosphate MLVs Investigated by EPR.................................. 91
1,2-Dipalmitoyl-sn-Glycero-3-Phosphocholine /Bis(monoacylglycerol)phosphate 31P
N M R ................... ............................................................. ................ 9 4

7 SUM M ARY AND CONCLUSIONS........................................................ ............. 115

L IST O F R E F E R E N C E S ..................................................................................... ..................117

B IO G R A PH IC A L SK E T C H ......................................................................... ... ..................... 122









LIST OF TABLES


Table page

4-1 Parameters defining order of the 5-DOXYL nitroxide spin-probe in lipid aggregates
at room tem perature. ......................... ......... .. .. .. ...... .. ............. 72

4-2 Parameters defining order of the 10-DOXYL nitroxide spin-probe in lipid aggregates
at room tem perature. ......................... ......... .. .. .. ...... .. ............. 72

5-1 Parameters defining order of the 5-DOXYL nitroxide spin-probe in POPC/BMP
mixed M LVs at room temperature....................... ...... ............................ 85

5-2 Parameters defining order of the 10-DOXYL nitroxide spin-probe in lipid aggregates
at room tem perature .............. .. .......................... .. .................... 86

5-3 Values of CSA span for POPC/BMP mixed MLVs at room temperature .......................87

6-1 Total order parameter values for terminal methyls of d62-DPPC in DPPC MLVs.........107

6-2 Total order parameter values for terminal methyls of d62-DPPC in DPPC/BMP (5-
m ol% ) m ixed M LV s ........ .... ........ ................................ .......... ......... 108

6-3 Parameters defining order of the 16-DOXYL nitroxide spin-probe in DPPC MLVs in
5 m M N a+ buffer.............. .................... ................. .... ......109

6-4 Parameters defining order of the 16-DOXYL nitroxide spin-probe in DPPC MLVs in
105 m M N a buffer. ...................... ...................................................... ....... 110

6-5 Pre-transition temperatures obtained from 16-DOXYL labeled lipid in various MLV
lam ellar structures......... .......................................................... ............................. 111

6-6 Parameters defining order of the 16-DOXYL nitroxide spin-probe in DPPC/BMP (5
mol%) mixed MLVs in 5 mM Na buffer. ........ .. ..... ... ................................... 112

6-7 Parameters defining order of the 16-DOXYL nitroxide spin-probe in DPPC/BMP (5
mol%) mixed MLVs in 105 mM Na+ buffer. ........................... .................... 113

6-8 Span values for DPPC MLVs and DPPC/BMP (5 mol%) mixed MLVs .......................114









LIST OF FIGURES


Figure page

1-1 Lipid classification................................................. ...... ............ 25

1-2 Anatomy of 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) and D-erythro-
sphingosine-1-phosphate ............. .... ........ .................... .. .... .. .. ............ 26

1-3 Cross-sectional representations of lipid polymorphic structures in aqueous
en v iro n m en ts ....................................................................................................2 7

1-4 Geometric shape approximations and lipid aggregates ........................................... 28

1-5 E xam ples of labeled lipids ......... .................................................................. ..... .... 29

1-6 B M P structural isom ers........................................................................... .....................30

1 -7 G M 1 lip id ..................................................................................................................... 3 0

1-8 Sandhoff-Kolter model for lysosomal membrane digestion and endocytosis .................31

3-1 Selected lipid motions and associated correlation times in hydrated lamellar
structures ................................. .................. ..... ......... ............................ 47

3-2 An order parameter as a function of the angular displacement of the plane containing
a specific carbon and two deuterium atoms from the bilayer normal.............................48

3-3 Comm on organic radical spin-labels. ........................................ .......................... 49

3-4 Energy level diagram illustrating the electronic Zeeman and electron-nuclear
hyperfine interactions and the resulting derivative of EPR transitions for a spin-probe
such as, T E M P O in solution ..................................................................... ...................50

3-5 4,4-dimethyloxazolidine-3-oxyl labeled 5-a-cholestane-3-one .............. ...... ........ 50

3-6 Spatial dependence of the coupling strength of the anisotropic hyperfine interaction...... 51

3-7 Theoretical nitroxide label EPR line shapes. ......................................... ...............52

3-8 Energy level diagram illustrating the nuclear Zeeman and quadruplar coupling
interaction of 2H in an applied field Bo and for the 1H decoupled chemical shift of 31P
in an applied field B...................... ......................... ........ 52

3-9 Theoretical quadrupolar echo powder spectrum of a single deuterium labeled site
(r 0) ............................................................................................. 53

3-10 Graphical representation of CSA and EFG the angular orientations with respect to
the applied field (B o) ............... .................. ............. ............... .. ...... 53









3-11 Theoretical powder spectra of various lipid aggregates ....................................................54

4-1 M odel m em brane perturbants ................................................. ............................... 64

4-2 Positional isomers of spin-labeled lipids. ............................................... ............... 65

4-3 Phase Diagram for SDS and POPC LUVS............................................. ............... 66

4-4 cw-EPR spectra of POPC LUVs with 5-DOXYL spin probe ......................................67

4-5 cw-EPR spectra of POPC LUVs with 10-DOXYL spin probe ............... .....................67

4-6 Various spectral parameters of 5-DOXYL labeled lipid incorporated into POPC
LUVs as a function of SDS/Lipid concentration ratio............... .....................68

4-7 Various spectral parameters of 10-DOXYL labeled lipid incorporated into POPC
LUVs as a function of SDS/Lipid concentration ratio....... .. ......................................... 69

4-8 cw-EPR spectra of BMP with 5-DOXYL spin probe.......................... ................. 70

4-9 Various spectral parameters of 5-DOXYL labeled lipid incorporated into POPC
L U V s and B M P aggregates ....................................................................... ..................7 1

5-1 31P NMR chemical shift spectrum of BMPs18: MLVs .................................. .............78

5-2 cw-EPR spectra of POPC/BMP mixed LUVs ..... ...................... .............79

5-3 AHpp and Si of 5 and 10-DOXYL labeled lipid (1 mol%) incorporated into
POPC/BMP14:0 mixed LUVs as a function of BMP14:0/Lipid concentration ratio.............80

5-4 AHpp and Si of 5 and 10-DOXYL labeled lipid (lmol%) incorporated into
POPC/BMPi8:s mixed LUVs as a function of BMPis8:/Lipid concentration ratio.............80

5-5 31P NMR chemical shift of POPC MLVs in 5 mM HEPES.............................................81

5-6 31P NMR spectra ofPOPC/BMP18:1 MLVs in 5 mM HEPES, 100 mM NaCl ................82

5-7 31P NM R spectra ofPOPC/BM P18:1 M LVs .............. ... .............. .....................83

5-8 CSA span of (POPC/BMP MLVs) as a function of BMP mole fraction......................84

6-1 2H NMR spectra: DPPC and 5 mol% BMP/DPPC mixed MLVs.............................. 97

6-2 2H NMR spectra of DPPC MLVs and BMP/DPPC (5 mol%) mixed MLVs the in a
5 mM HEPES, 100 mM NaC1, at pH 7.4 buffer. ....................................................... 97

6-3 Total order parameter and residual quadrupolar splitting of terminal methyl groups
as a function of tem perature .......................................... ..................... ............... 98









6-4 cw-EPR spectra ofDPPC MLVs with 16-DOXYL spin probe (1 mol%) as function
of tem p eratu re ............................... ..................................................... 9 9

6-5 cw-EPR spectra ofDPPC MLVs with 16-DOXYL spin probe (1 mol%) as function
of temperature: 100mM DPPC, 5mM HEPES, 100 mM NaC1, 0.1 mM EDTA, at
p H 7 .4 ................................................................................ .. 9 9

6-6 Various spectral parameters of 16-DOXYL labeled lipid incorporated into DPPC
M LVs as a function of tem perature. ................................................... ..................100

6-7 DPPC/BMP MLVs with 16-DOXYL spin probe as function of temperature at low
ionic strength ............................ .................... .................. ................. 10 1

6-8 DPPC/BMP MLVs with 16-DOXYL spin probe as function of temperature (near
biological ionic strength) .................................... .......... .. .............. 102

6-9 Various spectral parameters of 16-DOXYL labeled lipid incorporated into
DPPC/BMP (5 mol%) mixed MLVs as a function of temperature. ............. ...............103

6-10 Various spectral parameters of 16-DOXYL labeled lipid incorporated into DPPC
MLVs and DPPC/BMP (5 mol%) mixed MLVs as a function of temperature ...............104

6-11 31P NMR chemical shift of single component DPPC MLVs and DPPC/BMP (5
m ol% ) M L V s ................................105.............................

6-12 31P CSA span of DPPC MLVs and DPPC/BMP (5 mol%) mixed MLVs as a function
o f tem p eratu re ............ ...................................................... ................ 10 6









Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

BIS(MONOACYLGLYCEROL)PHOSPHATE EFFECT ON MODEL MEMBRANE
MORPHOLOGY: A MAGNETIC RESONANCE INVESTIGATION

By

Chad E. Mair

May 2008

Chair: Gail E. Fanucci
Major: Chemisrty


Bis(monoacylglycerol)phosphate (BMP) or (S,R Isomer)sn-(3-Oleoyl-2-Hydroxy)-

Glycerol-l-Phospho-sn-3'-(l'-Oleoyl-2'-Hydroxy)-Glycerol,Ammonium Salt is a phospholipid

found primarily in late endosomes, and has a unique structure due to single acyl chains located at

the 3 and 1' positions on the glycerol components. BMP is known to play an important role in

late-endosome sorting functions, and is also thought to be involved in glycosphingolipid

catobilism. When BMP is present in liposomes containing ganglioside GM1, the enzymatic

hydrolysis of GM1 to GM2, stimulated by activator proteins, is dramatically enhanced.

This work is focused on determining the effect BMP has upon model membrane lipid

morphology, and acyl chain dynamics, using magnetic resonance spectroscopy, in order to

further understand BMP's role in lipid catabolism and lysosomal storage disease.












CHAPTER 1
INTRODUCTION

Lipid Description and Classification

In general, lipids can be described as molecules that are composed of mainly hydrogen and

carbon atoms, they can be as simple as alkanes or as complex as lipopolysaccharides(1). Lipid

molecules isolated from biological sources are generally classified as either neutral storage lipids

or zwitterionic/charged membrane lipids (Figure 1-1)(1-3). The chemical composition of storage

lipids (triacylglycerides or triacylglycerols) includes a glycerol backbone with three fatty acid

molecules linked via ester bonds to three reactive hydroxide groups on the glycerol backbone(3).

Membrane lipid chemical structures contain either a glycerol or a sphingosine backbone with

fatty acids linked to two of the reactive groups of the glycerol or to the reactive amine group of

the sphingosine (Figure 1-2). The other reactive hydroxide moiety on each backbone is linked to

either a saccharide or more commonly, linked to a phosphate group.

Most biologically relevant membrane lipids are amphipathic in nature, having both a polar

and an apolar region(]-3). The polar portion of a lipid is usually referred to as the polar head

group region, which is defined as the volume containing the substituted phosphate/saccharide,

the glycerol or the sphingosine backbone, and the alpha carbons of the carboxylic acid chains.

The apolar portion is defined as the acyl chain tail region(], 2). Usually, the apolar tails are long

hydrocarbon chains covalently bound via ester bonds to glycerol or amide bonds to sphingosine

backbone moieties (Figure 1-2)(1, 2). According to Yeagle, a "long chain" is defined as a

hydrocarbon chain of 12 or more carbon atoms(]).









Typical Membrane Lipid Structure

The most abundant eukaryotic membrane lipids are glycerophospholipids(1, 2), named for

the glycerol backbone and the phosphate group. These particular lipids are derivatives of sn-

glycero-3-phosphoric acid, in which the sni and sn2 positions commonly contain esterified, long

chain, fatty acids(2). The stereospecific numbering (sn) of phospholipids is derived by drawing a

Fischer projection of glycerol with the P-hydroxl group on the left, the sni position is then

located at the top of this projection while the sn3 position is located at the bottom. Typically, in

lipids isolated from natural mammalian sources, the acyl chain located in the sni position is

saturated and the acyl chain in the sn2 position contains at least one site of unsaturation or

double bond(]). Chemical moieties attached to the phosphate, located at the sn3 position of

membrane lipids, are referred to as head groups, the chemical identities of which give rise to the

diverse functionality found in biologically important lipids. For example, in

glycerophospholipids, the phosphate group has a negative charge, thus the chemical

structure/identity of the head group imparts the characteristic charge or zwitterionic character,

apparent size of the head group region, and propensity to hydrogen bond to other membrane lipid

molecules. These head group moieties include, but are not limited to, glycerol, inositol,

ethanolamine, serine, choline, and carbohydrates(], 2).

Lipid Self-Assembly and Organization

Because of their amphiphilic character, lipids with polar heads and apolar tails self-

assemble according to their solvent environment. In aqueous environments, the entropic

hydrophobic effect causes apolar acyl chain regions to organize in such a manner as to minimize

the free energy of the system and exclude the maximum number of water molecules from the

lipid assembly(], 2). Several polymorphs/mesomorphs or macroscopic aggregation states are

possible for solvated lipids, e.g. lamellar, hexagonal, micellar and inverted micellar (Figure 1-









3)(1, 2, 4). The specific polymorphism formed can be influenced by many factors including

structural features of the lipid, such as chemical composition of the head group region, length

and degree of saturation of the acyl chains, and extrinsic parameters such as temperature,

pressure, and degree of hydration(], 2, 4). Each of these factors provides a means of controlling

the packing density (interaction energy) of the individual lipid molecules within an aggregate,

which in turn can have dramatic effects upon the physical and morphological properties of the

assembly(4).

Under physiological conditions, the most common type of lipid organization found in

living cells is a two-dimensional lamellar or bilayer structure(4, 5). A bilayer is formed by

stacking lipid molecules tail-to-tail bound on either side by polar head groups in a repeating

pattern (Figure 1-3 A and B)(5). Although the biologically relevant mesophase/polymorph is

lamellar, other types of mesophase structures can exist under certain circumstances, e.g.

increasing the temperature or changing the hydration state of lamellar assemblies can result in

the formation of a hexagonal aggregate structure (Figure 1-3 C and D)(4). Lipids organized in a

three-dimensional hexagonal arrangement either surround a cylindrical column of solvent with

polar groups (HII) or organize with polar head groups facing out toward the bulk solvent phase

(HI). Micelles and inverted micelles are organized in a similar fashion as the hexagonal "phases"

except the aggregates are spherical in shape. The specific assembly formed is influenced by the

solvent environment and the apparent molecular shape of the lipids.

Molecules that occupy a molecular cross-sectional area resembling a cone, inverted cone,

or cylinder are predicted to form hexagonal (HII), micellar, or lamellar aggregates,

respectively(4). A simple geometric model based on relative cross-sectional areas of head

groups and acyl chains can be used to predict aggregation states for hydrated lipid molecules(4).









Lipid molecules with a head group cross-sectional area greater than the cross-sectional area of

their acyl chains self-assemble into a micellar type structure. Those with head group area

approximately equal to the acyl chains area prefer to assemble into a lamellar type structure,

while lipids with head group area less than the area of the acyl chains assemble into a hexagonal

type structure. An illustration of the geometric shapes and representative lipids of each category

can be found in Figure 1-4. Sodium dodecyl sulfate (SDS) has a head group with a much larger

cross-sectional area than its single acyl chain, thus hydrated SDS molecules will polymorph into

micellar aggregates. A different situation arises for 1,2-dipalmitoyl-sn-glycero-3-

phosphocholine (DPPC). The phosphocholine head group area and the cross-sectional area of

the two palmitoyl chains are similar; therefore, DPPC is predicted to form a lamellar aggregate in

an aqueous environment. Phosphatidic acid (PA) in an acidic environment has a small head

group area with respect to the area occupied by the two acyl chains and forms a HII aggregate.

Two broad categories of "phase" transitions can be defined for lipid polymorphs or

aggregates: 1) those that result from changes in organization and packing of the acyl chains and

mean volume occupied per lipid molecule, and 2) those that alter the mesophase or polymorphic

structure of the lipid aggregate(], 4). Aggregation state, or polymorphic transitions, can be

altered in various ways including partitioning of a variety of molecules into the assembly,

changing temperature, degree of hydration and lipid composition of the assembly, and varying

ionic strength or pH of the solvent(], 2, 4, 6), or via changes in extrinsic variables such as

temperature and pressure. However, in biological systems, the pressure can often be assumed to

remain constant; hence, most studies of membrane biophysics focus on the effects of changes in

temperature, not pressure. A common phase transition in lamellar mesophases exploited within

the membrane biophysics field is the lamellar gel to lamellar liquid transition(], 2, 4). This









transition is illustrated in Figure 1-3. The gel phase (Figure 1-3 A) is characterized by closely

packed molecules and a more extended acyl chain conformation. Whereas the liquid state

(Figure 1-3 B)) retains a two-dimensional order but lipid molecules are diffusing axially and the

acyl chain region is less ordered (high probability of trans/gauche isomerization) when compared

to the gel phase(4, 7). Factors such as solvent ionic strength, degree of acyl chain saturation,

lipid composition (mixtures of lipid molecules) and addition of protein/peptides perturbants can

all affect the temperature at which this thermotropic phase transition occurs(], 2, 4). For

example, at temperatures below the main transition temperature (T,), fully hydrated DPPC exists

in either a lamellar gel state with saturated acyl chains extended and highly ordered (very few

gauche rotamers), or in a ripple phase (Pp') with lipid molecules slightly tilted with respect to the

bilayer normal(8) resulting in a decrease in bilayer thickness with respect to the lamellar gel

state. The ripple phase is characterized by tilted and extended acyl chains that appear to have a

symmetric rotational axis (defined in Chapter 3), thus the lipids are more disordered with respect

to the gel phase but more ordered than the liquid lamellar phase(]). However, at temperatures

above Tm the chains become less extended, thus reducing the bilayer thickness to a greater

extent, making the chains more disordered until a pure liquid crystalline phase (La) is present.

Also, fully hydrated, non-bilayer structures, e.g. HUI, can be induced by further increasing the

temperature(4, 9). An example of a lipid exhibiting a transition to the HI state from either the

lamellar liquid or lamellar gel state is phosphatidylehanolamine (PE)(4, 9).

Characterization of Model Membrane Structural Properties

X-ray diffraction, neutron scattering, solid state nuclear magnetic resonance (SSNMR),

and electron paramagnetic resonance (EPR) are techniques commonly used to investigate the

structure and dynamical properties of lipid assemblies(2, 10, 11). Although diffraction and









scattering methods provide a detailed description of the thermodynamic phase structure of lipid

assemblies and bilayer thicknesses, these techniques measure only a static macroscopic structure

with little information regarding molecular motion and require a periodic lattice(2, 12). A

periodic lattice or an array of a large number of molecules arranged in a periodic structure

typically occurs in very pure systems. However, biological membranes are inherently

heterogeneous; thus, diffraction techniques are limited to purified systems in which a well

organized lattice structure has been induced. In some cases, diffraction techniques can be used

to measure the macroscopic structure in less organized systems but they lack the resolution

obtainable by magnetic resonance and other spectroscopic techniques.

Investigations utilizing NMR and/or EPR provide detailed molecular level information

regarding the degree of order at individual chemical bonds for particular molecules(10, 11, 13-

16). The types of order/disorder commonly investigated using these spectroscopic techniques

include a emotionally averaged picture of phospholipid head group and acyl chain angular

orientation with respect to the bilayer normal(10, 16, 17). Because all phospholipids (by

definition) contain a phosphate group, the phosphorus atom provides a natural probe for 31P

NMR investigation of average head group orientation. Advantages of this technique include

100% natural abundance of the 31P isotope and a good sensitivity due to a relatively large

gyromagnetic ratio (y is about 40% of the 1H nucleus). It has been shown that solid-state 31P

NMR chemical shift line shapes can be used to differentiate between lamellar, hexagonal (HUI),

and isotropic polymorphic assemblies(4); however, line shape can only be consistent with a

particular mesomorphic state. X-ray or cryogenic electron microscopy measurements are needed

to fully confirm the aggregate morphology.









The degree of acyl chain organization of the bilayer interior can be investigated either by

2H NMR studies of lipids containing deuterated acyl chains or by EPR studies of lipid

dispersions containing 0.5 1 mol% of a lipid labeled with a paramagnetic nitroxide spin- probe

(Figure 1-5)(10, 11, 14, 18). A major disadvantage for both EPR and 2H NMR measurements is

the need for a synthetic probe. Additionally, 2H has a y value that is only 15% of the 1H nucleus

and requires large samples compared to 31P or EPR sample sizes due to its relatively low

sensitivity. Despite these constraints, magnetic resonance techniques are invaluable for

biophysical investigations of heterogeneous, non-periodic model and biological membranes.

Thus, they are the primary tools employed in our pursuit to describe the conformation and acyl

chain order of mixed lipids in a model membrane system.

Bis(monoacylglycerol)phosphate

Since first isolation from ovine lung homogenate in 1967, bis(monoacylglycero)phosphate

(BMP) (Figure 1-6), also known as lysobisphosphatidic acid (LBPA), has been found to

represent less than 1% of total phospholipid mass in most tissue and cell types(19, 20). Although

a majority of mammalian cells contain a small amount of BMP, its concentration in late

endosomes (LE) is elevated to near 15% of the total lipid content (21-23) and up to 70% of the

total lipid content of internal membrane domains within the LE(23).

Endosomes are intercellular organelles that act as a staging area for sorting endocytosed

material either back to the plasma membrane for recycling or to specialized organelles

(lysosomes) for degradation(20, 24). Endosomes can be identified either by the presence of

internal membranes, also known as multivesicular structures, inside the lumen of a limiting

(boundary) membrane or by lipid composition(22, 25). Early endosomes have a limiting

membrane with a lipid composition very similar to that of the plasma membrane, whereas late









endosomes are characterized by an absence of a significant amount of cholesterol and a relatively

high concentration of BMP(25). In certain situations, multilamellar internal membrane

structures (multivesicular bodies or MVBs) are present in late endosomes(26). A question that

naturally arises is: "Do elevated BMP levels in these cellular structures play a significant role in

controlling membrane organization?"

BMP is negatively charged and has an atypical sn-l-glycerophospho-sn-l'-glycerol

(sn l:sn ') stereoconfiguration (27) with respect to sn-3-glycerophosphate structures exhibited by

most other glycerophospholipids(], 2, 20). This unique structure and negative charge is likely to

have functional implications beyond its ability to resist degradation by most phospholipases due

to their stereospecific recognition of the sn-3 stereoconfiguration(25, 27).

Even though the snl:snl' and sn3:snl' stereoisomers are different, similar thermotropic

phase transition temperatures were measured by differential scanning calorimetry (DSC) for 1,3'

dimyristoyl snl:snl' (40C) and 1,3' dimyristoyl sn3:snl' (42C) BMP structural isomers(28),

justifying the use of the non-natural stereoisoform as a first approximation for characterization

purposes. However, it has been reported that sn-(3-hydroxyl-2-oleoyl)glycerol-1-phospho-sn-l'-

(3'-hydroxy-2'-oleoyl)glycerol may be the biologically relevant isoform(23). It has also been

shown by gas chromatography and mass spectrometry that the major fatty acid components of

BMP, isolated from baby hamster kidney (BHK) cells, are oleic acids (91%)(23).

Three structural isoforms are shown in Figure 1-6; the synthetic (S,R or 1,3' diacyl sn3,

snl') molecule was used in this work due to its commercial availability and is designated

throughout this dissertation as either BMPis:l, BMP14:0, depending on the fatty acid substituents,

or collectively BMP. Because each acyl chain of BMPis:1 contains an unsaturated site, this lipid

may increase the overall bilayer disorder of model and endosomal membranes due to









intermolecular packing constraints(29). Increased lipid disorder is most likely not the only role

BMP plays in modulating endosomal membrane morphology because other negatively charged

lipids containing unsaturated chains, such as phosphatidylinositol (PI), would most likely have a

similar effect on lipid order as BMP and thus affect a specific hydrolysis reaction in a similar

manner. This has been shown, but only at concentrations of PI much greater than that found in

vivo(30). Therefore, BMP's unique geometric structure must also be an important factor in

controlling endosomal morphology and molecular trafficking in late endosomal organelles.

According to literature, the geometric shape predicted for BMP is either a cone or inverted

cone(25, 26, 29, 31, 32). If the former were true, BMP would most likely exist in an HUl

aggregate or may induce HII aggregation as the relative BMP concentration is increased in model

or biological membranes. However, at pH 7.4 BMPis8: forms multilamellar vesicles and lacks

any three-dimensional structural changes according to fluorescence emission of pyrene labeled

lipids incorporated into BMP multilamellar vesicles (MLVs)(29). Also, studies of sn-(3-

hydroxyl-2-oleoyl)glycerol-l-phospho-sn-1'-(3'-hydroxy-2'-oleoyl)glycerol induces

multilamellar structure formation, as visualized by fluorescence and cryogenic electron

microscopy in the presence of a pH gradient in lipid mixtures having a similar composition to

that found in late endosomes(26). Moreover, the small and wide angle X-ray diffraction patterns

for both 1,3' dimyristoyl sn3:snl'and 3,3' dimyristoyl snl:snl' BMP are consistent with a

lamellar structure(28). There is not any indication that any isoform of BMP exists in or induces

either a HII or micellar structure in the current literature.

The electron microscopy and X-ray diffraction data described above provide a static

description of the lipid aggregate macromolecular structure. This dissertation reports results

obtained from morphological studies of BMP and BMP mixed with model membrane lipids









using magnetic resonance techniques to investigate the microscopic structure and dynamic

structural properties of BMP and BMP mixed with model membrane lipids.

Bis(monoacylglycerol)phosphate May Be Important to Ganglioside Catabolism

Gangliosides and glycosphingolipids (GSLs) are components of eukaryotic plasma

membranes and are involved in passing cellular signals from outside the cell to the cell

interior(2, 33). The degradation of these particular lipid molecules occurs in acidic cellular

compartments of lysosomes, specifically on the surface of intraendosomal and intralysosomal

vesicular structures (an example of a GSL in shown in Figure 1-7)(20, 34).

A model for endocytosis and GSL digestion, proposed by Sandhoff and Kolter, can be seen

in Figure 1-8(20, 34). Vesicles containing GSLs, which are destined for endosomal and

lysosomal compartments, begin as either invaginations or clathrin coated pits formed in the

plasma membrane(35). These vesicles fuse with early endosomes where some lipids and

proteins are shuttled back to the plasma membrane for recycling (36, 37), while others are sorted

from the limiting membrane of the early endosome and incorporated into intralysosomal

structures(20). Early endosomes mature into late endosomes that transiently fuse with lysosomes

where enzymatic digestion occurs(20). During this maturation process, the luminal pH of the

endosome decreases and the protein and lipid composition of the intralysosomal structures

change (20) becoming enriched in GSLs and BMP and depleted in cholesterol(20, 38).

Glycosphingolipid degradation occurs in the lysosome as a stepwise cleavage of

monosaccharide units from the oligosaccharide head group of the GSLs until the recyclable

biomolecule sphingosine is produced(20, 33, 34). In vivo, several accessory molecules are

needed to degrade glycoshingolipids containing head groups of four or fewer sugar molecules: a

water-soluble enzyme (the hydrolase), a sphingolipid activator protein (SAP), and possibly a

membrane surface including anionic phospholipids like BMP or PI(20, 33, 34). However, it has









been demonstrated that the enzymatic cleavage reaction does occur in vitro with micellar

ganglioside substrate in the absence of a membrane surface(30). SAPs are membrane binding

cofactors believed to have variable specificity for both membrane lipids and enzymes(20, 34).

These cofactors are required for the enzymatic cleavage and are believed to either extract the

ganglioside from the bilayer and present the lipid to the hydrolase for an aqueous reaction or to

lift the ganglioside slightly from the bilayer surface for hydrolysis (a membrane associated

reaction)(20, 33, 34).

Any dysfunction in the SAP, substrate, or hydrolytic enzyme, can lead to a particular form

of lysosomal storage disease. There are many diseases associated with storage of lipids in both

endosomes and lysosomes. These are classified by the non-degraded lipid or protein accumulated

in either the endosome or lysosome(20, 33, 34). Subclasses or variants of the storage diseases

are characterized by the particular molecule in which the defect occurred. For example, Tay-

Sachs, or GM2 gangliosidosis, is caused by mutations of the gene encoding for the enzyme P-

hexosaminidase A; GM1 gangliosidosis and Morquio Type B syndrome are caused by a

mutation in the enzyme GM1 3-galactosidase(20, 33).

It has already been shown that in the presence of BMP or PI in model membranes

(POPC/CHOL/GM2/BMP mixed lipid composition) the rate of GM2 hydrolysis by 3-

hexosaminidase A in the presence of the sphingolipid activator protein GM2-activator protein

(GM2AP) is more than two orders of magnitude faster than in BMP free liposomes(30). How

BMP influences the hydrolysis rate is still unknown, but it is possible that BMP modulates

surface charge, membrane order, vesicle size/shape, or any combination of these properties.

The following work presented in this dissertation focuses on model membrane

morphological perturbations caused by BMP. It is likely that perturbations such as modulating









membrane surface charge, changing membrane packing parameters, and preferential interaction

of BMP with GM2 are crucial for the interaction of the catalytic triad (P-hexosaminidase A,

GM2, and GM2AP) to function properly. BMP may, under certain conditions, preferentially

sequester various lipid substrates into small vesicular or multilamellar structures required for

transportation to the lysosome, which might provide the proper topology needed for efficient

enzymatic cleavage. In support of this hypothesis, there is not any evidence for microdomain

(raft) formation when BMP is present in POPC bilayers at either acidic or neutral pH(29).

Hence, it is likely that BMP may induce vesicle budding events, as opposed to microdomain

formation.

Biological Membranes

Biological membranes are important cellular structures, because they create selective

chemical and physical barriers between cells or organelles and their surroundings. Lipid

composition and geometrical packing patterns, along with incorporation of proteins and other

molecules into the lipid matrix, such as cholesterol, allow membranes to exhibit diverse

mesophases, surface properties, and permeabilities(1, 2, 4, 31, 39). The aforementioned

properties play a role in the ability of biological membranes to participate in small molecule,

protein and lipid trafficking, membrane fusion, and to act as platforms for catalysis(], 2, 40-42).

Biological membranes also adopt a variety of shapes, for example red blood cell membranes

resemble small biconcave discs, lysosomal membranes resemble hollow spheres, the

endoplasmic reticulum and the golgi apparatus membranes have very convoluted surfaces.

Differences in size and shape of biological membranes are likely to be related to function, and it

is known that lipid interactions with proteins and other molecules are often needed to modulate

membrane shape(], 2).









Dissertation Outline

The work presented in this dissertation focuses on the effect of BMP on the morphology of

model phosphoglycerol membranes. Magnetic resonance techniques were used to measure

phosphate head group orientation with respect to the bilayer normal and the average angular

excursion from the bilayer normal of either a nitroxide spin probe or a deuterated acyl chain as a

function of temperature and/or BMP concentration. By using both SSNMR (2H) and EPR, we

were able to monitor lipid acyl chain dynamics (order/disorder) on two different time scales:

slow motional fluctuations (-105 Hz) and rapid motional fluctuations (-108 Hz), respectively.

31P SSNMR measurements of chemical shift anisotropy can be correlated to head group order,

mesophase, and changes in phase transition temperature for several lipid aggregates.

After analysis of data collected with each of the previously mentioned techniques, we were

able to confirm the lamellar aggregation of pure BMP, and characterize some of the effects on

head group and acyl chain order caused by incorporation of BMP into model membranes.

General background information about lipids, lipid self-assembly, membrane characterization,

and specific information about BMP is found in Chapter 1. Chapter 2 reports the materials and

methods utilized throughout and experimental parameters needed to perform our investigations.

An overview of basic magnetic resonance applications and expected spectroscopic signals

obtained from membrane lipid dispersions are discussed in Chapter 3. The focus of Chapter 4 is

the solubilization of model (PC), pure BMP and PC/BMP mixed membranes. Chapters 5 and 6

report results on characterization of overall perturbations of the acyl chain and head group

regions of model membranes caused by BMP.

In the field of membrane biophysics, it is strongly desired to understand how proteins and

lipid molecules alter the physical shapes of membrane structures within cells(38). A detailed

understanding of how BMP modulates bilayer physical properties, at the molecular level, will










directly impact other research in the Fanucci group that is focusing on characterizing the

membrane binding interactions of GM2AP with POPC and POPC:GM2 containing vesicles.

Results from this work may also begin to explain why BMP alters the enzymatic rates of GM2

hydrolysis when it is incorporated into lipid vesicles. On a broader scale, the relatively high

levels of BMP within the late endosome have been shown using fluorescence microscopy to lead

to vesicle budding and multilamellar structure formation(26).


Storage
Lipids
(Neutral) Membrane Lipids (Polar)

Phospholipids Glycolipids


Triacylglycerols Glycerphospholipids Spingolipids Sphingolipids Galactolipids

Fa Acidl Iatt Acid Fa Acid
a Acid at Acid a Acid --|Fatty Aci Fa Acid
-Fat Acidl P c PO hol m -PO line m -\Saccharide -Saccharide SO


Figure 1-1. Lipid classification. Adapted from Lehninger, Principles of Biochemistry, 4th
edition(3).

















Head group


A)


- Polar region
r


sn3
Glycerol backbone
snl








snl chain


Figure 1-2. A) Anatomy of 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) and B) D-
erythro-sphingosine- l-phosphate.


sn2 chain






A) ffiffiff-ff
ayVJKK~


0 "


B)
B5 55ci%
%%%%%%R~~


D)


*C~


*

*


Figure 1-3. Cross-sectional representations of lipid polymorphic structures in aqueous
environments: A) lamellar gel (Lp'); B) lamellar liquid crystalline (La); C) hexagonal
(HII); D) hexagonal (HI). Adapted from Gruner et al. Ann. Rev. Biophys. Chem.
1985(4).










Lipid


Sodium dodecyl sulfate
(SDS)


1,2-Dipalmitoyl-sn-
Glycero-3-
Phosphocholine
(DPPC)


Geometric Shape


Associated
Mesophase
Micellar


Lamellar


Phosphatidic Acid Hexagonal (HII)
(pH < 3)


Figure 1-4. Geometric shape approximations and lipid aggregates. Adapted from Gruner et al.,
Ann. Rev. Biophys. Chem. 1985(4).











A) B)

O O
I I

o o





?CH3 Do O
>H H





3 D DD D
D D D D
D DD D
0D DD D
D DD D
D DD D
D D D D
D DD D
D D D D
D DD D



D D D
D D D



Figure 1-5. Examples of labeled lipids: A) 1-Palmitoyl-2-Stearoyl-(5-DOXYL)-sn-Glycero-3-
Phosphocholine; B) 1,2-Dipalmitoyl-D62-sn-Glycero-3-Phosphocholine.







































Figure 1-6. BMP structural isomers: A) (S,R isomer) sn-(3-oleoyl-2-hydroxy)-glycerol-1-
phospho-sn-3'-(l'-oleoyl-2'-hydroxy)-glycerol (ammonium salt), (1,3' dioleoyl
sn3:snl'or BMP8s: ); B) (S,S isomer) sn-(3-oleoyl-2-hydroxy)-glycerol- -phospho-
sn-3'-(1'-oleoyl-2'-hydroxy)-glycerol (ammonium salt) (3,3' dioleoyl snl:snl'); C)
(R,R isomer)sn-(3-oleoyl-2-hydroxy)-glycerol-l-phospho-sn-3'-(l'-oleoyl-2'-
hydroxy)-glycerol (ammonium salt) (1,1' dioleoyl sn3:sn3').

HO H
HMN 0 OH


D3 H 0H OH O
NHO M-ykN
HO
D--7 o OH~P~
NINH OH


Figure 1-7. GM1 lipid: GalBetal-3GalNAcBetal-4(NeuAcAlpha2-3)GalBetal-4GlcBetal-l'-
Cer (GM1 ganglioside).






















Endosome \ ER
Golgi

Lysosome Go i
S/ Temporal fusion
-' g and discharge
Vescicular ,
Transport? -

Glycocalix Lysosome


Figure -8. Sandhoff-Kolter model for lysosomal membrane digestion and endocytosis; Adapted
from Annu. Rev. Cell Dev. Biol., 2005(20).










CHAPTER 2
MATERIALS AND METHODS

Materials

The following lipids and lipid derivatives, dissolved in chloroform, were purchased from

Avanti Polar Lipids (Alabaster AL, USA), stored at -200C and used without further purification.

DPPC (1,2-Dipalmitoyl-sn-Glycero-3-Phosphocholine), d62-DPPC (1,2-Dipalmitoyl-D62-sn-

Glycero-3-Phosphocholine), POPC (1-Palmitoyl-2-Oleoyl-sn-Glycero-3-Phosphocholine), d31-

POPC (1-Palmitoyl(D31)-2-Oleoyl-sn-Glycero-3-Phosphocholine), BMP18:1 ((S,R Isomer)sn-

(3-Oleoyl-2-Hydroxy)-Glycerol- -Phospho-sn-3'-(l'-Oleoyl-2'-Hydroxy)-Glycerol,Ammonium

Salt)), 5-DOXYL (1-Palmitoyl-2-Stearoyl-(5-DOXYL)-sn-Glycero-3-Phosphocholine), 10-

DOXYL (1-Palmitoyl-2-Stearoyl-(5-DOXYL)-sn-Glycero-3-Phosphocholine), 16-DOXYL (1-

Palmitoyl-2-Stearoyl-(5-DOXYL)-sn-Glycero-3-Phosphocholine). HEPES (4-(2-hydroxyethyl,)-

1-piperazineethanesulfonic acid, CsH18N204S), EDTA (ethylenediamine tetraacetic acid,

C10H16N20s), sodium citrate (Na3C6H5O7), and NaCl (sodium chloride) were purchased from

Fisher Biotech (Pittsburgh, PA). CHC13 (chloroform), C6H12 (clyclohexane), CH3OH

(methanol), 12 (iodine), PrC13 praseodymiumm chloride), HC1 (hydrochloric acid), NaOH (sodium

hydroxide), H2SO4 (sulfuric acid), and NH4OH (ammonium hydroxide) were obtained from

Fisher Scientific (Pittsburgh, PA). Reagents for the phosphate assay were purchased as a kit

(QuantiChromTM Phosphate Assay Kit (DIPI-500) from BioAssay Systems (Hayward, CA).

Multilamellar Vesicle Preparation

Multilamellar vesicles (MLVs) for phosphorus (31P) and deuterium (2H) experiments were

prepared from mixtures of lipids dissolved in chloroform. Stock bottles containing lipids in

chloroform were allowed to reach room temperature before opening. Prior to vesicle









preparation, glassware was cleaned with methanol and chloroform; flasks and syringes were

rinsed at least once with HPLC grade methanol and washed three times with chloroform.

The desired amount of lipid stock solution was drawn into a clean, gastight, syringe

(Hamilton) and transferred to an appropriate container. If a mixture of two or more lipids was

prepared, the previous procedure was repeated for each component of the lipid mixture. After

use, the headspace of all stock bottles were covered with argon or nitrogen gas, the cap was

wrapped in Parafilm or Teflon tape and returned to the freezer. Mixtures were prepared in either

round bottom or pear-shaped flasks (the tapered bottom of the pear-shaped flask is helpful during

the hydration of the lipid film), and solvent was removed by rotary evaporation at room

temperature using a water-cooled solvent trap. Dry lipid films (both single component and

mixtures) were re-dissolved in 4:1 (v:v) co-solvent system of cyclohexane and methanol (HPLC

grade) to ensure complete mixing of the component lipids prior to lyophilization. This

dispersion was gently mixed and flash frozen by placing the flask into a container with liquid

nitrogen (1N2). Frozen lipids were lyophilized, typically overnight and subsequently hydrated

with an appropriate buffer (5:1 buffer:total lipid, mass:mass) in a 55C oven for at least one hour,

followed by gentle vortex mixing of the hydrated dispersion; however, samples containing

mostly POPC were hydrated at room temperature (POPC has a main phase transition near

0C)(43). General lipid handling and sample preparation procedures can be found at

www.avantilipids.com.

For nuclear magnetic resonance (NMR) measurements, approximately 200 ptL or enough

of the hydrated MLV dispersion to fill the tube was used for 31P experiments. For 2H

measurements 750 ptL was loaded into a custom cut (- 38 mm long for 2H), 5mm outer

diameter NMR tube. These tubes were either special ordered (Wilmad) or conventional tubes









were cut in our laboratory to fit inside the RF coil of the solid state probe. Prior to sample

loading the NMR tubes were washed with sample buffer. Buffers were first prepared using

nano-pure deionized water, and brought to the appropriate pH with either HC1 or NaOH.

Aliquots of each buffer were then flash frozen with N2 (/), lyophilized, and re-hydrated with an

equivalent volume of deuterium depleted water (Cambrige Isotope Laboratories)(44). Total lipid

concentration for NMR experiments was near 275 mM based on theoretical values. Either two

or four mol% of d62-DPPC was used in lipid samples prepared for 2H measurements.

MLVs and large unilamellar vesicles (LUVs) for electron paramagnetic resonance (EPR)

experiments were also prepared from mixtures of lipids dissolved in chloroform. However, due

to small sample volumes used for EPR measurements, lipid mixtures were prepared in a glass

culture tubes, and blown dry with a stream of nitrogen gas. Dry lipid films were placed under

vacuum overnight then hydrated with an appropriate buffer in a 55C oven for at least one hour.

Mixtures containing mostly POPC were hydrated at room temperature. For hydrated MLVs,

total lipid concentration was near 100 mM, (4-5 [tL) were loaded into a 1.8 mm inner diameter

(id.) borosilicate capillary tube purchased from Vitrocom. Typically, samples contained 1 mol%

of the DOXYL spin label for EPR measurements.

LUVs were prepared by extrusion through polycarbonate membranes. Gas tight syringes

were cleaned with chloroform, rinsed three times with methanol, and finally thoroughly rinsed

with sample buffer. The mechanical extrusion device was assembled according to the

manufacture's instructions (www.avantilipids.com). A volume of sample buffer was first

injected into the extrusion device to fill the void volume in the extrusion chamber. An

appropriate amount of lipid solution was loaded into one syringe and inserted into the extrusion

device and the solution was passed through a membrane (typically 100 nm pore size) a minimum









of 55 times, until the lipid suspension became transparent. One pass is defined as solution

moving from the starting syringe across the polycarbonate membrane and back to the starting

syringe.

Phosphate Assay

Both MLV and LUV suspensions were diluted to obtain a concentration near 30 ng/100

[IL. Three 100 |pL aliquots of each diluted sample were placed in separate labeled glass culture

tubes. Specific quantities of a phosphorus standard purchased from BioAssay Systems, ranging

between 0 and 100 nmol, were placed in separate tubes; each standard quantity was made in

triplicate in order to estimate the error of the standards. A volume (0.45 mL) of 8.9 N sulfuric

acid was added to each tube and the tubes were placed in an oven at 2200C for 25 minutes to

decompose the lipid and liberate the phosphate. Sample tubes were cooled to room temperature,

and 3.9 mL of 1 M NaOH was added to each tube (this reaction is exothermic). Tubes were

again cooled to room temperature and vortex mixed. A volume of 800 pL was transferred from

each tube to a standard, disposable, UV-vis cuvette and 200 pL of malachite green phosphate

analysis solution (BioAssay Systems) was added to each cuvette. The contents of each cuvette

was mixed by inversion, and allowed to equilibrate for 10 minutes. Cuvettes were placed in a

UV-vis spectrometer and the absorbance values at 650 nm were recorded. Unknown sample

absorbance values were compared to the calibration curve prepared from the absorbance values

of the standard solutions. This procedure for determination of phosphorus content has been

modified from the general procedure of Warner et. al., 1956(45).

Thin Layer Chromatography

Lipid integrity was verified by thin layer chromatography (TLC) before and after exposure

to magnetic resonance radio frequency pulsing as the ester bonds in the lipids are somewhat









labile. A very small amount (- 2 [LL) of lipid dissolved in chloroform or hydrated lipid

suspension was added to seven drops of chloroform and spotted on silica coated aluminum plates

(Whatman) purchased from Fisher Scientific. TLC plates were placed in a chamber containing a

mobile phase of 65:25:4 (by volume) chloroform:methanol:ammonium hydroxide (14 N) and the

solvent front was allowed to migrate to approximately 75% of the plate height. For quantitative

experiments, TLC plates were washed in the mobile phase, and air dried prior to spotting of the

sample. The stationary phase was allowed to dry and the slides were exposed to 12 vapor for

visualization of the lipid fractions.

Magnetic Resonance

Wide line 31P NMR experiments were performed using a Tecmag spectrometer operating

at a resonance frequency of 145.2 MHz. Spectra were acquired with a CP/MAS probe purchased

from Doty Scientific, Inc., with variable temperature capability. A two pulse Hahn echo pulse

sequence with CYCLOPS phase cycling was used with 1H decoupling. Details of the pulse

sequence include an echo pulse spacing of 40 his, a 4 |ts 31P pulse (76/2x), a 5 s recycle delay, 5 |[s

dwell time, 1024 8192 time domain data points. A minimum of 2048 transients were averaged

for each experiment.

2H experiments were performed on a Bruker Avance spectrometer operating at a resonance

frequency of 61.4 MHz. Spectra were acquired with a high power, broad band, and high

temperature probe manufactured by Bruker. A standard quadrupole echo pulse sequence was

used with 3.2 ts excitation pulses, 40 ts pulse spacing, a 500 |s recycle delay, and a 4 |s dwell

time. Typically 8192 time domain data points were collected and a minimum of 14400 transients

were averaged per spectrum.









Paramagnetic resonance experiments were completed using a modified ER200 (Bruker)

with an ER023M signal channel, an E032 field controller, SPEX data acquisition software and a

loop gap resonator (Medical Advances, Milwaukee, WI). Typical EPR experimental parameters

are a 100 G spectral width, 20 mW of average microwave power, and a 0.16 s time constant.

The sample temperature was controlled by flowing either compressed air or nitrogen through a

copper coil immersed in a circulating water bath and passing the gas over the sample tube.

Data Processing

Raw 31P free induction decays were base line corrected, zero filled to twice the number of

data points collected in the time domain, left shifted (if required), Fourier transformed, apodized

by exponential multiplication (100 Hz), and phase corrected using the NTNMR software

provided by Tecmag.

Raw 2H free induction decays were base line corrected, zero filled to twice the number of

data points collected in the time domain, left shifted (if required), apodized by exponential

multiplication (100 Hz), Fourier transformed and phase corrected using a Matlab routine

provided by the Long research group.

EPR line shapes were baseline and phase corrected and area normalized using Labview

software written by Dr. Christian Altenbach (UCLA, laboratory of Dr. Wayne Hubbell). Second

moments and peak-to-peak widths were also calculated by Labview software written by Dr.

Altenbach.










CHAPTER 3
MAGNETIC RESONANCE APPLICATIONS IN MEMBRANE BIOPHYSICS

Hydrated Lipid Motions and Order

Lamellar biological and model membrane systems are often described as two-dimensional

"fluids" where lipids, proteins, and other molecules are allowed to diffuse laterally within the

boundaries of a two-dimensional matrix(], 2). However, this dynamic description should also

include many other motional degrees of freedom for individual lipid molecules including

individual bond vibrations and rotations, trans/gauche isomerization of the acyl chains,

molecular rotation about and wobble of the symmetric, molecular long axis perpendicular to the

bilayer normal, and lipid exchange or "flip-flop" between the two monolayers of the bilayer

structure. The time scales of these motions span more than fifteen orders of magnitude, from

bond vibrations, on the order of femtoseconds to flip-flop motions on the order of seconds to

minutes(], 3). Figure 3-1 depicts a cross-section of a bilayer plane and illustrates three selected

motions and their respective reorientation times. Lipid molecules in Figure 3-1 are represented

by circles (polar head groups) and squiggle lines (apolar acyl chains). Figure 3-1 A illustrates

the molecular rotation and wobble motion of a lipid molecule in the top monolayer of the

membrane, B illustrates the molecular exchange that occurs during lateral diffusion, and C

illustrates the flip-flop exchange of lipid molecules between the two monolayers.

Spectroscopic methods such as NMR and EPR can provide information regarding the

aggregation state and the orientation of single bonds, or molecules within a membrane with

respect to a reference coordinate system, usually the long axis of the lipid molecule, or the

bilayer normal in lamellar systems(4-7). Average orientations are determined by mapping the

experimental observables such as the principal components of the chemical shift, the quadrupole

coupling, and the hyperfine tensor values onto the molecular frame. Specific details regarding









these tensors and the spectroscopic signals obtained for each method are discussed later in this

monograph.

Experimental signals acquired by these techniques are related to the time averaged

reorientation of the previously mentioned tensor values and are usually converted to order

parameters (Si) used to describe average tensor orientations with respect to a reference

coordinate system in the static limit. In general, order parameters for unoriented systems in a

uniaxial magnetic field are expressed as second order Legendre polynomials: (S = /2 (3 cos2() -

1)), and are functions of the average angle (0) between the molecular frame and the reference

frame (the magnetic field). Unfortunately, the experimental observables for membranes that are

not mechanically oriented with respect to the magnetic field provide a total order parameter that

is the product of two order parameters, one describing the orientation of the molecule or bond

with respect to the bilayer normal and the other describing the orientation of the bilayer normal

with respect to the magnetic field. Figure 3-2 shows the order parameter of a specific molecular

site (i) as a function of the angular orientation of that site with respect to the bilayer normal.

According to the mathematical expression for order parameters, the maximum occurs at an angle

of 0 and is assigned a value of +1, while the minimum occurs at an angle of 90 and is assigned

a value of- 12.

Two important pieces of information must be kept in mind when evaluating or reporting

order parameters: 1) A small order parameter value is not necessarily low because of rapid,

random motional averaging; for example, it may be low because the tensor is oriented at the

magic angle (54.70), which leads to an intrinsic order value of zero, and 2) Different

spectroscopic techniques are sensitive to different time scales of tensor reorientations. For

example, NMR reports motions reorienting faster than 10 |ts (such as trans/gauche









isomerization) as an averaged resonance weighted by the number of individual conformations

represented. Techniques such as infrared (IR) spectroscopy, however will report the same

motions as a superposition of individual resonances. Therefore, one should always keep in mind

the time bases of the experimental methods when comparing order parameters.

Nitroxide Spin-Probes

Stable free radical spin probes have been used for many years to investigate the dynamics

of model and biological membranes assemblies(4, 8-15). A variety of hydrophobic and

amphiphilic organic spin-probes have been employed such as, 2,2,6,6,-tetramethylpiperidine-1-

oxyl (TEMPO), paramagnetic oxazolidine ring labeled fatty acids, and paramagnetic oxazolidine

ring labeled lipids(13-15). The structures of selected spin-labels are shown in Figure 3-3; A

depicts a hydrophobic nitroxide spin-label (TEMPO), B) depicts the ring structure incorporated

into the acyl chain regions of the fatty acid and the labeled lipid shown in C and D, while C and

D are amphiphilic nitroxide spin-labels.

Description of Nitroxide Spin-Label Order Parameter in Hydrated Lipid Bilayer
Assemblies Obtained by Electron Paramagnetic Resonance

Electron paramagnetic resonance (EPR) spectra of spin = /2 molecules, such as nitroxide

radicals, trapped as impurities in crystals with fixed orientation can be described by the basic

spin Hamiltonian of McConnell and McFarland (Eqn. 3-1), neglecting proton hyperfine

interactions and including only the g (electron screening tensor) and T hyperfinee coupling

tensor).(13) The hyperfine tensor is a result of coupling between the electron angular

momentum operator (S) and the nuclear angular momentum operator (I). For a nitroxide radical

(S = 1/2 and I = 1), and the hyperfine interaction will give rise to three absorption lines, thus the

EPR derivative signal has three peaks. The exact details of the EPR transitions and line shapes

are discussed later. 3 and PN are the Bohr magnetons for the electron and nucleus respectively, g









is the electron screening tensor (analogous to the NMR chemical shift), T is the hyperfine

coupling tensor, S and I are spin angular momentum operators, and h is Planck's constant.

H, = /PS.g.H, +hS.T.I-gNNIHN (3.1)

An energy level diagram illustrating electronic Zeeman and hyperfine interactions of a

representative spin-probe in an external field along with the resulting derivative EPR spectrum is

illustrated in Figure 3-4. When an unpaired electron is placed in an external field (Bo) the

degeneracy of electron spin angular momentum is lifted, which is referred to as the electronic

Zeeman interaction. For a nitroxide radical, the electron spin angular momentum and nuclear

spin angular momentum are coupled through a hyperfine interaction tensor (T), due to the

interaction of the electron angular momentum and the 14N nucleus with a magnetic moment,

splitting each electronic spin state into (21 + 1) states. The hyperfine interaction is not dependent

upon field strength, but has an anisotropic spatial dependence that determines the energy splitting

(spacing between peaks in derivative mode) of the three allowed transitions (see Figure 3-5).

Selection rules are (Ams = +/- 1) and (Ami = 0) for a one photon process. The resulting energy

transitions can be probed with microwaves oscillating at the correct frequency.

However, as discussed earlier in this chapter, self-assembled, spin-labeled fatty acids and

phospholipids in an aqueous environment are neither solution nor static structures but undergo

rapid, anisotropic molecular motions about the molecular long axis. This motion must be

described by a time dependent equation(4, 13). This time dependent Hamiltonian can be

separated into two parts: 1) a time-independent, effective Hamiltonian (H,) (Eqn. 3-2) and 2) a

Hamiltonian that is a function of time. As long as the time dependent fluctuations are

sufficiently averaged; H, will adequately describe the system(4, 13).

H= pllS.g'HHe, +hS.T'.I-gNNI.HN (3.2)









Unfortunately, experimental powder line shapes of hydrated lipids only yield information

about molecular orientation with respect to the laboratory reference frame and must be related to

the molecular axis system using an oriented reference molecule. The reference molecule used

for understanding spin-labeled lipids was spin-labeled 5-a-cholestan-3-one (Figure 3-5), with g

values (gx,gy,gz (2.0089, 2.0058, 2.0021 +/- 0.001)) and T values (Tx,Ty,Tz (5.8, 5.8, 30.8 +/- 0.5

G)) calculated by Hubbell(13, 16). However, if g' and T' tensors in Eqn. 3-2 are appropriate

time averages of g and T from Eqn. 3-1, the energies and thus the two effective tensors, can be

related to calculated tensor values of the trapped radical through the averaged squared directed

cosine projected onto the principal axis system of the trapped nitroxide(4, 13). Equations 3-3

and 3-4 show the energy obtained from the effective Hamiltonian in the laboratory frame and in

terms of the basic Hamiltonian, respectively. From Eqn. 3-3 and 3-4,using the relation c2 + 32 +

72 = 1, and since Txx and Tyy are equal for 4,4-dimethyloxazolidine-3-oxyl,(13) the value of Tzz,'

can be derived and is shown in Eqn. 3-5. In Eqn. 3-5, a is the isotropic component of the

hyperfine tensor (a = 1/3 Tr(T) = 1/3 Tr(T')) and S is the order parameter. It can be shown that

S can be related to calculated values of Txx and Tzz and experimental values of and T' as seen

in Eqn. 3-6. Because the hyperfine interaction is dependent on solvent polarity the order

parameter requires a slight correction which can be seen in equation 3-7(9, 13).

E= f g SgH, H, +hT S I, g, I, *H, (3.3)

E = (a2 g + (22g, + ,2 gzz)S, H +h(a2T +pf2T + r2T)Sz I, g l z, Hz, (3.4)


T' = a2TX + 2T= a + -T )S (3.5)
S 2- 3

1 T'-T
S (32 1)= (3.6)
2 TZ T










S = (3.7)


Orientation of the Nitroxide Spin-Label in Hydrated Lipid Bilayer Assemblies and
Expected EPR Lineshapes

Referring to Figure 3-3 C and D and recalling the orientation of individual hydrated lipid

molecules in a bilayer aggregate, one can see that the nitrogen-oxygen bond is oriented parallel

to the bilayer surface, making the orbital designated pz perpendicular to the bilayer surface and

parallel to the acyl chain long axis. The electron spin angular momentum is most strongly

coupled to the 14N nuclear angular momentum in the direction of pz and the hyperfine coupling is

minimal in the plane perpendicular to this orbital; therefore, the largest component of the

hyperfine coupling tensor is Tz and Tz > Txx= Tyy(4, 17). This is again illustrated in Figure 3-6

where the z-axis represents the nitrogen pz orbital and the x-axis represents the N-O bond

vector(4, 17). Even though X-band (9 10 GHz) EPR is dominated by the hyperfine interaction,

it is important to note that the g tensor is also anisotropic with the largest value in the direction of

the N-O bond (because g is a first order function of field it becomes significant in high field

EPR)(17). Figure 3-7 illustrates two representative line shapes: spectrum A represents an axially

symmetric, nitroxide label with restricted motion (broad spectral lines) and visible splitting of the

low and high field lines, while spectrum B) represents an axially symmetric, nitroxide label

without restricted motion (narrow spectral lines).

The central resonance line is invariant to changes in the hyperfine interaction(1 7);

therefore, the central peak-to-peak line width (AHpp Figure 3-7 B)) is also a good approximation

of relative mobility of the spin-label. There are two other methods commonly used to determine

mobility of nitroxide spin labels: normalized fractional intensity (fi), and second moment (M2)

analysis(4, 13, 18). Normalized fractional intensity, as a function of a dependent variable (e.g.









temperature), tracks the change in intensity (I) of a specific spectral position with respect to an

intensity designated as final (If) divided by the difference in an intensity designated initial (Ii)

and the final intensity (If) (Eqn. 3-8)(19). This is an adequate method for comparing mobility

within a set of similar experiments, but for small changes in spectral intensity the level of noise

must be near zero.

I-II
f = -- (3.8)
I -I

The coupling strength or spitting of the z component of T' (Tz'z,') measured for a specific

nitroxide labeled site yields information about the relative order of an aggregated lipid system in

the vicinity of that site. This coupling strength is directly related to angular deviations (y) of the

spin-probe (pz orbital) from a plane perpendicular to the bilayer normal, and indirectly related to

local environmental order. These angular excursions for the bilayer normal are mainly caused by

trans\gauche isomerizations in the local vicinity of the probe(5, 13, 20). The probabilities of

these isomerizations for free polymer chains and lipids in an assembly are a function of the total

internal energy of the hydrocarbon chain in a specific conformation(5, 20). However, due to

interaction with other molecules the total internal energy for assemblies must include both

intramolecular and intermolecular energies(5, 20). Examples of interaction energies that can be

included in the intermolecular group include van der Waals interactions, electrostatic repulsions

and thermodynamic energies from hydrophobic forces. Considerations must also be made for

restricted motion near the head groups (conical boundary condition)(5, 20). Moreover, in

assembled aggregates the isomerization probability (chain order) is a function of depth (or

location) in the hydrophobic domain of the aggregate establishing an average, site-dependent

order(13, 20). Although the EPR active spin-labeled lipids may perturb lipid aggregate









structures they still report at least qualitative information concerning the order/disorder of the

hydrated lipid aggregates as they would exist without the spin-label.

Solid State 3P and 2H NMR of Hydrated Lipid Aggregates

The following general discussion of NMR will focus on two specific nuclei: 2H and 31p.

2H is an I = 1 nucleus with a dominant quadrupolar coupling interaction, and 31P is an I = /2

nucleus with a dominant anisotropic chemical shift (CSA) interaction, under the condition of full

proton decoupling. There are two allowed transitions in the radio frequency energy range for I=

1 (2H) nuclei and a single allowed transition for I = 1 (31P) nuclei; both transitions obey the

selection rule Ami= 1. An energy diagram illustrating the quadrupolar splitting and allowed

transitions for an I = 1 nucleus can be seen in Figure 3-8. Application of a strong external

magnetic field lifts the degeneracy of the nuclear spin angular momentum of an I =1 nucleus;

therefore, three (21 + 1) nuclear Zeeman states, designated +1, 0, and -1 are present. Further

perturbations of these nuclear Zeeman states are caused by an interaction of the nuclear spin

angular momentum with the electric field gradient (EFG) at the quadrupolar nucleus. Energies

for these transitions are given in Figure 3-8, where eq is the electric field gradient, Q is the

quadrupole moment, e is the elementary charge, 0 is the angle of the principle axis of the efg

with respect to the applied field, q is the angle in the x,y plane, and r is the asymmetry

parameter(5, 21, 22). The value of 7 is defined as (|vxxl IvYY|)/|vzzl(5, 21), where vii are

diagonal elements of the electric field tensor V.

Lipid MLVs in an aqueous environment have all possible spatial orientations with respect

to the applied field, and yield the classic Pake powder spectrum (Figure 3-9), which is narrowed

and axially symmetric (r = 0) because of rapid rotational motion around the long axis of the lipid

molecule(5, 7, 21, 23). There are three distinct features of the Pake-doublet when all possible









values of 0 and q are considered: steps (Avzz), shoulders (Avyy), and singularities (Avxx).

Figure 3-10 shows the average angular orientation of the elements of both the electric field

tensor (V) and chemical shift tensor a with respect to the applied field (Bo) (5). Also note that

the quadrupolar splitting is invariant with respect to the applied filed. The difference in

frequency units between each location (u = -+ (3 cos2 0 -1- r/(- cos2 0)(cos 20))) of the


two maxima of a specific Pake doublet defines the averaged quadrupolar coupling constant and

the orientation (order parameter SCD) of a specific deuteron with respect to the applied field(5,

21). The value of the coupling constant (5, 21) can be compared to the static limit and related to

angular deviations and local order as discussed in the previous section for EPR probes.

Magnetic Resonance Line Shapes and Order in Hydrated Lipids

Spectroscopic information such as line shape and residual coupling strengths obtained

from the natural 31P and virtually non-perturbing 2H NMR experiments allow for indirect

aggregated structural assignments and determination of relative order/disorder in the head group

and acyl chain regions of the aggregates. 2H and 31P spectra of lipid aggregates have

characteristic line shapes, which can be associated with aggregation states and degree of order of

hydrated lipids. As discussed previously, order/disorder is defined as the degree of motional

averaging relative to the static limit.

Typical 31P and 2H line shapes and their corresponding aggregate types for hydrated lipids

are shown in Figure 3-11. Each carbon atom within the acyl chain region is deuterated and the

resulting 2H spectrum will be a superposition of quadrupole splitting from each site, but for

clarity only one site is shown. Visual inspection of 31P line shapes yields information about

aggregation type: lamellar gel phase spectra have a broad, asymmetric chemical shift anisotropy

(CSA) line shape; lamellar liquid crystalline spectra have a narrowed axially symmetric CSA









lineshape; very small MLVs, micelles, and cubic phases yield isotropic peaks; as more motional

averaging narrows the NMR spectral width(3, 21-23, 25). It is important to note that 31P powder

NMR cannot differentiate between aggregates with signals that appear isotropic, because the

chemical shift of each of these aggregation states is the isotropic value Oiso, see Figure 3-11.

Hexagonal phases have a 31P line shape that is opposite in sign and half the numerical span value

of the corresponding lamellar line shape, because the CSA is averaged in a second dimension,

with respect to lamellar liquid crystalline spectra.(26) 2H spectra for the lamellar gel phase are

broad and unresolved; similar to the span of 31P spectra, the quadrupole splitting of the

hexagonal aggregate is also half as broad as the lamellar liquid phase(5).

The NMR data collected from hydrated lipid aggregates not only afford aggregate

structural assignments but also allow for comparisons of the relative order/disorder among

different individual locations, albeit on a different time scale, to those obtained from EPR

measurements. Comparisons of relative order will be used to determine the morphological

effects of incorporating the negatively charged, atypically shaped BMP molecule into model

membranes.

A) B) 000 O






S- 100 ps 10 ns z 100 ns 1 ms Tz seconds minutes

Figure 3-1. Selected lipid motions and associated correlation times in hydrated lamellar
structures: A) rotational diffusion and long axis wobble; B) lateral diffusion; C) flip-
flop.











Bilayer Normal


0 < 0 < 90






Bilayer Normal


0 = 90


-80 -60 -40 -20 0 20 40 60 80
0 (degrees)


Figure 3-2. An order parameter as a function of the angular displacement of the plane containing
a specific carbon and two deuterium atoms from the bilayer normal.


























' CH3
lj-CH,


O'

N CHa

OY- CH3


Figure 3-3. Common organic radical spin-labels: A) 2,2,6,6-tetramethylpiperidine-l-oxyl; B)
4,4-dimethyloxazolidine-3-oxyl; C) (5-DOXYL) steric acid ; D) 1-palmitoyl-2-
stearoyl-(5-DOXYL)-sn-glycero-3-phosphocholine.



















E 12AE = hvm
-1/2




+1




B,,
Bo





Figure 3-4. Energy level diagram illustrating the electronic Zeeman and electron-nuclear
hyperfine interactions and the resulting derivative of EPR transitions for a spin-probe
such as, TEMPO in solution.


Figure 3-5. 4,4-dimethyloxazolidine-3-oxyl labeled 5-ac-cholestane-3-one.
















2,2,6,6-tetramethylpiperidine-1-oxyl


Figure 3-6. Spatial dependence of the coupling strength of the anisotropic hyperfine interaction;
where the z-axis represents the nitrogen pz orbital and the xy-axis represents the N-O
bond plane. The individual couplings can be obtained by rotating the uniaxial
magnetic field so that it is coincident with each axis.





























Figure 3-7. Theoretical nitroxide label EPR line shapes: A) Representation of partially
immobilized environment; B) Representation of an isotropic solution environment.


S AF7hBo
/' 0


AE = yhB
\ --,


A= hB 3e2qQ(3cos2 -1 -

AE = yhB 3e2qQ(3cos2
* 3e~qQ(3cos20
A = yhB---


-7(1 cos2 )(cos 20))
8


-1 7(1 cos2 0)(cos 20))
8


Nuclear
Zeeman
+1/2


Quadrupole
Splitting


E= yh(1 )B


-1/2


Figure 3-8. A) Energy level diagram illustrating the nuclear Zeeman and quadruplar coupling
interaction of 2H in an applied field Bo. B) Energy level diagram for the 1H decoupled
chemical shift of 31P in an applied field Bo.










Avxx


Avzz =Av.

A vr = 2Av(1+ )
1
A vxx = ~2AvQ(1 -q)


AVzz
I I I I
-100 -50 0 50
Frequency (kHz)


Figure 3-9. Theoretical quadrupolar echo powder spectrum of a single deuterium labeled site
( 0 o).

33 VZZ
A) B)
Bo BO


2 31P
0 0



ozz = o l sin2 0 cos2 0 + 722 sin2 sin2 0 + o33 COS2 0

Figure 3-10. Graphical representation of A) CSA and B) EFG the angular orientations with
respect to the applied field (Bo). This figure has been adapted from Santos(24).















A)


-60 -40 -20 0 20
chemical shift (ppm)

SO >"11 22 +33 ) a11 Ao
a11 Span ActC a11-a33


Lamellar




Hexagonal


II


Micellar, Small Vesicle


Avxx = Ayy
.---------


Vzz


-VZ


-100 -50 0 50 100
Frequency (kHz)

Av, = Av -
1 -
Av/=Avr. =AvQ


Figure 3-11. Theoretical powder spectra of various lipid aggregates: A) axially symmetric 31P
and B) a single axially symmetric 2H labeled site.










CHAPTER 4
MONITORING MODEL BILAYER SOLUBILIZATION BY DETERGENT MOLECULES
USING EPR SPECTROSCOPY

Model Membrane Solubilization

The underlying question addressed within this dissertation is the investigation of the

structure of BMP vesicles and the effect this lipid has upon model membrane morphology of

POPC/BMP or DPPC/BMP mixed vesicles. Structure shall be defined as head group and acyl

chain packing, as well as the morphology of lipid aggregation states. Using the geometrical

shape approximations discussed in Chapter 1, BMP is assumed to have a polar region (defined as

the phosphate group and both glycerol moeities) that occupies a larger cross-sectional area than

the apolar, acyl chain region (refer to Figure 1-6 or Figure 4-1 B). According to the simple

geometrical packing model (described in Chapter 1), we can predict that hydrated BMP lipids

may form micellar or highly curved vesicular types of assemblies. Our first hypothesis, based on

simple geometric constraints alone, is that BMP could have detergent-like properties similar to

sodium dodecyl sulfate (SDS or the DS- anion), such as micellar aggregation and the ability to

solubilize lipid bilayer membrane aggregates. The structures of SDS and the S,R isomer of BMP

are shown in Figure 4-1.

According to literature, there are three stages for pure membrane solubilization by

detergent molecules(], 2): 1) detergent monomer molecules bind membranes and partition into

the membrane structure by inserting their acyl chain into the hydrophobic region of the lamellar

structure, thus increasing the overall size of the aggregate structure; 2) membranes become

saturated with detergent molecules and a two phase equilibrium exists between saturated

membranes and detergent/lipid mixed micelles; and 3) bilayer structures become fully

solubilized and lipid molecules are incorporated into detergent micelles (termed mixed micelles).









The molecular species and or aggregates present in each stage of the solubilization process are as

follows: 1) detergent monomer, and lipid/detergent membrane aggregate; 2) detergent monomer,

saturated lipid/detergent membrane aggregate, and detergent/lipid mixed micelles; 3) detergent

monomer and detergent/lipid mixed micelles.

Both micelle formation and membrane solubilization can be investigated using cw-EPR

spectroscopy with positional isomers of spin-labeled lipids (Figure 4-2). If BMP does form

micellar structures, the fast rotational correlation time of the assembly will emotionally average

(narrow) the line shapes of the DOXYL spin-probes incorporated into thes micelles (3) when

compared to the line shapes obtained from relatively immobilized labels in bilayer aggregates(4).

Typical EPR experimental conditions utilize as little as 0.5 mol% of a spin labeled fatty

acid or spin labeled phospholipid to probe structure and dynamics in model membrane

systems(5). Again, motional averaging will narrow the overall EPR line shapes, compared to the

static powder pattern. Spectral features such as the peak-to-peak width of the central resonance

line (AHp), the normalized spectral fractional intensity (fi), order parameter (S,), and second

moment (M2) can be used to compare the degree of motional averaging and relative order of

these spin-probes incorporated into lipid assemblies. The values of these parameters can then be

compared among those obtained from probes incorporated into mixed micellar, lipid/SDS

aggregates, mixed lipid/SDS bilayer structures, and single component, bilayer mesotructures of

model membrane lipids made of POPC and DPPC. Similarly, the effect of BMP upon the

organization, morphology and acyl chain dynamics of POPC and DPPC bilayers can be

investigated using the same strategy, by precisely incorporating 1 mol% of the label into the

assemblies and varying the position of the DOXYL spin-label along the acyl chain of the

phospholipid.









The solubilization of POPC by SDS has been studied in detail by light scattering, NMR

and isothermal titration calorimetry (ITC)(2, 6, 7). Critical values used to map phase boundaries

are obtained by calculating the points of inflection from either right angle light scattering

intensity or normalized heats of reaction (ITC) as a function of detergent concentration(2).

These same critical values and phase boundaries can also be determined by monitoring the

isotropic 31P NMR signal intensity from lipids in mixed micellar aggregates(6). From these

types of data, Seelig and coworkers constructed a phase diagram with SDS concentration (C )

as the dependent variable plotted as a function of total lipid concentration (CQ). The linear

phase boundaries described by Eqns. 4-1 and 4-2 as determined by Seelig and coworkers, have

been reproduced in Figure 4-3(6). The approximate boundaries between the saturated bilayer

and the mixed micelle/bilayer coexistence (Eqn. 4-1) and the mixed micelle/bilayer coexistence

and the mixed micellar regions (Eqn. 4-2) are indicated by solid black lines. The intercept of

each line corresponds to the minimum concentration of detergent needed to either saturate or

completely solubilize the POPC vesicles at 560C(6). According to this phase diagram for SDS

partitioning and micellization of POPC membranes, we sampled the bilayer, micelle/bilayer

coexistence, and micellar regions in our EPR investigation. In Figure 4-3 open circles represent

various POPC vesicle samples prepared with the 10-DOXYL labeled lipid in the presence of

SDS, and closed squares represent various SDS POPC vesicle samples prepared with the 5-

DOXYL labeled lipid in the presence of SDS. It should be noted that our EPR experiments were

carried out at room temperature, whereas the Seelig investigations were performed at 560C.

However, given that POPC is in the La phase above 0C, we anticipate that the experimentally

determined phase boundaries would show little temperature dependence

CD = 0.283C +2.2 (4.1)









C0 = 2.2CL +1.69 (4.2)

Characterization of the EPR Line Shapes of Spin-probes Located in Bilayer Aggregates in
the Presence of an Anionic Detergent

Control experiments were performed that characterized the EPR line shapes of DOXYL

labeled lipids during partitioning of the anionic detergents SDS into lamellar POPC model

membranes and consequent micellization of these membranes. EPR spectra were collected for

DOXYL lipids (1 mol%) incorporated into POPC large unilamaller vesicles (LUVs) that were

made by extrusion through polycarbonate membranes with a 100 nm pore size. POPC and

POPC/BMP LUVs were prepared according to the procedure in Chapter 2 and hydrated in a

buffer containing 20 mM HEPES, 100 mM NaCl and 0.02% NaN3, at pH 7.4 or 5 mM HEPES,

100 mM NaC1, 0. ImM EDTA, at pH 7.4. The concentration of the DOXYL labeled lipid was 1

mol% of the total lipid fraction for all samples unless otherwise indicated. Samples containing

SDS were allowed to equilibrate for at least 30 minutes prior to measurements. EPR

measurements were collected using a 100 G spectral width with 20 mW of microwave power and

a 0.16 s time constant, at room temperature unless otherwise indicated.

In order to characterize the effects that detergent partioning and subsequent micellization

has upon the EPR line shape of DOXYL labels in model POPC membranes, SDS was titrated

into POPC vesicles containing the DOXYL spin-probe. Two sites within the bilayer structures

were examined, carbon positions 5 and 10 on the sn2, steric acid chain of the labeled lipid

(Figure 4-2 A and B respectively). Changing the position of the nitroxide probe provides

information about acyl chain order both near the lipid head group and well within the

hydrophobic region of the molecular assemblies.

As detergent is incorporated into a membrane-like structure, the bilayer packing is

disrupted, thus changing the interactions between lipid molecules and consequently the micro-









environment of the spin-probe. The order parameter of each labeled lipid is expected to decrease

as the system moves through each stage of the solubilization process (1 > 2 > 3)(8, 9). This

change in order should be apparent in EPR line-shapes as the residual anisotropic, hyperfine

coupling strength is affected by local micro-environment(4, 5, 10-14).

Both 5 and 10-DOXYL labeled lipids report observable changes, such as, a smaller order

parameter and a smaller fractional intensity as the concentration of SDS is increased, which is

consistent with SDS partitioning into the bilayer and subsequent micellization. These changes in

acyl chain order can be characterized by any of four numerical values: the second moment, the

peak-to-peak central line width, the normalized fractional intensity or the order parameter

values(4, 11, 15). These values have been defined previously in Chapter 3. Values of the second

moment, order parameters (Si), normalized fractional intensity (fi), and AHpp obtained from EPR

line shapes for the lipid solubilization experiments are listed in Tables 4-1 and 4-2. Figures 4-4

and 4-5 display EPR line shapes as a function of detergent concentration for nitroxide labeled

positions 5 and 10 respectively. From these spectra, it is easy to see the line shape changes upon

addition of SDS, and that the DOXYL label in position 5 is more sensitive to membrane

micellization.

The EPR line shape shown in Figures 4-4 (spectrum a) is for 5-DOXYL labeled lipid in

POPC LUVs with no SDS and shows a typical anisotropic powder-like pattern expected for a

spin-probe with restricted motion intercalated in a membrane structure(4, 11). This line shape

reflects the most immobilized spin-label/lipid motion in this series as indicated by the largest

splitting of the hyperfine interaction tensor component ( ;), the largest value of the second

moment, the largest fractional intensity, and the largest order parameter (26.4 G, 202 G2, 1, and

0.68 respectively). Moreover, the order parameter for spectrum a corresponds to motional









averaging with an angular deviation of 280 from the bilayer normal. Spectrum d shows a line

shape that is the most mobile for the series with values of the hyperfine splitting, second

moment, fractional intensity, and order parameter determined to be 20.7 G, 198 G2, 0.0, and 0.40

respectively, corresponding to motions leading to an average angular deviation of 390 from the

bilayer normal. Assuming the boundaries of the phase diagram in Figure 4-3, this line shape can

be understood in terms of solubilization of the POPC bilayer by SDS resulting in the formation

of mixed micelles. Hence some of the 5-DOXYL lipid is now in a micellar environment and the

line shape parameters reflect the increased correlation time of the smaller spherical micelle

compared to the POPC vesicle. The values of angular deviation from the bilayer normal (0)

were calculated using Eqn. 3-6 along with the order parameter obtained from Eqn. 3-7. Also

note the variable used to represent the average orientation with respect to the bilayer normal is

changed from y to 0. Spectra b and c in Figures 4-4 and 4-5 are EPR line shapes obtained from

the region on the phase diagram in Figure 4-3 between the two phase boundaries but are not

simple superpositions of line shapes obtained from labeled lipids in saturated lamellar bilayer

aggregates and those located mixed SDS/lipid micelles. This may be because spectrum d in

Figure 4-4 was obtained from a location very near the phase boundary, therefore the label is

reporting motion from both mixed micelles and saturated lamellar structural environments.

For tracking changes in acyl chain order it is most useful, at least for this investigation, to

compare spectral parameters such as the fractional intensities, the peak-to-peak widths, and the

order parameters, because the changes in second moment appear to be extremely sensitive to

baseline correction errors. Figures 4-6 and 4-7 show plots of each of these values as a function

of the ratio of SDS to lipid concentrations (CSDS/Cpopc) for samples incorporating a 5 and 10-

DOXYL labeled lipid, respectively. Recall, a less ordered location in the bilayer, such as









position 10, has a narrower line shape because the degree of motional averaging is greater for

spin probes closer to the center of the bilayer (a larger probability for trans/gauche isomerization,

(16) thus more disorder). Unfortunately, number of data points collected for the spin-labeled

lipids, incorporated into POPC LUVs, is too small to draw any quantitative conclusions with

regards to utilizing EPR line shapes of incorporated spin probes to define the phase boundaries

for the solubilization of POPC LUVs by SDS. However, we have measured the characteristic

line shapes and corresponding trends in fractional intensity, peak-to-peak line width, second

moment, and order parameters, for two spin probe positions during the solubilization of POPC

bilayers by SDS detergent.

Hydrated Bis(monoacylglycerol)phosphate Assemblies Solubilized by Sodium Dodecyl
Sulfate

In order to initially characterize the effects that detergent partitioning and subsequent

micellization has upon the EPR line shape of DOXYL labels in BMP aggregates and compare

them to the results obtained for POPC LUVs, SDS was titrated into extruded BMP structures

containing the 5-DOXYL PC. Due to the relative monetary expense of purchasing BMP and the

exploratory nature of our investigation only a single site within the aggregated structure was

examined, carbon position 5 on the sn2, steric acid chain of the labeled lipid.

The EPR spectrum (Figure 4-8 spectrum a) for 99.5% BMPis:i mesostructures (extruded

through 100nm diameter membranes) doped 0.5% 5-DOXYL spin-labeled lipid in an aqueous

environment, and in the absence of SDS is broad (T = 24.6 G and T7 = 9.0 G) and has similar

structural features as the line shapes obtained from the lamellar POPC assemblies in seen in

Figure 4-4 (7, = 26.4 G and T7 = 8.8 G). Spectra (Figure 4-8 b i) are line shapes illustrating the

effect of SDS partitioning and solubilization of BMP aggregates. These spectra show a similar

trend in spectral line narrowing previously seen in POPC lamellar assemblies as SDS









concentration is increased. The similar line shapes and spectral parameters (in the absence of

SDS) indicate that the acyl chain environment in BMP dispersions is similar to that in POPC

LUVs. In addition, the trend in variation of the line shape parameters as SDS solubilization

occurs for BMP is analogous to those observed upon SDS solubilization of POPC LUVs. The

second moment, AHpp, normalized fractional intensity, order parameter, and angular deviation

values obtained from analysis of these line shapes of 5-DOXYL PC in BMP as a function of

SDS concentration are listed in Table 4-1.

Figure 4-9 shows a comparison of the values of peak-to-peak width, second moment, order

parameter, and normalized fractional intensity as a function of SDS/Lipid concentration ratio for

BMP and POPC aggregates obtained using the 5-DOXYL labeled lipid. Again, it is most

instructive to examine the trends in AHpp, Si, and fi because the second moment value is very

sensitive to baseline correction errors. By visual inspection it is clear that each of these

parameters shows an increase in disorder as SDS concentration increases, indicating both

partitioning and solubilization of the lipid structures by SDS. The order parameter for pure BMP

(0.62) is lower than pure POPC (0.68) and could be explained by BMP's second unsaturated

chain; this would probably result in a larger volume requirement per lipid molecule and reduce

steric constraints for isomerization. The values used to track partitioning and solubilization of

the BMP aggregates follow the same general trend as POPC LUVs. The similarity of these

trends may be another positive indication that BMP assembles into a lamellar aggregate with

acyl chain packing and dynamics similar to those in POPC lamellar structures.

Since more EPR data was collected for the BMP solubilization study than for the POPC

solubilization study it may be possible to make more quantitative statements concerning the

sensitivity of spin-probes to the solubilization process. For example, according to Blume and









coworkers a limiting ratio of-1.5 SDS molecules to 1 POPC molecule corresponds to the

saturation point of POPC vesicles and onset of solubilization at 650C(2). A similar ratio is

detected by our EPR investigation of BMP aggregates; clearly indicated by the discontinuity

between concentration ratios 1.3 and 1.7 of plots C and D of Figure 4-9. This is not in agreement

with the value of 0.28 reported earlier in equation 4-2 reported by Seelig for data collected at

56C. However, in the low concentration range (Figure 3C of (6)) light scattering data at 20C

and ITC data at 560C report a much steeper slope corresponding to a ratio of approximately 1.4

or 1.5(2). This ratio is in good agreement with our experimental value of -1.5 SDS molecules to

1 POPC molecule. There is another possible discontinuity present in our EPR data for the region

between concentration ratios 2.5 and 3.5 corresponding to the beginning of solubilization of the

aggregate structure; this discontinuity is more obvious in Figure 4-9 plot D. Seelig and Blume

report ratios of 2.2 and 2.7 SDS molecules per lipid molecule needed to initiate solubilization of

POPC membranes at 56C and 65C, respectively(2, 6).

According to the previous observations and the assumption that BMP has a similar

solubilization diagram as POPC, we believe we have sampled each region of the phase diagram

for the solubilization of BMP by SDS detergent, and have obtained line shapes that are

characteristic of each region(3, 9). Therefore, the spectra for BMP dispersions found in Figure

4-8 are assigned to the following regions of a solubilization diagram a lamellar BMP, b-d SDS

momomer/SDS partitioned into BMP bilayers, e-g SDS momomer/SDS saturated BMP

bilayers/mixed SDS BMP micelles, and h-i SDS momomer/mixed SDS BMP micelles.

Given that our EPR results for BMP solubilization behavior are consistent with those

obtained by other techniques for bilayer forming POPC; the acyl chain packing in BMP

dispersions likely adopts a similar lamellar structure. Hence, these data are not consistent with a








model of BMPis:i as a micellar aggregate. This conclusion is in accordance with light scattering,

electron microscopy and fluorescence results indicating BMPis:1 forms MLV assemblies(17),

and X-ray and molecular dynamics data obtained by Kobayashi showing BMP14:0 also forms a

stable lamellar aggregate(18).




o-NH4(s
A) o- Na B) O= -
S0 HO H O
R H
o "'OH
O
O O




















Figure 4-1. Model membrane perturbants: A) Sodium dodecyl sulfate; B) (S,R isomer) sn-(3-
oleoyl-2-hydroxy)-glycerol-l-phospho-sn-3'-(l'-oleoyl-2'-hydroxy)-glycerol
(ammonium salt).












































Figure 4-2. Positional isomers of spin-labeled lipids: A) 1-palmitoyl-2-stearoyl-(5-DOXYL)-sn-
glycero-3-phosphocholine (5-DOXYL); B) 1-palmitoyl-2-stearoyl-(10-DOXYL)-sn-
glycero-3-phosphocholine (10-DOXYL).
















45

40

35

30

25
Uo
20 -0

15

10

5


0 10 20 30 40

Cm mM




Figure 4-3. Phase Diagram for SDS and POPC LUVS at 560C reproduced from linear regression
analysis by Seelig and coworkers. Solid lines represent phase boundaries: the symbol
(m) indicates samples made with POPC LUVs containing 5-DOXYL labeled lipid and
(o) indicates samples made with POPC LUVs containing 10-DOXYL labeled lipid to
collect representative spectra for specific regions of the phase diagram in our
solubilization experiments.















d)


b)

a)
20 G



Figure 4-4. cw-EPR spectra ofPOPC LUVs with 5-DOXYL spin probe (1 mol%) at room
temperature in 20mM HEPES, 100mM NaC1, 0.02% NaN3 at pH 7.4: a) 27 mM
POPC OmM SDS; b) 22 mM POPC 6mM SDS; c) 11 mM POPC 18mM SDS; d) 7
mM POPC 22mM SDS.

2





d)


b)







Figure 4-5. cw-EPR spectra of POPC LUVs with 10-DOXYL spin probe (1 mol%) at room
temperature in 20mM HEPES, 100mM NaC1, 0.02% NaN3 at pH 7.4: a) 27 mM
POPC OmM SDS; b) 19 mM POPC 8mM SDS; c) 10 mM POPC 19mM SDS; d) 8
mM POPC 21mM SDS.

















4U

38

36

34

32 -

330

28

26

24

22

00 05 10 15 20 25 30 35


0 75 p

070

065

060 -

055

050

045

040

035
00 05 10 15 20 25 30 35


205









200









195




1 0


08


06


CS/C
*-W Lipid


*












00 05 10 15 20 25 30 3:
C/C




MS L,, Id


00 05 10 15 20
C C...


25 30


Figure 4-6. Various spectral parameters of 5-DOXYL labeled lipid incorporated into POPC

LUVs as a function of SDS/Lipid concentration ratio at room temperature in 20mM

HEPES, 100mM NaCl and 0.02% NaN3 and pH 7.4: A) Peak-to-peak width of central

derivative line; B) Second spectral moment; C) Order parameter; D) Average

normalized fractional intensity. Order parameter error bars are estimated by

assuming a 1 G error in the difference between the parallel and perpendicular

components of the hyperfine tensor.































68


C















A) B)
40 200

38
0 195 -o
36
190
34

32 185


a 180
28
26 o- 175
260
24 170

2.2
165
00 05 10 15 20 25 30 35 00 05 10 15 20 25 30 35

C) c/c cS ,,. D)
0.50
10 -
045

040 08
C 0
035 06

030

0 25
T02
0.20 -
O0 o
0.15 I I I 0
00 05 10 15 20 25 30 35 00 05 10 15 20 25 30 35
CMs/CLa Cs /CL,a




Figure 4-7. Various spectral parameters of 10-DOXYL labeled lipid incorporated into POPC

LUVs as a function of SDS/Lipid concentration ratio at room temperature in 20mM

HEPES, 100mM NaCl and 0.02% NaN3 and pH 7.4: A) Peak-to-peak width of central

derivative line; B) Second spectral moment; C) Order parameter; D) Average

normalized fractional intensity. Order parameter error bars are estimated by

assuming a 1 G error in the difference between the parallel and perpendicular

components of the hyperfine tensor.





























69




















20 G


Figure 4-8. cw-EPR spectra of BMP with 5-DOXYL spin probe (1 mol%) at room temperature
in 5 mM HEPES,100 mM NaC1, and pH 7.4; a) 40mM total lipid OmM SDS; b)
40mM total lipid 17mM SDS; c) 40mM total lipid 35mM SDS; d) 40mM total lipid
52mM SDS; e) 40mM total lipid 69mM SDS; f) 40mM total lipid 87mM SDS; g)
40mM total lipid 104mM SDS; h) 40mM total lipid 139mM SDS; i) 20mM total lipid
104mM SDS.





















40

38

36

34

32

30

28

26

24

22



C)
0.75
0.70
0.65
0.60
0.55
0.50
0.45
0.40
0.35
0.30
0.25
0.20
0.15


- -

- *









O D
-
-D




-

-

00 05 10 15 20 25 30 35 40 45 50 55






^ n-





-









O00 0.5 10 15 '210 '2.5 '310 '315 '40 '45 '50 '55
C-.C. ,


00 05 10 15 20 25 30 35 40 45
C_/C ,


50 55


Figure 4-9. Various spectral parameters of 5-DOXYL labeled lipid incorporated into POPC

LUVs (m) and BMP aggregates (o) as a function of SDS/Lipid concentration ratio at

room temperature in 20mM HEPES, 100mM NaC1, 0.02% NaN3 at pH 7.4 for POPC

LUVs and 5 mM HEPES,100 mM NaCI, at pH 7.4 for BMP aggregates: A) Peak-to-

peak width of central derivative line; B) Second spectral moment; C) Order

parameter; D) Average normalized fractional intensity. Order parameter error bars

are estimated by assuming a 1 G error in the difference between the parallel and

perpendicular components of the hyperfine tensor.


5







3 D






00 05 10 15 20 25 30 3.5 40 45 50 55


C /C
SDsI Liid








F1-













Table 4-1. Parameters defining order of the 5-DOXYL nitroxide spin-probe in lipid aggregates
at room temperature.
[Total Lipid] [SDS] AHP, G 2nd Moment Order Normalized Average angle to
mM mM G2 Parameter Fractional Intensity bilayer normal (0)
(S) (f)
5-DOXYL 27 0 3.9 202 0.68 1 280
label/POPC
LUVS
22 6 3.6 201 0.62 0.83 300


11 18 3.2 197 0.51 0.34 350

7 22 2.9 198 0.40 0 390


5-DOXYL 40 0 2.9 189 0.62 1 300
label/BMP
40 17 2.7 185 0.60 0.89 310

40 35 2.7 190 0.58 0.89 320

40 52 2.7 190 0.58 0.89 320

40 69 2.6 185 0.50 0.56 350

40 87 2.6 189 0.49 0.44 360

40 104 2.5 179 0.45 0.44 370

40 139 2.4 179 0.42 0.33 390

20 104 2.2 180 0.19 0 470



Table 4-2. Parameters defining order of the 10-DOXYL nitroxide spin-probe in lipid aggregates
at room temperature.
[Total Lipid] mM [SDS] mM t, G 2nd Moment G2 Order Normalized Average angle


10-DOXYL
label/POPC
LUVs


Parameter S, Fractional Intensity to bilayer normal
(f) (6)
0 3.3 182 0.43 1 380


8 3.2 181 0.43


0.97


0.047

0










CHAPTER 5
PERTURBATIONS OF LAMELLAR LIQUID CRYSTALLINE ORDER BY
BIS(MONOACYLGLYCEROL)PHO SPHATE

Solid State Phosphorus-31 NMR Investigation of Hydrated
Bis(monoacylglycerol)phosphate Aggregation State

Solid state 31P NMR experiments were performed at 145.2 MHz using a two pulse Hahn

echo sequence with full proton decoupling on single component BMPPi:1 MLV samples in an

aqueous environment. The 31P NMR measurements allowed us to investigate the head group

order and the mesophase structure of hydrated BMP MLV dispersions. Visual inspection of the

line shapes of phosphorus NMR spectra provides information about head group orientation and

information needed to discriminate among lamellar, HII, and isotropic aggregation states(16, 59,

66). Both changes in the span of the chemical shift anisotropy and in the overall line shapes are

governed by individual molecular motions, local environment of phosphorus nuclei, and overall

aggregate tumbling rate, each of which affect the orientation of the 31P chemical shift tensor in

the phosphate head group with respect to the applied field(16, 67).

The 31P NMR spectrum of the BMPPi8: MLV dispersion shown in Figure 5-1 is narrow (AC

=-11.6 ppm) and has an overall line shape characteristic of a lamellar mesophase. This finding

is in agreement with our previous conclusion in Chapter 4 that BMP forms a stable lamellar

structure. This being said, BMP would also then be expected to form stable lamellar structures

when mixed with model lipids such as POPC and DPPC. All of the EPR and NMR evidence

presented thus far indicates that BMP adopts a lamellar aggregate structure and thus should not

exhibit detergent behavior. However, the effect BMP has on model membrane morphology or

any possible role BMP may play in late endosomal lipid trafficking or in the hydrolysis of GSLs

has yet to be addressed.









Acyl Chain Order of 1-Palmitoyl-2-Oleoyl-sn-Glycero-3-Phosphocholine
/Bis(monoacylglycerol)phosphate Mixed Vesicles Determined by Electron Paramagnetic
Spectroscopy

In this section of the dissertation, EPR results are presented regarding the interaction of

BMP with model membrane lipids in the La phase. POPC and BMP LUVs of mixed

composition were prepared with POPC, varying amounts of BMPis:1 or BMP14:0 and 1 mol% of

spin-labeled lipid (5-DOXYL or 10-DOXYL). These experiments parallel the previous EPR

investigations reported in Chapter 4 with single component, lipid LUVs and SDS detergent.

EPR spectra shown in Figure 5-2 obtained from POPC/BMP mixed LUVs have the typical

anisotropic powder patterns seen previously in Chapter 4. However, the distinct lines shape

changes seen in EPR spectra for the 5-DOXYL labeled lipid that occurs when SDS partitions

into lamellar aggregates are not detected in POPC/BMP mixed vesicles, indicating that BMP

does not solubilize POPC membranes.

To track the changes that occur in the EPR line shapes of the doxyl labeled lipids as a

function of BMP concentration, values of AHpp and the order parameter (Si) were determined

from the EPR line shapes. These data are plotted as a function of the concentration ratio of BMP

to POPC (CBMp/CpoPc) in Figures 5-3 and 5-4, respectively, and values are listed in Table 5-1.

Figure 5-3 shows that the peak-to-peak width of the central derivative line and the order

parameter values are constant for both the 5 and 10-DOXYL labeled lipids in POPC/BMP14:0

mixed LUVs for concentrations up 20 mol% BMP14:0. Figure 5-4 shows the values of AHpp and

order parameter for both the 5 and 10-DOXYL labeled lipids in POPC/BMPis:1 mixed LUVs are

also constant, within experimental error, over the same concentration range. It is also important

to note the average value of the order parameters for the 5 and 10-DOXYL positions are very

similar for both BMPis:1 and BMP14:0. Moreover, these results indicate that mixed composition









LUVs of POPC/BMP (less than 20 mol % BMP) have the same degree of acyl chain order as

pure POPC (within experimental error) at both label positions regardless of BMP's degree of

acyl chain saturation or chain length. These results further support our previous claim that BMP

does not have classic detergent properties, and indicates BMP does not perturb the acyl chain

order of POPC LUVs, at least not on the time scale observed by EPR.

31P NMR of 1-Palmitoyl-2-Oleoyl-sn-Glycero-3-Phosphocholine
/Bis(monoacylglycerol)phosphate mixed MLVs

The conclusions drawn in the previous section indicate that incorporation of BMP into

POPC vesicles does not alter the order or packing of the acyl chains as can be observed on the

EPR timescale. Here, the effects of BMP on head group orientation were investigated using

solid-state 31P NMR to characterize the interaction of BMP with POPC in hydrated MLVs. Solid

state Phosphorus-31 NMR experiments obtained at 145.2 MHz using a two pulse Hahn echo

sequence with full proton decoupling were performed on mixed POPC/BMPis:1 MLV samples in

an aqueous environment to investigate the head group order and mesophase structure. POPC and

POPC doped with BMPis:1 MLV samples were prepared according to the procedure for LUV

preparation but were not extruded (see Chapter 2). All 31P spectra were aligned so that isotropic

value of either axially symmetric fully hydrated POPC MLVs or axially symmetric fully

hydrated DPPC MLVs was assigned to 0 ppm, see Figure 3-11.

Recalling Figure 3-11, 31P spectra for hydrated lamellar aggregates have characteristic

powder averaged chemical shift line shapes. The solid state 31P NMR spectrum of the hydrated

single lipid component POPC MLVs at 370C is shown in Figure 5-5. This spectrum exhibits a

classic axially symmetric lamellar line shape with a span of -46 1 ppm, which is in excellent

agreement with the reported literature value (- 46 1 ppm)(68-70). Line shape simulation via

software (dmfit) (71) did not offer any advantage over direct measurement from the experimental









line shapes in estimation of 31P chemical shift span, therefore all reported span values were

measured directly from the experimental data.

According to the model of glycerophospholipid orientation with respect to the normal of a

bilayer for DPPC and DPPE; the order parameter of the C1C2 bond vector (S(cl)-(2) from (72))

obtained from H NMR (Ci and C2 are glycerol carbons enumerated using the sn naming system

see Figure 1-2), Scic2 is 0.66 and defines the tilt angle of the CiC2 axis with respect to the bilayer

normal.(72, 73) This wobble freedom reduces the span of the strict axially symmetric powder

average for 31P (AG -124 ppm) to a maximum of -82 ppm for hydrated MLVs (16) (AG = (C11

- C33)*SC1C2 = -82 ppm). However, the experimental value (-46 ppm) is clearly narrower

indicating an additional angular offset of the CSA axis from the bilayer normal. The

experimental principal value ol is -30 ppm for pure POPC (the trace of the CSA is invariant to

motion so a = 16 ppm) and yields an angle of about 33 from the normal of the bilayer. Again

the order parameters are represented as second order Legendre polynomials (S = /2 (3 cos2() -

1)). Both of the previous values are in good agreement with current literature; Lorigan and

coworkers measured ol = 30 1 ppm and o- = 15 1 ppm from oriented POPC MLVs(70),

and an angle with respect to the bilayer normal of 300 has been reported for hydrated

dipalmitoyllecithin(8, 9).

Firgure 5-6 shows 31P line shapes for various mixtures ofPOPC/BMP MLVs. Two

important observations about mixed POPC/BMPis:1 MLVs can be made by visual inspection of

the line shapes: 1) the CSA span of the POPC line shape decreases as the negatively charged

BMP concentration increases when compared to the span of single component POPC (spectrum

a 2) the spectra of the mixed lipids do not appear to be a simple superposition of the single

component POPC and single component BMP CSA line shapes. Each of the previous









observations, more clearly illustrated in Figure 5-7, indicates the lipids are interacting (mixing)

in the same lamellar structures.

According to work by Seelig and Scherer, the orientation that the phosphate/choline group

has with respect to the bilayer normal is sensitive to membrane surface charge(68, 74).

Quadrupolar splitting of the ac and 3 carbons located on the choline head group of deuterium

labeled POPC and the CSA span of POPC were measured as a function of positive charged

dioctadecyldimethylammonium-bromide (2C18N2C1+Br-) and negative charged sodium

didodecyl phosphate (2Cl2PO4-Na+) amphiphilic molecules incorporated into POPC MLVs. A

linear relationship relating CSA span and mole fraction of 2C12PO4-Na+ was determined to an

amphiphile mole fraction (Xb) of 0.3, (Ac = 45.6 + 18.7Xb ppm ) and a plateau value of CSA

span 36 ppm at high amphiphile concentrations(68). Increased disorder in the

phosphate/choline region, indicated by a decrease in the value of the 31P CSA span, implies a

larger angular deviation from the bilayer normal(68).

Figure 5-8 shows the change in span of the POPC rich line shape of POPC/BMP mixed

MLVs up to a mole fraction BMP of 0.3 (closed squares) and the best linear fit of that data

(Ac = 45.6 + 16.0XBMP (dashed line)), however the limiting span value observed by Seelig and

Scherer is not obvious in our data due to the noise level near CH1. Line shape simulations may be

helpful in determining the limiting span in this case. Comparison of the two regression lines

indicates that the phosphate groups of MLVs containing POPC/BMP are interacting and ordered

in a similar fashion to the MLVs with the model anionic amphiphile used by Seelig and Scherer.

At a mole fraction BMP of 0.3 both regressions predicts an angle of inclination to the bilayer

normal of -36 and a value of AG of -40 ppm, in good agreement with our experimental values

(0 = 35 and AG = -41 ppm)(63). These results indicate that BMP, when mixed with the model









membrane lipid POPC up to 30 mol% imparts a surface charge similar to that of 2C12PO4-Na

and has a similar effect on the average orientation of the head group with the bilayer normal.

As for the very narrow span (-11.6 ppm) observed for pure BMP structures there are at

least two possible explanations: 1) BMP forms LUVs that have a small diameter (highly curved

surface), or 2) the phosphorus has a large tilt angle with respect to the normal of the bilayer.

Molecular dynamics (MD) simulations indicate that BMP14:0 has a bilayer thickness on the order

of 4.2 nm and a 200 tilt angle with respect to the bilayer normal(28). According to the MD it

seems possible that the lamellar structures formed by single component BMP may have a small

diameter and thus a narrowed CSA span. This claim should be further substantiated by dynamic

light scattering experiments.




















-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30
chemical shift (ppm)

Figure 5-1. 31P NMR chemical shift spectrum ofBMPis:1 MLVs: in 5 mM HEPES, 100 mM
NaC1, 0.1 mM EDTA, pH 7.4, at T = 37 C. The CSA is referenced to the isotropic
value of axially symmetric, hydrated POPC MLVs.











5-DOXYL 18:1 B)


C) 10-DOXYL 14:0 10-DOXYL 18:1 D)



C)-- C)







Figure 5-2. cw-EPR spectra of POPC/BMP mixed LUVs containing 1 mol% of either 5 or 10-
DOXYL spin probe in 5 mM HEPES,100 mM NaC1, at pH 7.4; A) and C) a) 23 mM
POPC 0 mM BMP14:o; b) 23 mM POPC 0.2 mM BMP14:0; c) 22 mM POPC ImM
BMP14:o; d) 21 mM POPC 2 mM BMP14:o; e) 19 mM POPC 5 mM BMP14:o; B) and
D) a) 20 mM total lipid 0 mM BMP1i:1; b) 23 mM POPC 0.2 mM BMP18:1; c) 22 mM
POPC 1 mM BMP18:1; d) 21 mM POPC 2 mM BMP18:1; e) 19 mM POPC 5 mM
BMPis:1.


5-DOXYL 14:0















3.90 070
3.85 065
3.80
3.75
0 60
3.70
S3.65 055
3.60
3. 050
3.55
3.50 0 45
3.45 -
3.40 L 040
000 005 010 015 020 025 030 000 005 010 015 020 025 030
Cs/Cpc Cs/Cpc





Figure 5-3. AHpp and Si of 5 (m) and 10-DOXYL (o) labeled lipid (1 mol%) incorporated into
POPC/BMP14:0 mixed LUVs as a function of BMP14:0/Lipid concentration ratio at
room temperature in 5mM HEPES, 100mM NaCl and pH 7.4: A) Peak-to-peak width
of central derivative line; B) Order parameter. Order parameter error bars are
estimated by assuming a 1 G error in the difference between the parallel and

perpendicular components of the hyperfine tensor. The error in AHpp is + 0.2 G based
on three independent measurements of single component POPC LUVs.



A) B)
410 070-0
405 -
400 -E
395 0 65 -
3 95
385 9
385 060
380 -
375 -
0 370 0 55
3 65
360
355 050
350 00 0 O
345 0 0 45
3 40 -
335
330 I O 040
000 005 010 015 020 025 030 000 005 010 015 020 025 030
CBMp /Co C /C c




Figure 5-4. AHpp and Si of 5 (m) and 10-DOXYL (o) labeled lipid (lmol%) incorporated into
POPC/BMPi8:s mixed LUVs as a function of BMPis8:/Lipid concentration ratio at
room temperature in 5mM HEPES, 100mM NaCl and pH 7.4: A) Peak-to-peak width
of central derivative line; B) Order parameter. Order parameter error bars are
estimated by assuming a 1 G error in the difference between the parallel and

perpendicular components of the hyperfine tensor. The error in AHpp is 0.2 G three
independent measurements of single component POPC LUVs.







80











AcT = C,, C33 = 46 1 ppm


-40 -35 -30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30 35 40
ChemicalShift (ppm)

Figure 5-5. 31P NMR chemical shift ofPOPC MLVs in 5 mM HEPES, 100 mM NaC1, 0.1 mM
EDTA, pH 7.4, at T = 37 C. The CSA is aligned according to the isotropic value of
fully hydrated POPC MLVs.


























-40 -30 -20 -10 0 10 20 30
Chemical Shift (ppm)


-40 -30 -20 -10 0 10 20 30
Chemical Shift (ppm)

Figure 5-6. 31P NMR spectra of POPC/BMP18:1 MLVs in 5 mM HEPES, 100 mM NaC1, 0.1
mM EDTA, pH 7.4, at T = 37 C. A) a) 278 mM POPC 0 mM BMP; b) 276 mM
POPC 1.8 mM BMP; c) 275 mM POPC 2.8 mM BMP; d) 270 mM POPC 8.3 mM
BMP; e) 264 mM POPC 13.7 mM BMP f) 249 mM POPC 27.8 mM BMP; g) 224
mM POPC 52.3 mM BMP; h) 192 mM POPC 82 mM BMP; i) 139 mM POPC 133
mM BMPj) 53.8 mM POPC 215 mM BMP; k) POPC 0 mM 42 mM BMP B) shows
the same spectra as A) but in reverse order.























SI'
I '
I '
I'(

/
/


-20 0 20I
-20 0 20


chemical shift (ppm)


S-20 0 20I
40 -20 0 20


chemical shift (ppm)


Figure 5-7. 31P NMR spectra of POPC/BMP18:1 MLVs in 5 mM HEPES, 100 mM NaC1, 0.1
mM EDTA, pH 7.4, at T = 37 C. A) 278 mM POPC 0 mM BMP (solid line) and 192
mM POPC 82 mM BMP (dashed line); B) 192 mM POPC 82 mM BMP (dashed line)
and a linear combination of two line shapes with a contribution of 70% pure POPC
line shape and 30% pure BMP line shape.
















-47


-46


-45 i .


E- -44


-43


-42 "-


-41 -


-40 I I I I I
0.00 0.05 0.10 0.15 0.20 0.25 0.30


BMP 18:1


Figure 5-8. CSA span of (POPC/BMP MLVs) as a function of BMP mole fraction (i) and linear
fit (Ac = 45.6 + 16.0XBP (dashed line)).


































84


IIIIIIIIIIIII









Table 5-1 Parameters defining order of the 5-DOXYL nitroxide spin-probe in POPC/BMP
mixed MLVs at room temperature.
[Total Lipid] [BMP] AH, G Order
mM mM Parameter
(Si)


5-DOXYL label
POPC/BMP14:0
LUVS











5-DOXYL label
POPC/BMPi,:1
LUVS


0.66



0.66


0.65

0.65

0.66

0.65



0.66

0.64

0.64

0.64









Table 5-2 Parameters defining order of the 10-DOXYL nitroxide spin-probe in lipid aggregates
at room temperature.
[Total Lipid] [BMP] AH, pG Order
mM mM Parameter
(Si)


10-DOXYL label
POPC/BMP14:0
LUVS












10-DOXYL label
POPC/BMP18.1
LUVS


0.46



0.46


0.46

0.45

0.44

0.47



0.46

0.47

0.46

0.46









Table 5-3 Values of CSA span for POPC/BMP mixed MLVs at room temperature.
[POPC] mM [BMPl8:I] mM XBMP Ao ( 1 ppm)

278 0 0 -46


276 1.8 0.0065 -45


275 2.8 0.0100 -46


270 8.3 0.030 -45


264 13.7 0.049 -45


249 10 0.039 -44


224 52.3 0.19 -43


192 82 0.30 -41


139 133 0.5 -41


53.8 215 0.80 -20


0 42 1.00 -11.6










CHAPTER 6
PERTURBATIONS OF THERMOTROPIC PHASE TRANSITIONS DUE TO
BIS(MONOACYLGLYCEROL)PHOSPHATE

Thermotropic Phase Behavior of 1,2-Dipalmitoyl-sn-Glycero-3-Phosphocholine
/Bis(monoacylglycerol)phosphate MLVs Investigated by H NMR

Incorporation of a lipid deuterated in the acyl chain region allows order/disorder to be

measured along the length of the apolar region of a MLV assembly. We chose perdeuterated

DPPC for this investigation in order to simultaneously probe the effects of BMP on the

molecular order at each carbon position along the acyl chain region as well as its effects on

thermotropic phase transition temperatures of DPPC(17). Investigations of a thermotropic phase

transition are not practical using POPC because the main phase transition of POPC occurs at 0C

and is near the freezing point of water, which makes this type of measurement difficult(43). The

following experimental 2H NMR data affords information concerning the order of the acyl chain

region and is analogous to that obtained from EPR measurements with the n-DOXYL labeled

lipids discussed earlier. However, H NMR operates on a much slower time scale, thus allowing

more averaging by the lipid motions. Also, recall that substitution of deuterium for hydrogen

atoms in the acyl chain region is only a small structural perturbation when compared to the large

nitroxide labels used for EPR measurements.

2H NMR line shapes as a function of temperature were obtained for DPPC MLVs and

DPPC doped with BMPis:1 (5 mol%) MLVs in low ionic strength buffer (Im < 0.05 m) and are

shown in Figures 6-1 and 6-2. The main phase transition (Tm) for DPPC MLVs occurs at ~

41.5C as determined by DSC measurements(29); however, this transition is depressed to ~

37.7C for perdeuterated DPPC MLVs, also determined by DSC(75). A comparison of the line

shapes in Figure 6-1 A and B shows evidence that incorporation of BMP into the lamellar

structure increases the onset temperature of a thermotropic phase transition relative to single









component DPPC MLVs. This is easily visualized in spectra at 35 37 C in Figure 6-1 A and

B. NMR line shapes for MLVs containing BMP are broad and unresolved at 35 and 36C,

indicative of a lamellar gel phase but the spectrum of single component DPPC has already

started to narrow and shows resolved terminal methyl peaks (the Pake doublet with the smallest

quadrupole splitting) at 360C, indicating more motional freedom.

At temperatures above the main phase transition of single component DPPC, 2H line

shapes (Figure 6-1 B) obtained from mixed DPPC/BMP (5 mol%) MLVs show a broadening of

the terminal methyl signal relative to that of DPPC MLVs. Initially we were inclined to interpret

this broadening as an indication of restricted mobility and evidence of an interdigitated phase.

However, results obtained from EPR experiments, with 16-DOXYL labeled lipid (discussed in

the next section), did not confirm our hypothesis and the cause of the broad, unresolved peaks

and a molecular level understanding of the structure or dynamics that leads to the broadening of

the terminal methyl signal remain unknown.

2H NMR measurements were also performed on DPPC and DPPC/BMP mixed MLVs at

near biological ionic strength (Im 100 m), and the results show quite different effects of BMP

on the onset of a thermotropic transition temperature. When the number of counter ions in the

buffer is much greater than the number of negative BMP molecules (-15000:1) a decrease in the

phase transition temperature is observed when compared to BMP effects in low ionic strength

buffer and when compared to single component DPPC in near biological ionic strength buffer.

The former can be visualized by comparing the spectra at 35C of Figure 6-2 A and B. It is clear

that for MLVs doped with BMP under physiological buffer conditions a significantly larger

portion of the acyl chains are disordered when compared to single component DPPC at 35C.









These effects on acyl chain order can be seen more clearly by comparing the order

parameter (Stotai = Smoi*SLD) values (Tables 6-1 and 6-2) for the terminal methyl groups as a

function of temperature or the residual quadrupolar splttings as a function of temperature in

Figure 6-3. Referring to Chapter 3 the quadrupolar splitting, Av, is defined for axially symmetric

3
motion about the lipid's long molecular axis as (Av = XQStot, ), and -, is 167 kHz.(76) Smois



assumed to be positive and equal to -3SCD and SCD is PD .(46, 77) The angle OPD


describes the orientation angle of the C-D bond vector with respect to the director axis, and SLD

takes the same functional form as SCD but describes the orientation angle OLD of the director with

respect to the applied magnetic field. Computational methods are available to "dePake" or

approximate SLD (46, 78, 79) but the signal to noise level of our data is not high enough to obtain

this value. However, Stotal should be sufficient to show trends related to the relative order of a

specific site in the bilayer.

Figure 6-3 show plots of the terminal methyl order parameter and residual methyl splitting

for MLV dispersions of DPPC( filled symbols) and for DPPC/BMP (5 mol%)(open symbols)

Data were collected for temperatures ranging from 35 to 430C, but order parameters cannot be

determined for those spectra in the gel phase. A discontinuity is observed in the trend of order

parameter values and residual quadrupolar splitting when plotted as a function of temperature

for the terminal methyl peaks in single component DPPC dispersions. This discontinuity occurs

for temperatures between 37 and 38C in Figure 6-3 A and B, and its value coincides nicely with

the detected phase transition temperature from the gel to the La phase of perdeuterated DPPC in

50 mM phosphate buffer observed by Davis, which were measured by both DSC and 2H

NMR.(75) A comparison of the temperature dependent residual quadrupolar splitting between









single component DPPC and DPPC/BMP mixed MLVs obtained in low salt buffer (5 mM Na ,

Figure 6-3 A) shows that BMP does not affect the order in the acyl chain region at temperatures

above 37C. Additionally, the main phase transition temperature for DPPC/BMP (5 mol%)

mixed MLVs is elevated in comparison to that of single component DPPC in low salt buffer.

The order parameter and residual quadrupolar splitting profiles for MLVs in near

biological ionic strength buffer (105 mM Na Figure 6-3 B) show that in the presence of BMP

the phase transition is significantly depressed but the overall order is similar to single component

DPPC at temperatures above 37 C. Additionally, the terminal methyl group signal is not

broadened, thus the broadening of the terminal methyl intensity previously observed at low ionic

strength is not an effect of BMP alone. In order to determine the onset temperature of the phase

transition for DPPC/BMP (5 mol%) we must collect data for several temperatures below 35 C.

Overall the total order parameter and residual quadrupolar splitting results lead to values

for thermotropic phase transitions that are consistent with literature values; however, the 2H data

should be recollected to obtain higher quality line shapes, and to cover larger temperature range.

This will not only afford a better estimation of phase transition temperatures and experimental

error but the spectra can be "dePaked" and an order parameter that is independent of orientation

with respect to the magnetic field can be assigned to each site along the acyl chains. Some of

this work has already begun by others in the Fanucci research group, and the broadening of the

terminal methyl signal at low ionic strength has been reproduced but the cause of the broadening

is still unexplained.

Thermotropic Phase Behavior of 1,2-Dipalmitoyl-sn-Glycero-3-Phosphocholine
/Bis(monoacylglycerol)phosphate MLVs Investigated by EPR

The effects of BMPis:1 on thermotropic phase transitions of lamellar assemblies was also

monitored by EPR spectroscopy with the n-DOXYL labeled lipids incorporated into a model









membrane phospholipid containing fully saturated acyl chains, (DPPC). The lipid containing a

single unsaturated acyl chain (POPC) could not be used for these experiments because the main

phase transition is near 0C(43). Given that water freezes near this temperature, DPPC was a

more appropriate model lipid for this investigation. 16-DOXYL spin-labeled lipids were used to

obtain EPR data parallel to the previous H NMR study regarding changes in acyl chain order

near the center of the bilayer.

The EPR spectra of DPPC MLVs (containing 1% 16 Doxyl PC) at two different buffer

ionic strengths are shown in Figures 6-4 (low with 5 mM Na ) and 6-5 (near biological ionic

strength with 105 mM Na+), and demonstrate the temperature dependent order profile of the

nitroxide label located at carbon position 16 of the steric acid chain of the label in DPPC MLVs.

The broadest splitting of the parallel component of the hyperfine interaction is observed when

the hydrated lipids are in the gel phase, indicating the most ordered environment (largest order

parameter) of the series. As the temperature is increased, the lipid molecules obtain more kinetic

energy, allowing for more motional freedom, which results in a narrowing of the EPR lineshape.

Plots of spectral parameters such as the fractional intensity, the peak-to peak width, and the order

parameter as a function of temperature are expected to give characteristic sigmoidal shaped

profiles for smooth thermotropic phase transitions(80-83). The phase transition temperature is

defined as the inflection point of the sigmoidal shaped line.

A comparison of the temperature dependence of the peak-to-peak width, the second

moment, and the order parameter values (Tables 6-3 to 6-4) for DPPC MLVs (1% 16-DOXYL)

at low ionic strength (5 mM Na+) and DPPC MLVs (1% 16-DOXYL) at near biological ionic

strength (105mM Na ) are shown in Figure 6-6. It is clear that the acyl chain order is almost

identical for all temperatures regardless of ionic strength, including the thermotropic phase









transition occurring near 35C. This transition is assigned as the pre-transition of single

component DPPC MLVs, because the breadth of this transition is much larger than the breadth

(0.5C) of the main phase transition obtained by Davis (75, 84). Figure 6-6 D and E show the

first derivative of the sigmoidal fit used to determine the inflection point and thus the transition

temperature, and the sigmoidal fit overlain on the plot of peak-to-peak line width, respectively.

The values of the pre-transition temperatures calculated from each of the parameters defining

bilayer order are listed in Table 6-5. Additional data should be collected in the range between 25

and 35C to verify the shape of the sigmoidal curve and the transitions temperatures.

EPR spectra for 16-DOXYL PC (1 mol%) incorporated into DPPC/BMPis:1 MLV

dispersions prepared in a buffer with low ionic strength (Figure 6-7) have similar shapes as those

seen previously for this spin probe in single component DPPC dispersions. In additions, values

of the peak-to-peak line width, the second moment, and the order parameters are similar to those

obtained for single component DPPC. From these data we can conclude that little to no change

in order of the acyl chains near the center of the bilayer is caused by BMPis:1 in low ionic

strength buffer.

Analysis of the EPR line shapes from 16-DOXYL in DPPC/BMPis:1 MLVs prepared in

buffer that mimics biological ionic strength show a lowering (compared to DPPC) of the pre-

transition temperature when compared to the value obtained for MLVs in low ionic strength

buffer. Specifically, this conclusion can been seen by comparing the line shapes between 35C

and 37C in Figures 6-7 and 6-8, and it is more clearly evident in the plots of AHpp, second

moment, and Si values as a function of temperature (Tables 6-6 and 6-7) in Figure 6-9. Clearly,

this thermotropic phase transition is altered for DPPC/BMP (5 mol%) mixed MLVs in near

biological ionic strength buffer with respect to DPPC/BMP (5 mol%) mixed MLVs in low ionic









strength buffer (Figure 6-9 C). This possible depression of a thermotropic transition is consistent

with the result determined earlier in this chapter by 2H measurements under similar conditions.

The degree of order is similar at 25 C for the MLVs containing BMP (5 mol%) in near

biological ionic strength buffer as can be seen by comparing the peak-to-peak width values in

Figure 6-10 C, indicating a similar packing arrangement of the acyl chains in the gel phase.

However, a comparison of the peak-to-peak width and order parameter values in Figure 6-10 A,

B, and D indicate that DPPC/BMP (5 mol%) mixed MLVs in near biological ionic strength

buffer are less ordered than those in low ionic strength buffer. Clearly, experiments should be

repeated at 250C and data should be obtained in the temperature region between 25 and 35C in

order to verify either observation. At temperatures above the chain melting transition (Tm) of

single component DPPC (- 41.5C) (75, 84) the order parameters are identical within

experimental error. This result is consistent with the observation that the order parameters for 5

and 10-DOXYL labels in POPC/BMP (5 mol%) mixed MLVs in the La phase are not affected by

inclusion of BMP into the MLVs.

1,2-Dipalmitoyl-sn-Glycero-3-Phosphocholine /Bis(monoacylglycerol)phosphate 31P NMR

Additional information concerning the effects that BMP has on lipid molecular order can

be obtained by examining changes in the 31P NMR chemical shift line shapes. Experiments

performed with T>Tm provide information concerning the order of the PC head group in the La

phase and can be compared to results obtained in Chapter 5 for mixed POPC/BMP MLVs.

Figure 6-11 shows a comparison between the 31P chemical shift line shapes of single

component DPPC and DDPC/BMP (5 mol%) MLVs at temperatures above (43C) and below

(35C) the main phase transition temperature of fully hydrated DPPC MLVs. The 31P spectra

obtained from MLVs containing DPPC/BMP are composed of at least two line shapes but are not









simple superpositions of the individual components; as was also seen in the spectra obtained for

POPC/BMP mixed MLVs in Chapter 5.

Plots A and B of Figure 6-11 and values listed in Table 6-8 show that both lamellar CSA

patterns for DPPC/BMP (5 mol%) MLVs at either low (5 mM Na ), or near biological (105 mM

Na+) ionic strength are narrowed with respect to pure DPPC MLVs at 43 C. This result is

consistent with 31P chemical shift measurements reported earlier for POPC/BMP MLVs in La

type assemblies, indicating the choline head group is tilted more toward the bilayer plane (larger

angle from the bilayer normal when compared to single component DPPC) in the presence of

negative charged amphiphiles. This result also provides more evidence that BMPis:1, even at

concentrations below those found in late endosomes, changes the packing parameters of the head

group region above the gel to liquid lamellar phase transition temperature of pure DPPC.

However, for chemical shift line shapes obtained at T< Tm (Figure 6-10 C and D) the head

group region of mixed DPPC/BMP (5 mol%) MLVs is only slightly more disordered in low

ionic strength buffer but is significantly more disordered in near biological strength buffer.

These results are more easily visualized by comparing the plots in Figure 6-12 and values listed

in Table 6-8. The increased concentration of positive counter ions at the buffer bilayer interface

in near biological ionic strength buffers may in fact increase the choline head group tilt induced

by incorporation of BMP into zwitterionic model bilayers.

Figure 6-12 A reports the span of single component DPPC as a function of temperature in

low and near biological ionic strength buffers; the reported span values are in agreement within +

2 ppm of literature values for DPPC MLVs hydrated with water. Griffin and coworkers report

values of-54 ppm (T = 37C) and -48 ppm (T = 45C)(85), while Seelig reports values of-54

ppm (T = 38C) and -49 ppm (T = 44C). (73) The values of the CSA span obtained from DPPC









MLVs as a function of ionic strength do not change; hence, these data indicate that there is no

significant change in head group orientation in DPPC for the two ionic strength buffers used in

our investigation.

A comparison of the span as a function of temperature for DPPC MLVs and DPPC doped

with BMP (5 mol%) MLVs in low and at near biological ionic strength buffers are shown in

Figures 6-12 B and C, respectively. It is evident that BMP induces a change in the PC head

group orientation from that in single component DPPC for both buffer conditions and at all

temperatures investigated. Moreover for DPPC/BMP (5 mol%) mixed MLVs in buffer near

biological ionic strength the PC head group orientation is more affected than for DPPC/BMP (5

mol%) mixed MLVs at low ionic strength. This conclusion is drawn from the smaller CSA span

values in Figure 6-11 D for the MLVs in near biological ionic strength buffer at all temperatures

investigated. Further inspection Figure 6-11 D also shows evidence of a phase transition for

DPPC/BMP mixed MLVs beginning at 370C. It is clear that the chemical shift span and thus the

head group order is decreasing in a linear fashion for mixed MLVs in both low and near

biological strength buffer (the span deceases 1 ppm per C over the temperature range 37 to

41C. The onset of this transition is consistent with that obtained from the total order parameter

of the terminal methyl group as a function of temperature for mixed MLVs under similar

conditions seen in Figure 6-3 B. However, more data points above 43C are needed to fully

characterize this thermotropic phase transition.

It is also interesting to note that as the temperature approaches Tm the values of the CSA

span for the PC head group in DPPC/BMP (5 mol%) MLVs appear to rapidly approach -44 ppm.

Recall the linear relationship between POPC span and mole fraction BMP obtained previously in

Chapter 5 for POPC/BMP mixed MLVs; the predicted span for the POPC CSA span for XBMP of











5 mol% is -45 ppm. However, additional data points at temperatures above 43 C are needed to


confirm this limiting value. Further investigations at temperatures above the main phase


transition and with mole fractions of BMP up to 0.3 would also substantiate the linear


relationship between the 31P chemical shift span and mol fraction of negative amphiphile


obtained for POPC/BMP mixed MLVs.


-50 -40 -30 -20 -10 0 10
Frequency (kHz)


T oC
43

41

I ~--+ '. ~39
38
37

.,35 . .
20 30 40 50 -50 -40 -30 -20 -10 0 10 20 30 40 50
Frequency (kHz)


Figure 6-1 2H NMR spectra of A) DPPC and B) DPPC/BMP (5 mol%) mixed MLVs showing
the temperature dependence of the phase transition. Lipid samples (275 mM total
lipid) were hydrated in a 5 mM HEPES, 0.1 mM EDTA, at pH 7.4.


A) B)
T oC










-50 -40 -30 -20 -10 0 10 20 30 40 50 -50 -40 -30 -20 -10 0 10 20 30 40 50
Frequency (kHz) Frequency (kHz)



Figure 6-2. 2H NMR spectra of A) DPPC MLVs and B) DPPC/BMP (5 mol%) mixed MLVs
showing the temperature dependence of the phase transition. Lipid samples (275 mM
total lipid) were hydrated in a 5 mM HEPES, 100 mM NaC1, at pH 7.4.

















015 A

A
014

A
013
A

012 A
A

011


010 -------------
34 35 36 37 38 39 40 41 42 43 44
4 2 TC


40
38
36
34
S32
2 30
30 -
28
S26
24
22
20 -
34


35 36 37 38 39 40
TC


41 42 43 44


0.015


0014

o
0013 -
0

0012


0011-
0011


0010 --------------I --I -- I i- I = I i I -
34 35 36 37 38 39 40 41 42 43 4'
4 TC
a?


40
38
36
34
,32
S30
28
2 26
24
S22
20 -
34


35 36 37 38 39
T C


40 41 42 43 44


Figure 6-3. Total order parameter and residual quadrupolar splitting of terminal methyl groups

as a function of temperature: A) and C) pure DPPC MLVs (A) and DPPC/BMP

mixed MLVs (A) in 5 mM HEPES, 0.1 mM EDTA, t pH 7.4; B) and D) pure DPPC

MLVs (e) and DPPC/BMP mixed MLVs (o) in 5 mM NaHEPES, 0.1 mM EDTA,

100 mM NaCl at pH 7.4. The error bars are estimated as approximately the FWHM

(+ 300 Hz) of the terminal methyl horn.


0.


0

0
m"




|

0


1:ih


t ~ ~ I


m
ju
a
&
^


I


4













T OC
50
45
44
43
42
41
40
39
38
37
36
35
25
20 G



Figure 6-4. cw-EPR spectra ofDPPC MLVs with 16-DOXYL spin probe (1 mol%) as function
of temperature: 100mM DPPC, 5mM HEPES, 0.1 mM EDTA, at pH 7.4.



T OC
50
45
44
43
42
41
40
39
38
37
36
35
25
20 G



Figure 6-5. cw-EPR spectra of DPPC MLVs with 16-DOXYL spin probe (1 mol%) as function
of temperature: 100mM DPPC, 5mM HEPES, 100 mM NaCI, 0.1 mM EDTA, at pH
7.4.
























35


3
30 -


aA
25
*


20
AAA
*
15
25 30 35 40 45 50

C) Temperature C

035 A


030


0.25
A

0.20 -


015


010 A- ElI ,
n n I I I ,


25 30 35 40

Temperature C


45 50


005
000
-0 05
S-010
S-015
-020
-0 25
-030
-0 35
40

35

30

25

20

15


25 30 35 40

Temoerathre C


25 30 35 40

Temperature C


Figure 6-6. Various spectral parameters of 16-DOXYL labeled lipid incorporated into DPPC

MLVs as a function of temperature: A) peak-to-peak width; B) second moment; C)

order parameter; D) first derivative of sigmoidal fit to AHpp(T) (low ionic strength);

and E) AHpp(T) with sigmoidal fit (dashed line) to AHpp(T) (low ionic strength)

MLVs in 5mM HEPES, 0.1 mM EDTA, at pH 7.4 (A); MLVs in 5mM HEPES, 0.1

mM EDTA, 100 mM NaCl at pH 7.4 (e).










































100


A





it

0
A






A^^ A
~tr'&


45 50


45 50


24 26 28 30 32 34 36 38 40 42 44 46 48 50 52






-
------- -
... i i kA





AI i A_ .....A


-


-
-
-
-
-
-

















T OC
50
45
44
43
42
41
40
39
38
37
36
35


20 G




Figure 6-7. DPPC/BMP MLVs with 16-DOXYL spin probe as function of temperature at low
ionic strength: sample contains 95mM DPPC, 5mM BMP18:1 5mM HEPES, 0.1 mM
EDTA pH 7.4.

































101

















T OC
50
45
44
43
42
41
40
39
38
37
36
35
25
20 G
^^^ IS=


Figure 6-8. DPPC/BMP MLVs with 16-DOXYL spin probe as function of temperature (near
biological ionic strength) 95mM DPPC, 5mM BMPs18: 5mM HEPES, 0.1 mM EDTA
pH 7.4.



























102




















35


30



25

oA
20
20 o


15
25 30 35 40 45 50

C) Temperature C

035


030
0

025


020


015
0
010 00

nooooo n
o is


25 30 35 40
Temperature C


45 50


190


180 0


170


160

0
150
O1O
140 -
l*l l i


25 30 35 40
Temperature C


45 50


000 -
-005
-010
S-015
S-020
S-025
-030
S-035
S-040
-045
24 26 28 30 32 34 36 38 40 42 44 46 48 50


25 30 35 40
Temperature C


45 50


JT U)


SE)


Figure 6-9. Various spectral parameters of 16-DOXYL labeled lipid incorporated into

DPPC/BMP (5 mol%) mixed MLVs as a function of temperature: A) peak-to-peak

width; B) second moment; C) order parameter; D) first derivative of sigmoidal fit to

AHpp(T) (low ionic strength); and E) AHpp(T) with sigmoidal fit (dashed line) to

AHpp(T) (low ionic strength) MLVs in 5mM HEPES, 0.1 mM EDTA, at pH 7.4 (A);

MLVs in 5mM HEPES, 0.1 mM EDTA, 100 mM NaCl at pH 7.4 (o).


































103


52


4 o0 1 1 ,i


































25 30 35 40 45 50
Temperature 'C
















ooooooooo o
0 0





0 00 000 0 0


25 30 35 40
Temperature C


45 50


035


030

025


020

015


010 -

005
25 30 35 40 45 50
Temperature C

035

030


025


020

015


010 O B a

nn I I I


25 30 35 40
Temperature C


45 50


Figure 6-10. Various spectral parameters of 16-DOXYL labeled lipid incorporated into DPPC

MLVs and DPPC/BMP (5 mol%) mixed MLVs as a function of temperature: A)

peak-to-peak width of DPPC MLVs (A) and DPPC/BMP MLVs (A) in 5mM

HEPES, 0.1 mM EDTA, at pH 7.4 ; B) order parameter of DPPC MLVs (A) and

DPPC/BMP MLVs (A) in 5mM HEPES, 0.1 mM EDTA, at pH 7.4; C) peak-to-peak

width of DPPC MLVs (e) and DPPC/BMP MLVs (o) in 5mM HEPES, 0.1 mM

EDTA, ,100 mM NaCl at pH 7.4; D) order parameter of DPPC MLVs (e) and

DPPC/BMP MLVs (o) in 5mM HEPES, 0.1 mM EDTA, ,100 mM NaCl at pH 7.4.


























104












B)

T = 43 oC ,


-60 -40 -20 0 20 40 60 -60 -40 -20 0 20 40 60


-60 -40 -20 0 20 40 60 -60 -40 -20
ppm


0 20 40 60


Figure 6-11. 31P NMR chemical shift of single component DPPC MLVs (solid line),
DPPC/BMP (5 mol%) MLVs (dashed line): A) and C) 5 mM HEPES, 0.1 mM
EDTA, at pH 7.4, 5; B) and D) 5 mM HEPES, 0.1 mM EDTA, 100 mM NaC1, at pH
7.4.


- T= 35 oC















A) B)
-56 -56 B)

-54 A -54 A

52 -52


-50 -50 -
I* S
-48 -48 -

-46 -46

-44 -44 -

-42 I I I I I -42
34 35 36 37 38 39 40 41 42 43 44 34 35 36 37 38 39 40 41 42 43 44
C) T C T C D)
-56 , , ,-- -56 ,---,---,---,---,---,---,---,---,---,

-54 -54

-52 -52 -
S-S
-50 -50 I-

b -48 -

-46 -46- o

-44 o o -44- 0

-42 ~ I L -424
34 35 36 37 38 39 40 41 42 43 44 34 35 36 37 38 39 40 41 42 43 44
T C T C




Figure 6-12. 31P CSA span of DPPC MLVs and DPPC/BMP (5 mol%) mixed MLVs as a

function of temperature: A) DPPC low (A) and high (A) ionic strength; B) DPPC

(A) and mixed DPPC/BMP (5 mol%) (*) both in low ionic strength buffer; C) DPPC

(A) and mixed DPPC/BMP (5 mol%) (o) both in high ionic strength buffer; D) Mixed

DPPC/BMP (5 mol%)low (e) and high (o) ionic strength buffer Error in span = + 1

ppm.





























106









Table 6-1 Total order parameter values for terminal methyls of d62-DPPC in DPPC MLVs.
T oC Order Parameter Order Parameter
(Stotal) 5 mM Na (Stotal) 100 mM Na+
35 N/A N/A


36 0.015 N/A


37 0.015 0.014


38 0.014 0.013


39 0.013 0.012


40 0.013 0.012


41 0.012 0.012


42 0.012 0.011


43 0.012 0.011









Table 6-2 Total order parameter values for terminal methyls of d62-DPPC in DPPC/BMP (5-
mol%) mixed MLVs.
T oC Order Parameter Order Parameter
(Stotai) 5 mM Na (Stota) 100 mM Na
35 N/A 0.013


36 N/A 0.013


37 0.012 0.013


38 0.012 0.013


39 0.012 0.013


40 0.012 0.012


41 0.012 0.012


42 0.012 0.012

43 0.011 0.011









Table 6-3 Parameters defining order of the 16-DOXYL nitroxide spin-probe in DPPC MLVs in
5 mM Na+ buffer.
T C AHpp 2nd Moment Si (0)

25 3.8 191 0.34

35 2.6 175 0.20

36 2.4 163 0.18

37.5 1.9 144 0.12

38 1.9 144 0.11

39 1.9 1434 0.11

40 1.8 145 0.10

41 1.8 143 0.10

42 1.8 145 0.10

43 1.8 144 0.10

44 1.8 142 0.10

45 1.8 142 0.010

50 1.8 141 0.095









Table 6-4 Parameters defining order of the 16-DOXYL nitroxide spin-probe in DPPC MLVs in
105 mM Na buffer.
T C AHpp 2nd Moment Si (0)

25 3.3 193 0.35

35 2.4 173 0.23

36 2.1 167 0.20

37 1.6 143 0.11

38 1.7 143 0.11

39 1.6 144 0.10

40 1.7 141 0.10

41 1.6 142 0.10

42 1.6 142 0.10

43 1.7 144 0.10

44 1.6 144 0.10

45 1.6 142 0.10

50 1.6 141 0.095









Table 6-5 Pre-transition temperatures obtained from 16-DOXYL labeled lipid in various MLV
lamellar structures.
AHpp 2nd Moment Si (0)

DPPC (5 mM Na) 34 C 36 C 35 C

DPPC (105 mMNa+) 35 oC 36 oC 35 C


DPPC/BMP (5 mol%) (5 mM Na') 35 oC 35 oC 35 C


DPPC/BMP (5 mol%) (105 mM Na') 34 oC 33 oC 29 C









Table 6-6 Parameters defining order of the 16-DOXYL nitroxide spin-probe in DPPC/BMP (5
mol%) mixed MLVs in 5 mM Na+ buffer.
T oC AHpp 2nd Moment Si (0)

25 3.2 187 0.30

35 2.5 168 0.21

36 2.2 158 0.14

37 1.9 142 0.11

38 1.9 143 0.11

39 1.9 142 0.11

40 1.8 143 0.11

41 1.8 138 0.11

42 1.8 139 0.11

43 1.8 139 0.11

44 1.8 138 0.11

45 1.8 139 0.11

50 1.8 137 0.09









Table 6-7 Parameters defining order of the 16-DOXYL nitroxide spin-probe in DPPC/BMP (5
mol%) mixed MLVs in 105 mM Na+ buffer.
T oC AHpp 2nd Moment Si (0)

25 3.2 187 0.30

35 2.5 168 0.21

36 2.2 158 0.14

37 1.9 142 0.11

38 1.9 143 0.11

39 1.9 142 0.11

40 1.8 143 0.11

41 1.8 138 0.11

42 1.8 139 0.11

43 1.8 139 0.11

44 1.8 138 0.11

45 1.8 139 0.11

50 1.8 137 0.090










Table 6-8 Span values for DPPC MLVs and DPPC/BMP (5 mol%) mixed MLVs.
T oC Ao (ppm) AG (ppm) AG (ppm) AC (ppm)
DPPC DPPC DPPC/BMP DPPC/BMP
5 mM Na 105 mM Na+ 5 mM Na 105 mM Na
35 -55 -55 -51 -48


36 -54 -54 N/A -47


37 -52 -54 -51 -48


38 -52 -54 -50 -46


39 -52 -53 -49 -45


40 -53 -53 -49 -44


41 -52 -54 -47 -43


42 -52 -54 -45 -44


43 -51 -53 -44 -44









CHAPTER 7
SUMMARY AND CONCLUSIONS

The question posed in the beginning of this dissertation was "What effect does BMP have

on model membrane morphology?" In order to answer this question membrane solubilization

was first investigated using SDS detergent, since our original hypothesis was that BMP might

have detergent properties, such as micelle formation and bilayer solubilization. However, we

have presented data that contradicts our original hypothesis and further corroborates

experimental results obtained by others that BMP self-assembles into a lamellar aggregate

structure in an aqueous environment. Furthermore, our solubilization results indicate that the

acyl chain packing motifs and the order parameters of BMPis:1 are similar to those found in

POPC MLVs in that BMP is solubilized by SDS in a similar fashion. Also, the line shapes

obtained from the 5-DOXYL labeled lipid are almost identical in either single component POPC

or BMP LUVs indicating that the acyl chains pack in a similar manner.

Initial characterization of the interaction of the BMP lipid incorporated into model PC

membrane structures indicated that negatively charged BMP causes choline head groups to tilt

away from the bilayer normal as a function of BMP concentration as seen in other negatively

charged amphiphile/PC membrane mixtures. This is a direct indication that BMP modulates the

surface charge and molecular interactions in the vicinity of the polar region of these mixed

model membrane systems.

Considering the results presented in this body of work there are some key experiments that

should be revisited in order to further our understanding of lipid membrane structures and give

us greater detail regarding the current observations.

2H NMR experiments should be repeated for both DPPC and DPPC/BMP MLVs as a

function of temperature with a larger deuterated lipid component. Increasing the number of









deuterium-enriched lipid molecules will increase the signal to noise ratio, allowing the spectra to

be dePaked so the full order parameter profiles for each system can be reported. Also, spectra

should be collected at a temperature well above Tm for DPPC, e.g. 550C, to ensure that a single

Lc component is present. This is necessary for proper dePaking and peak assignment.

Data points in the temperature region between 25 and 35C should be acquired for all of

the EPR experiments involving the 16-DOXYL spin-labeled lipids. These data points will be

useful in obtaining a better fit of the sigmoidal line shape and thus a better estimate of the

thermotropic phase transition temperatures. Also, selected data points should be recollected to

establish the error in the measurements using the 16-DOXYL probe.

Each EPR solubilization experiment should be repeated initially using only the 5-DOXYL

labeled lipid. The preliminary results of these experiments seem to be very significant and the

data should be obtained in triplicate and extended to both smaller and larger SDS/Lipid ratios.

Other useful experiments would include systematic studies of size distributions between

POPC and POPC/BMP mixed vesicles and MLVs using light scattering techniques, confirmation

of mesophase by cryogenic electron microscopy, and a comparison of these and results obtained

previously with those obtained from mixed MLVs and LUVs containing negatively charged

phosphatidylinositol.













LIST OF REFERENCES

1. P. L. Yeagle, Ed., The Structure ofBiological Membranes (CRC Press 2004).

2. R. B. Gennis, C. R. Cantor, Ed., Biomembranes Molecular Structure and Function.
(Springer-Verlag 1989).

3. D. L. Nelson, M. M. Cox, Lehninger Principles ofBiochemistry, Fourth Edition. (W. H.
Freeman 2004).

4. S. M. Gruner, P. R. Cullis, M. J. Hope, C. P. Tilcock, Annu. Rev. Biophy.s Biophys.
Chem. 14, 211 (1985).

5. E. Gorter, F. Grendel, Journal ofExperimental Medicine 41 (1925).

6. J. R. Silvius, M. Lyons, P. L. Yeagle, T. J. O'Leary, Biochemistry 24, 5388
(Sep. 24, 1985).

7. H. Schindler, J. Seelig, Biochemistry 14, 2283 (Jun. 3, 1975).

8. M. J. Janiak, D. M. Small, G. G. Shipley, Biochemistry 15, 4575 (Oct. 19, 1976).

9. A. M. Thayer, S. J. Kohler, Biochemistry 20, 6831 (Nov 24. 1981).

10. W. L. Hubbell, H. M. McConnell, Pro.c Nat.lAcad. Sci. USA 64, 20 (Sep. 1969).

11. J. Seelig, Q. Rev. Biophys. 10, 353 (Aug.1977).

12. G. G. Shipley, L. S. Avecilla, D. M. Small, J. LipidRes. 15, 124 (Mar. 1974).

13. J. R. Trudell, W. L. Hubbell, E. N. Cohen, Annals New York Academy of Science, 222,
530 (Dec. 1973).

14. W. L. Hubbell, H. M. McConnell, J. Am. Chem. Soc. 93, 314 (Jan. 27, 1971).

15. J. M. Boggs, J. C. Hsia, Proc. Natl. Acad. Sci. USA 70, 1406 (May. 1973).

16. J. Herzfeld, R. G. Griffin, R. A. Haberkorn, Biochemistry 17, 2711 (1978).

17. A. Seelig, J. Seelig, Biochemistry 13, 4839 (Nov. 5, 1974).

18. M. Auger, Biophys. Chem. 68, 233 (Oct. 1997).

19. D. R. Body, G. M. Gray, Chemistry andPhysics ofLipids 1, 254 (1967).










20. T. Kolter, K. Sandhoff, Annu. Rev. CellDev. Biol. 21, 81 (2005).


21. T. Kobayashi, K. Startchev, A. J. Whitney, J. Gruenber, Biol. Chem. 382, 483
(Mar. 2001).

22. T. Kobayashi, E. Stang, K.S. Fang, P. de Moerloose, R. G. Parton, J.
Gruenberg, Nature 392, 193 (Mar. 12, 1998).

23. T. Kobayashi, M. Beuchat, J. H.Chevallier, A. Makino, N. Mayran, J.Escola, C,
M.Lebrand, P. Cosson, J. Gruenberg, J. Biol. Chem. 277, 32157 (Aug. 30, 2002).

24. B. M. Mullock, R. H. Hinton, J. V. Peppard, J. W. Slot, J. P. Luzio, CellBiochem. Funct.
5, 235 (Oct. 1987).

25. T. Kobayashi, F. Gu, J. Gruenberg, Semin. CellDev. Biol. 9, 517 (Oct, 1998).

26. H. Matsuo,J. Chevallier, N. Mayran, I. Le Blanc, C. Ferguson, J. Faure, N. S. Blanc,
S. Matile, J. Dubochet, R. Sadoul, R. G. Parton, F. Vilbois, J.Gruenberg, Science 303,
531 (Jan. 23, 2004).

27. J. Brothers, O. Renkonen, W. Fischer, J. Herrmann, Chem. Phys. Lipids 13, 178
(Oct. 1974).

28. T. Hayakawa, Hirano, Y., A. Makino, S. Michaud, M. Lagarde, J. F. Pageaux, A.
Doutheau, K.Ito, T. Fujisawa, H. Takahashi, T.. Kobayashi, Biochemistry 45, 9198
(Aug. 1, 2006).

29. J. M. Holopainen Soderlund, T., J. M.Alakoskela, M. Saily, O. Eriksson,Kinnunen, P.
K., Chem. Phys. Lipids 133, 51 (Jan. 2005).

30. N. Werth, C. G. Schuette, G. Wilkening, T. Lemm, K. Sandhoff, J. Biol. Chem. 276,
12685 (Apr. 20, 2001).

31. C. D. Stubbs, S. J. Slater, Chem. Phys. Lipids 81, 185 (Jul.15, 1996).

32. I. Dikic, Bioessays 26, 604 (Jun. 2004).

33. T. Kolter, K. Sandhoff, Angew. Chem. Int. Ed. 38, 1532 (1999).

34. K. Sandhoff, T. Kolter, Trends CellBiol. 6, 98 (Mar. 1996).

35. F. R. Maxfield, T. E. McGraw, Nat. Rev. Mol. Cell Biol. 5, 121 (Feb. 2004).

36. M. Koval, R. E. Pagano, J. CellBiol. 108, 2169 (Jun. 1989).









37. M. Koval, R. E. Pagano, J. CellBiol. 111, 429 (Aug. 1990).

38. K. Laulagnier Motta, C. S. Hamdi, S. Roy, F. Fauvelle, J. F. Pageaux, T. Kobayashi, J.
P. Salles, B. Perret, C. Bonnerot, M. Record, Biochemistry Journal, 161 (2004).

39. R. D. Singh Liu, Y. C. L. Wheatley, E. L. Holicky, A. Makino, D. L. Marks, T.
Kobayashi, G. Subramaniam, R. Bittman,Pagano, R. E., J. Biol. Chem. 281, 30660
(Oct. 13, 2006).

40. F. Aussenac, M. Tavares, E. J. Dufourc, Biochemistry 42, 1383 (Feb. 18, 2003).

41. R. E. Glover, R. R. Smith, M. V. Jones, S. K. Jackson, C. C. Rowlands, FEMSMicrobiol.
Lett. 177, 57 (Aug. 1, 1999).

42. N. Van Eps, W. M. Oldham, H. E. Hamm, W. L. Hubbell, Proc. Natl. Acad. Sci. USA
103, 16194 (Oct. 31, 2006).

43. P. J. Davis, B. D. Fleming, K. P. Coolbear, K. M. Keough, Biochemistry 20, 3633
(Jun. 9, 1981).

44. K. Gawrisch. Buffer preparations in D20 (Private Communication 2007).

45. H. Warner, T. Y. Toribara, P. S. Chen, Analytical Chemistry 28, 1756 (1956).

46. K. Gawrisch. The Dynamics of Lipid Membranes (Private Communication 2007).

47. M. D. King, D. Marsh, Biochemistry 26, 1224 (Mar. 10, 1987).

48. D. Marsh, Journal ofMagnetic Resonance 157, 114 (2002).

49. K. Schorn, D. Marsh, Biochemistry 35, 3831 (Mar. 26, 1996).

50. H. M. McConnell, B. G. McFarland, Ann. N.Y. Acad. Sci. 195, 207 (Jun. 20, 1972).

51. H. M. McConnell, K. L. Wright, B. G. McFarland, Biochem. Biophys. Res. Commun. 47,
273 (Apr. 14, 1972).

52. W. L. Hubbell, H. M. McConnell, Proc. Natl. Acad. Sci. USA 61, 12 (Sep. 1968).

53. W. L. Hubbell, J. C. Metcalfe, S. M. Metcalfe, H. M. McConnell, Biochim. Biophys. Acta
219, 415 (Dec. 1, 1970).

54. W. L. Hubbell, Stanford University (1969).

55. P. Poole, Electron Spin Resonance. (John Wiley and Sons, Inc., 1983).









56. H. S. McHaourab, M. A. Lietzow, K. Hideg, W. L. Hubbell, Biochemistry 35, 7692
(Jun. 18, 1996).

57. J. H. Davis, Biochim. Biophys. Acta 737, 117 (Mar. 21, 1983).

58. A. Abragam, Principles ofNuclear Magnetism (Clarendon Press, 1961).

59. R. G. Griffin, L. Powers, P. S. Pershan, Biochemistry 17, 2718 (1978).

60. J. S. Santos, The University of Michigan (2001).

61. M. le Maire, P. Champeil, J. V. Moller, Biochim. Biophys. Acta 1508, 86
(Nov. 23, 2000).

62. S. Keller, H. Heerklotz, N. Jahnke, A. Blume, Biophys. J 90, 4509 (Jun. 15, 2006).

63. A. Tan, A. Ziegler, B. Steinbauer, J. Seelig, Biophys. J. 83, 1547 (Sep. 2002).

64. P. R. Mahji, A. Blume, Langmuir 17, 3844 (2001).

65. N. Deo, P. Somasundaran, K. Subramanyan, K. P. Ananthapadmanabhan, Journal of
Colloid and Interface Science 256, 100 (2002).

66. R. G. Griffin, J. Am. Chem. Soc. 98, 851 (Feb. 4, 1976).

67. J. A. Berden, P. R. Cullis, D. I. Hoult, A. C. McLaughlin, G. K.Radda,Richards, R. E.,
FEBSLett. 46, 55 (Sep. 15, 1974).

68. P. G. Scherer, J. Seelig, Biochemistry 28, 7720 (Sep. 19, 1989).

69. L. K. Tamm, J. Seelig, Biochemistry 22, 1474 (1983).

70. S. Abu-Baker, X. Qi, G. A. Lorigan, Biophysical Journal 93, 3480 (2007).

71. D. Massiot, F. Fayon, M. Capron, I. King, S. Le Calve, B. Alonso, J-O Durand, B. Bujoli,
Z. Gan, G. Hoatson, Magnetic Resonance in Chemistry 40, 70 (2002).

72. J. Seelig, H. Gally, Biochemistry 15, 5199 (Nov. 30, 1976).

73. W. Neiderberger, J. Seelig, J. Am. Chem. Soc. 98, 3704 (Jun. 9, 1976).

74. J. Seelig, P. M. Macdonald, P. G. Scherer, Biochemistry 26, 7535 (Dec. 1, 1987).

75. J. H. Davis, M. R. Vist, Biochemistry 29, 451 (1990).

76. J. H. Davis, Biochimica et. Biophysicsa Acta 737, 117 (1983).










77. M. F. Brown, S. Lope-Piedrafita, G. V. Martinez, H. I. Petrache, in Modern Magnetic
Resonance (Springer, 2006).

78. E. Sternin, in Membrane Lipid Protocols (Humana Press, Inc., 2005).

79. M. Bloom, J. H. Davis, A. L. MacKay, Chem. Phys. Lett. 80, 198 (1981).

80. J. Dey, R. R. Nayak, S. Roy, Polymer 46, 12401 (2005).

81. T. M. Bayerl, T. Brumm, H. Reinl, Biophysics Journal 61, 1025 (1992).

82. M. Korkmaz, C. Aydas, Journal of the Korean Physical Society 45, 975 (2004).

83. D. G. Cameron, H. L. Casal, H. H. Mantsch, Biochemistry 19, 3665 (Aug. 5, 1980).

84. T. P. McMullen, R. N. McElhaney, Biochim. Biophys. Acta 1234, 90 (Mar. 8, 1995).

85. R. G. Griffin, M. J. Ruocco, D. J. Siminovitch, Biochemistry 24, 2406 (1985).









BIOGRAPHICAL SKETCH

Chad Mair was born on December 29, 1975 in Evansville, Indiana. He spent his childhood

years in Owensville, Indiana where he attended Owensville Elementary. After completing

secondary education, at Gibson Southern High School, he began his undergraduate education in

Evansville at the University of Southern Indiana. Chad received a Bachelor of Science degree in

chemistry from USI in 2001. His graduate education started at the University of Florida in the

area of physical chemistry under the supervision of Dr. Valeria Kleiman. He spent four years

working in the field of ultrafast laser spectroscopy on hyperconjugated polymers. Chad then

began working on magnetic resonance spectroscopy of model biological membranes under the

supervision of Dr. Gail E. Fanucci.





PAGE 1

BIS(MONOACYLGLYCEROL)PHOSPHATE EFFECT ON MODEL MEMBRANE MORPHOLOGY: A MAGNETIC RE SONANCE INVESTIGATION By CHAD E. MAIR A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2008 1

PAGE 2

2008 CHAD E. MAIR 2

PAGE 3

To my mother, and father Cyra B. and Robert V. Mair 3

PAGE 4

ACKNOWLEDGMENTS I first thank my advisor Dr. Gail E. Fanucci fo r allowing me to complete my graduate work in her research group and for her patience as a me ntor. I consider myself fortunate for the two years of scientific interacti on with, in my opinion, one of th e top scientific minds in our department. I also thank my supervisory committee Profe ssors John Eyler, Clifford Bowers, Joanna Long, and Tom Lyons. Special thanks are extended to Dr. Long for her help with the setup of the NMR instrumentation, insightful discussions, a nd use of her laboratory facilities and to Dr. Bowers for use of his laboratory facilities. I thank the University of Florida machine shop, especially Mr. Todd Prox for his patience and help in design and machining of custom parts. Many fellow graduate students have played an important role in my graduate education. Evrim Atas and Daniel Kuroda were my soundi ng boards while I was a member of the Kleiman group. Luis Galliano has been my scientific conf idant and became a close friend after I joined the Fanucci group. However, I am most grateful to Dr. Lindsay Hardison she has become my best friend and support group since he r arrival at UF in 2002. I woul d not have finished the PhD. without her! Finally, I thank my parents for their encour agement and support, not only during my academic career but throughout my life. 4

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........7 LIST OF FIGURES.........................................................................................................................8 ABSTRACT...................................................................................................................................11 Chapter 1 INTRODUCTION................................................................................................................. .12 Lipid Description and Classification......................................................................................12 Typical Membrane Lipid Structure........................................................................................13 Lipid Self-Assembly and Organization..................................................................................13 Characterization of Model Memb rane Structural Properties..................................................16 Bis(monoacylglycerol)phosphate...........................................................................................18 Bis(monoacylglycerol)phosphate May Be Important to Ganglioside Catabolism.................21 Biological Membranes........................................................................................................... .23 Dissertation Outline........................................................................................................... .....24 2 MATERIALS AND METHODS...........................................................................................32 Materials.................................................................................................................................32 Multilamellar Vesicle Preparation..........................................................................................32 Phosphate Assay.....................................................................................................................35 Thin Layer Chromatography..................................................................................................35 Magnetic Resonance............................................................................................................. ..36 Data Processing......................................................................................................................37 3 MAGNETIC RESONANCE APPLICATIO NS IN MEMBRANE BIOPHYSICS...............38 Hydrated Lipid Motions and Order........................................................................................38 Nitroxide Spin-Probes.......................................................................................................... ..40 Description of Nitroxide Spin-Label Orde r Parameter in Hydrated Lipid Bilayer Assemblies Obtained by Electron Paramagnetic Resonance..............................................40 Orientation of the Nitroxide Spin-Label in Hydrated Lipid Bilayer Assemblies and Expected EPR Lineshapes..................................................................................................43 Solid State 31P and 2H NMR of Hydrated Lipid Aggregates..................................................45 Magnetic Resonance Line Shapes and Order in Hydrated Lipids..........................................46 5

PAGE 6

6 4 MONITORING MODEL BILAYER SO LUBILIZATION BY DETERGENT MOLECULES USING EP R SPECTROSCOPY...................................................................55 Model Membrane Solubilization............................................................................................55 Characterization of the EPR Line Shapes of Spin-probes Located in Bilayer Aggregates in the Presence of an Anionic Detergent.............................................................................58 Hydrated Bis(monoacylglycerol)phosphate Assemblies Solubilized by Sodium Dodecyl Sulfate.................................................................................................................................61 5 PERTURBATIONS OF LAMELLAR LIQUID CHRYSTALLINE ORDER BY BIS(MONOACYLGLYCEROL)PHOSPHATE....................................................................73 Solid State Phosphorus-31 NMR Investigation of Hydrated Bis(monoacylglycerol)phosphate Aggregation State..........................................................73 Acyl Chain Order of 1-Palmitoyl-2 -Oleoyl-sn-Glycero-3-Phosphocholine /Bis(monoacylglycerol)phosphate Mixe d Vesicles Determined by Electron Paramagnetic Spectroscopy................................................................................................74 31P NMR of 1-Palmitoyl-2-Oleoyl -sn-Glycero-3-Phosphocholine /Bis(monoacylglycerol)phosphate mixed MLVs................................................................75 6 PERTURBATIONS OF THERMOTROPIC PHASE TRANSITIONS DUE TO BIS(MONOACYLGLYCEROL)PHOSPHATE....................................................................88 Thermotropic Phase Behavior of 1,2-Di palmitoyl-sn-Glycero-3-Phosphocholine /Bis(monoacylglycerol)phosphate MLVs Investigated by 2H NMR..................................88 Thermotropic Phase Behavior of 1,2-Di palmitoyl-sn-Glycero-3-Phosphocholine /Bis(monoacylglycerol)phosphate MLVs Investigated by EPR.........................................91 1,2-Dipalmitoyl-sn-Glycero-3-Phosphocho line /Bis(monoacylglycerol)phosphate 31P NMR....................................................................................................................................94 7 SUMMARY AND CONCLUSIONS...................................................................................115 LIST OF REFERENCES.............................................................................................................117 BIOGRAPHICAL SKETCH.......................................................................................................122

PAGE 7

LIST OF TABLES Table page 4-1 Parameters defining order of the 5-DOXYL nitroxide spin -probe in lipid aggregates at room temperature...........................................................................................................7 2 4-2 Parameters defining order of the 10-DOXYL nitroxide spin -probe in lipid aggregates at room temperature...........................................................................................................7 2 5-1 Parameters defining order of the 5-DOX YL nitroxide spin-probe in POPC/BMP mixed MLVs at room temperature.....................................................................................85 5-2 Parameters defining order of the 10-DOXYL nitroxide spin -probe in lipid aggregates at room temperature...........................................................................................................8 6 5-3 Values of CSA span for POPC/BMP mixed MLVs at room temperature.........................87 6-1 Total order parameter values for term inal methyls of d62-DPPC in DPPC MLVs.........107 6-2 Total order parameter values for termin al methyls of d62-DPPC in DPPC/BMP (5mol%) mixed MLVs........................................................................................................108 6-3 Parameters defining order of the 16-DOXYL nitroxide spin-probe in DPPC MLVs in 5 mM Na+ buffer..............................................................................................................109 6-4 Parameters defining order of the 16-DOXYL nitroxide spin-probe in DPPC MLVs in 105 mM Na+ buffer..........................................................................................................110 6-5 Pre-transition temperatures obtained fr om 16-DOXYL labeled lipid in various MLV lamellar structures............................................................................................................ 111 6-6 Parameters defining order of the 16-DOXYL nitroxide spin-probe in DPPC/BMP (5 mol%) mixed MLVs in 5 mM Na+ buffer......................................................................112 6-7 Parameters defining order of the 16-DOXYL nitroxide spin-probe in DPPC/BMP (5 mol%) mixed MLVs in 105 mM Na+ buffer..................................................................113 6-8 Span values for DPPC MLVs and DPPC/BMP (5 mol%) mixed MLVs........................114 7

PAGE 8

LIST OF FIGURES Figure page 1-1 Lipid classification....................................................................................................... ......25 1-2 Anatomy of 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) and D-erythrosphingosine-1-phosphate....................................................................................................26 1-3 Cross-sectional representations of lipid polymorphic structures in aqueous environments................................................................................................................... ...27 1-4 Geometric shape approxima tions and lipid aggregates.....................................................28 1-5 Examples of labeled lipids................................................................................................. 29 1-6 BMP structural isomers..................................................................................................... .30 1-7 GM1 lipid.................................................................................................................. .........30 1-8 Sandhoff-Kolter model for lysosomal membrane digestion and endocytosis...................31 3-1 Selected lipid motions and associated correlation times in hydrated lamellar structures..................................................................................................................... .......47 3-2 An order parameter as a function of the angular displacement of the plane containing a specific carbon and two deuterium atoms from the bilayer normal................................48 3-3 Common organic radical spin-labels.................................................................................49 3-4 Energy level diagram illustrating the electronic Zeeman and electron-nuclear hyperfine interactions and the resulting derivative of EPR transitions for a spin-probe such as, TEMPO in solution..............................................................................................50 3-5 4,4-dimethyloxazolidine-3-oxyl labeled 5-cholestane-3-one........................................50 3-6 Spatial dependence of the coupling strength of the anisot ropic hyperfine interaction......51 3-7 Theoretical nitroxide label EPR line shapes......................................................................52 3-8 Energy level diagram illustrating the nuclear Zeeman and quadruplar coupling interaction of 2H in an applied field Bo and for the 1H decoupled chemical shift of 31P in an applied field Bo..........................................................................................................52 3-9 Theoretical quadrupolar echo powder spect rum of a single deuterium labeled site ( 0)..................................................................................................................................53 3-10 Graphical representation of CSA and EFG the angular orientations with respect to the applied field (Bo)..........................................................................................................53 8

PAGE 9

3-11 Theoretical powder spectra of various lipid aggregates....................................................54 4-1 Model membrane perturbants............................................................................................64 4-2 Positional isomers of spin-labeled lipids...........................................................................65 4-3 Phase Diagram for SDS and POPC LUVS........................................................................66 4-4 cw-EPR spectra of POPC LUVs with 5-DOXYL spin probe............................................67 4-5 cw-EPR spectra of POPC LUVs with 10-DOXYL spin probe..........................................67 4-6 Various spectral parameters of 5-DOXYL labeled lipid incorporated into POPC LUVs as a function of SDS/ Lipid concentration ratio.......................................................68 4-7 Various spectral parameters of 10-DOXYL labeled lipid incor porated into POPC LUVs as a function of SDS/ Lipid concentration ratio.......................................................69 4-8 cw-EPR spectra of BMP with 5-DOXYL spin probe........................................................70 4-9 Various spectral parameters of 5-DOXYL labeled lipid incorporated into POPC LUVs and BMP aggregates...............................................................................................71 5-1 31P NMR chemical shift spectrum of BMP18:1 MLVs.......................................................78 5-2 cw-EPR spectra of POPC/BMP mixed LUVs...................................................................79 5-3 Hpp and Si of 5 and 10-DOXYL labeled lipi d (1 mol%) incorporated into POPC/BMP14:0 mixed LUVs as a function of BMP14:0/Lipid concentration ratio.............80 5-4 Hpp and Si of 5 and 10-DOXYL labeled lipi d (1mol%) incorporated into POPC/BMP18:1 mixed LUVs as a function of BMP18:1/Lipid concentration ratio.............80 5-5 31P NMR chemical shift of POPC MLVs in 5 mM HEPES..............................................81 5-6 31P NMR spectra of POPC/BMP18:1 MLVs in 5 mM HEPES, 100 mM NaCl................82 5-7 31P NMR spectra of POPC/BMP18:1 MLVs.....................................................................83 5-8 CSA span of (POPC/BMP MLVs) as a function of BMP mole fraction...........................84 6-1 2 H NMR spectra: DPPC and 5 mo l% BMP/DPPC mixed MLVs.....................................97 6-2 2 H NMR spectra of DPPC MLVs and BMP/ DPPC (5 mol%) mixed MLVs the in a 5 mM HEPES, 100 mM NaCl, at pH 7.4 buffer................................................................97 6-3 Total order parameter and residual quadrupol ar splittings of terminal methyl groups as a function of temperature...............................................................................................98 9

PAGE 10

10 6-4 cw-EPR spectra of DPPC MLVs with 16-DOXYL spin probe (1 mol%) as function of temperature................................................................................................................. ...99 6-5 cw-EPR spectra of DPPC MLVs with 16-DOXYL spin probe (1 mol%) as function of temperature: 100mM DPPC, 5mM HEPES, 100 mM NaCl, 0.1 mM EDTA, at pH 7.4.................................................................................................................................99 6-6 Various spectral parameters of 16-DOXYL labeled lipid incor porated into DPPC MLVs as a function of temperature.................................................................................100 6-7 DPPC/BMP MLVs with 16-DOXYL spin pr obe as function of temperature at low ionic strength....................................................................................................................101 6-8 DPPC/BMP MLVs with 16-DOXYL spin pr obe as function of temperature (near biological ionic strength).................................................................................................102 6-9 Various spectral parameters of 16-DOXYL labeled lipid incorporated into DPPC/BMP (5 mol%) mixed MLVs as a function of temperature.................................103 6-10 Various spectral parameters of 16-DOXYL labeled lipid incor porated into DPPC MLVs and DPPC/BMP (5 mol%) mixed MLVs as a function of temperature...............104 6-11 31P NMR chemical shift of single co mponent DPPC MLVs and DPPC/BMP (5 mol%) MLVs...................................................................................................................105 6-12 31P CSA span of DPPC MLVs and DPPC/BMP (5 mol%) mixed MLVs as a function of temperature................................................................................................................. .106

PAGE 11

Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy BIS(MONOACYLGLYCEROL)PHOSPHATE EFFECT ON MODEL MEMBRANE MORPHOLOGY: A MAGNETIC RE SONANCE INVESTIGATION By Chad E. Mair May 2008 Chair: Gail E. Fanucci Major: Chemisrty Bis(monoacylglycerol)phosphate (BMP) or (S,R Isomer)sn-(3-Oleoyl-2-Hydroxy)Glycerol-1-Phospho-sn-3'-(1'-Oleoyl-2'-Hydroxy )-Glycerol,Ammonium Sa lt is a phospholipid found primarily in late endosomes, and has a unique st ructure due to single acyl chains located at the 3 and 1 positions on the glycerol components. BMP is known to play an important role in late-endosome sorting functions, and is also thought to be involved in glycosphingolipid catobilism. When BMP is present in liposom es containing ganglioside GM1, the enzymatic hydrolysis of GM1 to GM2, stimulated by activat or proteins, is dramatically enhanced. This work is focused on determining the effect BMP has upon model membrane lipid morphology, and acyl chain dynamics, using magne tic resonance spectroscopy, in order to further understand BMPs role in lipid cat abolism and lysosomal storage disease. 11

PAGE 12

12 CHAPTER 1 INTRODUCTION Lipid Description and Classification In general, lipids can be described as molecu les that are composed of mainly hydrogen and carbon atoms, they can be as simple as al kanes or as complex as lipopolysaccharides( 1 ). Lipid molecules isolated from biological sources are generally classified as either neutral storage lipids or zwitterionic/charged memb rane lipids (Figure 1-1)(1-3 ). The chemical composition of storage lipids (triacylglycerides or tri acylglycerols) includes a glycerol backbone with three fatty acid molecules linked via ester bonds to three react ive hydroxide groups on the glycerol backbone( 3 ). Membrane lipid chemical structures contain eith er a glycerol or a sphi ngosine backbone with fatty acids linked to two of the reactive groups of the glycerol or to the reactive amine group of the sphingosine (Figure 1-2). Th e other reactive hydroxide moiety on each backbone is linked to either a saccharide or more commonly, linked to a phosphate group. Most biologically relevant me mbrane lipids are amphipathic in nature, having both a polar and an apolar region( 1-3 ). The polar portion of a lipid is us ually referred to as the polar head group region, which is defined as the volume containing the substituted phosphate/saccharide, the glycerol or the sphingosin e backbone, and the alpha carbons of the carboxylic acid chains. The apolar portion is defined as the acyl chain tail region( 1, 2 ). Usually, the apolar tails are long hydrocarbon chains covalently bound via ester bonds to glycerol or amide bonds to sphingosine backbone moieties (Figure 1-2)( 1, 2). According to Yeagle, a long chain is defined as a hydrocarbon chain of 12 or more carbon atoms( 1).

PAGE 13

Typical Membrane Lipid Structure The most abundant eukaryotic membra ne lipids are glycerophospholipids( 1, 2), named for the glycerol backbone and the phosphate group. Thes e particular lipids ar e derivatives of snglycero-3-phosphoric acid, in which the sn1 and sn2 positions commonly c ontain esterified, long chain, fatty acids( 2 ). The stereospecific numbering (sn) of phospholipids is derived by drawing a Fischer projection of glycerol with the -hydroxl group on the left, the sn1 position is then located at the top of this projection while the s n3 position is located at th e bottom. Typically, in lipids isolated from natural mammalian sources, the acyl chain located in the sn1 position is saturated and the acyl chain in the sn2 position c ontains at least one site of unsaturation or double bond(1). Chemical moieties attached to the phosphate, located at the sn3 position of membrane lipids, are referred to as head groups, the chemical identiti es of which give rise to the diverse functionality found in biologically important lipids. For example, in glycerophospholipids, the phosphate group has a negative charge, thus the chemical structure/identity of the head group imparts the ch aracteristic charge or zwitterionic character, apparent size of the head group region, and propen sity to hydrogen bond to other membrane lipid molecules. These head group moieties include but are not limited t o, glycerol, inositol, ethanolamine, serine, ch oline, and carbohydrates( 1, 2). Lipid Self-Assembly and Organization Because of their amphiphilic character, lipids with polar heads and apolar tails selfassemble according to their solvent environmen t. In aqueous environments, the entropic hydrophobic effect causes apolar acyl chain regions to organize in such a manner as to minimize the free energy of the system and exclude the maximum number of water molecules from the lipid assembly( 1, 2). Several polymorphs/mesomorphs or macroscopic aggregation states are possible for solvated lipids, e .g. lamellar, hexagonal, micellar and inverted micellar (Figure 113

PAGE 14

14 3)( 1, 2, 4). The specific polymorphism formed can be influenced by many factors including structural features of the lipid, such as chem ical composition of the head group region, length and degree of saturation of the acyl chains, and extrinsic parameters such as temperature, pressure, and degree of hydration( 1, 2, 4 ). Each of these factors provides a means of controlling the packing density (interaction en ergy) of the individua l lipid molecules within an aggregate, which in turn can have dramatic effects upon th e physical and morphological properties of the assembly( 4). Under physiological conditions, the most co mmon type of lipid organization found in living cells is a two-dimensional lamellar or bilayer structure( 4, 5). A bilayer is formed by stacking lipid molecules tail-to-tail bound on either side by pola r head groups in a repeating pattern (Figure 1-3 A and B)( 5). Although the biologically re levant mesophase/polymorph is lamellar, other types of mesophase structures can exist under certain circumstances, e.g. increasing the temperature or changing the hydrati on state of lamellar assemblies can result in the formation of a hexagonal aggregat e structure (Figur e 1-3 C and D)( 4). Lipids organized in a three-dimensional hexagonal arrangement either surround a cylindrical column of solvent with polar groups (HII) or organize with polar head groups f acing out toward the bulk solvent phase (HI). Micelles and inverted micelles are organized in a similar fashion as the hexagonal phases except the aggregates are spherical in shape. The specific assembly formed is influenced by the solvent environment and the apparent molecular shape of the lipids. Molecules that occupy a molecular cross-sect ional area resembling a cone, inverted cone, or cylinder are predicted to form hexagonal (HII), micellar, or lamellar aggregates, respectively( 4). A simple geometric model based on relative cross-sectio nal areas of head groups and acyl chains can be used to predict aggregation states for hydrated lipid molecules( 4 ).

PAGE 15

Lipid molecules with a head group cross-sectional area greater than the cross-sectional area of their acyl chains se lf-assemble into a micellar type stru cture. Those with head group area approximately equal to the acyl chains area prefer to assemble into a lamellar type structure, while lipids with head group area less than the area of the acyl ch ains assemble into a hexagonal type structure. An illustration of the geometric shapes and representative lipids of each category can be found in Figure 1-4. Sodium dodecyl sulfat e (SDS) has a head group with a much larger cross-sectional area than its si ngle acyl chain, thus hydrated SD S molecules will polymorph into micellar aggregates. A different situati on arises for 1,2-dipalmitoyl-sn-glycero-3phosphocholine (DPPC). The phosphocholine head gr oup area and the crosssectional area of the two palmitoyl chains are similar; therefore, D PPC is predicted to form a lamellar aggregate in an aqueous environment. Phosphatidic acid (PA) in an acidic environment has a small head group area with respect to the area occupied by the two acyl chains and forms a HII aggregate. Two broad categories of phase transitions can be defined for lipid polymorphs or aggregates: 1) those that result from changes in organization and packing of the acyl chains and mean volume occupied per lipid molecule, and 2) those that alter the mesophase or polymorphic structure of the lipid aggregate( 1, 4). Aggregation state, or polymorphic transitions, can be altered in various ways including partitioning of a variety of molecule s into the assembly, changing temperature, degree of hydration and lip id composition of the assembly, and varying ionic strength or pH of the solvent( 1, 2, 4, 6 ), or via changes in extrinsic variables such as temperature and pressure. However, in biological systems, the pressure can often be assumed to remain constant; hence, most studies of membra ne biophysics focus on the effects of changes in temperature, not pressure. A common phase tr ansition in lamellar mes ophases exploited within the membrane biophysics field is the la mellar gel to lamella r liquid transition( 1, 2, 4). This 15

PAGE 16

transition is illustrated in Figure 1-3. The gel pha se (Figure 1-3 A) is characterized by closely packed molecules and a more extended acyl chain conformation. Whereas the liquid state (Figure 1-3 B)) retains a two-dime nsional order but lipid molecules are diffusing axially and the acyl chain region is less ordered (h igh probability of trans/gauche isomerization) when compared to the gel phase( 4, 7). Factors such as solv ent ionic strength, degree of acyl chain saturation, lipid composition (mixtures of lipid molecules) a nd addition of protein/pe ptides perturbants can all affect the temperature at which this thermotropic phase transition occurs( 1, 2, 4). For example, at temperatures below the main transition temperature ( Tm), fully hydrated DPPC exists in either a lamellar gel state with saturated acyl chains exte nded and highly ordered (very few gauche rotamers), or in a ripple phase (P) with lipid molecules slightly tilted with respect to the bilayer normal( 8) resulting in a decrease in bilayer thic kness with respect to the lamellar gel state. The ripple phase is characterized by tilted and extended acyl chains that appear to have a symmetric rotational axis (defined in Chapter 3), thus the lipids are more disordered with respect to the gel phase but more ordered than the liquid lamellar phase( 1). However, at temperatures above Tm the chains become less extended, thus re ducing the bilayer th ickness to a greater extent, making the chains more disorder ed until a pure liquid crystalline phase (L) is present. Also, fully hydrated, non-bil ayer structures, e.g. HII, can be induced by further increasing the temperature( 4, 9). An example of a lipid exhibiting a transition to the HII state from either the lamellar liquid or lamellar gel st ate is phosphatidylehanolamine (PE)(4, 9). Characterization of Model Membrane Structural Properties X-ray diffraction, neutron scattering, solid st ate nuclear magnetic resonance (SSNMR), and electron paramagnetic resonance (EPR) are techniques commonly used to investigate the structure and dynamical properties of lipid assemblies(2, 10, 11). Although diffraction and 16

PAGE 17

scattering methods provide a detailed description of the thermodynamic phase structure of lipid assemblies and bilayer thicknesses, these techniques measure only a static macroscopic structure with little information regarding molecu lar motion and require a periodic lattice( 2, 12). A periodic lattice or an array of a large number of molecules ar ranged in a periodic structure typically occurs in very pur e systems. However, biologi cal membranes are inherently heterogeneous; thus, diffraction techniques ar e limited to purified systems in which a well organized lattice structure has b een induced. In some cases, di ffraction techniques can be used to measure the macroscopic structure in less organized systems but they lack the resolution obtainable by magnetic resonance an d other spectroscopic techniques. Investigations utilizing NMR and/or EPR provide detailed molecular level information regarding the degree of orde r at individual chemical bond s for particular molecules( 10, 11, 1316). The types of order/disorder commonly inve stigated using these sp ectroscopic techniques include a motionally averaged picture of phospholipid head group and acyl chain angular orientation with respect to the bilayer normal( 10, 16, 17). Because all phospholipids (by definition) contain a phosphate group, the phosphorus atom provides a natural probe for 31P NMR investigation of average head group orienta tion. Advantages of this technique include 100% natural abundance of the 31P isotope and a good sensitivity due to a relatively large gyromagnetic ratio ( is about 40% of the 1H nucleus). It has been shown that solid-state 31P NMR chemical shift line shapes can be used to differentiate between lamellar, hexagonal (HII), and isotropic polymorphic assemblies( 4); however, line shape can only be consistent with a particular mesomorphic state. X-ray or cryoge nic electron microscopy measurements are needed to fully confirm the aggregate morphology. 17

PAGE 18

The degree of acyl chain organization of the bila yer interior can be i nvestigated either by 2H NMR studies of lipids contai ning deuterated acyl chains or by EPR studies of lipid dispersions containing 0.5 1 mol% of a lipid labeled with a pa ramagnetic nitroxide spinprobe (Figure 1-5)( 10, 11, 14, 18). A major disadvantage for both EPR and 2H NMR measurements is the need for a synthetic probe. Additionally, 2H has a value that is only 15% of the 1H nucleus and requires large samples compared to 31P or EPR sample sizes due to its relatively low sensitivity. Despite these constraints, magne tic resonance techniques are invaluable for biophysical investigations of heterogeneous, non-periodic model and biological membranes. Thus, they are the primary tools employed in ou r pursuit to describe the conformation and acyl chain order of mixed lipids in a model membrane system. Bis(monoacylglycerol)phosphate Since first isolation from ovine lung hom ogenate in 1967, bis(monoacylglycero)phosphate (BMP) (Figure 1-6), also known as lysobisphos phatidic acid (LBPA), has been found to represent less than 1% of total phospho lipid mass in most tissue and cell types(19, 20). Although a majority of mammalian cells contain a small amount of BMP, its concentration in late endosomes (LE) is elevated to near 15% of the tota l lipid content ( 21-23 ) and up to 70% of the total lipid content of internal membrane domains within the LE( 23). Endosomes are intercellular organelles that act as a staging area fo r sorting endocytosed material either back to the plasma membrane for recycling or to specialized organelles (lysosomes) for degradation(20, 24). Endosomes can be identified either by the presence of internal membranes, also known as multivesicul ar structures, inside the lumen of a limiting (boundary) membrane or by lipid composition( 22, 25 ). Early endosomes have a limiting membrane with a lipid composition very similar to that of the plasma membrane, whereas late 18

PAGE 19

endosomes are characterized by an absence of a significant amount of cholesterol and a relatively high concentration of BMP( 25). In certain situations, multilamellar internal membrane structures (multivesicular bodies or MVBs) are present in late endosomes( 26). A question that naturally arises is: Do elevated BMP levels in th ese cellular structures play a significant role in controlling membrane organization? BMP is negatively charged and has an atypical sn-1-glycerophospho-sn-1-glycerol (sn1:sn1) stereoconfiguration (27) with respect to sn-3-glycer ophosphate structures exhibited by most other glycerophospholipids( 1, 2, 20 ). This unique structure and negative charge is likely to have functional implications beyond its ability to resist degradation by most phospholipases due to their stereospecific recognition of the sn-3 stereoconfiguration(25, 27). Even though the sn1:sn1 and sn3:sn1 ster eoisomers are different, similar thermotropic phase transition temperatures were measured by differential scanning calorimetry (DSC) for 1,3 dimyristoyl sn1:sn1 (40oC) and 1,3 dimyrist oyl sn3:sn1 (42oC) BMP structural isomers( 28 ), justifying the use of the non-natu ral stereoisoform as a first a pproximation for characterization purposes. However, it has been reported that sn-(3-hydroxyl-2-oleoy l)glycerol-1-phospho-sn-1 (3 -hydroxy-2 -oleoyl)glycerol may be the biologically relevant isoform( 23). It has also been shown by gas chromatography and mass spectrometry that the major fatty acid components of BMP, isolated from baby hamster kidney (BHK) cells, are oleic acids (91%)( 23 ). Three structural isoforms are shown in Figur e 1-6; the synthetic (S ,R or 1,3 diacyl sn3, sn1) molecule was used in this work due to its commercial availability and is designated throughout this disserta tion as either BMP18:1, BMP14:0, depending on the fatty acid substituents, or collectively BMP. Because each acyl chain of BMP18:1 contains an unsaturat ed site, this lipid may increase the overall bilayer disorder of model and endosomal membranes due to 19

PAGE 20

intermolecular packing constraints( 29 ). Increased lipid disorder is most likely not the only role BMP plays in modulating endosomal membrane morphology because other negatively charged lipids containing unsaturated chains, such as phos phatidylinositol (PI), woul d most likely have a similar effect on lipid order as BMP and thus affect a specific hydrolysis reaction in a similar manner. This has been shown, but only at concen trations of PI much greater than that found in vivo ( 30 ). Therefore, BMPs unique geometric struct ure must also be an important factor in controlling endosomal morphology a nd molecular trafficking in late endosomal organelles. According to literature, the geom etric shape predicted for BMP is either a cone or inverted cone( 25, 26, 29, 31, 32 ). If the former were true, BM P would most likely exist in an HII aggregate or may induce HII aggregation as the relative BMP concentration is increased in model or biological membranes. However, at pH 7.4 BMP18:1 forms multilamellar vesicles and lacks any three-dimensional structural changes accord ing to fluorescence emission of pyrene labeled lipids incorporated into BMP multilamellar vesicles (MLVs)( 29). Also, studies of sn-(3hydroxyl-2-oleoyl)glycerol-1-phospho-sn-1 -(3 -hydroxy-2 -oleoyl)glycerol induces multilamellar structure formation, as visualized by fluorescence and cryogenic electron microscopy in the presence of a pH gradient in lipid mixtures having a similar composition to that found in late endosomes( 26). Moreover, the small and wide angle X-ray diffraction patterns for both 1,3 dimyristoyl sn3:s n1and 3,3 dimyristoyl sn1:sn1 BMP are consistent with a lamellar structure( 28). There is not any indication that a ny isoform of BMP exis ts in or induces either a HII or micellar structure in the current literature. The electron microscopy and X-ray diffraction data descri bed above provide a static description of the lipid aggregat e macromolecular structure. Th is dissertation reports results obtained from morphological studies of BMP and BMP mixed with model membrane lipids 20

PAGE 21

using magnetic resonance techni ques to investigate the microscopic structure and dynamic structural properties of BMP and BMP mixed with model membrane lipids. Bis(monoacylglycerol)phosphate May Be Important to Ganglioside Catabolism Gangliosides and glycosphingolipids (GSLs) are components of eukaryotic plasma membranes and are involved in passing cellular signals from outside the cell to the cell interior(2, 33 ). The degradation of these particular lipid molecules occurs in acidic cellular compartments of lysosomes, specifically on the surface of intraendosomal and intralysosomal vesicular structures (an example of a GSL in shown in Figure 1-7)( 20, 34 ). A model for endocytosis and GSL digestion, pr oposed by Sandhoff and Kolter, can be seen in Figure 1-8(20, 34). Vesicles containi ng GSLs, which are destined for endosomal and lysosomal compartments, begin as either invaginations or clathrin coated pits formed in the plasma membrane( 35 ). These vesicles fuse with ear ly endosomes where some lipids and proteins are shuttled back to the plasma membrane for recycling ( 36, 37), while others are sorted from the limiting membrane of the early endos ome and incorporated into intralysosomal structures( 20 ). Early endosomes mature into late endos omes that transiently fuse with lysosomes where enzymatic digestion occurs( 20 ). During this maturation process, the luminal pH of the endosome decreases and the protein and lipid co mposition of the intralysosomal structures change ( 20 ) becoming enriched in GSLs and BMP and depleted in cholesterol( 20, 38 ). Glycosphingolipid degradation occurs in th e lysosome as a stepwise cleavage of monosaccharide units from the oligosaccharide head group of the GSLs until the recyclable biomolecule sphingosine is produced( 20, 33, 34). In vivo several accessory molecules are needed to degrade glycoshingolipid s containing head groups of four or fewer sugar molecules: a water-soluble enzyme (the hydrolase), a sphingoli pid activator protein (SAP), and possibly a membrane surface including anionic phospholipids like BMP or PI( 20, 33, 34 ). However, it has 21

PAGE 22

been demonstrated that the enzymatic cleavage reaction does occur in vitro with micellar ganglioside substrate in the absence of a membrane surface( 30 ). SAPs are membrane binding cofactors believed to have variable specifi city for both membrane lipids and enzymes( 20, 34). These cofactors are required for the enzymatic cl eavage and are believed to either extract the ganglioside from the bilayer and present the lipid to the hydrolase for an aqueous reaction or to lift the ganglioside slightly fr om the bilayer surface for hydrol ysis (a membrane associated reaction)( 20, 33, 34 ). Any dysfunction in the SAP, substrate, or hydro lytic enzyme, can lead to a particular form of lysosomal storage disease. There are many diseases associated with storage of lipids in both endosomes and lysosomes. These are classified by the non-degraded lipid or protein accumulated in either the endosome or lysosome( 20, 33, 34). Subclasses or variants of the storage diseases are characterized by the particular molecule in which the defect occurred. For example, TaySachs, or GM2 gangliosidosis, is caused by mutations of the gene encoding for the enzyme hexosaminidase A; GM1 gangliosidosis and Morquio Type B syndrome are caused by a mutation in the enzyme GM1 -galactosidase(20, 33 ). It has already been shown that in the pr esence of BMP or PI in model membranes (POPC/CHOL/GM2/BMP mixed lipid compos ition) the rate of GM2 hydrolysis by hexosaminidase A in the presence of the sphingo lipid activator protein GM2-activator protein (GM2AP) is more than two or ders of magnitude faster than in BMP free liposomes( 30 ). How BMP influences the hydrolysis rate is still unknown, but it is possibl e that BMP modulates surface charge, membrane order, vesicle size/sha pe, or any combination of these properties. The following work presented in this dissertation focuses on model membrane morphological perturbations caused by BMP. It is likely that perturbati ons such as modulating 22

PAGE 23

membrane surface charge, changing membrane pack ing parameters, and pref erential interaction of BMP with GM2 are crucial for the interaction of the catalytic triad ( -hexosaminidase A, GM2, and GM2AP) to function properly. BMP may, under certain conditions, preferentially sequester various lipid substrates into small ve sicular or multilamellar structures required for transportation to the lysosome, which might provide the proper topology needed for efficient enzymatic cleavage. In support of this hypothesi s, there is not any evidence for microdomain (raft) formation when BMP is present in POPC bilayers at either acidic or neutral pH( 29). Hence, it is likely that BMP may induce ve sicle budding events, as opposed to microdomain formation. Biological Membranes Biological membranes are important cellular structures, because they create selective chemical and physical barriers between cells or organelles and their surroundings. Lipid composition and geometrical packing patterns, al ong with incorporation of proteins and other molecules into the lipid matrix, such as chol esterol, allow membranes to exhibit diverse mesophases, surface properties, and permeabilities( 1, 2, 4, 31, 39 ). The aforementioned properties play a role in the ability of biologi cal membranes to participate in small molecule, protein and lipid trafficking, membrane fusion, and to act as platforms for catalysis( 1, 2, 40-42). Biological membranes also adop t a variety of shapes, for example red blood cell membranes resemble small biconcave discs, lysosoma l membranes resemble hollow spheres, the endoplasmic reticulum and the golgi apparatus membranes have very convoluted surfaces. Differences in size and shape of biological memb ranes are likely to be re lated to function, and it is known that lipid interactions with proteins and other molecule s are often needed to modulate membrane shape(1, 2). 23

PAGE 24

Dissertation Outline The work presented in this dissertation fo cuses on the effect of BMP on the morphology of model phosphoglycerol membranes. Magnetic re sonance techniques were used to measure phosphate head group orientation with respect to the bilayer normal and the average angular excursion from the bilayer normal of either a nitroxi de spin probe or a deuterated acyl chain as a function of temperature and/or BMP c oncentration. By using both SSNMR (2H) and EPR, we were able to monitor lipid acyl chain dynamics (order/disorder) on two different time scales: slow motional fluctuations (~105 Hz) and rapid motional fluctuations (~108 Hz), respectively. 31P SSNMR measurements of chemi cal shift anisotropy can be corre lated to head group order, mesophase, and changes in phase transition temperature for several lipid aggregates. After analysis of data collected with each of the previously mentioned techniques, we were able to confirm the lamellar aggregation of pur e BMP, and characterize some of the effects on head group and acyl chain order caused by incor poration of BMP into model membranes. General background information about lipids, lip id self-assembly, membrane characterization, and specific information about BMP is found in Chapter 1. Chapter 2 reports the materials and methods utilized throughout and experimental para meters needed to perform our investigations. An overview of basic magnetic resonance app lications and expected spectroscopic signals obtained from membrane lipid dispersions are disc ussed in Chapter 3. The focus of Chapter 4 is the solubilization of model (PC), pure BMP and PC/BMP mixed membranes. Chapters 5 and 6 report results on characterizati on of overall perturbations of the acyl chain and head group regions of model membranes caused by BMP. In the field of membrane biophysics, it is st rongly desired to unders tand how proteins and lipid molecules alter the phys ical shapes of membrane structures within cells( 38). A detailed understanding of how BMP modulates bilayer physical properties, at the molecular level, will 24

PAGE 25

directly impact other research in the Fanucci group that is focusi ng on characterizing the membrane binding interactions of GM2AP with POPC and POPC:GM2 containing vesicles. Results from this work may also begin to expl ain why BMP alters the enzymatic rates of GM2 hydrolysis when it is incorporated into lipid ve sicles. On a broader scale, the relatively high levels of BMP within the late endosome have been shown using fluorescence microscopy to lead to vesicle budding and multilamellar structure formation( 26). Storage Lipids (Neutral) Triacylglycerols Glycerphospholipids GlycerolFatty Acid Fatty Acid Fatty Acid Fatty AcidGlycerolFatty Acid PO4 Alcohol Choline Fatty Acid PO4Sphingosine Sphingolipids Glycolipids Phospholipids Sphingolipids Galactolipids SO4Membrane Lipids (Polar) GlycerolFatty Acid Fatty Acid Saccharide Fatty AcidSphingosineSaccharide Storage Lipids (Neutral) Triacylglycerols Glycerphospholipids GlycerolFatty Acid Fatty Acid Fatty Acid GlycerolFatty Acid Fatty Acid Fatty Acid Fatty AcidGlycerolFatty Acid PO4 Fatty AcidGlycerolFatty Acid PO4 Alcohol Choline Fatty Acid PO4Choline Fatty Acid PO4Sphingosine Sphingolipids Glycolipids Phospholipids Sphingolipids Galactolipids SO4Membrane Lipids (Polar) GlycerolFatty Acid Fatty Acid Saccharide GlycerolFatty Acid Fatty Acid Saccharide Fatty AcidSphingosineSaccharide Figure 1-1. Lipid classificati on. Adapted from Lehninger, Pr inciples of Biochemistry, 4th edition( 3). 25

PAGE 26

Figure 1-2. A) Anatomy of 1,2-dipalmitoyl-s n-glycero-3-phosphocho line (DPPC) and B) Derythro-sphingosine-1-phosphate. 26

PAGE 27

A) B) C) D) A) B) C) D) A) B) C) D) A) B) C) D) Figure 1-3. Cross-sectional re presentations of lipid polymorphic structures in aqueous environments: A) lamellar gel (L lamellar liquid crystalline (L); C) hexagonal (HII); D) hexagonal (HI). Adapted from Gruner et al. Ann. Rev. Biophys. Chem. 1985( 4). 27

PAGE 28

Geometric Shape Micellar Sodium dodecylsulfate (SDS) Hexagonal (HII) PhosphatidicAcid (pH < 3) Lamellar 1,2-Dipalmitoyl-snGlycero-3Phosphocholine (DPPC) Associated Mesophase Lipid Geometric Shape Micellar Sodium dodecylsulfate (SDS) Hexagonal (HII) PhosphatidicAcid (pH < 3) Lamellar 1,2-Dipalmitoyl-snGlycero-3Phosphocholine (DPPC) Associated Mesophase Lipid Figure 1-4. Geometric shape approximations and lipid aggregates. Adapted from Gruner et al., Ann. Rev. Biophys. Chem. 1985( 4). 28

PAGE 29

P OO O N+ O H O O N O O CH3 CH3 O O P OO O N+ O H D D D D D D D D D D D D D D D D D D D D D D D D D D D O D D D D O D D D D D D D D D D D D D D D D D D D D D D D D D D D O D D D D O A) B)P OO O N+ O H O O N O O CH3 CH3 O O P OO O N+ O H D D D D D D D D D D D D D D D D D D D D D D D D D D D O D D D D O D D D D D D D D D D D D D D D D D D D D D D D D D D D O D D D D O A) B) Figure 1-5. Examples of labeled lipids: A) 1-Palmitoyl-2-Stearoyl -(5-DOXYL)-sn-Glycero-3Phosphocholine; B) 1,2-Dipalmitoyl -D62-sn-Glycero-3-Phosphocholine. 29

PAGE 30

Figure 1-6. BMP structural is omers: A) (S,R isomer) sn-(3-oleoyl-2-hydroxy)-glycerol-1phospho-sn-3'-(1'-oleoyl-2'-hydroxy)-glycero l (ammonium salt), (1,3 dioleoyl sn3:sn1or BMP18:1); B) (S,S isomer) sn-(3-oleoyl-2-hydroxy)-glycerol-1-phosphosn-3'-(1'-oleoyl-2'-hydroxy)-glycerol (ammoni um salt) (3,3 dioleo yl sn1:sn1); C) (R,R isomer)sn-(3-oleoy l-2-hydroxy)-glycerol-1-phospho-sn-3'-(1'-oleoyl-2'hydroxy)-glycerol (ammonium salt) (1,1 dioleoyl sn3:sn3). Figure 1-7. GM1 lipid: GalBeta1-3GalNAcBeta 1-4(NeuAcAlpha2-3)GalBeta1-4GlcBeta1-1'Cer (GM1 ganglioside). 30

PAGE 31

Plasma Membrane GSL Coated Pit Sorting Early Endosome Late Endosome Sorting Sorting Caveola Cavesome ER Golgi Lysosome Lysosome Vescicular Transport? Temporal fusion and discharge Glycocalix Plasma Membrane GSL Coated Pit Sorting Early Endosome Late Endosome Sorting Sorting Caveola Cavesome ER Golgi Lysosome Lysosome Vescicular Transport? Temporal fusion and discharge Glycocalix Figure1-8. Sandhoff-Kolter model for lysosomal membrane digestion and endocytosis; Adapted from Annu. Rev. Cell Dev. Biol., 2005 ( 20). 31

PAGE 32

CHAPTER 2 MATERIALS AND METHODS Materials The following lipids and lipid derivatives, dissolved in chloroform, were purchased from Avanti Polar Lipids (Alabaster AL, USA), stored at -20oC and used without further purification. DPPC (1,2-Dipalmitoyl-sn-Glycero-3-Phosphoc holine), d62-DPPC (1,2-Dipalmitoyl-D62-snGlycero-3-Phosphocholine), POPC (1-Palmit oyl-2-Oleoyl-sn-Glycero-3-Phosphocholine), d31POPC (1-Palmitoyl(D31)-2-Oleoyl-sn-Glycero3-Phosphocholine), BMP18:1 ((S,R Isomer)sn(3-Oleoyl-2-Hydroxy)-Glycerol-1-Phospho-sn-3'(1'-Oleoyl-2'-Hydroxy)-Glycerol,Ammonium Salt)), 5-DOXYL (1-Palmitoyl-2-Stearoyl -(5-DOXYL)-sn-Glycero-3-Phosphocholine), 10DOXYL (1-Palmitoyl-2-Stearoyl-(5-DOXYL)-sn -Glycero-3-Phosphocholine), 16-DOXYL (1Palmitoyl-2-Stearoyl-(5-DOXYL )-sn-Glycero-3-Phosphocholine). HEPES (4-(2-hydroxyethyl,)1-piperazineethanesulfonic acid, C8H18N2O4S), EDTA (ethylenediamine tetraacetic acid, C10H16N2O8), sodium citrate (Na3C6H5O7), and NaCl (sodium chloride) were purchased from Fisher Biotech (Pittsburgh, PA). CHCl3 (chloroform), C6H12 (clyclohexane), CH3OH (methanol), I2 (iodine), PrCl3 (praseodymium chloride), HCl (hydrochloric acid), NaOH (sodium hydroxide), H2SO4 (sulfuric acid), and NH4OH (ammonium hydroxide) were obtained from Fisher Scientific (Pittsburgh, PA). Reagents for the phosphate assay were purchased as a kit (QuantiChromTM Phosphate Assay Kit (DIPI-500) from BioAssay Systems (Hayward, CA). Multilamellar Vesicle Preparation Multilamellar vesicles (MLVs) for phosphorus (31P) and deuterium (2H ) experiments were prepared from mixtures of lipid s dissolved in chloroform. St ock bottles containing lipids in chloroform were allowed to reach room temp erature before opening. Prior to vesicle 32

PAGE 33

preparation, glassware was cleaned with metha nol and chloroform; flasks and syringes were rinsed at least once with HPLC grade metha nol and washed three times with chloroform. The desired amount of lipid stock solution was drawn into a clean, gastight, syringe (Hamilton) and transferred to an appropriate contai ner. If a mixture of two or more lipids was prepared, the previous procedure was repeated fo r each component of the lipid mixture. After use, the headspace of all stock bottles were covered with argon or nitrogen gas, the cap was wrapped in Parafilm or Teflon tape and returned to the freezer. Mixtures were prepared in either round bottom or pear-shaped flasks (the tapered bottom of the pear -shaped flask is helpful during the hydration of the lipid film ), and solvent was removed by rotary evaporation at room temperature using a water-cooled solvent trap. Dry lipid films (both single component and mixtures) were re-dissolved in 4:1 (v:v) co-sol vent system of cyclohexane and methanol (HPLC grade) to ensure complete mixing of the com ponent lipids prior to lyophilization. This dispersion was gently mixed and flash frozen by placing the flask into a container with liquid nitrogen (lN2). Frozen lipids were ly ophilized, typically overnight and subsequently hydrated with an appropriate buffer (5:1 buffer:total lipid, mass:mass) in a 55oC oven for at least one hour, followed by gentle vortex mixing of the hydrat ed dispersion; however samples containing mostly POPC were hydrated at room temperat ure (POPC has a main phase transition near 0oC)( 43). General lipid handling and sample preparation procedur es can be found at www.avantilipids.com. For nuclear magnetic resonance (NMR ) measurements, approximately 200 L or enough of the hydrated MLV dispersion to fill the tube was used for 31P experiments. For 2H measurements ~ 750 L was loaded into a custom cut (~ 38 mm long for 2H), 5mm outer diameter NMR tube. These tubes were either special ordered (Wilmad) or conventional tubes 33

PAGE 34

were cut in our laboratory to fit inside the RF coil of the solid state pr obe. Prior to sample loading the NMR tubes were washed with sample buffer. Buffers were first prepared using nano-pure deionized water, and brought to the appropriate pH with ei ther HCl or NaOH. Aliquots of each buffer were then flash frozen with N2 ( l ), lyophilized, and re -hydrated with an equivalent volume of deuterium depleted water (Cambrige Isotope Laboratories)( 44). Total lipid concentration for NMR experiments was near 275 mM based on theoretical values. Either two or four mol% of d62-DPPC was used in lipid samples prepared for 2H measurements. MLVs and large unilamellar vesicles (LUVs) for electron paramagnetic resonance (EPR) experiments were also prepared from mixtures of lipids dissolved in chloroform. However, due to small sample volumes used for EPR measuremen ts, lipid mixtures were prepared in a glass culture tubes, and blown dry with a stream of nitrogen gas. Dry lipid films were placed under vacuum overnight then hydrated wi th an appropriate buffer in a 55oC oven for at least one hour. Mixtures containing mostly POPC were hydrated at room temperature. For hydrated MLVs, total lipid concentration was near 100 mM, (4-5 L) were loaded into a 1.8 mm inner diameter (id.) borosilicate capillary tube purchased from Vitrocom. Typically, sa mples contained 1 mol% of the DOXYL spin label for EPR measurements. LUVs were prepared by extrusion through polycarbonate membranes. Gas tight syringes were cleaned with chloroform, rinsed three times with methanol, and finally thoroughly rinsed with sample buffer. The mechanical extrusion device was assembled according to the manufactures instructions ( www.avantilipids.com). A volume of sample buffer was first injected into the extrusion device to fill the void volume in the extrusion chamber. An appropriate amount of lipid soluti on was loaded into one syringe a nd inserted into the extrusion device and the solution was passed through a memb rane (typically 100 nm pore size) a minimum 34

PAGE 35

of 55 times, until the lipid suspension became transparent. One pass is defined as solution moving from the starting syringe across the polycarbonate membra ne and back to the starting syringe. Phosphate Assay Both MLV and LUV suspensions were dilute d to obtain a concentr ation near 30 ng/100 L. Three 100 L aliquots of each diluted sample were placed in separate labeled glass culture tubes. Specific quantities of a phosphorus stan dard purchased from BioAssay Systems, ranging between 0 and 100 nmol, were placed in separate tubes; each standard quantity was made in triplicate in order to estimate th e error of the standards. A volume (0.45 mL) of 8.9 N sulfuric acid was added to each tube and the t ubes were placed in an oven at 220oC for 25 minutes to decompose the lipid and liberate the phosphate. Sample tubes were cooled to room temperature, and 3.9 mL of 1 M NaOH was added to each tube (this reaction is exothermic). Tubes were again cooled to room temperatur e and vortex mixed. A volume of 800 L was transferred from each tube to a standard, dis posable, UV-vis cuvette and 200 L of malachite green phosphate analysis solution (BioAssay Systems) was added to each cuvette. The contents of each cuvette was mixed by inversion, and allowe d to equilibrate for 10 minutes. Cuvettes were placed in a UV-vis spectrometer and the absorbance values at 650 nm were recorded. Unknown sample absorbance values were compared to the calibra tion curve prepared from the absorbance values of the standard solutions. This procedure fo r determination of phosphorus content has been modified from the general proc edure of Warner et. al., 1956( 45 ). Thin Layer Chromatography Lipid integrity was verified by thin layer chromatography (TLC ) before and after exposure to magnetic resonance radio frequency pulsing as the ester bonds in the lipids are somewhat 35

PAGE 36

labile. A very small amount (~ 2 L) of lipid dissolved in ch loroform or hydrated lipid suspension was added to seven drops of chloroform and spotted on silica coated aluminum plates (Whatman) purchased from Fisher Scientific. TL C plates were placed in a chamber containing a mobile phase of 65:25:4 (by volume) chloroform :methanol:ammonium hydroxide (14 N) and the solvent front was allowed to migr ate to approximately 75 % of the plate height. For quantitative experiments, TLC plates were washed in the mob ile phase, and air dried pr ior to spotting of the sample. The stationary phase was allowed to dry and the slides were exposed to I2 vapor for visualization of the lipid fractions. Magnetic Resonance Wide line 31P NMR experiments were performed using a Tecmag spectrometer operating at a resonance frequency of 145.2 MHz. Spectra were acquired with a CP/MAS probe purchased from Doty Scientific, Inc., with variable te mperature capability. A two pulse Hahn echo pulse sequence with CYCLOPS phase cycling was used with 1H decoupling. Details of the pulse sequence include an echo pulse spacing of 40 s, a 4 s 31P pulse ( x), a 5 s recycle delay, 5 s dwell time, 1024 8192 time domain data points. A minimum of 2048 transients were averaged for each experiment. 2H experiments were performed on a Bruker Av ance spectrometer operating at a resonance frequency of 61.4 MHz. Spectra were acquire d with a high power, broad band, and high temperature probe manufactured by Bruker. A standard quadrupole echo pulse sequence was used with 3. s excitation pulses, 40 s pulse spacing, a 500 s recycle delay, and a 4 s dwell time. Typically 8192 time domain data points we re collected and a minimum of 14400 transients were averaged per spectrum. 36

PAGE 37

Paramagnetic resonance experiments were completed using a modified ER200 (Bruker) with an ER023M signal channel, an E032 field c ontroller, SPEX data acquisition software and a loop gap resonator (Medical Adva nces, Milwaukee, WI). Typical EPR experimental parameters are a 100 G spectral width, 20 mW of average mi crowave power, and a 0.16 s time constant. The sample temperature was controlled by flowi ng either compressed air or nitrogen through a copper coil immersed in a circulating water ba th and passing the gas over the sample tube. Data Processing Raw 31P free induction decays were base line corr ected, zero filled to twice the number of data points collected in the time domain, left shifted (if required ), Fourier transformed, apodized by exponential multiplication (100 Hz), and pha se corrected using the NTNMR software provided by Tecmag. Raw 2H free induction decays were base line corr ected, zero filled to twice the number of data points collected in the tim e domain, left shifted (if required), apodized by exponential multiplication (100 Hz), Fourier transformed a nd phase corrected using a Matlab routine provided by the Long research group. EPR line shapes were baseline and phase corrected and area normalized using Labview software written by Dr. Christia n Altenbach (UCLA, laboratory of Dr. Wayne Hubbell). Second moments and peak-to-peak widths were also ca lculated by Labview software written by Dr. Altenbach. 37

PAGE 38

38 CHAPTER 3 MAGNETIC RESONANCE APPLICATIONS IN MEMBRANE BIOPHYSICS Hydrated Lipid Motions and Order Lamellar biological and model membrane system s are often described as two-dimensional fluids where lipids, proteins, and other molecules are allowed to diffuse laterally within the boundaries of a two-dimensional matrix( 1, 2). However, this dynami c description should also include many other motional degrees of freedom for individual lipi d molecules including individual bond vibrations a nd rotations, trans/gauche isomerization of the acyl chains, molecular rotation about and wobbl e of the symmetric, molecular long axis perpendicular to the bilayer normal, and lipid excha nge or flip-flop between the two monolayers of the bilayer structure. The time scales of these motions span more than fifteen orders of magnitude, from bond vibrations, on the order of femtoseconds to flip-flop motions on the order of seconds to minutes( 1, 3). Figure 3-1 depicts a crosssection of a bilayer plane a nd illustrates three selected motions and their respective reorientation times. Lipid molecules in Figure 3-1 are represented by circles (polar head gr oups) and squiggle lines (apolar acyl ch ains). Figure 3-1 A illustrates the molecular rotation and wobble motion of a lip id molecule in the top monolayer of the membrane, B illustrates the molecular exchange that occurs during lateral diffusion, and C illustrates the flip-flop exchange of lipid molecules between the two monolayers. Spectroscopic methods such as NMR and EP R can provide information regarding the aggregation state and the orient ation of single bonds, or molecu les within a membrane with respect to a reference coordina te system, usually the long axis of the lipid molecule, or the bilayer normal in lamellar systems( 4-7 ). Average orientations are determined by mapping the experimental observables such as the principal components of the chemical shift, the quadrupole coupling, and the hyperfine tensor values onto th e molecular frame. Specific details regarding

PAGE 39

these tensors and the spectroscopic signals obtain ed for each method are discussed later in this monograph. Experimental signals acquired by these tech niques are related to the time averaged reorientation of the previously mentioned tens or values and are usua lly converted to order parameters (Si) used to describe average tensor orie ntations with respect to a reference coordinate system in the static limit. In gene ral, order parameters for unoriented systems in a uniaxial magnetic field are expressed as sec ond order Legendre polynomials: (S = (3 cos2( ) 1)), and are functions of the average angle ( ) between the molecular frame and the reference frame (the magnetic field). Unfo rtunately, the experimental obser vables for membranes that are not mechanically oriented with respect to the ma gnetic field provide a total order parameter that is the product of two order parameters, one describing the orientation of the molecule or bond with respect to the bilayer normal and the other describing the orientati on of the bilayer normal with respect to the magnetic field. Figure 3-2 sh ows the order parameter of a specific molecular site (i) as a function of the angular orientation of that site with respect to the bilayer normal. According to the mathematical expression for order parameters, the maximum occurs at an angle of 0o and is assigned a value of +1, while the minimum occurs at an angle of 90o and is assigned a value of Two important pieces of information must be kept in mind when evaluating or reporting order parameters: 1) A small order parameter value is not necessarily low because of rapid, random motional averaging; for example, it may be low because the tensor is oriented at the magic angle (54.70), which leads to an intrinsic or der value of zero, and 2) Different spectroscopic techniques are sensitive to different time scales of tensor reorientations. For example, NMR reports motions reorienting faster than 10 s (such as trans/gauche 39

PAGE 40

40 isomerization) as an averaged resonance weighted by the number of individual conformations represented. Techniques such as infrared (IR) spectroscopy, however will report the same motions as a superposition of individual resonances. Therefore, one should always keep in mind the time bases of the experimental methods when comparing order parameters. Nitroxide Spin-Probes Stable free radical spin probes have been us ed for many years to investigate the dynamics of model and biological membranes assemblies( 4, 8-15). A variety of hydrophobic and amphiphilic organic spin-probes have been em ployed such as, 2,2,6,6,-tetramethylpiperidine-1oxyl (TEMPO), paramagnetic oxazo lidine ring labeled fatty acid s, and paramagnetic oxazolidine ring labeled lipids( 13-15 ). The structures of selected sp in-labels are shown in Figure 3-3; A depicts a hydrophobic nitroxide spin -label (TEMPO), B) depicts th e ring structure incorporated into the acyl chain regions of the fatty acid and the labeled lipid shown in C and D, while C and D are amphiphilic nitroxide spin-labels. Description of Nitroxide Spi n-Label Order Parameter in Hydrated Lipid Bilayer Assemblies Obtained by Elect ron Paramagnetic Resonance Electron paramagnetic resonance (EPR) spectra of spin = molecules, such as nitroxide radicals, trapped as impurities in crystals with fixed orientation can be described by the basic spin Hamiltonian of McConnell and McFarland (Eqn. 3-1), neglecting proton hyperfine interactions and including only the g (electron screening tensor) and T (hyperfine coupling tensor).( 13) The hyperfine tensor is a result of coupling between th e electron angular momentum operator ( S ) and the nuclear angula r momentum operator ( I ). For a nitroxide radical (S = 1/2 and I = 1), and the hyperf ine interaction will give rise to three absorption lines, thus the EPR derivative signal has three peaks. The exact details of the EPR tran sitions and line shapes are discussed later. and are the Bohr magnetons for the electron and nucleus respectively, g

PAGE 41

is the electron screening tensor (a nalogous to the NMR chemical shift), T is the hyperfine coupling tensor, S and I are spin angular momentum operators, and h is Plancks constant. -TeN NHhg SN g HSTIIH (3.1) An energy level diagram illustrating electroni c Zeeman and hyperfine interactions of a representative spin-probe in an external field along with the resulting derivative EPR spectrum is illustrated in Figure 3-4. When an unpaired el ectron is placed in an external field (Bo) the degeneracy of electron spin angul ar momentum is lifted, which is referred to as the electronic Zeeman interaction. For a nitroxide radical, the electron spin angular momentum and nuclear spin angular momentum ar e coupled through a hyperfin e interaction tensor ( T ), due to the interaction of the electron angular momentum and the 14N nucleus with a magnetic moment, splitting each electronic spin state into (2I + 1) states. The hyperfine interaction is not dependent upon field strength, but has an anisotropic spatial dependence that determin es the energy splitting (spacing between peaks in derivative mode) of the three allowed transitions (see Figure 3-5). Selection rules are ( ms = +/1) and ( mI = 0) for a one photon process. The resulting energy transitions can be probed with microwav es oscillating at the correct frequency. However, as discussed earlier in this chapte r, self-assembled, spin-labeled fatty acids and phospholipids in an aqueous environment are neith er solution nor static structures but undergo rapid, anisotropic molecular motions about the molecular long axis. This motion must be described by a time dependent equation( 4, 13). This time dependent Hamiltonian can be separated into two parts: 1) a time -independent, effective Hamiltonian ( ) (Eqn. 3-2) and 2) a Hamiltonian that is a function of time. As long as the time dependent fluctuations are sufficiently averaged; will adequately describe the system(4, 13). TH' TH '''-TeN NHhg SN g HSTIIH (3.2) 41

PAGE 42

Unfortunately, experimental powder line shapes of hydrated lipids only yield information about molecular orientation with respect to the la boratory reference frame and must be related to the molecular axis system using an oriented re ference molecule. The reference molecule used for understanding spin-labeled lipids was spin-labeled 5-cholestan-3-one (Figure 3-5), with g values (gx,gy,gz (2.0089, 2.0058, 2.0021 +/0.001)) and T values (Tx,Ty,Tz (5.8, 5.8, 30.8 +/0.5 G)) calculated by Hubbell(13, 16). However, if g and T tensors in Eqn. 3-2 are appropriate time averages of g and T from Eqn. 3-1, the energies and thus the two effective tensors, can be related to calculated tensor values of the trapped radical throu gh the averaged squared directed cosine projected onto the principal axis system of the trapped nitroxide(4, 13). Equations 3-3 and 3-4 show the energy obtained from the effective Hamiltonian in the laboratory frame and in terms of the basic Hamiltonian, respectively. From Eqn. 3-3 and 3-4,using the relation 2 + 2 + 2 = 1, and since Txx and Tyy are equal for 4,4-dimethyloxazolidine-3-oxyl,(13) the value of Tzz can be derived and is shown in Eqn. 3-5. In Eqn. 3-5, a is the isotropic component of the hyperfine tensor (a = 1/3 Tr( T ) = 1/3 Tr(T )) and S is the order parameter. It can be shown that S can be related to calculated values of Txx and Tzz and experimental values of and 'T 'T as seen in Eqn. 3-6. Because the hyperfine interaction is dependent on solvent polarity the order parameter requires a slight correction which can be seen in equation 3-7(9, 13). ,,,,,,,, ,,NN zzzzzzzz zz E gSHhTSIgIH (3.3) ,, ,,, ,(3.4) 222 222()()xxyyzz xxyyzz NN zz zzzzEgggSHhTTTSIgIH ,,'2222 ( 3xxyyzz zzxx zzTTTTaTT ) S (3.5) '' 21 (31) 2zzxxTT S TT (3.6) 42

PAGE 43

'' zzxxTT a S TTa (3.7) Orientation of the Nitroxide Spin-Label in Hydrated Lipid Bilayer Assemblies and Expected EPR Lineshapes Referring to Figure 3-3 C and D and recalling th e orientation of individual hydrated lipid molecules in a bilayer aggregate, one can see that the nitrogenoxygen bond is oriented parallel to the bilayer surface, making the orbital designated pz perpendicular to the bilayer surface and parallel to the acyl chain long axis. The electron spin angular momentum is most strongly coupled to the 14N nuclear angular momentum in the direction of pz and the hyperfine coupling is minimal in the plane perpendicular to this orbital; therefore, the la rgest component of the hyperfine coupling tensor is Tzz and Tzz > Txx = Tyy( 4, 17). This is again illustrated in Figure 3-6 where the z-axis represents the nitrogen pz orbital and the x-axis represents the N-O bond vector( 4, 17 ). Even though X-band (9 10 GHz) EPR is dominated by the hyperfine interaction, it is important to note that the g tensor is also anisotropic with the largest value in the direction of the N-O bond (because g is a first order function of fiel d it becomes significant in high field EPR)( 17). Figure 3-7 illustrates two representative line shapes: spectrum A represents an axially symmetric, nitroxide label with re stricted motion (broad spectral lines) and visible splitting of the low and high field lines, while spectrum B) represents an axially symmetric, nitroxide label without restricted motion (n arrow spectral lines). The central resonance line is invariant to changes in the hyperfine interaction( 17); therefore, the central peak-to-peak line width ( Hpp Figure 3-7 B)) is also a good approximation of relative mobility of the spin-label. There are two other methods commonly used to determine mobility of nitroxide spin labels: normalized fractional intensity (fI), and second moment (M2) analysis( 4, 13, 18 ). Normalized fractional intensity, as a function of a dependent variable (e.g. 43

PAGE 44

temperature), tracks the change in intensity (I) of a specific spectral position with respect to an intensity designated as final (If) divided by the difference in an intensity designated initial (Ii) and the final intensity (If) (Eqn. 3-8)( 19). This is an adequate method for comparing mobility within a set of similar experiments, but for small changes in spectral intensity the level of noise must be near zero. f I if I I f I I (3.8) The coupling strength or spitt ing of the z component of T (Tzz) measured for a specific nitroxide labeled site yields information about the relative order of an aggr egated lipid system in the vicinity of that site. This coupling strength is directly related to angular deviations ( ) of the spin-probe (pz orbital) from a plane perpendicular to the bilayer normal, and indirectly related to local environmental order. These angular excurs ions for the bilayer normal are mainly caused by trans\gauche isomerizations in the local vicinity of the probe( 5, 13, 20). The probabilities of these isomerizations for free polymer chains and li pids in an assembly are a function of the total internal energy of the hydrocarbon chain in a specific conformation( 5, 20). However, due to interaction with other molecules the total inte rnal energy for assemblies must include both intramolecular and intermolecular energies( 5, 20 ). Examples of interaction energies that can be included in the intermolecular group include van de r Waals interactions, electrostatic repulsions and thermodynamic energies from hydrophobic forces. Considerations must also be made for restricted motion near the head groups (conical boundary condition)(5, 20 ). Moreover, in assembled aggregates the isomerization probability (chain order) is a function of depth (or location) in the hydrophobic domain of the aggreg ate establishing an aver age, site-dependent order(13, 20 ). Although the EPR active spin-labeled lipids may perturb lipid aggregate 44

PAGE 45

structures they still report at least qualitative information concerning the order/disorder of the hydrated lipid aggregates as they would exist without the spin-label. Solid State 31P and 2H NMR of Hydrated Lipid Aggregates The following general discussion of NMR will focus on two specific nuclei: 2H and 31P. 2H is an I = 1 nucleus with a dominan t quadrupolar coupling interaction, and 31P is an I = nucleus with a dominant anisotropic chemical shif t (CSA) interaction, under the condition of full proton decoupling. There are two allowed transiti ons in the radio freque ncy energy range for I = 1 (2H) nuclei and a single allowed transition for I = (31P) nuclei; both transitions obey the selection rule mI = An energy diagram illustrati ng the quadrupolar splitting and allowed transitions for an I = 1 nucleus can be seen in Figure 3-8. Applicati on of a strong external magnetic field lifts the degeneracy of the nuclear spin angular momentum of an I =1 nucleus; therefore, three (2I + 1) nuclear Zeeman states, designated +1, 0, and -1 are present. Further perturbations of these nuclear Zeeman states ar e caused by an interactio n of the nuclear spin angular momentum with the electri c field gradient (EFG) at the quadrupolar nucleus. Energies for these transitions are given in Figure 3-8, where eq is the electric field gradient, Q is the quadrupole moment, e is the elementary charge, is the angle of the principle axis of the efg with respect to the applied field, is the angle in the x,y plane, and is the asymmetry parameter( 5, 21, 22 ). The value of is defined as (| XX| | YY|)/| ZZ|( 5, 21 ), where ii are diagonal elements of the electric field tensor V. Lipid MLVs in an aqueous environment have a ll possible spatial orient ations with respect to the applied field, and yield the classic Pake powder spectrum (Figure 3-9), which is narrowed and axially symmetric ( = 0) because of rapid rotational motion around the long axis of the lipid molecule(5, 7, 21, 23 ). There are three distin ct features of the Pake -doublet when all possible 45

PAGE 46

values of and are considered: steps (ZZ), shoulders ( YY), and singularities ( XX). Figure 3-10 shows the average angular orientati on of the elements of both the electric field tensor (V) and chemical shift tensor with respect to the applied field (Bo) (5 ). Also note that the quadrupolar splitting is invariant with resp ect to the applied filed. The difference in frequency units between each location ( 223 (3cos1(1cos)(cos2)) 4QQ ) of the two maxima of a specific Pake doublet defines the averaged quadrupolar coupling constant and the orientation (order parameter SCD) of a specific deuteron with respect to the applied field(5, 21). The value of the coupling constant (5, 21) can be compared to the static limit and related to angular deviations and local or der as discussed in the prev ious section for EPR probes. Magnetic Resonance Line Shapes a nd Order in Hydrated Lipids Spectroscopic information such as line sh ape and residual coupli ng strengths obtained from the natural 31P and virtually non-perturbing 2H NMR experiments allow for indirect aggregated structural assignments and determinati on of relative order/disorder in the head group and acyl chain regions of the aggregates. 2H and 31P spectra of lipid aggregates have characteristic line shapes, which can be associated with aggregation states and degree of order of hydrated lipids. As discussed pr eviously, order/disorder is de fined as the degree of motional averaging relative to the static limit. Typical 31P and 2H line shapes and their corresponding a ggregate types for hydrated lipids are shown in Figure 3-11. Each carbon atom within the acyl chain region is deuterated and the resulting 2H spectrum will be a superposition of quadrupole splittings from each site, but for clarity only one site is s hown. Visual inspection of 31P line shapes yields information about aggregation type: lamellar gel phas e spectra have a broad, asymmetric chemical shift anisotropy (CSA) line shape; lamellar liquid crystalline sp ectra have a narrowed axially symmetric CSA 46

PAGE 47

lineshape; very small MLVs, micelles, and cubic phases yield isotropic peaks; as more motional averaging narrows the NMR spectral width(3, 21-23, 25). It is important to note that 31P powder NMR cannot differentiate between aggregates with signals that appear isotropic, because the chemical shift of each of these aggreg ation states is the isotropic value iso, see Figure 3-11. Hexagonal phases have a 31P line shape that is opposite in sign and half the numerical span value of the corresponding lamellar line shape, because the CSA is averaged in a second dimension, with respect to lamellar liquid crystalline spectra.(26) 2H spectra for the lamellar gel phase are broad and unresolved; similar to the span of 31P spectra, the quadrupole splitting of the hexagonal aggregate is also half as broad as the lamellar liquid phase(5). The NMR data collected from hydrated lipid aggregates not only afford aggregate structural assignments but also allow for comparisons of the relative order/disorder among different individual locations, al beit on a different time scale, to those obtained from EPR measurements. Comparisons of relative order will be used to determine the morphological effects of incorporating the ne gatively charged, atypically shaped BMP molecule into model membranes. ~ 100 ns -1 ms ~ 100 ps-10 ns ~ seconds -minutes A) B) C) ~ 100 ns -1 ms ~ 100 ps-10 ns ~ seconds -minutes ~ 100 ns -1 ms ~ 100 ps-10 ns ~ seconds -minutes A) B) C) Figure 3-1. Selected lipid motions and asso ciated correlation times in hydrated lamellar structures: A) rotational diffu sion and long axis wobble; B) lateral diffusion; C) flipflop. 47

PAGE 48

-80-60-40-20020406080 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 Si() (degrees) = 90 BilayerNormal = 90 = 90 BilayerNormal 0 < < 90 BilayerNormal 0 < < 90 0 < < 90 BilayerNormal Figure 3-2. An order parameter as a function of the angular displacement of the plane containing a specific carbon and two deuterium atoms from the bilayer normal. 48

PAGE 49

Figure 3-3. Common orga nic radical spin-labels: A) 2,2,6, 6-tetramethylpiperidine-1-oxyl; B) 4,4-dimethyloxazolidine-3-oxyl; C) (5DOXYL) steric acid ; D) 1-palmitoyl-2stearoyl-(5-DOXYL)-sn-gl ycero-3-phosphocholine. 49

PAGE 50

+1 0 -1 -1 0 +1mIms +1/2 -1/2E Bo E = h mw +1 0 -1 -1 0 +1mIms +1/2 -1/2E Bo +1 0 -1 -1 0 +1mIms +1/2 -1/2 +1 0 -1 -1 0 +1 +1 0 -1 +1 0 -1 -1 0 +1 -1 0 +1mIms +1/2 -1/2E Bo E = h mw Figure 3-4. Energy level diagram illustrating the electronic Zeeman and electron-nuclear hyperfine interactions and the resulting deri vative of EPR transitions for a spin-probe such as, TEMPO in solution. Figure 3-5. 4,4-dimethylox azolidine-3-oxyl labeled 5-cholestane-3-one. 50

PAGE 51

Figure 3-6. Spatial dependence of the coupling strength of the anisotropic hyperfine interaction; where the z-axis represents the nitrogen pz orbital and the xy-axis represents the N-O bond plane. The individual couplings can be obtained by rotating the uniaxial magnetic field so that it is coincident with each axis. 51

PAGE 52

'2 T'2 T H ppA) B) '2 T'2 T H pp '2 T'2 T '2 T'2 T H pp H ppA) B) Figure 3-7. Theoretical nitr oxide label EPR line shapes: A) Representation of partially immobilized environment; B) Representation of an isotropic solution environment. oEhB 22 23(3cos1(1cos)(cos2)) 8oeqQ EhB 22 23(3cos1(1cos)(cos2)) 8oeqQ EhB oEhBNuclear Zeeman +1 -1 0 Quadrupole Splitting (1)oEhB+1/2 -1/2 A) B) oEhB 22 23(3cos1(1cos)(cos2)) 8oeqQ EhB 22 23(3cos1(1cos)(cos2)) 8oeqQ EhB oEhBNuclear Zeeman +1 -1 0 Quadrupole Splitting (1)oEhB+1/2 -1/2 oEhB 22 23(3cos1(1cos)(cos2)) 8oeqQ EhB 22 23(3cos1(1cos)(cos2)) 8oeqQ EhB oEhBNuclear Zeeman +1 -1 0 Quadrupole Splitting oEhB 22 23(3cos1(1cos)(cos2)) 8oeqQ EhB 22 23(3cos1(1cos)(cos2)) 8oeqQ EhB oEhBNuclear Zeeman +1 -1 0 Quadrupole Splitting (1)oEhB+1/2 -1/2 (1)oEhB+1/2 -1/2 A) B) Figure 3-8. A) Energy level diagram illustrati ng the nuclear Zeeman and quadruplar coupling interaction of 2H in an applied field Bo. B) Energy level diagram for the 1H decoupled chemical shift of 31P in an applied field Bo. 52

PAGE 53

(1) (1)1 2 1 2ZZ Q YY Q XX Q -100-50 0 50100 Frequency (kHz)XX ZZ YY (1) (1)1 2 1 2ZZ Q YY Q XX Q -100-50 0 50100 Frequency (kHz)XX ZZ YY -100-50 0 50100 Frequency (kHz)XX ZZ YY Figure 3-9. Theoretical quadr upolar echo powder spectrum of a single deuterium labeled site ( 0). 332211 Bo 31P Bo C D D ZZXXYY22222 11 22 33sincossinsincoszz A) B) 332211 Bo 31P Bo C D D ZZXXYY22222 11 22 33sincossinsincoszz 332211 Bo 31P332211 Bo 31P Bo Bo 31P Bo C D D ZZXXYY22222 11 22 33sincossinsincoszz A) B) Figure 3-10. Graphical representation of A) CSA and B) EFG th e angular orientations with respect to the applied field (Bo). This figure has been adapted from Santos(24). 53

PAGE 54

-60 -40 -20 0 20 ppm 1122 = 33iso Lamellar Hexagonal Micellar, Small Vesicle-100-50 0 50100 Frequency (kHz) XX= YY ZZ 11223311 112233 1133iso=12 + + 33 Span== = 1 2ZZ Q XXYY Q chemical shift (ppm) Frequency (kHz)A) B)-60 -40 -20 0 20 ppm 1122 = 33iso Lamellar Hexagonal Micellar, Small Vesicle-100-50 0 50100 Frequency (kHz) XX= YY ZZ 11223311 112233 1133iso=12 + + 33 Span== = 1 2ZZ Q XXYY Q chemical shift (ppm) Frequency (kHz)-60 -40 -20 0 20 ppm 1122 = 33iso -60 -40 -20 0 20 ppm 1122 = 33iso-60 -40 -20 0 20 ppm -60 -40 -20 0 20 ppm -60 -40 -20 0 20 ppm 1122 = 33iso Lamellar Hexagonal Micellar, Small Vesicle Lamellar Hexagonal Micellar, Small Vesicle-100-50 0 50100 Frequency (kHz) XX= YY ZZ -100-50 0 50100 Frequency (kHz) XX= YY ZZ -100-50 0 50100 Frequency (kHz) -100-50 0 50100 Frequency (kHz) -100-50 0 50100 Frequency (kHz) XX= YY ZZ 11223311 112233 1133iso=12 + + 33 Span== = 1 2ZZ Q XXYY Q chemical shift (ppm) Frequency (kHz)A) B) Figure 3-11. Theoretical powder spectra of va rious lipid aggregates: A) axially symmetric 31P and B) a single axially symmetric 2H labeled site. 54

PAGE 55

CHAPTER 4 MONITORING MODEL BILAYER SOLUBILI ZATION BY DETERGENT MOLECULES USING EPR SPECTROSCOPY Model Membrane Solubilization The underlying question addressed within this dissertation is the i nvestigation of the structure of BMP vesicles a nd the effect this lipid has upon model membrane morphology of POPC/BMP or DPPC/BMP mixed ve sicles. Structure shall be defined as head group and acyl chain packing, as well as the morphology of lipid aggregation st ates. Using the geometrical shape approximations discussed in Chapter 1, BMP is assumed to have a polar region (defined as the phosphate group and both glycerol moeities) that occupies a la rger cross-sectional area than the apolar, acyl chain region (refer to Figure 16 or Figure 4-1 B). According to the simple geometrical packing model (descr ibed in Chapter 1), we can pr edict that hydrated BMP lipids may form micellar or highly curved vesicular t ypes of assemblies. Our first hypothesis, based on simple geometric constraints alone, is that BM P could have detergent-lik e properties similar to sodium dodecyl sulfate (SDS or the DSanion), such as micellar a ggregation and the ability to solubilize lipid bilayer membrane aggregates. The structures of SDS and the S,R isomer of BMP are shown in Figure 4-1. According to literature, there are three stages for pure membrane solubilization by detergent molecules(1, 2): 1) detergent monomer molecule s bind membranes and partition into the membrane structure by inse rting their acyl chai n into the hydrophobic region of the lamellar structure, thus increasing the overall size of the aggregate structure; 2) membranes become saturated with detergent molecules and a two phase equilibrium exists between saturated membranes and detergent/lipid mixed micelles; and 3) bilayer structures become fully solubilized and lipid molecules are incorporated into detergent micelles (termed mixed micelles). 55

PAGE 56

The molecular species and or aggregates present in each stage of the solubilization process are as follows: 1) detergent monomer, and lipid/detergen t membrane aggregate; 2) detergent monomer, saturated lipid/detergent membrane aggregate, an d detergent/lipid mixed micelles; 3) detergent monomer and detergent/lipid mixed micelles. Both micelle formation and membrane solubi lization can be inves tigated using cw-EPR spectroscopy with positional isomers of spin-lab eled lipids (Figure 42). If BMP does form micellar structures, the fast rotational correla tion time of the assembly will motionally average (narrow) the line shapes of the DOXYL spin -probes incorporated into thes micelles (3) when compared to the line shapes obtained from rela tively immobilized labels in bilayer aggregates(4). Typical EPR experimental condit ions utilize as little as 0.5 mol% of a spin labeled fatty acid or spin labeled phospholipid to probe structure and dynamics in model membrane systems(5). Again, motional averaging will narrow the overall EPR line shapes, compared to the static powder pattern. Spectral features such as the peak-to-peak width of the central resonance line (Hpp), the normalized spectral fractional intensity (fI), order parameter (Si), and second moment (M2) can be used to compare the degree of motional averaging and relative order of these spin-probes incorporated into lipid assemblies The values of these parameters can then be compared among those obtained from probes inco rporated into mixed micellar, lipid/SDS aggregates, mixed lipid/SDS bilayer structures, and single component, bila yer mesotructures of model membrane lipids made of POPC and DP PC. Similarly, the effect of BMP upon the organization, morphology and acyl chain dynamics of POPC and DPPC bilayers can be investigated using the same stra tegy, by precisely incorporating 1 mol% of the label into the assemblies and varying the position of the DOX YL spin-label along the acyl chain of the phospholipid. 56

PAGE 57

The solubilization of POPC by SDS has been studied in detail by light scattering, NMR and isothermal titration calorimetry (ITC)(2, 6, 7). Critical values used to map phase boundaries are obtained by calculating the points of inflectio n from either right angle light scattering intensity or normalized heats of reaction (ITC ) as a function of de tergent concentration(2). These same critical values and phase boundaries can also be determined by monitoring the isotropic 31P NMR signal intensity from lipids in mixed micellar aggregates(6). From these types of data, Seelig and coworkers construc ted a phase diagram with SDS concentration (o D C) as the dependent variable plotted as a function of total lipid concentration ( ). The linear phase boundaries described by Eqns. 4-1 and 4-2 as determined by Seelig and coworkers, have been reproduced in Figure 4-3(6). The approximate boundaries between the saturated bilayer and the mixed micelle/bilayer coexistence (Eqn. 4-1) and the mixed mice lle/bilayer coexistence and the mixed micellar regions (Eqn. 4-2) are in dicated by solid black lines. The intercept of each line corresponds to the minimum concentration of detergent needed to either saturate or completely solubilize the POPC vesicles at 56oC(6). According to this phase diagram for SDS partitioning and micellization of POPC membranes, we sampled the bilayer, micelle/bilayer coexistence, and micellar regions in our EPR investigation. In Figure 4-3 open circles represen various POPC vesicle samples prepared with the 10-DOXYL labeled lipid in the presence of SDS, and closed squares represent various SD S POPC vesicle samples prepared with the 5DOXYL labeled lipid in the presen ce of SDS. It should be note d that our EPR experiments wer carried out at room temperatur e, whereas the Seelig investig ations were performed at 56oC. However, given that POPC is in the L phase above 0oC, we anticipate that the experimentall determined phase boundaries would show little temperao LC t e y ture dependence (4.1) 0.2832.2oo DLCC 57

PAGE 58

(4.2) 2.21.69oo DLCCCharacterization of the EPR Line Shapes of Spin-probes Located in Bilayer Aggregates in the Presence of an Anionic Detergent Control experiments were performed that ch aracterized the EPR line shapes of DOXYL labeled lipids during partitioning of the anioni c detergents SDS into lamellar POPC model membranes and consequent micellization of these membranes. EPR spectra were collected for DOXYL lipids (1 mol%) incorporated into POPC large unilamaller vesicles (LUVs) that were made by extrusion through polycarbonate memb ranes with a 100 nm pore size. POPC and POPC/BMP LUVs were prepared according to the procedure in Chapter 2 and hydrated in a buffer containing 20 mM HEPES, 100 mM NaCl and 0.02% NaN3, at pH 7.4 or 5 mM HEPES, 100 mM NaCl, 0.1mM EDTA, at pH 7.4. The c oncentration of the DOXYL labeled lipid was 1 mol% of the total lipid fraction for all samples unless otherwise indicated. Samples containing SDS were allowed to equilibrate for at least 30 minutes prior to measurements. EPR measurements were collected using a 100 G spect ral width with 20 mW of microwave power and a 0.16 s time constant, at room temperature unless otherwise indicated. In order to characterize the effects that de tergent partioning and subsequent micellization has upon the EPR line shape of DOXYL labels in model POPC membranes, SDS was titrated into POPC vesicles containing the DOXYL spin-pr obe. Two sites within the bilayer structures were examined, carbon positions 5 and 10 on the sn2, steric acid chain of the labeled lipid (Figure 4-2 A and B respectivel y). Changing the position of the nitroxide probe provides information about acyl chain order both near the lipid head group and well within the hydrophobic region of the molecular assemblies. As detergent is incorporated into a membrane-like structure, the bilayer packing is disrupted, thus changing the inte ractions between lipid molecules and consequently the micro58

PAGE 59

environment of the spin-probe. The order paramete r of each labeled lipid is expected to decrease as the system moves through each stage of the solubilization process (1 > 2 > 3)(8, 9). This change in order should be apparent in EPR lineshapes as the residual anisotropic, hyperfine coupling strength is affected by local micro-environment(4, 5, 10-14). Both 5 and 10-DOXYL labeled lipids report obser vable changes, such as, a smaller order parameter and a smaller fractional intensity as th e concentration of SDS is increased, which is consistent with SDS partitioning into the bilayer and subsequent micellization. These changes in acyl chain order can be characterized by any of four numerical values: the second moment, the peak-to-peak central line width, the normalized fractional intensity or the order parameter values(4, 11, 15). These values have been defined previ ously in Chapter 3. Values of the second moment, order parameters (Si), normalized fractional intensity (fI), and Hpp obtained from EPR line shapes for the lipid solubilization experiments are listed in Tables 41 and 4-2. Figures 4-4 and 4-5 display EPR line shapes as a function of detergent concentration for nitroxide labeled positions 5 and 10 respectively. From these spec tra, it is easy to see the line shape changes upon addition of SDS, and that the DOXYL label in position 5 is more sensitive to membrane micellization. The EPR line shape shown in Figures 4-4 (spe ctrum a) is for 5-DOXYL labeled lipid in POPC LUVs with no SDS and s hows a typical anisotropic powde r-like pattern expected for a spin-probe with restricted motion intercalated in a membrane structure(4, 11). This line shape reflects the most immobilized spin -label/lipid motion in this seri es as indicated by the largest splitting of the hyperfine interaction tensor component ( ), the largest value of the second moment, the largest fractional intensity, a nd the largest order parameter (26.4 G, 202 G2, 1, and 0.68 respectively). Moreover, the order pa rameter for spectrum a corresponds to motional 'T 59

PAGE 60

averaging with an angular deviation of 28o from the bilayer normal. Spectrum d shows a line shape that is the most mobile for the series with values of the hyperfine splitting, second moment, fractional intensity, and order parameter determined to be 20.7 G, 198 G2, 0.0, and 0.40 respectively, corresponding to motions leading to an aver age angular deviation of 39o from the bilayer normal. Assuming the boundaries of the phase diagram in Figure 4-3, this line shape can be understood in terms of solub ilization of the POPC bilayer by SDS resulting in the formation of mixed micelles. Hence some of the 5-DOXYL lipid is now in a micellar environment and the line shape parameters reflect the increased co rrelation time of the smaller spherical micelle compared to the POPC vesicle. The values of angular deviation from the bilayer normal ( ) were calculated using Eqn. 3-6 along with the or der parameter obtained from Eqn. 3-7. Also note the variable used to represen t the average orientation with re spect to the bilayer normal is changed from to Spectra b and c in Figures 4-4 and 4-5 are EPR line shapes obtained from the region on the phase diagram in Figure 4-3 between the two phase boundaries but are not simple superpositions of line shapes obtained from labeled lipids in saturated lamellar bilayer aggregates and those located mixed SDS/lipid micelles. This may be because spectrum d in Figure 4-4 was obtained from a location very n ear the phase boundary, therefore the label is reporting motion from both mixed micelles and saturated lamellar structural environments. For tracking changes in acyl chain order it is most useful, at least for this investigation, to compare spectral parameters such as the fractiona l intensities, the peak-to-peak widths, and the order parameters, because the changes in second moment appear to be extremely sensitive to baseline correction errors. Figures 4-6 and 4-7 s how plots of each of these values as a function of the ratio of SDS to lipid concentrations (CSDS/CPOPC) for samples incorporating a 5 and 10DOXYL labeled lipid, respectively. Recall, a le ss ordered location in the bilayer, such as 60

PAGE 61

position 10, has a narrower line shape because th e degree of motional averaging is greater for spin probes closer to the center of the bilayer (a larger probability for trans/gauche isomerization, (16) thus more disorder). Unfortunately, numbe r of data points collected for the spin-labeled lipids, incorporated into POPC LUVs, is too small to draw an y quantitative conclusions with regards to utilizing EPR line shapes of incorporated spin probes to define the phase boundaries for the solubilization of POPC LUVs by SDS. However, we have measured the characteristic line shapes and corresponding tre nds in fractional intensity, p eak-to-peak line width, second moment, and order parameters, for two spin probe positions during the so lubilization of POPC bilayers by SDS detergent. Hydrated Bis(monoacylglycerol)phosphate A ssemblies Solubilized by Sodium Dodecyl Sulfate In order to initially characterize the effects that detergen t partitioning and subsequent micellization has upon the EPR line shape of DOXYL labels in BMP aggregates and compare them to the results obtained for POPC LUVs, SDS was titrated into extruded BMP structures containing the 5-DOXYL PC. Due to the relative monetary expense of purchasing BMP and the exploratory nature of our inves tigation only a single site within the aggregated structure was examined, carbon position 5 on the sn2, ster ic acid chain of the labeled lipid. The EPR spectrum (Figure 4-8 spectrum a) for 99.5% BMP18:1 mesostructures (extruded through 100nm diameter membranes) doped 0.5% 5-DOXYL spin-labeled lipid in an aqueous environment, and in the absence of SDS is broad ( 24.6T G and 9.0T G) and has similar structural features as the line shapes obtained from the lamellar POPC assemblies in seen in Figure 4-4 ( G and G). Spectra (Figure 4-8 b i) are line shapes illustrating the effect of SDS partitioning and solubilization of BMP aggregates. These spectra show a similar trend in spectral line narrowing previously seen in POPC lamellar assemblies as SDS 26.4T8.8T 61

PAGE 62

concentration is increased. The similar line shapes and spectral parameters (in the absence of SDS) indicate that the acyl ch ain environment in BMP dispersions is similar to that in POPC LUVs. In addition, the trend in variation of the line shape parameters as SDS solubilization occurs for BMP is analogous to those observe d upon SDS solubilization of POPC LUVs. The second moment, Hpp, normalized fractional intensity, orde r parameter, and angular deviation values obtained from analysis of these line sh apes of 5-DOXYL PC in BMP as a function of SDS concentration are listed in Table 4-1. Figure 4-9 shows a comparison of the values of peak-to-peak width, second moment, order parameter, and normalized fractional intensity as a function of SDS/Lipid concentration ratio for BMP and POPC aggregates obtained using th e 5-DOXYL labeled lipid. Again, it is most instructive to examine the trends in Hpp, Si, and fI because the second moment value is very sensitive to baseline correction errors. By visual inspection it is clear that each of these parameters shows an increase in disorder as SDS concentration increases, indicating both partitioning and solubilization of the lipid struct ures by SDS. The order parameter for pure BMP (0.62) is lower than pure POPC (0.68) and could be explained by BM Ps second unsaturated chain; this would probably result in a larger volume requirement per lipid molecule and reduce steric constraints for isomerization. The values used to track partitioning and solubilization of the BMP aggregates follow the same general tr end as POPC LUVs. The similarity of these trends may be another positive indication that BMP assembles into a lamellar aggregate with acyl chain packing and dynamics similar to those in POPC lamellar structures. Since more EPR data was collected for the BMP solubilization study than for the POPC solubilization study it may be possible to make more quantitative statements concerning the sensitivity of spin-probes to the solubilization process. For example, according to Blume and 62

PAGE 63

coworkers a limiting ratio of ~1 .5 SDS molecules to 1 POPC molecule corresponds to the saturation point of POPC vesicles and onset of solubilization at 65oC(2). A similar ratio is detected by our EPR investigation of BMP aggr egates; clearly indicate d by the discontinuity between concentration ratio s 1.3 and 1.7 of plots C and D of Figur e 4-9. This is not in agreement with the value of 0.28 reported earlier in equation 4-2 reported by Seelig for data collected at 56oC. However, in the low concen tration range (Figure 3C of (6)) light scattering data at 20oC and ITC data at 56oC report a much steeper slope correspon ding to a ratio of approximately 1.4 or 1.5(2). This ratio is in good agreement with our experimental value of ~1.5 SDS molecules to 1 POPC molecule. There is another possible disc ontinuity present in our EPR data for the region between concentration ratios 2.5 and 3.5 corresponding to the begi nning of solubilization of the aggregate structure; this disconti nuity is more obvious in Figure 4-9 plot D. Seelig and Blume report ratios of 2.2 and 2.7 SDS molecules per lipid molecule needed to initiate solubilization of POPC membranes at 56oC and 65oC, respectively(2, 6). According to the previous observations and the assumption that BMP has a similar solubilization diagram as POPC, we believe we have sampled each region of the phase diagram for the solubilization of BMP by SDS detergent, and have obtained li ne shapes that are characteristic of each region(3, 9). Therefore, the spectra for BMP dispersions found in Figure 4-8 are assigned to the following regions of a solubilization diagram a lamellar BMP, b-d SDS momomer/SDS partitioned into BMP bilayers, e-g SDS momomer/SDS saturated BMP bilayers/mixed SDS BMP micelles, and hi SDS momomer/mixed SDS BMP micelles. Given that our EPR results for BMP solubili zation behavior are c onsistent with those obtained by other techniques for bilayer form ing POPC; the acyl chain packing in BMP dispersions likely adopts a similar lamellar structur e. Hence, these data are not consistent with a 63

PAGE 64

model of BMP18:1 as a micellar aggregate. This conclusi on is in accordance with light scattering, electron microscopy and fluores cence results indicating BMP18:1 forms MLV assemblies(17), and X-ray and molecular dynamics data obtained by Kobayashi showing BMP14:0 also forms a stable lamellar aggregate(18). Figure 4-1. Model membrane pe rturbants: A) Sodium dodecyl sulfate; B) (S,R isomer) sn-(3oleoyl-2-hydroxy)-glycerol-1-phosphosn-3'-(1'-oleoyl-2'-hydroxy)-glycerol (ammonium salt). 64

PAGE 65

Figure 4-2. Positional isomers of spin-label ed lipids: A) 1-palmit oyl-2-stearoyl-(5-DOXYL)-snglycero-3-phosphocholine (5-DOXYL); B) 1-palmitoyl-2-stearoyl-(10-DOXYL)-snglycero-3-phosphocholine (10-DOXYL). 65

PAGE 66

01 02 03 04 0 0 5 10 15 20 25 30 35 40 45 50 Co L mM Co D mM Figure 4-3. Phase Diagram fo r SDS and POPC LUVS at 56oC reproduced from linear regression analysis by Seelig and coworkers. Solid lines represent phase boundaries: the symbol ( ) indicates samples made with POPC LUVs containing 5-DOXYL labeled lipid and ( ) indicates samples made with POPC LUVs containing 10-DOXYL labeled lipid to collect representative spectra for specific regions of the phase diagram in our solubilization experiments. 66

PAGE 67

a) b) c) d) 20 G 12 a) b) c) d) 20 G a) b) c) d) a) b) c) d) 20 G 20 G 12 Figure 4-4. cw-EPR spectra of POPC LUVs with 5-DOXYL spin probe (1 mol%) at room temperature in 20mM HEPES, 100mM NaCl, 0.02% NaN3 at pH 7.4: a) 27 mM POPC 0mM SDS; b) 22 mM POPC 6mM SD S; c) 11 mM POPC 18mM SDS; d) 7 mM POPC 22mM SDS. 1 2 20 Ga) b) c) d) 1 2 20 G 1 2 1 2 20 G 20 Ga) b) c) d) Figure 4-5. cw-EPR spectra of POPC LUVs with 10-DOXYL spin probe (1 mol%) at room temperature in 20mM HEPES, 100mM NaCl, 0.02% NaN3 at pH 7.4: a) 27 mM POPC 0mM SDS; b) 19 mM POPC 8mM SD S; c) 10 mM POPC 19mM SDS; d) 8 mM POPC 21mM SDS. 67

PAGE 68

0.00.51.01.52.02.53.03.5 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 Hpp(G)CSDS/CLipid0.00.51.01.52.02.53.03.5 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 Order Parameter (Si)CSDS/CLipid0.00.51.01.52.02.53.03.5 195 200 205 Second Moment (M2)CSDS/CLipid0.00.51.01.52.02.53.03.5 0.0 0.2 0.4 0.6 0.8 1.0 Normalized Fractional Intensity (fI)CSDS/CLipidA) D) B) C)0.00.51.01.52.02.53.03.5 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 Hpp(G)CSDS/CLipid0.00.51.01.52.02.53.03.5 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 Order Parameter (Si)CSDS/CLipid0.00.51.01.52.02.53.03.5 195 200 205 Second Moment (M2)CSDS/CLipid0.00.51.01.52.02.53.03.5 0.0 0.2 0.4 0.6 0.8 1.0 Normalized Fractional Intensity (fI)CSDS/CLipid0.00.51.01.52.02.53.03.5 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 Hpp(G)CSDS/CLipid0.00.51.01.52.02.53.03.5 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 Order Parameter (Si)CSDS/CLipid0.00.51.01.52.02.53.03.5 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 Hpp(G)CSDS/CLipid0.00.51.01.52.02.53.03.5 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 Order Parameter (Si)CSDS/CLipid0.00.51.01.52.02.53.03.5 195 200 205 Second Moment (M2)CSDS/CLipid0.00.51.01.52.02.53.03.5 0.0 0.2 0.4 0.6 0.8 1.0 Normalized Fractional Intensity (fI)CSDS/CLipidA) D) B) C) Figure 4-6. Various spectral parameters of 5-DOXYL labeled lip id incorporated into POPC LUVs as a function of SDS/Lipid concentr ation ratio at room temperature in 20mM HEPES, 100mM NaCl and 0.02% NaN3 and pH 7.4: A) Peak-to-peak width of central derivative line; B) Second spectral moment; C) Order parameter; D) Average normalized fractional intensity. Order parameter error bars are estimated by assuming a 1 G error in the difference between the parallel and perpendicular components of the hyperfine tensor. 68

PAGE 69

0.00.51.01.52.02.53.03.5 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 Order Parameter (Si)CSDS/CLipid0.00.51.01.52.02.53.03.5 165 170 175 180 185 190 195 200 Second Moment (M2)CSDS/CLipid0.00.51.01.52.02.53.03.5 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 Hpp(G)CSDS/CLipid0.00.51.01.52.02.53.03.5 0.0 0.2 0.4 0.6 0.8 1.0 Normalized Fractional Intensity (fI)CSDS/CLipidA) D) B) C)0.00.51.01.52.02.53.03.5 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 Order Parameter (Si)CSDS/CLipid0.00.51.01.52.02.53.03.5 165 170 175 180 185 190 195 200 Second Moment (M2)CSDS/CLipid0.00.51.01.52.02.53.03.5 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 Hpp(G)CSDS/CLipid0.00.51.01.52.02.53.03.5 0.0 0.2 0.4 0.6 0.8 1.0 Normalized Fractional Intensity (fI)CSDS/CLipid0.00.51.01.52.02.53.03.5 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 Order Parameter (Si)CSDS/CLipid0.00.51.01.52.02.53.03.5 165 170 175 180 185 190 195 200 Second Moment (M2)CSDS/CLipid0.00.51.01.52.02.53.03.5 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 Hpp(G)CSDS/CLipid0.00.51.01.52.02.53.03.5 0.0 0.2 0.4 0.6 0.8 1.0 Normalized Fractional Intensity (fI)CSDS/CLipidA) D) B) C) Figure 4-7. Various spectral parameters of 10-DOXYL labeled lipid incorporat ed into POPC LUVs as a function of SDS/Lipid concentr ation ratio at room temperature in 20mM HEPES, 100mM NaCl and 0.02% NaN3 and pH 7.4: A) Peak-to-peak width of central derivative line; B) Second spectral moment; C) Order parameter; D) Average normalized fractional intensity. Order parameter error bars are estimated by assuming a 1 G error in the difference between the parallel and perpendicular components of the hyperfine tensor. 69

PAGE 70

20 Ga) b) c) d) e) f) g) h) i) 20 G 20 G 20 Ga) b) c) d) e) f) g) h) i) a) b) c) d) e) f) g) h) i) Figure 4-8. cw-EPR spectra of BMP with 5-DOXYL spin probe (1 mol%) at room temperature in 5 mM HEPES,100 mM NaCl, and pH 7.4; a) 40mM total lipid 0mM SDS; b) 40mM total lipid 17mM SDS; c) 40mM total lipid 35mM SDS; d) 40mM total lipid 52mM SDS; e) 40mM total lipid 69mM SDS; f) 40mM total lipid 87mM SDS; g) 40mM total lipid 104mM SDS; h) 40mM to tal lipid 139mM SDS; i) 20mM total lipid 104mM SDS. 70

PAGE 71

0.00.51.01.52.02.53.03.54.04.55.05.5 175 180 185 190 195 200 205 Second Moment (M2)CSDS/CLipid0.00.51.01.52.02.53.03.54.04.55.05.5 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 Hpp(G)CSDS/CLipid0.00.51.01.52.02.53.03.54.04.55.05.5 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 Order Parameter (Si)CSDS/CLipid0.00.51.01.52.02.53.03.54.04.55.05.5 0.0 0.2 0.4 0.6 0.8 1.0 Normalized Fractional Intensity (fI)CSDS/CLipidA) D) B) C)0.00.51.01.52.02.53.03.54.04.55.05.5 175 180 185 190 195 200 205 Second Moment (M2)CSDS/CLipid0.00.51.01.52.02.53.03.54.04.55.05.5 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 Hpp(G)CSDS/CLipid0.00.51.01.52.02.53.03.54.04.55.05.5 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 Order Parameter (Si)CSDS/CLipid0.00.51.01.52.02.53.03.54.04.55.05.5 0.0 0.2 0.4 0.6 0.8 1.0 Normalized Fractional Intensity (fI)CSDS/CLipidA) D) B) C) Figure 4-9. Various spectral parameters of 5-DOXYL labeled lip id incorporated into POPC LUVs ( ) and BMP aggregates ( ) as a function of SDS/Lipi d concentration ratio at room temperature in 20mM HE PES, 100mM NaCl, 0.02% NaN3 at pH 7.4 for POPC LUVs and 5 mM HEPES,100 mM NaCl, at pH 7.4 for BMP aggregates: A) Peak-topeak width of central derivative line; B) Second spectral moment; C) Order parameter; D) Average normalized fractiona l intensity. Order parameter error bars are estimated by assuming a 1 G error in the difference between the parallel and perpendicular components of the hyperfine tensor. 71

PAGE 72

Table 4-1. Parameters defining order of the 5-DOXYL nitroxide sp in-probe in lipid aggregates at room temperature. 0 0.33 0.44 0.44 0.56 0.89 0.89 0.89 1 0 0.34 0.83 1 Normalized Fractional Intensity (fI) 36o0.49 189 2.6 87 40 35o0.50 185 2.6 69 40 32o0.58 190 2.7 52 40 37o0.45 179 2.5 104 40 32o0.58 190 2.7 35 40 31o0.60 185 2.7 17 40 39o0.42 179 2.4 139 40 30o0.62 189 2.9 0 40 5-DOXYL label/BMP 39o0.40 198 2.9 22 7 0.19 0.51 0.62 0.68 Order Parameter (Si) 180 197 201 202 2ndMoment G247o35o30o28oAverage angle to bilayernormal ( 2.2 3.2 3.6 3.9 HppG 0 27 5-DOXYL label/POPC LUVS 104 20 18 11 6 22 [SDS] mM [Total Lipid] mM 0 0.33 0.44 0.44 0.56 0.89 0.89 0.89 1 0 0.34 0.83 1 Normalized Fractional Intensity (fI) 36o0.49 189 2.6 87 40 35o0.50 185 2.6 69 40 32o0.58 190 2.7 52 40 37o0.45 179 2.5 104 40 32o0.58 190 2.7 35 40 31o0.60 185 2.7 17 40 39o0.42 179 2.4 139 40 30o0.62 189 2.9 0 40 5-DOXYL label/BMP 39o0.40 198 2.9 22 7 0.19 0.51 0.62 0.68 Order Parameter (Si) 180 197 201 202 2ndMoment G247o35o30o28oAverage angle to bilayernormal ( 2.2 3.2 3.6 3.9 HppG 0 27 5-DOXYL label/POPC LUVS 104 20 18 11 6 22 [SDS] mM [Total Lipid] mM Table 4-2. Parameters defining order of the 10-DOXYL nitroxide sp in-probe in lipid aggregates at room temperature. 0 0.047 0.97 1Normalized Fractional Intensity (fI)0.21 0.26 0.43 0.43Order Parameter Si168 172 181 1822ndMoment G247o45o38o38oAverage angle to bilayernormal ( 2.6 2.7 3.2 3.3 HppG0 2710-DOXYL label/POPC LUVs21 8 19 10 8 19[SDS] mM [Total Lipid] mM 0 0.047 0.97 1Normalized Fractional Intensity (fI)0.21 0.26 0.43 0.43Order Parameter Si168 172 181 1822ndMoment G247o45o38o38oAverage angle to bilayernormal ( 2.6 2.7 3.2 3.3 HppG0 2710-DOXYL label/POPC LUVs21 8 19 10 8 19[SDS] mM [Total Lipid] mM 72

PAGE 73

73 CHAPTER 5 PERTURBATIONS OF LAMELLAR LIQUID CHRYSTALLINE ORDER BY BIS(MONOACYLGLYCEROL)PHOSPHATE Solid State Phosphorus-31 NMR Investigation of Hydrated Bis(monoacylglycerol)phosphate Aggregation State Solid state 31P NMR experiments were performed at 145.2 MHz using a two pulse Hahn echo sequence with full proton d ecoupling on single component BMP18:1 MLV samples in an aqueous environment. The 31P NMR measurements allowed us to investigate the head group order and the mesophase structur e of hydrated BMP MLV dispersions. Visual inspection of the line shapes of phosphorus NMR spectra provides information about head group orientation and information needed to discriminate among lamellar, HII, and isotropic aggregation states(16, 59, 66). Both changes in the span of the chemical shift anisotropy and in the overall line shapes are governed by individual molecula r motions, local environment of phosphorus nuclei, and overall aggregate tumbling rate, each of whic h affect the orientation of the 31P chemical shift tensor in the phosphate head group with respect to the applied field(16, 67). The 31P NMR spectrum of the BMP18:1 MLV dispersion shown in Figure 5-1 is narrow ( = -11.6 ppm) and has an overall li ne shape characteristic of a lamellar mesophase. This finding is in agreement with our previous conclusion in Chapter 4 that BMP forms a stable lamellar structure. This being said, BMP would also then be expected to form stable lamellar structures when mixed with model lipids such as POPC and DPPC. All of the EPR and NMR evidence presented thus far indicates that BMP adopts a lamellar aggregate structure and thus should not exhibit detergent behavior. However, the ef fect BMP has on model membrane morphology or any possible role BMP may play in late endosomal lipid trafficki ng or in the hydrolysis of GSLs has yet to be addressed.

PAGE 74

Acyl Chain Order of 1-Palmitoyl-2 -Oleoyl-sn-Glycero-3-Phosphocholine /Bis(monoacylglycerol)phosphate Mixed Vesicles Determined by Electron Paramagnetic Spectroscopy In this section of the dissertation, EPR resu lts are presented regarding the interaction of BMP with model membrane lipids in the L phase. POPC and BMP LUVs of mixed composition were prepared with POPC, varying amounts of BMP18:1 or BMP14:0 and 1 mol% of spin-labeled lipid (5-DOXYL or 10-DOXYL). Th ese experiments parall el the previous EPR investigations reported in Chap ter 4 with single component, lipid LUVs and SDS detergent. EPR spectra shown in Figure 5-2 obtained from POPC/BMP mixed LUVs have the typical anisotropic powder patterns seen previously in Chapter 4. However, the distinct lines shape changes seen in EPR spectra for the 5-DOXYL la beled lipid that occurs when SDS partitions into lamellar aggregates are not detected in POPC/BMP mixed vesicles indicating that BMP does not solubilize POPC membranes. To track the changes that occur in the EPR line shapes of the doxyl labeled lipids as a function of BMP concentration, values of Hpp and the order parameter (Si) were determined from the EPR line shapes. These data are plotted as a function of the con centration ratio of BMP to POPC (CBMP/CPOPC) in Figures 5-3 and 5-4, respectively, and values are listed in Table 5-1. Figure 5-3 shows that the peak-to-peak width of the central deriva tive line and the order parameter values are constant for both the 5 and 10-DOXYL labele d lipids in POPC/BMP14:0 mixed LUVs for concentrations up 20 mol% BMP14:0. Figure 5-4 shows the values of Hpp and order parameter for both the 5 and 10-DOXYL labeled lipids in POPC/BMP18:1 mixed LUVs are also constant, within experimental error, over the same concentration range. It is also important to note the average value of the order parame ters for the 5 and 10-DOXYL positions are very similar for both BMP18:1 and BMP14:0. Moreover, these results indicate that mixed composition 74

PAGE 75

75 LUVs of POPC/BMP (less than 20 mol % BMP) have the same degree of acyl chain order as pure POPC (within experimental error) at both label positions regardless of BMPs degree of acyl chain saturation or chain length. These results further support our previous claim that BMP does not have classic detergent properties, a nd indicates BMP does not perturb the acyl chain order of POPC LUVs, at least not on the time scale observed by EPR. 31P NMR of 1-Palmitoyl-2-Oleoyl-sn-Glycero-3-Phosphocholine /Bis(monoacylglycerol)phosphate mixed MLVs The conclusions drawn in the previous secti on indicate that incorporation of BMP into POPC vesicles does not alter the order or packin g of the acyl chains as can be observed on the EPR timescale. Here, the effects of BMP on h ead group orientation were investigated using solid-state 31P NMR to characterize the in teraction of BMP with POPC in hydrated MLVs. Solid state Phosphorus-31 NMR experiments obtained at 145.2 MHz using a two pulse Hahn echo sequence with full proton decouplin g were performed on mixed POPC/BMP18:1 MLV samples in an aqueous environment to investigate the head group order and mesophase structure. POPC and POPC doped with BMP18:1 MLV samples were prepared acc ording to the procedure for LUV preparation but were not ex truded (see Chapter 2). All 31P spectra were aligned so that isotropic value of either axially symme tric fully hydrated POPC MLVs or axially symmetric fully hydrated DPPC MLVs was assigned to 0 ppm, see Figure 3-11. Recalling Figure 3-11, 31P spectra for hydrated lamellar a ggregates have characteristic powder averaged chemical shift line shapes. The solid state 31P NMR spectrum of the hydrated single lipid component POPC MLVs at 37oC is shown in Figure 5-5. This spectrum exhibits a classic axially symmetric lamellar line shape with a span of -46 1 ppm, which is in excellent agreement with the reported literature value (46 1 ppm)(68-70). Line shape simulation via software (dmfit) (71) did not offer any advantage over direct measurement from the experimental

PAGE 76

line shapes in estimation of 31P chemical shift span, therefore all reported span values were measured directly from the experimental data. According to the model of glycerophospholipid orientation with respect to the normal of a bilayer for DPPC and DPPE; th e order parameter of the C1C2 bond vector (S(c1)-(c2) from (72)) obtained from 2H NMR (C1 and C2 are glycerol carbons enumerat ed using the sn naming system see Figure 1-2), SC1C2 is 0.66 and defines th e tilt angle of the C1C2 axis with respect to the bilayer normal.(72, 73) This wobble freedom reduces the span of the strict axially symmetric powder average for 31P ( ~ -124 ppm) to a maximum of -82 ppm for hydrated MLVs (16) ( = ( 11 33)*SC1C2 = -82 ppm). However, the experimental value (-46 ppm) is clearly narrower indicating an additional angular offset of the CSA axis from the bilayer normal. The experimental principal value is ~ -30 ppm for pure POPC (the trace of the CSA is invariant to motion so = 16 ppm) and yields an angle of about 33o from the normal of the bilayer. Again the order parameters are represented as s econd order Legendre polynomials (S = (3 cos2( ) 1)). Both of the previous values are in good agreement with current literature; Lorigan and coworkers measured = 30 1 ppm and = 15 1 ppm from oriented POPC MLVs(70), and an angle with respect to the bilayer normal of ~ 30o has been reported for hydrated dipalmitoyllecithin(8, 9). Firgure 5-6 shows 31P line shapes for various mixt ures of POPC/BMP MLVs. Two important observations about mixed POPC/BMP18:1 MLVs can be made by visual inspection of the line shapes: 1) the CSA span of the POPC line shape decreases as the negatively charged BMP concentration increases when compared to the span of single component POPC (spectrum a 2) the spectra of the mixed lipids do not appe ar to be a simple superposition of the single component POPC and single component BMP CS A line shapes. Each of the previous 76

PAGE 77

observations, more clearly illustrated in Figure 5-7, indicates the lipids are interacting (mixing) in the same lamellar structures. According to work by Seelig and Scherer, th e orientation that th e phosphate/choline group has with respect to the b ilayer normal is sensitive to membrane surface charge(68, 74). Quadrupolar splitting of the and carbons located on the cholin e head group of deuterium labeled POPC and the CSA span of POPC were measured as a function of positive charged dioctadecyldimethylammonium-bromide (2C18N2C1 +Br-) and negative charged sodium didodecyl phosphate (2Cl2PO4 -Na+) amphiphilic molecules incorpor ated into POPC MLVs. A linear relationship relating CSA span and mole fraction of 2Cl2PO4 -Na+ was determined to an amphiphile mole fraction ( b) of 0.3, ( b = 45.6 + 18.7 ppm ) and a plateau value of CSA span ~ -36 ppm at high amphiphile concentrations(68). Increased disorder in the phosphate/choline region, indicated by a decrease in the value of the 31P CSA span, implies a larger angular deviation from the bilayer normal(68). Figure 5-8 shows the change in span of the POPC rich li ne shape of POPC/BMP mixed MLVs up to a mole fraction BMP of 0.3 (closed s quares) and the best lin ear fit of that data ( (dashed line)), however the limiti ng span value observed by Seelig and Scherer is not obvious in our data due to the noise level near 11. Line shape simulations may be helpful in determining the limiting span in this case. Comparison of the two regression lines indicates that the phosphate groups of MLVs co ntaining POPC/BMP are interacting and ordered in a similar fashion to the MLVs with the model anionic amphiphile used by Seelig and Scherer. At a mole fraction BMP of 0.3 bot h regressions predicts an angle of inclination to the bilayer normal of ~36o and a value of of -40 ppm, in good agreement with our experimental values ( and = ppm)(63). These results indicate that BMP, when mixed with the model BMP = 45.6 + 16.0 77

PAGE 78

membrane lipid POPC up to 30 mol% imparts a surface charge similar to that of 2Cl2PO4 -Na+ and has a similar effect on the average orientat ion of the head group w ith the bilayer normal. As for the very narrow span (-11.6 ppm) observed for pure BMP structures there are at least two possible explana tions: 1) BMP forms LUVs that have a small diameter (highly curved surface), or 2) the phosphorus has a large tilt angl e with respect to the normal of the bilayer. Molecular dynamics (MD) simulations indicate that BMP14:0 has a bilayer thickness on the order of 4.2 nm and a 20o tilt angle with respect to the bilayer normal(28). According to the MD it seems possible that the lamellar structures fo rmed by single component BMP may have a small diameter and thus a narrowed CSA span. This claim should be further substantiated by dynamic light scattering experiments. -30-25-20-15-10-5051015202530 chemical shift (ppm) Figure 5-1. 31P NMR chemical shift spectrum of BMP18:1 MLVs: in 5 mM HEPES, 100 mM NaCl, 0.1 mM EDTA, pH 7.4, at T = 37 oC. The CSA is referenced to the isotropic value of axially symmetric, hydrated POPC MLVs. 78

PAGE 79

20 GA) B) C) D)a) b) c) d) e) a) b) c) d) e) a) b) c) d) e) a) b) c) d) e)5-DOXYL 14:0 10-DOXYL 14:0 5-DOXYL 18:1 10-DOXYL 18:1 20 G 20 GA) B) C) D)a) b) c) d) e) a) b) c) d) e) a) b) c) d) e) a) b) c) d) e) a) b) c) d) e) a) b) c) d) e) a) b) c) d) e) a) b) c) d) e)5-DOXYL 14:0 10-DOXYL 14:0 5-DOXYL 18:1 10-DOXYL 18:1 Figure 5-2. cw-EPR spectra of POPC/BMP mixed LUVs containi ng 1 mol% of either 5 or 10DOXYL spin probe in 5 mM HEPES,100 mM Na Cl, at pH 7.4; A) and C) a) 23 mM POPC 0 mM BMP14:0; b) 23 mM POPC 0.2 mM BMP14:0; c) 22 mM POPC 1mM BMP14:0; d) 21 mM POPC 2 mM BMP14:0; e) 19 mM POPC 5 mM BMP14:0; B) and D) a) 20 mM total lipid 0 mM BMP18:1; b) 23 mM POPC 0.2 mM BMP18:1; c) 22 mM POPC 1 mM BMP18:1; d) 21 mM POPC 2 mM BMP18:1; e) 19 mM POPC 5 mM BMP18:1. 79

PAGE 80

0.000.050.100.150.200.250.30 3.40 3.45 3.50 3.55 3.60 3.65 3.70 3.75 3.80 3.85 3.90 Hpp (G)CBMP/CPOPC 0.000.050.100.150.200.250.30 0.40 0.45 0.50 0.55 0.60 0.65 0.70 Order Parameter (Si)CBMP/CPOPCA) B)0.000.050.100.150.200.250.30 3.40 3.45 3.50 3.55 3.60 3.65 3.70 3.75 3.80 3.85 3.90 Hpp (G)CBMP/CPOPC 0.000.050.100.150.200.250.30 0.40 0.45 0.50 0.55 0.60 0.65 0.70 Order Parameter (Si)CBMP/CPOPC0.000.050.100.150.200.250.30 3.40 3.45 3.50 3.55 3.60 3.65 3.70 3.75 3.80 3.85 3.90 Hpp (G)CBMP/CPOPC 0.000.050.100.150.200.250.30 0.40 0.45 0.50 0.55 0.60 0.65 0.70 Order Parameter (Si)CBMP/CPOPCA) B) Figure 5-3. Hpp and Si of 5 ( ) and 10-DOXYL () labeled lipid (1 mol%) incorporated into POPC/BMP14:0 mixed LUVs as a function of BMP14:0/Lipid concentration ratio at room temperature in 5mM HEPES, 100mM Na Cl and pH 7.4: A) Peak-to-peak width of central derivative line; B) Order parameter. Order parameter error bars are estimated by assuming a 1 G error in the difference between the parallel and perpendicular components of the hy perfine tensor. The error in Hpp is 0.2 G based on three independent measurements of single component POPC LUVs. 0.000.050.100.150.200.250.30 0.40 0.45 0.50 0.55 0.60 0.65 0.70 Order Parameter (Si)CBMP/CPOPC0.000.050.100.150.200.250.30 3.30 3.35 3.40 3.45 3.50 3.55 3.60 3.65 3.70 3.75 3.80 3.85 3.90 3.95 4.00 4.05 4.10 Hpp (G)CBMP/CPOPCA) B) 0.000.050.100.150.200.250.30 0.40 0.45 0.50 0.55 0.60 0.65 0.70 Order Parameter (Si)CBMP/CPOPC0.000.050.100.150.200.250.30 3.30 3.35 3.40 3.45 3.50 3.55 3.60 3.65 3.70 3.75 3.80 3.85 3.90 3.95 4.00 4.05 4.10 Hpp (G)CBMP/CPOPCA) B) Figure 5-4. Hpp and Si of 5 ( ) and 10-DOXYL () labeled lipid (1mol%) incorporated into POPC/BMP18:1 mixed LUVs as a function of BMP18:1/Lipid concentration ratio at room temperature in 5mM HEPES, 100mM Na Cl and pH 7.4: A) Peak-to-peak width of central derivative line; B) Order parameter. Order parameter error bars are estimated by assuming a 1 G error in the difference between the parallel and perpendicular components of the hy perfine tensor. The error in Hpp is 0.2 G three independent measurements of single component POPC LUVs. 80

PAGE 81

= 11 33= -46 1 ppm 3311-40-35-30-25-20-15-10-50510152025303540 ChemicalShift (ppm) = 11 33= -46 1 ppm 3311-40-35-30-25-20-15-10-50510152025303540 ChemicalShift (ppm) Figure 5-5. 31P NMR chemical shift of POPC MLVs in 5 mM HEPES, 100 mM NaCl, 0.1 mM EDTA, pH 7.4, at T = 37 oC. The CSA is aligned accordi ng to the isotropic value of fully hydrated POPC MLVs. 81

PAGE 82

-40-30-20-100102030 Chemical Shift (ppm) -40-30-20-100102030 Chemical Shift (ppm)A) B)a) k) k) a) -40-30-20-100102030 Chemical Shift (ppm) -40-30-20-100102030 Chemical Shift (ppm) -40-30-20-100102030 Chemical Shift (ppm) -40-30-20-100102030 Chemical Shift (ppm)A) B) A) B)a) k) k) a) Figure 5-6. 31P NMR spectra of POPC/BMP18:1 MLVs in 5 mM HEPES, 100 mM NaCl, 0.1 mM EDTA, pH 7.4, at T = 37 oC. A) a) 278 mM POPC 0 mM BMP; b) 276 mM POPC 1.8 mM BMP; c) 275 mM POPC 2. 8 mM BMP; d) 270 mM POPC 8.3 mM BMP; e) 264 mM POPC 13.7 mM BMP f) 249 mM POPC 27.8 mM BMP; g) 224 mM POPC 52.3 mM BMP; h) 192 mM PO PC 82 mM BMP; i) 139 mM POPC 133 mM BMP j) 53.8 mM POPC 215 mM BMP; k) POPC 0 mM 42 mM BMP B) shows the same spectra as A) but in reverse order. 82

PAGE 83

-40 -20 0 20 chemical shift (ppm)-40 -20 0 20 chemical shift (ppm)A) B)-40 -20 0 20 chemical shift (ppm)-40 -20 0 20 chemical shift (ppm)-40 -20 0 20 chemical shift (ppm)-40 -20 0 20 chemical shift (ppm)A) B) Figure 5-7. 31P NMR spectra of POPC/BMP18:1 MLVs in 5 mM HEPES, 100 mM NaCl, 0.1 mM EDTA, pH 7.4, at T = 37 oC. A) 278 mM POPC 0 mM BMP (solid line) and 192 mM POPC 82 mM BMP (dashed line); B) 192 mM POPC 82 mM BMP (dashed line) and a linear combination of two line shapes with a contributio n of 70% pure POPC line shape and 30% pure BMP line shape. 83

PAGE 84

0.000.050.100.150.200.250.30 -40 -41 -42 -43 -44 -45 -46 -47 -48 (ppm)BMP18:1 Figure 5-8. CSA span of (POPC/BMP MLVs ) as a function of BMP mole fraction ( ) and linear fit ( (dashed line)). BMP = 45.6 + 16.0 84

PAGE 85

Table 5-1 Parameters defining order of th e 5-DOXYL nitroxide spin-probe in POPC/BMP mixed MLVs at room temperature. 0.65 3.7 2 21 0.64 3.4 5 19 0.64 3.7 2 21 0.64 3.7 1 22 0.66 3.6 0.2 23 0.65 4.0 0 23 5-DOXYL label POPC/BMP18:1LUVS 0.66 3.7 5 19 0.65 0.66 0.66 Order Parameter (Si) 3.7 3.8 3.7 HppG 0 23 5-DOXYL label POPC/BMP14:0LUVS 1 22 0.2 23 [BMP] mM [Total Lipid] mM 0.65 3.7 2 21 0.64 3.4 5 19 0.64 3.7 2 21 0.64 3.7 1 22 0.66 3.6 0.2 23 0.65 4.0 0 23 5-DOXYL label POPC/BMP18:1LUVS 0.66 3.7 5 19 0.65 0.66 0.66 Order Parameter (Si) 3.7 3.8 3.7 HppG 0 23 5-DOXYL label POPC/BMP14:0LUVS 1 22 0.2 23 [BMP] mM [Total Lipid] mM 85

PAGE 86

Table 5-2 Parameters defining order of the 10DOXYL nitroxide spin-probe in lipid aggregates at room temperature. 0.45 3.4 2 21 0.46 3.5 5 19 0.46 3.3 2 21 0.47 3.4 1 22 0.46 3.5 0.2 23 0.47 3.5 0 23 10-DOXYL label POPC/BMP18:1LUVS 0.44 3.5 5 19 0.46 0.46 0.46 Order Parameter (Si) 3.4 3.5 3.4 HppG 0 23 10-DOXYL label POPC/BMP14:0LUVS 1 22 0.2 23 [BMP] mM [Total Lipid] mM 0.45 3.4 2 21 0.46 3.5 5 19 0.46 3.3 2 21 0.47 3.4 1 22 0.46 3.5 0.2 23 0.47 3.5 0 23 10-DOXYL label POPC/BMP18:1LUVS 0.44 3.5 5 19 0.46 0.46 0.46 Order Parameter (Si) 3.4 3.5 3.4 HppG 0 23 10-DOXYL label POPC/BMP14:0LUVS 1 22 0.2 23 [BMP] mM [Total Lipid] mM 86

PAGE 87

Table 5-3 Values of CSA span for POPC /BMP mixed MLVs at room temperature. -20 0.80 215 53.8 -11.6 1.00 42 0 -41 0.30 82 192 0.5 0.19 0.039 0.049 0.030 0.0100 0.0065 0 BMP-41 133 139 -43 -44 -45 -45 -46 -45 -46 1 ppm) 0 278 8.3 270 2.8 275 1.8 276 52.3 224 10 249 13.7 264 [BMP18:1] mM [POPC] mM -20 0.80 215 53.8 -11.6 1.00 42 0 -41 0.30 82 192 0.5 0.19 0.039 0.049 0.030 0.0100 0.0065 0 BMP-41 133 139 -43 -44 -45 -45 -46 -45 -46 1 ppm) 0 278 8.3 270 2.8 275 1.8 276 52.3 224 10 249 13.7 264 [BMP18:1] mM [POPC] mM 87

PAGE 88

CHAPTER 6 PERTURBATIONS OF THERMOTROPIC PHASE TRANSITIONS DUE TO BIS(MONOACYLGLYCEROL)PHOSPHATE Thermotropic Phase Behavior of 1,2-Di palmitoyl-sn-Glycero-3-Phosphocholine /Bis(monoacylglycerol)phosphate MLVs Investigated by 2H NMR Incorporation of a lipid deuter ated in the acyl chain region a llows order/disorder to be measured along the length of the apolar region of a MLV assembly. We chose perdeuterated DPPC for this investigation in order to si multaneously probe the effects of BMP on the molecular order at each carbon position along the acyl chain region as well as its effects on thermotropic phase transiti on temperatures of DPPC(17). Investigations of a thermotropic phase transition are not practical using POPC because th e main phase transition of POPC occurs at 0oC and is near the freezing point of water, whic h makes this type of measurement difficult(43). The following experimental 2H NMR data affords information con cerning the order of the acyl chain region and is analogous to that obtained from EPR measurements with the n-DOXYL labeled lipids discussed earlier. However, 2H NMR operates on a much slower time scale, thus allowing more averaging by the lipid motions. Also, reca ll that substitution of deuterium for hydrogen atoms in the acyl chain region is only a small stru ctural perturbation when compared to the large nitroxide labels used for EPR measurements. 2H NMR line shapes as a function of temper ature were obtained for DPPC MLVs and DPPC doped with BMP18:1 (5 mol%) MLVs in low ionic strength buffer (Im < 0.05 m) and are shown in Figures 6-1 and 6-2. The main phase transition (Tm) for DPPC MLVs occurs at ~ 41.5oC as determined by DSC measurements(29); however, this transi tion is depressed to ~ 37.7oC for perdeuterated DPPC MLVs, also determined by DSC(75). A comparison of the line shapes in Figure 6-1 A and B shows evidence th at incorporation of BMP into the lamellar structure increases the onset temperature of a thermotropic phase transition relative to single 88

PAGE 89

component DPPC MLVs. This is easily visualized in spectra at 35 37 oC in Figure 6-1 A and B. NMR line shapes for MLVs containing BMP are broad and unr esolved at 35 and 36oC, indicative of a lamellar gel phase but the spec trum of single component DPPC has already started to narrow and shows resolved terminal met hyl peaks (the Pake doublet with the smallest quadrupole splitting) at 36oC, indicating more motional freedom. At temperatures above the main phase transition of singl e component DPPC, 2H line shapes (Figure 6-1 B) obtained from mixed DPPC/BMP (5 mol%) MLVs show a broadening of the terminal methyl signal relative to that of DPPC MLVs. Initially we were inclined to interpret this broadening as an indication of restricted mobility and eviden ce of an interdigitated phase. However, results obtained from EPR experime nts, with 16-DOXYL labeled lipid (discussed in the next section), did not conf irm our hypothesis and the cause of the broad, unresolved peaks and a molecular level understandi ng of the structure or dynamics that leads to the broadening of the terminal methyl signal remain unknown. 2H NMR measurements were also performe d on DPPC and DPPC/BMP mixed MLVs at near biological ionic strength (Im ~ 100 m), and the results show quite different effects of BMP on the onset of a thermotropic tran sition temperature. When the number of counter ions in the buffer is much greater than th e number of negative BMP molecule s (~15000:1) a decrease in the phase transition temperature is observed when co mpared to BMP effects in low ionic strength buffer and when compared to single component D PPC in near biological ionic strength buffer. The former can be visualized by comparing the spectra at 35oC of Figure 6-2 A and B. It is clear that for MLVs doped with BMP under physiologi cal buffer conditions a significantly larger portion of the acyl chains are disordered when compared to single component DPPC at 35oC. 89

PAGE 90

These effects on acyl chain order can be seen more clearly by comparing the order parameter (Stotal = Smol*SLD) values (Tables 6-1 and 6-2) for the terminal methyl groups as a function of temperature or the residual quadrupo lar splttings as a function of temperature in Figure 6-3. Referring to Chapte r 3 the quadrupolar splitting, is defined for axially symmetric motion about the lipids l ong molecular axis as ( 3 2QtotalS ), and Q is 167 kHz.(76) Smol is assumed to be positive and equal to -3SCD and SCD is 2 PD3cos( )-1 2.(46, 77) The angle PD describes the orientation angle of the C-D bond vect or with respect to th e director axis, and SLD takes the same functional form as SCD but describes the orientation angle LD of the director with respect to the applied magnetic field. Computational methods are available to dePake or approximate SLD (46, 78, 79) but the signal to noise level of our data is not high enough to obtain this value. However, Stotal should be sufficient to show trends related to the re lative order of a specific site in the bilayer. Figure 6-3 show plots of the terminal methyl order parameter and residual methyl splittings for MLV dispersions of DPPC( filled symbols) and for DPPC/BMP (5 mol%)(open symbols) Data were collected for temperatures ranging from 35 to 43oC, but order parameters cannot be determined for those spectra in the gel phase. A discontinuity is observed in the trend of order parameter values and residual qu adrupolar splittings when plotte d as a function of temperature for the terminal methyl peaks in single component DPPC dispersions. This discontinuity occurs for temperatures between 37 and 38oC in Figure 6-3 A and B, and its value coincides nicely with the detected phase transition temp erature from the gel to the L phase of perdeuterated DPPC in 50 mM phosphate buffer observed by Davis, which were measured by both DSC and 2H NMR.(75) A comparison of the temperature dependent residual quadrupolar splittings between 90

PAGE 91

single component DPPC and DPPC/BMP mixed MLVs obtained in low salt buffer (5 mM Na+, Figure 6-3 A) shows that BMP does not affect the order in the acyl chain region at temperatures above 37oC. Additionally, the main phase trans ition temperature for DPPC/BMP (5 mol%) mixed MLVs is elevated in comparison to that of single component DPPC in low salt buffer. The order parameter and residual quadrupolar splitting profiles for MLVs in near biological ionic stre ngth buffer (105 mM Na+, Figure 6-3 B) show that in the presence of BMP the phase transition is significan tly depressed but the overall order is similar to single component DPPC at temperatures above 37 oC. Additionally, the terminal methyl group signal is not broadened, thus the broadening of the terminal me thyl intensity previously observed at low ionic strength is not an effect of BMP alone. In orde r to determine the onset temperature of the phase transition for DPPC/BMP (5 mol%) we must co llect data for several temperatures below 35 oC. Overall the total order parameter and residual quadrupolar splitting results lead to values for thermotropic phase transitions that are c onsistent with literature values; however, the 2H data should be recollected to obtain highe r quality line shapes, and to c over larger temperature range. This will not only afford a better estimation of phase transition temperatures and experimental error but the spectra can be dePaked and an orde r parameter that is independent of orientation with respect to the magnetic field can be assigne d to each site along the acyl chains. Some of this work has already begun by others in the Fa nucci research group, and the broadening of the terminal methyl signal at low ionic strength has been reproduced but the cause of the broadening is still unexplained. Thermotropic Phase Behavior of 1,2-Di palmitoyl-sn-Glycero-3-Phosphocholine /Bis(monoacylglycerol)phosphate MLVs Investigated by EPR The effects of BMP18:1 on thermotropic phase transitions of lamellar assemblies was also monitored by EPR spectroscopy with the n-DOXYL labeled lipids incorporated into a model 91

PAGE 92

membrane phospholipid containing fully saturated acyl chains, (D PPC). The lipid containing a single unsaturated acyl chain (POP C) could not be used for these experiments because the main phase transition is near 0oC(43). Given that water freezes near this temperature, DPPC was a more appropriate model lipid for this investigatio n. 16-DOXYL spin-labeled lipids were used to obtain EPR data parallel to the previous 2H NMR study regarding changes in acyl chain order near the center of the bilayer. The EPR spectra of DPPC MLVs (containing 1% 16 Doxyl PC) at two different buffer ionic strengths are shown in Fi gures 6-4 (low with 5 mM Na+) and 6-5 (near biological ionic strength with 105 mM Na+), and demonstrate the temperatur e dependent order profile of the nitroxide label located at carbon position 16 of the st eric acid chain of the label in DPPC MLVs. The broadest splitting of the parallel component of the hyperfine interaction is observed when the hydrated lipids are in the gel phase, indicating the most ordered environment (largest order parameter) of the series. As the temperature is increased, the lipid molecules obtain more kinetic energy, allowing for more motiona l freedom, which results in a narrowing of the EPR lineshape. Plots of spectral parameters such as the fractiona l intensity, the peak-to pe ak width, and the order parameter as a function of temperature are expect ed to give characteristic sigmoidal shaped profiles for smooth thermo tropic phase transitions(80-83). The phase transition temperature is defined as the inflection point of the sigmoidal shaped line. A comparison of the temperature dependence of the peak-to-peak width, the second moment, and the order parameter values (Tab les 6-3 to 6-4) for DPPC MLVs (1% 16-DOXYL) at low ionic strength (5 mM Na+) and DPPC MLVs (1% 16-DOXYL) at near biological ionic strength (105mM Na+) are shown in Figure 6-6. It is clear that the acyl chain order is almost identical for all temperatures regardless of ionic strength, including the thermotropic phase 92

PAGE 93

transition occuring near 35oC. This transition is assigned as the pre-tran sition of single component DPPC MLVs, because the breadth of this transition is much larger than the breadth (0.5oC) of the main phase transition obtained by Davis (75, 84). Figure 6-6 D and E show the first derivative of the sigmoidal fit used to determine the inflection point and thus the transition temperature, and the sigmoidal fit overlain on the plot of peak-to-peak line width, respectively. The values of the pre-transition temperatures ca lculated from each of the parameters defining bilayer order are listed in Table 65. Additional data should be collected in the range between 25 and 35oC to verify the shape of the sigmoidal curve and the transitions temperatures. EPR spectra for 16-DOXYL PC (1 mo l%) incorporated into DPPC/BMP18:1 MLV dispersions prepared in a buffer w ith low ionic strength (Figure 6-7) have similar shapes as those seen previously for this spin probe in single component DPPC dispersions. In additions, values of the peak-to-peak line width, the second moment and the order parameters are similar to those obtained for single component DPPC. From these data we can conclude that little to no change in order of the acyl chains near the ce nter of the bilayer is caused by BMP18:1 in low ionic strength buffer. Analysis of the EPR line shapes from 16-DOXYL in DPPC/BMP18:1 MLVs prepared in buffer that mimics biological ioni c strength show a lowering (c ompared to DPPC) of the pretransition temperature when compared to the value obtained for MLVs in low ionic strength buffer. Specifically, this conclusion can b een seen by comparing the line shapes between 35oC and 37oC in Figures 6-7 and 6-8, and it is mo re clearly evident in the plots of Hpp, second moment, and Si values as a function of temperature (Tab les 6-6 and 6-7) in Figure 6-9. Clearly, this thermotropic phase transition is altered for DPPC/BMP (5 mol%) mixed MLVs in near biological ionic strength buffer with respect to DPPC/BMP (5 mol%) mixed MLVs in low ionic 93

PAGE 94

strength buffer (Figure 6-9 C). This possible depression of a ther motropic transition is consistent with the result determined earlier in this chapter by 2H measurements under similar conditions. The degree of order is similar at 25 oC for the MLVs containing BMP (5 mol%) in near biological ionic strength buffer as can be seen by comparing the peak-to-peak width values in Figure 6-10 C, indicating a simila r packing arrangement of the acyl chains in the gel phase. However, a comparison of the peak-to-peak width and order parameter values in Figure 6-10 A, B, and D indicate that DPPC/BMP (5 mol%) mi xed MLVs in near biol ogical ionic strength buffer are less ordered than those in low ionic strength buffer. Clearly experiments should be repeated at 25oC and data should be obtained in th e temperature region between 25 and 35oC in order to verify either observation. At te mperatures above the chain melting transition (Tm) of single component DPPC (~ 41.5oC) (75, 84) the order parameters are identical within experimental error. This result is consistent with the observation that th e order parameters for 5 and 10-DOXYL labels in POPC/BMP (5 mol%) mixed MLVs in the L phase are not affected by inclusion of BMP into the MLVs. 1,2-Dipalmitoyl-sn-Glycero-3-Phosphocholine /Bis(monoacylglycerol)phosphate 31P NMR Additional information concerning the effects that BMP has on lipid molecular order can be obtained by examining changes in the 31P NMR chemical shift line shapes. Experiments performed with T>Tm provide information concerning the or der of the PC head group in the L phase and can be compared to results obtai ned in Chapter 5 for mixed POPC/BMP MLVs. Figure 6-11 shows a comparison between the 31P chemical shift line shapes of single component DPPC and DDPC/BMP (5 mol% ) MLVs at temperatures above (43oC) and below (35oC) the main phase transition temperature of fully hydrated DPPC MLVs. The 31P spectra obtained from MLVs cont aining DPPC/BMP are composed of at least two line shapes but are not 94

PAGE 95

simple superpositions of the individual components; as was also seen in the spectra obtained for POPC/BMP mixed MLVs in Chapter 5. Plots A and B of Figure 6-11 and values liste d in Table 6-8 show that both lamellar CSA patterns for DPPC/BMP (5 mol%) MLVs at either low (5 mM Na+), or near biological (105 mM Na+) ionic strength are narrowed with respect to pure DPPC MLVs at 43 oC. This result is consistent with 31P chemical shift measurements reporte d earlier for POPC/BMP MLVs in L type assemblies, indicating the choline head group is tilted more toward the bilayer plane (larger angle from the bilayer normal when compared to single component DPPC) in the presence of negative charged amphiphiles. This resu lt also provides more evidence that BMP18:1, even at concentrations below those found in late endosomes, changes the packing parameters of the head group region above the gel to liquid lamellar phase transition temperature of pure DPPC. However, for chemical shift line shapes obtained at T< Tm (Figure 6-10 C and D) the head group region of mixed DPPC/BMP (5 mol%) MLVs is only slightly more disordered in low ionic strength buffer but is sign ificantly more disordered in n ear biological strength buffer. These results are more easily visualized by compar ing the plots in Figure 6-12 and values listed in Table 6-8. The increased concentration of pos itive counter ions at th e buffer bilayer interface in near biological ionic strengt h buffers may in fact increase the choline head group tilt induced by incorporation of BMP into zwitterionic model bilayers. Figure 6-12 A reports the span of single com ponent DPPC as a function of temperature in low and near biological ionic strength buffers; the reported span values are in agreement within 2 ppm of literature values for DPPC MLVs hydrat ed with water. Griffin and coworkers report values of -54 ppm (T = 37oC) and -48 ppm (T = 45oC)(85), while Seelig repor ts values of -54 ppm (T = 38oC) and -49 ppm (T = 44oC). (73) The values of the CSA span obtained from DPPC 95

PAGE 96

MLVs as a function of ionic stre ngth do not change; hence, these data indicate that there is no significant change in head group orientation in DPPC for the two ionic strength buffers used in our investigation. A comparison of the span as a function of temperature for DPPC MLVs and DPPC doped with BMP (5 mol%) MLVs in low and at near biological ionic strength buffers are shown in Figures 6-12 B and C, respectivel y. It is evident that BMP induces a change in the PC head group orientation from that in single component DPPC for both buffer conditions and at all temperatures investigated. Moreover for DPPC /BMP (5 mol%) mixed MLVs in buffer near biological ionic strength the PC h ead group orientation is more a ffected than for DPPC/BMP (5 mol%) mixed MLVs at low ionic strength. This conclusion is drawn from the smaller CSA span values in Figure 6-11 D for the MLVs in near biol ogical ionic strength buffer at all temperatures investigated. Further inspection Figure 6-11 D also shows evidence of a phase transition for DPPC/BMP mixed MLVs beginning at 37oC. It is clear that the chem ical shift span and thus the head group order is decreasing in a linear fashion for mixed MLVs in both low and near biological strength buffer (the span deceases ~ 1 ppm per oC over the temperature range 37 to 41oC. The onset of this transition is consistent wi th that obtained from the total order parameter of the terminal methyl group as a function of temperature for mixed MLVs under similar conditions seen in Figure 6-3 B. However, more data points above 43oC are needed to fully characterize this thermotropic phase transition. It is also interesting to note th at as the temperature approaches Tm the values of the CSA span for the PC head group in DPPC/BMP (5 mo l%) MLVs appear to rapidly approach -44 ppm. Recall the linear relationship between POPC span and mole fraction BMP obtained previously in Chapter 5 for POPC/BMP mixed MLVs; the pr edicted span for the POPC CSA span for BMP of 96

PAGE 97

5 mol% is -45 ppm. However, additional data points at temperatures above 43 oC are needed to confirm this limiting value. Further investigat ions at temperatures above the main phase transition and with mole fractions of BMP up to 0.3 would also substantiate the linear relationship between the 31P chemical shift span and mol fraction of negative amphiphile obtained for POPC/BMP mixed MLVs. -50-40-30-20-1001020304050 Frequency (kHz) -50-40-30-20-1001020304050 Frequency (kHz)A) B)T oC35 36 37 38 39 40 41 42 43 -50-40-30-20-1001020304050 Frequency (kHz) -50-40-30-20-1001020304050 Frequency (kHz)A) B)T oC35 36 37 38 39 40 41 42 43 35 36 37 38 39 40 41 42 43 Figure 6-1 2H NMR spectra of A) DPPC and B) D PPC/BMP (5 mol%) mixed MLVs showing the temperature dependence of the phase transition. Lipid samples (275 mM total lipid) were hydrated in a 5 mM HE PES, 0.1 mM EDTA, at pH 7.4. -50-40-30-20-1001020304050 Frequency (kHz) -50-40-30-20-1001020304050 Frequency (kHz)35 36 37 38 39 40 41 42 43T oCA) B) -50-40-30-20-1001020304050 Frequency (kHz) -50-40-30-20-1001020304050 Frequency (kHz)35 36 37 38 39 40 41 42 43T oC -50-40-30-20-1001020304050 Frequency (kHz) -50-40-30-20-1001020304050 Frequency (kHz)35 36 37 38 39 40 41 42 43 35 36 37 38 39 40 41 42 43T oCA) B) Figure 6-2. 2H NMR spectra of A) DPPC MLVs and B) DPPC/BMP (5 mol%) mixed MLVs showing the temperature dependence of the phase transition. Li pid samples (275 mM total lipid) were hydrated in a 5 mM HEPES, 100 mM NaCl, at pH 7.4. 97

PAGE 98

3435363738394041424344 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 4.2 Residual Terminal Methyl Splitting (kHz)T oC3435363738394041424344 0.010 0.011 0.012 0.013 0.014 0.015 Orderer Parameter (Stotal)T oC3435363738394041424344 0.010 0.011 0.012 0.013 0.014 0.015 Orderer Parameter (Stotal)T oCA) B) C) D)3435363738394041424344 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 4.2 Residual Terminal Methyl Splitting (kHz)T oC3435363738394041424344 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 4.2 Residual Terminal Methyl Splitting (kHz)T oC3435363738394041424344 0.010 0.011 0.012 0.013 0.014 0.015 Orderer Parameter (Stotal)T oC3435363738394041424344 0.010 0.011 0.012 0.013 0.014 0.015 Orderer Parameter (Stotal)T oC3435363738394041424344 0.010 0.011 0.012 0.013 0.014 0.015 Orderer Parameter (Stotal)T oC3435363738394041424344 0.010 0.011 0.012 0.013 0.014 0.015 Orderer Parameter (Stotal)T oCA) B) C) D)3435363738394041424344 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 4.2 Residual Terminal Methyl Splitting (kHz)T oC Figure 6-3. Total order parameter and residual quadrupolar splittings of terminal methyl groups as a function of temperature: A) and C) pure DPPC MLVs ( ) and DPPC/BMP mixed MLVs ( ) in 5 mM HEPES, 0.1 mM EDTA, t pH 7.4; B) and D) pure DPPC MLVs ( ) and DPPC/BMP mixed MLVs ( ) in 5 mM NaHEPES, 0.1 mM EDTA, 100 mM NaCl at pH 7.4. The error bars are estimated as approximately the FWHM ( 300 Hz) of the terminal methyl horn. 98

PAGE 99

25 36 38 40 42 44 50 35 37 39 41 43 45 T oC 20 G 25 36 38 40 42 44 50 35 37 39 41 43 45 35 37 39 41 43 45 T oC 20 G 20 G Figure 6-4. cw-EPR spectra of DPPC MLVs with 16-DOXYL spin probe (1 mol%) as function of temperature: 100mM DPPC, 5mM HE PES, 0.1 mM EDTA, at pH 7.4. 25 36 38 40 42 44 50 35 37 39 41 43 45 T oC 20 G 25 36 38 40 42 44 50 35 37 39 41 43 45 35 37 39 41 43 45 T oC 20 G 20 G Figure 6-5. cw-EPR spectra of DPPC MLVs with 16-DOXYL spin probe (1 mol%) as function of temperature: 100mM DPPC, 5mM HEPES, 100 mM NaCl, 0.1 mM EDTA, at pH 7.4. 99

PAGE 100

253035404550 1.5 2.0 2.5 3.0 3.5 4.0 Hpp (G)Temperature oC253035404550 140 150 160 170 180 190 200 Second Moment (M2)Temperature oC253035404550 0.05 0.10 0.15 0.20 0.25 0.30 0.35 Order Parameter (Si)Temperature oC253035404550 1.5 2.0 2.5 3.0 3.5 4.0 Hpp (G)Temperature oC242628303234363840424446485052 -0.35 -0.30 -0.25 -0.20 -0.15 -0.10 -0.05 0.00 0.05 Amplitude (a.u.)A) B) D) C) E)253035404550 1.5 2.0 2.5 3.0 3.5 4.0 Hpp (G)Temperature oC253035404550 140 150 160 170 180 190 200 Second Moment (M2)Temperature oC253035404550 0.05 0.10 0.15 0.20 0.25 0.30 0.35 Order Parameter (Si)Temperature oC253035404550 1.5 2.0 2.5 3.0 3.5 4.0 Hpp (G)Temperature oC242628303234363840424446485052 -0.35 -0.30 -0.25 -0.20 -0.15 -0.10 -0.05 0.00 0.05 Amplitude (a.u.)253035404550 1.5 2.0 2.5 3.0 3.5 4.0 Hpp (G)Temperature oC253035404550 140 150 160 170 180 190 200 Second Moment (M2)Temperature oC253035404550 0.05 0.10 0.15 0.20 0.25 0.30 0.35 Order Parameter (Si)Temperature oC253035404550 1.5 2.0 2.5 3.0 3.5 4.0 Hpp (G)Temperature oC242628303234363840424446485052 -0.35 -0.30 -0.25 -0.20 -0.15 -0.10 -0.05 0.00 0.05 Amplitude (a.u.)A) B) D) C) E) Figure 6-6. Various spectral parameters of 16-DOXYL labeled lipid incorporat ed into DPPC MLVs as a function of temperature: A) peak-to-peak wi dth; B) second moment; C) order parameter; D) first de rivative of sigmoidal fit to Hpp(T) (low ionic strength); and E) Hpp(T) with sigmoidal f it (dashed line) to Hpp(T) (low ionic strength) MLVs in 5mM HEPES, 0.1 mM EDTA, at pH 7.4 ( ); MLVs in 5mM HEPES, 0.1 mM EDTA, 100 mM NaCl at pH 7.4 ( ). 100

PAGE 101

25 36 38 40 42 44 50 T oC 35 37 39 41 43 45 20 G 25 36 38 40 42 44 50 T oC 35 37 39 41 43 45 25 36 38 40 42 44 50 T oC 35 37 39 41 43 45 20 G 20 G Figure 6-7. DPPC/BMP MLVs with 16-DOXYL spin probe as function of temperature at low ionic strength: sample contains 95mM DPPC, 5mM BMP18:1 5mM HEPES, 0.1 mM EDTA pH 7.4. 101

PAGE 102

25 36 38 40 42 44 50 T oC 35 37 39 41 43 45 20 G 25 36 38 40 42 44 50 T oC 35 37 39 41 43 45 25 36 38 40 42 44 50 T oC 35 37 39 41 43 45 20 G 20 G Figure 6-8. DPPC/BMP MLVs with 16-DOXYL sp in probe as function of temperature (near biological ionic strength) 95mM DPPC, 5mM BMP18:1 5mM HEPES, 0.1 mM EDTA pH 7.4. 102

PAGE 103

253035404550 1.5 2.0 2.5 3.0 3.5 4.0 Hpp (G)Temperature oC253035404550 140 150 160 170 180 190 200 Second Moment (M2)Temperature oC253035404550 0.05 0.10 0.15 0.20 0.25 0.30 0.35 Order Parameter (Si)Temperature oC253035404550 1.5 2.0 2.5 3.0 3.5 4.0 Hpp (G)Temperature oC242628303234363840424446485052 -0.45 -0.40 -0.35 -0.30 -0.25 -0.20 -0.15 -0.10 -0.05 0.00 0.05 Amplitude (a.u.)A) B) D) C) E)253035404550 1.5 2.0 2.5 3.0 3.5 4.0 Hpp (G)Temperature oC253035404550 140 150 160 170 180 190 200 Second Moment (M2)Temperature oC253035404550 0.05 0.10 0.15 0.20 0.25 0.30 0.35 Order Parameter (Si)Temperature oC253035404550 1.5 2.0 2.5 3.0 3.5 4.0 Hpp (G)Temperature oC242628303234363840424446485052 -0.45 -0.40 -0.35 -0.30 -0.25 -0.20 -0.15 -0.10 -0.05 0.00 0.05 Amplitude (a.u.)253035404550 1.5 2.0 2.5 3.0 3.5 4.0 Hpp (G)Temperature oC253035404550 140 150 160 170 180 190 200 Second Moment (M2)Temperature oC253035404550 0.05 0.10 0.15 0.20 0.25 0.30 0.35 Order Parameter (Si)Temperature oC253035404550 1.5 2.0 2.5 3.0 3.5 4.0 Hpp (G)Temperature oC242628303234363840424446485052 -0.45 -0.40 -0.35 -0.30 -0.25 -0.20 -0.15 -0.10 -0.05 0.00 0.05 Amplitude (a.u.)A) B) D) C) E) Figure 6-9. Various spectral parameters of 16-DOXYL labe led lipid incorporated into DPPC/BMP (5 mol%) mixed MLVs as a func tion of temperature: A) peak-to-peak width; B) second moment; C) order parameter; D) first derivative of sigmoidal fit to Hpp(T) (low ionic strength); and E) Hpp(T) with sigmoidal fit (dashed line) to Hpp(T) (low ionic strength) MLVs in 5m M HEPES, 0.1 mM EDTA, at pH 7.4 ( ); MLVs in 5mM HEPES, 0.1 mM ED TA, 100 mM NaCl at pH 7.4 ( ). 103

PAGE 104

253035404550 1.5 2.0 2.5 3.0 3.5 4.0 Hpp (G)Temperature oC253035404550 0.05 0.10 0.15 0.20 0.25 0.30 0.35 Order Parameter (Si)Temperature oC253035404550 1.5 2.0 2.5 3.0 3.5 4.0 Hpp (G)Temperature oC253035404550 0.05 0.10 0.15 0.20 0.25 0.30 0.35 Order Parameter (Si)Temperature oCA) D) B) C)253035404550 1.5 2.0 2.5 3.0 3.5 4.0 Hpp (G)Temperature oC253035404550 0.05 0.10 0.15 0.20 0.25 0.30 0.35 Order Parameter (Si)Temperature oC253035404550 1.5 2.0 2.5 3.0 3.5 4.0 Hpp (G)Temperature oC253035404550 0.05 0.10 0.15 0.20 0.25 0.30 0.35 Order Parameter (Si)Temperature oC253035404550 1.5 2.0 2.5 3.0 3.5 4.0 Hpp (G)Temperature oC253035404550 0.05 0.10 0.15 0.20 0.25 0.30 0.35 Order Parameter (Si)Temperature oC253035404550 1.5 2.0 2.5 3.0 3.5 4.0 Hpp (G)Temperature oC253035404550 0.05 0.10 0.15 0.20 0.25 0.30 0.35 Order Parameter (Si)Temperature oCA) D) B) C) Figure 6-10. Various spectral parameters of 16DOXYL labeled lipid inco rporated into DPPC MLVs and DPPC/BMP (5 mol%) mixed MLVs as a function of temperature: A) peak-to-peak width of DPPC MLVs ( ) and DPPC/BMP MLVs ( ) in 5mM HEPES, 0.1 mM EDTA, at pH 7.4 ; B) order parameter of DPPC MLVs ( ) and DPPC/BMP MLVs ( ) in 5mM HEPES, 0.1 mM EDTA, at pH 7.4; C) peak-to-peak width of DPPC MLVs ( ) and DPPC/BMP MLVs () in 5mM HEPES, 0.1 mM EDTA, ,100 mM NaCl at pH 7.4; D) order parameter of DPPC MLVs ( ) and DPPC/BMP MLVs ( ) in 5mM HEPES, 0.1 mM EDTA, ,100 mM NaCl at pH 7.4. 104

PAGE 105

-60-40-200204060 ppm-60-40-200204060 ppmB) A) C) D)-60-40-200204060 ppm-60-40-200204060 ppmT = 43 oC T = 35 oC -60-40-200204060 ppm-60-40-200204060 ppmB) A) C) D)-60-40-200204060 ppm-60-40-200204060 ppmT = 43 oC T = 43 oC T = 35 oC T = 35 oC Figure 6-11. 31P NMR chemical shift of single component DPPC MLVs (solid line), DPPC/BMP (5 mol%) MLVs (dashed line) : A) and C) 5 mM HEPES, 0.1 mM EDTA, at pH 7.4, 5; B) and D) 5 mM HEPES, 0.1 mM EDTA 100 mM NaCl, at pH 7.4. 105

PAGE 106

3435363738394041424344 -42 -44 -46 -48 -50 -52 -54 -56 (ppm)T oC3435363738394041424344 -42 -44 -46 -48 -50 -52 -54 -56 (ppm)T oC3435363738394041424344 -42 -44 -46 -48 -50 -52 -54 -56 (ppm)T oC3435363738394041424344 -42 -44 -46 -48 -50 -52 -54 -56 (ppm)T oCA) B) C) D)3435363738394041424344 -42 -44 -46 -48 -50 -52 -54 -56 (ppm)T oC3435363738394041424344 -42 -44 -46 -48 -50 -52 -54 -56 (ppm)T oC3435363738394041424344 -42 -44 -46 -48 -50 -52 -54 -56 (ppm)T oC3435363738394041424344 -42 -44 -46 -48 -50 -52 -54 -56 (ppm)T oC3435363738394041424344 -42 -44 -46 -48 -50 -52 -54 -56 (ppm)T oC3435363738394041424344 -42 -44 -46 -48 -50 -52 -54 -56 (ppm)T oC3435363738394041424344 -42 -44 -46 -48 -50 -52 -54 -56 (ppm)T oC3435363738394041424344 -42 -44 -46 -48 -50 -52 -54 -56 (ppm)T oCA) B) A) B) C) D) Figure 6-12. 31P CSA span of DPPC MLVs and DPPC/BMP (5 mol%) mixed MLVs as a function of temperature: A) DPPC low ( ) and high ( ) ionic strength; B) DPPC ( ) and mixed DPPC/BMP (5 mol%) ( ) both in low ionic strength buffer; C) DPPC ( ) and mixed DPPC/BMP (5 mol%) ( ) both in high ionic stre ngth buffer; D) Mixed DPPC/BMP (5 mol%)low ( ) and high () ionic strength buffer Error in span = 1 ppm. 106

PAGE 107

Table 6-1 Total order parameter values for te rminal methyls of d62-DPPC in DPPC MLVs. 0.012 0.012 0.012 0.013 0.013 0.014 0.015 0.015 N/A Order Parameter (Stotal) 5 mMNa+43 42 41 40 39 38 37 36 35 oC 0.011 0.011 0.012 0.012 0.012 0.013 0.014 N/A N/A Order Parameter (Stotal) 100 mMNa+ 0.012 0.012 0.012 0.013 0.013 0.014 0.015 0.015 N/A Order Parameter (Stotal) 5 mMNa+43 42 41 40 39 38 37 36 35 oC 0.011 0.011 0.012 0.012 0.012 0.013 0.014 N/A N/A Order Parameter (Stotal) 100 mMNa+ 107

PAGE 108

Table 6-2 Total order parameter values for te rminal methyls of d62-DPPC in DPPC/BMP (5mol%) mixed MLVs. 0.011 0.012 0.012 0.012 0.012 0.012 0.012 N/A N/A Order Parameter (Stotal) 5 mMNa+43 42 41 40 39 38 37 36 35 oC 0.012 0.011 0.012 0.012 0.013 0.013 0.013 0.013 0.013 Order Parameter (Stotal) 100 mMNa+ 0.011 0.012 0.012 0.012 0.012 0.012 0.012 N/A N/A Order Parameter (Stotal) 5 mMNa+43 42 41 40 39 38 37 36 35 oC 0.012 0.011 0.012 0.012 0.013 0.013 0.013 0.013 0.013 Order Parameter (Stotal) 100 mMNa+ 108

PAGE 109

Table 6-3 Parameters defining order of the 16DOXYL nitroxide spin-probe in DPPC MLVs in 5 mM Na+ buffer. 0.10 145 1.8 42 0.10 144 1.8 43 0.10 142 1.8 44 0.010 142 1.8 45 141 143 145 1434 144 144 163 175 191 2ndMoment 0.095 0.10 0.10 0.11 0.11 0.12 0.18 0.20 0.34 Si( ) 1.9 37.5 1.9 38 1.9 39 1.8 40 1.8 41 1.8 2.4 2.6 3.8 Hpp25 50 36 35 T oC 0.10 145 1.8 42 0.10 144 1.8 43 0.10 142 1.8 44 0.010 142 1.8 45 141 143 145 1434 144 144 163 175 191 2ndMoment 0.095 0.10 0.10 0.11 0.11 0.12 0.18 0.20 0.34 Si( ) 1.9 37.5 1.9 38 1.9 39 1.8 40 1.8 41 1.8 2.4 2.6 3.8 Hpp25 50 36 35 T oC 109

PAGE 110

Table 6-4 Parameters defining order of the 16DOXYL nitroxide spin-probe in DPPC MLVs in 105 mM Na+ buffer. 0.10 142 1.6 42 0.10 144 1.7 43 0.10 144 1.6 44 0.10 142 1.6 45 141 142 141 144 143 143 167 173 193 2ndMoment 0.095 0.10 0.10 0.10 0.11 0.11 0.20 0.23 0.35 Si( ) 1.6 37 1.7 38 1.6 39 1.7 40 1.6 41 1.6 2.1 2.4 3.3 Hpp25 50 36 35 T oC 0.10 142 1.6 42 0.10 144 1.7 43 0.10 144 1.6 44 0.10 142 1.6 45 141 142 141 144 143 143 167 173 193 2ndMoment 0.095 0.10 0.10 0.10 0.11 0.11 0.20 0.23 0.35 Si( ) 1.6 37 1.7 38 1.6 39 1.7 40 1.6 41 1.6 2.1 2.4 3.3 Hpp25 50 36 35 T oC 110

PAGE 111

Table 6-5 Pre-transition temp eratures obtained from 16-DOXYL labeled lipid in various MLV lamellar structures. DPPC/BMP (5 mol%) (105 mMNa+) DPPC/BMP (5 mol%) (5 mMNa+) DPPC (105 mMNa+) DPPC (5 mMNa+) 33 oC 35 oC 36 oC 36 oC 2ndMoment 29 oC 35 oC 35 oC 35 oC Si( ) 34 oC 35 oC 35 oC 34 oC Hpp DPPC/BMP (5 mol%) (105 mMNa+) DPPC/BMP (5 mol%) (5 mMNa+) DPPC (105 mMNa+) DPPC (5 mMNa+) 33 oC 35 oC 36 oC 36 oC 2ndMoment 29 oC 35 oC 35 oC 35 oC Si( ) 34 oC 35 oC 35 oC 34 oC Hpp 111

PAGE 112

Table 6-6 Parameters defining order of the 16-DOXYL nitroxide spin-probe in DPPC/BMP (5 mol%) mixed MLVs in 5 mM Na+ buffer. 0.11 139 1.8 42 0.11 139 1.8 43 0.11 138 1.8 44 0.11 139 1.8 45 137 138 143 142 143 142 158 168 187 2ndMoment 0.09 0.11 0.11 0.11 0.11 0.11 0.14 0.21 0.30 Si( ) 1.9 37 1.9 38 1.9 39 1.8 40 1.8 41 1.8 2.2 2.5 3.2 Hpp25 50 36 35 T oC 0.11 139 1.8 42 0.11 139 1.8 43 0.11 138 1.8 44 0.11 139 1.8 45 137 138 143 142 143 142 158 168 187 2ndMoment 0.09 0.11 0.11 0.11 0.11 0.11 0.14 0.21 0.30 Si( ) 1.9 37 1.9 38 1.9 39 1.8 40 1.8 41 1.8 2.2 2.5 3.2 Hpp25 50 36 35 T oC 112

PAGE 113

Table 6-7 Parameters defining order of the 16-DOXYL nitroxide spin-probe in DPPC/BMP (5 mol%) mixed MLVs in 105 mM Na+ buffer. 0.11 139 1.8 42 0.11 139 1.8 43 0.11 138 1.8 44 0.11 139 1.8 45 137 138 143 142 143 142 158 168 187 2ndMoment 0.090 0.11 0.11 0.11 0.11 0.11 0.14 0.21 0.30 Si( ) 1.9 37 1.9 38 1.9 39 1.8 40 1.8 41 1.8 2.2 2.5 3.2 Hpp25 50 36 35 T oC 0.11 139 1.8 42 0.11 139 1.8 43 0.11 138 1.8 44 0.11 139 1.8 45 137 138 143 142 143 142 158 168 187 2ndMoment 0.090 0.11 0.11 0.11 0.11 0.11 0.14 0.21 0.30 Si( ) 1.9 37 1.9 38 1.9 39 1.8 40 1.8 41 1.8 2.2 2.5 3.2 Hpp25 50 36 35 T oC 113

PAGE 114

114 Table 6-8 Span values for DPPC MLVs and DPPC/BMP (5 mol%) mixed MLVs. -53 -54 -54 -53 -53 -54 -54 -54 -55 (ppm) DPPC 105 mMNa+-44 -45 -47 -49 -49 -50 -51 N/A -51 (ppm) DPPC/BMP 5 mMNa+-51 -52 -52 -53 -52 -52 -52 -54 -55 (ppm) DPPC 5 mMNa+43 42 41 40 39 38 37 36 35 oC -44 -44 -43 -44 -45 -46 -48 -47 -48 (ppm) DPPC/BMP 105 mMNa+ -53 -54 -54 -53 -53 -54 -54 -54 -55 (ppm) DPPC 105 mMNa+-44 -45 -47 -49 -49 -50 -51 N/A -51 (ppm) DPPC/BMP 5 mMNa+-51 -52 -52 -53 -52 -52 -52 -54 -55 (ppm) DPPC 5 mMNa+43 42 41 40 39 38 37 36 35 oC -44 -44 -43 -44 -45 -46 -48 -47 -48 (ppm) DPPC/BMP 105 mMNa+

PAGE 115

CHAPTER 7 SUMMARY AND CONCLUSIONS The question posed in the beginning of this dissertation was What effect does BMP have on model membrane morphology? In order to answer this question membrane solubilization was first investigated using SDS detergent, si nce our original hypothesi s was that BMP might have detergent properties, such as micelle formation and bilayer solubilization. However, we have presented data that c ontradicts our original hypothe sis and further corroborates experimental results obtained by others that BMP self-assembles into a lamellar aggregate structure in an aqueous environm ent. Furthermore, our solubili zation results indicate that the acyl chain packing motifs and the order parameters of BMP18:1 are similar to those found in POPC MLVs in that BMP is solubilized by SDS in a similar fashion. Also, the line shapes obtained from the 5-DOXYL labeled lipid are almost identical in either single component POPC or BMP LUVs indicating that the acyl chains pack in a similar manner. Initial characterization of the interaction of the BMP lipid incorporated into model PC membrane structures indicated that negatively charged BMP causes choline head groups to tilt away from the bilayer normal as a function of BMP concentration as seen in other negatively charged amphiphile/PC membrane mixtures. This is a direct indication that BMP modulates the surface charge and molecular inte ractions in the vicinity of the polar region of these mixed model membrane systems. Considering the results presented in this body of work there are some key experiments that should be revisited in order to further our unders tanding of lipid membrane structures and give us greater detail regardin g the current observations. 2H NMR experiments should be repeated fo r both DPPC and DPPC/BMP MLVs as a function of temperature with a larger deuterated lipid compone nt. Increasing the number of 115

PAGE 116

deuterium-enriched lipid molecules will increase th e signal to noise ratio, allowing the spectra to be dePaked so the full order parameter profiles for each system can be reported. Also, spectra should be collected at a temperature well above Tm for DPPC, e.g. 550C, to ensure that a single LC component is present. This is necessa ry for proper dePaking and peak assignment. Data points in the temperature region between 25 and 35oC should be acquired for all of the EPR experiments involving the 16-DOXYL spin -labeled lipids. Thes e data points will be useful in obtaining a better fit of the sigmoi dal line shape and thus a better estimate of the thermotropic phase transition temperatures. Also, selected data points should be recollected to establish the error in the measurements using the 16-DOXYL probe. Each EPR solubilization experiment should be repeated initially using only the 5-DOXYL labeled lipid. The preliminary results of these experiments seem to be very significant and the data should be obtained in trip licate and extended to both smalle r and larger SDS/Lipid ratios. Other useful experiments would include systematic studies of size distributions between POPC and POPC/BMP mixed vesicles and MLVs using light scattering techniques, confirmation of mesophase by cryogenic electron microscopy, and a comparison of these and results obtained previously with those obtained from mixed MLVs and LUVs containing negatively charged phosphatidylinositol. 116

PAGE 117

117 LIST OF REFERENCES 1. P. L. Yeagle, Ed., The Structure of Biological Membranes (CRC Press 2004). 2. R. B. Gennis, C. R. Cantor, Ed., Biomembranes Molecular Structure and Function. (Springer-Verlag 1989). 3. D. L. Nelson, M. M. Cox, Lehninger Principles of Biochemistry, Fourth Edition. (W. H. Freeman 2004). 4. S. M. Gruner, P. R. Cullis, M. J. Hope, C. P. Tilcock, Annu. Rev. Biophy.s Biophys. Chem. 14, 211 (1985). 5. E. Gorter, F. Grendel, Journal of Experimental Medicine 41 (1925). 6. J. R. Silvius, M. Lyons, P. L. Yeagle, T. J. O'Leary, Biochemistry 24, 5388 (Sep. 24, 1985). 7. H. Schindler, J. Seelig, Biochemistry 14, 2283 (Jun. 3, 1975). 8. M. J. Janiak, D. M. Small, G. G. Shipley, Biochemistry 15, 4575 (Oct. 19, 1976). 9. A. M. Thayer, S. J. Kohler, Biochemistry 20, 6831 (Nov 24. 1981). 10. W. L. Hubbell, H. M. McConnell, Pro.c Nat.l Acad. Sci. U S A 64, 20 (Sep. 1969). 11. J. Seelig, Q. Rev. Biophys. 10, 353 (Aug.1977). 12. G. G. Shipley, L. S. Avecilla, D. M. Small, J. Lipid Res. 15, 124 (Mar. 1974). 13. J. R. Trudell, W. L. Hubbell, E. N. Cohen, Annals New York Academy of Science, 222, 530 (Dec. 1973). 14. W. L. Hubbell, H. M. McConnell, J. Am. Chem. Soc. 93, 314 (Jan. 27, 1971). 15. J. M. Boggs, J. C. Hsia, Proc. Natl. Acad. Sci. U S A 70, 1406 (May. 1973). 16. J. Herzfeld, R. G. Griffin, R. A. Haberkorn, Biochemistry 17, 2711 (1978). 17. A. Seelig, J. Seelig, Biochemistry 13, 4839 (Nov. 5, 1974). 18. M. Auger, Biophys. Chem. 68, 233 (Oct. 1997). 19. D. R. Body, G. M. Gray, Chemistry and Physics of Lipids 1, 254 (1967).

PAGE 118

20. T. Kolter, K. Sandhoff, Annu. Rev. Cell Dev. Biol. 21, 81 (2005). 21. T. Kobayashi, K. Startchev, A. J. Whitney, J. Gruenber, Biol. Chem. 382, 483 (Mar. 2001). 22. T. Kobayashi, E. Stang, K.S. Fang, P. de Moerloose, R. G. Parton, J. Gruenberg, Nature 392, 193 (Mar. 12, 1998). 23. T. Kobayashi, M. Beuchat, J. H.Chevallier, A. Makino, N. Mayran, J.Escola, C, M.Lebrand, P. Cosson, J. Gruenberg, J. Biol. Chem. 277, 32157 (Aug. 30, 2002). 24. B. M. Mullock, R. H. Hinton, J. V. Peppard, J. W. Slot, J. P. Luzio, Cell Biochem. Funct. 5, 235 (Oct. 1987). 25. T. Kobayashi, F. Gu, J. Gruenberg, Semin. Cell Dev. Biol. 9, 517 (Oct, 1998). 26. H. Matsuo,J. Chevallier, N. Mayran, I. Le Blanc, C. Ferguson, J. Faure, N. S. Blanc, S. Matile, J. Dubochet, R. Sadoul, R. G. Parton, F. Vilbois, J.Gruenberg, Science 303, 531 (Jan. 23, 2004). 27. J. Brotherus, O. Renkonen, W. Fischer, J. Herrmann, Chem. Phys. Lipids 13, 178 (Oct. 1974). 28. T. Hayakawa, Hirano, Y., A. Makino, S. Michaud, M. Lagarde, J. F. Pageaux, A. Doutheau, K.Ito, T. Fujisawa, H. Takahashi, T.. Kobayashi, Biochemistry 45, 9198 (Aug. 1, 2006). 29. J. M. Holopainen Soderlund, T., J. M.Al akoskela, M. Saily, O. Eriksson,Kinnunen, P. K., Chem. Phys. Lipids 133, 51 (Jan. 2005). 30. N. Werth, C. G. Schuette, G. Wilkening, T. Lemm, K. Sandhoff, J. Biol. Chem. 276, 12685 (Apr. 20, 2001). 31. C. D. Stubbs, S. J. Slater, Chem. Phys. Lipids 81, 185 (Jul.15, 1996). 32. I. Dikic, Bioessays 26, 604 (Jun. 2004). 33. T. Kolter, K. Sandhoff, Angew. Chem. Int. Ed. 38, 1532 (1999). 34. K. Sandhoff, T. Kolter, Trends Cell Biol. 6, 98 (Mar. 1996). 35. F. R. Maxfield, T. E. McGraw, Nat. Rev. Mol. Cell Biol. 5, 121 (Feb. 2004). 36. M. Koval, R. E. Pagano, J. Cell Biol. 108, 2169 (Jun. 1989). 118

PAGE 119

119 37. M. Koval, R. E. Pagano, J. Cell Biol. 111, 429 (Aug. 1990). 38. K. Laulagnier Motta, C. S. Hamdi, S. Roy, F. Fauvelle, J. F. Pageaux, T. Kobayashi, J. P. Salles, B. Perret, C. Bonnerot, M. Record, Biochemistry Journal, 161 (2004). 39. R. D. Singh Liu, Y. C. L. Wheatley, E. L. Holicky, A. Makino, D. L. Marks, T. Kobayashi, G. Subramaniam, R. Bittman,Pagano, R. E., J. Biol. Chem. 281, 30660 (Oct.13, 2006). 40. F. Aussenac, M. Tavares, E. J. Dufourc, Biochemistry 42, 1383 (Feb. 18, 2003). 41. R. E. Glover, R. R. Smith, M. V. Jones, S. K. Jackson, C. C. Rowlands, FEMS Microbiol. Lett. 177, 57 (Aug. 1, 1999). 42. N. Van Eps, W. M. Oldham H. E. Hamm, W. L. Hubbell, Proc. Natl. Acad. Sci. U S A 103, 16194 (Oct. 31, 2006). 43. P. J. Davis, B. D. Fleming, K. P. Coolbear, K. M. Keough, Biochemistry 20, 3633 (Jun. 9, 1981). 44. K. Gawrisch. Buffer preparations in D2O (Private Communication 2007). 45. H. Warner, T. Y. Toribara, P. S. Chen, Analytical Chemistry 28, 1756 (1956). 46. K. Gawrisch. The Dynamics of Lipid Membranes (Private Communication 2007). 47. M. D. King, D. Marsh, Biochemistry 26, 1224 (Mar. 10, 1987). 48. D. Marsh, Journal of Magnetic Resonance 157, 114 (2002). 49. K. Schorn, D. Marsh, Biochemistry 35, 3831 (Mar. 26, 1996). 50. H. M. McConnell, B. G. McFarland, Ann. N.Y. Acad. Sci. 195, 207 (Jun. 20, 1972). 51. H. M. McConnell, K. L. Wright, B. G. McFarland, Biochem. Biophys. Res. Commun. 47, 273 (Apr. 14, 1972). 52. W. L. Hubbell, H. M. McConnell, Proc. Natl. Acad. Sci. U S A 61, 12 (Sep. 1968). 53. W. L. Hubbell, J. C. Metcalfe, S. M. Metcalfe, H. M. McConnell, Biochim. Biophys. Acta 219, 415 (Dec. 1, 1970). 54. W. L. Hubbell, Stanford University (1969). 55. P. Poole, Electron Spin Resonance. (John Wiley and Sons, Inc., 1983).

PAGE 120

56. H. S. McHaourab, M. A. Liet zow, K. Hideg, W. L. Hubbell, Biochemistry 35, 7692 (Jun. 18, 1996). 57. J. H. Davis, Biochim. Biophys. Acta 737, 117 (Mar. 21, 1983). 58. A. Abragam, Principles of Nuclear Magnetism (Clarendon Press, 1961). 59. R. G. Griffin, L. Powers, P. S. Pershan, Biochemistry 17, 2718 (1978). 60. J. S. Santos, The University of Michigan (2001). 61. M. le Maire, P. Champeil, J. V. Moller, Biochim. Biophys. Acta 1508, 86 (Nov. 23, 2000). 62. S. Keller, H. Heerklotz, N. Jahnke, A. Blume, Biophys. J. 90, 4509 (Jun. 15, 2006). 63. A. Tan, A. Ziegler, B. Steinbauer, J. Seelig, Biophys. J. 83, 1547 (Sep. 2002). 64. P. R. Mahji, A. Blume, Langmuir 17, 3844 (2001). 65. N. Deo, P. Somasundaran, K. Subr amanyan, K. P. Ananthapadmanabhan, Journal of Colloid and Interface Science 256, 100 (2002). 66. R. G. Griffin, J. Am. Chem. Soc. 98, 851 (Feb. 4, 1976). 67. J. A. Berden, P. R. Cullis, D. I. Hoult, A. C. McLaughlin, G. K.Radda,Richards, R. E., FEBS Lett. 46, 55 (Sep. 15, 1974). 68. P. G. Scherer, J. Seelig, Biochemistry 28, 7720 (Sep. 19, 1989). 69. L. K. Tamm, J. Seelig, Biochemistry 22, 1474 (1983). 70. S. Abu-Baker, X. Qi, G. A. Lorigan, Biophysical Journal 93, 3480 (2007). 71. D. Massiot, F. Fayon, M. Capr on, I. King, S. Le Calve, B. Alonso, J-O Durand, B. Bujoli, Z. Gan, G. Hoatson, Magnetic Resonance in Chemistry 40, 70 (2002). 72. J. Seelig, H. Gally, Biochemistry 15, 5199 (Nov. 30, 1976). 73. W. Neiderberger, J. Seelig, J. Am. Chem. Soc. 98, 3704 (Jun. 9, 1976). 74. J. Seelig, P. M. Macdonald, P. G. Scherer, Biochemistry 26, 7535 (Dec. 1, 1987). 75. J. H. Davis, M. R. Vist, Biochemistry 29, 451 (1990). 76. J. H. Davis, Biochimica et. Biophysicsa Acta 737, 117 (1983). 120

PAGE 121

77. M. F. Brown, S. Lope-Piedrafita, G. V. Martinez, H. I. Petrache, in Modern Magnetic Resonance (Springer, 2006). 78. E. Sternin, in Membrane Lipid Protocols (Humana Press, Inc., 2005). 79. M. Bloom, J. H. Davis, A. L. MacKay, Chem. Phys. Lett. 80, 198 (1981). 80. J. Dey, R. R. Nayak, S. Roy, Polymer 46, 12401 (2005). 81. T. M. Bayerl, T. Brumm, H. Reinl, Biophysics Journal 61, 1025 (1992). 82. M. Korkmaz, C. Aydas, Journal of the Korean Physical Society 45, 975 (2004). 83. D. G. Cameron, H. L. Casal, H. H. Mantsch, Biochemistry 19, 3665 (Aug. 5, 1980). 84. T. P. McMullen, R. N. McElhaney, Biochim. Biophys. Acta 1234, 90 (Mar. 8, 1995). 85. R. G. Griffin, M. J. Ruocco, D. J. Siminovitch, Biochemistry 24, 2406 (1985). 121

PAGE 122

122 BIOGRAPHICAL SKETCH Chad Mair was born on December 29, 1975 in Ev ansville, Indiana. He spent his childhood years in Owensville, Indiana where he attended Owensville Elementary. After completing secondary education, at Gibson Southern High School, he began his undergraduate education in Evansville at the University of Southern Indiana. Chad received a Bachel or of Science degree in chemistry from USI in 2001. His graduate educati on started at the University of Florida in the area of physical chemistry under th e supervision of Dr. Valeria Kl eiman. He spent four years working in the field of ultrafast laser spect roscopy on hyperconjugated polymers. Chad then began working on magnetic resonance spectrosc opy of model biological membranes under the supervision of Dr. Gail E. Fanucci.