n . raga .
W?

It: .. 5’...
HI ..
1%me .x,
,1. w «fl . FM“.
4w. 3.4.

“fa.

vfi

. s I‘

15.15.. i
5.1
_ s

.5 5...-
it:
in is
it»)...

I.

:

in...:€\:.
xix-1%..
.1. a:
.l I...
9“"... .924.
. :1.
ask” r00:
1.. .3.
.5.-
:3:
.9
:

:2!!-

.3291?

a)?! I |
.9: ‘0 v 9.

21:?! inn.

5... 1111...“: a...

t.
.Nt:
11.2. .i...
'1...'l;.l
,s‘ T 0-: ‘

.\ yréov .

 

1505i

 

LIBRARY
Michigan State

U!!!“ lnrfi:hl
I IIVCI OIL]

 

 

 

This is to certify that the
dissertation entitled

SOLID-STATE NUCLEAR MAGNETIC RESONANCE
STUDIES OF THE STRUCTURE AND MEMBRANE
INSERTION OF HIV FUSION PEPTIDES

presented by

Wei Qiang

has been accepted towards futfillment
of the requirements for the

Ph.D. degree in Department of Chemistry

 

 

M VII/00};

Mbjor Professor’s Signature

W 2-6ch11)?

Date

MSU is an Afiinnative ActiorVEquaI Opportunity Employer

 

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

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 KslProj/AchresICIRC/DateDueindd

 

SOLID-STATE NUCLEAR MAGNETIC RESONANCE STUDIES OF THE
STRUCTURE AND MEMBRANE INSERTION OF HIV FUSION PEPTIDES

By

Wei Qiang

A DISSERTATION

Submitted to
Michigan State University
In partial fulfillment of the requirements
For the degree of

DOCTOR OF PHILOSOPHY
Chemistry

2009

an e
IHFF

[hfee

ABSTRACT

SOLID-STATE NUCLEAR MAGNETIC RESONANCE STUDIES OF THE
STRUCTURE AND MEMBRANE INSERTION OF HIV FUSION PEPTIDES

By
Wei Qiang

Fusion between the viral and target cell membranes plays an important
role in the infection of human immunodeficiency virus (HIV). The ~20-residue
hydrophobic N-terminal HIV fusion peptide (HFP) catalyzes the membrane fusion
by interacting with the cell membranes. In the present studies, a series of the
variant peptides of HF P were synthesized. The first part of the results describes
an efficient synthetic scheme for HFP oligomers, in particular for HFP trimers
(HFPtr). Compared with previous schemes, the present method shows at least
three-fold increase in the overall yield as well as a great enhancement in purity.

Solid-state nuclear magnetic resonance (NMR) was applied to the
membrane-associated HFP systems because such systems were neither soluble '
nor crystalline. The residue-specific secondary structure was obtained through
13C chemical shift measurements. For HFP monomer (HFPmn), dimer (HFPdm)
and HFPtr, both a-helical and B-strand conformations were observed in the
membrane without cholesterol while only B—strand conformation was detected in
the membrane with cholesterol. For the V2E mutated HFP monomer
(HFPmn_mut), a mixture of a-helix and B—strand were observed in both

membranes. These observations indicated the conformation of HFP was

de;

56C

me'

me:

Witt

dependent on the membrane composition and in particular the existence of
cholesterol; however, there was not an obvious correlation between the
secondary structure and the fusion activity of HFPs.

The tertiary structure of HFPmn associated with cholesterol-containing
membrane was studied using rotational-echo double resonance (REDOR)
method. Two specific anti-parallel B—sheet registries were identified for HFPmn
with overlapping of the N-terminal 16 or 17 residues. 50 -60 % of the membrane-
associated HFPmn adopted these two registries. The study provided an
applicable approach to quantify the registries for the membrane-associated B-
sheet HFPs.

A systematic study of the membrane location of specifically-labeled
residues in HFPmn_mut, HFPmn and HFPtr was conducted using 13C-31P and
13C-‘S’F REDOR approaches in both cholesterol-containing and non-cholesterol
membranes. It was observed that in the cholesterol-containing membranes, the
membrane insertion depth followed the trend HFPmn_mut < HFPmn < HFPtr. In
the non-cholesteroI-containing membranes where the peptides adopted both a-
helical and B—strand conformations, both the membrane insertion depth and the
deeply-inserted population followed the same trend as in cholesterol-containing
membranes regardless of the secondary structures. These results suggested that
there is a positive correlation between the fusion activity of a HFP construct and
the insertion depth and deeply-inserted population of the construct, which may be

independent on the secondary structure of the HFPs.

Dedicated to my beloved wife, Yan Sun

sup
heir

soii

COII

Ga=

me

Fis

Drc

ACKOWNLEDGEMENTS

I would like to at first thank my advisor Dr. David Weliky for his instruction,
support and help during the five years. It is his encouragement and patience that
help me to become interested in and to decide to stick at the area of biological
solid-state Nuclear Magnetic Resonance.

I would like to give my thankfulness to the professors, seniors and
colleagues who are always supportive to my research. Especially I would like to
thank my committee members Dr. John McCracken, Dr. Michael Feig and'Dr.
Gavin Reid for their valuable suggestions on the thesis, my former group
members Dr. Zhaoxiong Zheng (Norm), Dr. Michelle Bodner and Dr. Rong Yang
for their kindness and help when I first joined the group and my current
colleagues Scott Schmick, Dr. Kelly Sackett, Dr. Charles Gabrys, Jaime Curtis-
Fisk, Matt Nethercott and Erica Schwander for their useful discussions about my
projects.

I would also like to thank my family and friends, in particular my parents
for their edification and stimulation not only during my graduate studies, but also
throughout my entire education.

Finally, I will express my greatest appreciation to my beloved wife, Yan
Sun, who is also my lab colleague. It is her support both inside and outside work
that makes me successfully complete my graduate study with much less tough

experience than what I heard at the beginning.

Refe

Char
Back

RESU

TABLE OF CONTENTS

List of Tables ...................................................................................... viii
List of Figures ...................................................................................... ix
List of Symbols and Abbreviations ......................................................... xviii

Chapter I. Introduction
Background ........................................................................................... 1
Reference .......................................................................................... 10

Chapter II. Optimization of the Synthesis of HIV Fusion Peptide Oligomers

Background ....................................................................................... 16
Materials ............................................................................................. 1 7
Results and Discussion
Optimization of Coupling Time ....................................................... 19
Synthesis of HF Pdm ................................................................... 20
Synthesis of HFPtr ...................................................................... 21
Conclusion .......................................................................................... 23
Reference .......................................................................................... 32

Chapter III. Optimization fo Solid-state NMR Experiments
Theoretical Background

Average Hamiltonian Theory (AHT) ................................................ 33

Theory of REDOR Experiments ..................................................... 39

Double CP (DCP) Experiments ...................................................... 44
Experimental Optimization

‘30-3‘P REDOR Experiment .......................................................... 49

13C-19F REDOR Experiment .......................................................... 52

DCP experiment ......................................................................... 53
Reference ........................................................................................... 56

Chapter IV. Secondary Structures of Membrane-Associated HIV Fusion Peptides

Background ........................................................................................ 59
Materials and Methods

NMR Sample Preparation ............................................................ 61

Solid-state NMR Experiments ....................................................... 62
Results and Discussion

Residue Specific Chemical Shifts ................................................... 67

vi

Cont
Reta

Chap
Back
Mate

Resu

Conc
REE:

Chap
Back
IlIaIEI

Resu

Concl
Refin

Chanl
REIen

AFDSr
ApDEr
DDEr

.En

up
€5_
(:1

Cholesterol-dependence of the HF P Conformation ............................ 69

Construct-dependence of the HFP conformation ............................... 71
Conclusion .......................................................................................... 76
Reference .......................................................................................... 77

Chapter V. Tertiary Structure of Membrane-Associated HIV Fusion Peptides

Background ........................................................................................ 81
Materials and Methods
Peptide ..................................................................................... 83
NMR Sample Preparation ............................................................ 83
1"’C-"SN REDOR Experiments and Simulations .................................. 84
Results and Discussion
Anti-parallel B—sheet Registries for HFPmn ....................................... 90
Quantitative Anti-parallel B—sheet Registry Models ............................. 96
Conclusion ....................................................................................... 104
Reference ......................................................................................... 106

Chapter VI. Membrane Insertion of HIV Fusion Peptides

Background ....................................................................................... 1 10
Materials and Methods
Peptides ................................................................................. 113
Lipid Mixing Induced by HFPs ...................................................... 114
Synthesis of the Precursor of 5-19F-DPPC ...................................... 118
Solid-state NMR Samples Preparation and Experiments ................... 119
Natural Abundance Correction for (AS/So)” ............................ , ....... 122
Results and Discussion
Fusion Activities of Different HFP Constructs .................................. 126
Membrane Insertion of TripIy-Iabeled HFPs ..................................... 127
Membrane Insertion of B-strand HFPs ........................................... 139
Membrane Insertion of a-helical HFPs ........................................... 159
Conclusion ......................................................................................... 194
Reference .......................................................................................... 196

Chapter VII. Summary and Future Work

Reference .......................................................................................... 208

Appendix 1. NMR Files Checklist ........................................................... 209

Appendix 2. Trouble Shooting for MAS Probes ......................................... 215

Appendix 3. F MOC Manual Peptide Synthesis .......................................... 219

Appendix 4. Introduction to NMR ............................................................ 225
vii

Tat

Tat

Tab
the

LIST OF TABLES

Table 1 Names and sequences of the HIV fusion peptides. .......................... 17
Table 2 Hamiltonians in REDOR So and S1 spectra. .................................... 39

Table 3 13C chemical shift Assignments for (a) HFPmn in PC:PG, (b) HFPdm in
PC:PG, (c) HFPtr in PC:PG, (d) HFPmn in PC:PG:CHOL, (e) HFPdm in
PC:PG:CHOL, (f) HFPtr in PC:PG:CHOL, (g) HFPmn_mut in PC:PG and (h)
HFPmn_mut in PC:PG:CHOL. ................................................................ 65

Table 4 Peptide sequences and labeling schemes. ..................................... 80

Table 5 The parameters of (a) (AS/So)”"(HFP) and (b) (AS/So)"a(DPPC) used for

the REDOR natural abundance correction. ............................................. 121
Table 6 Fitting parameters for the lipid mixing kinetics at 35°C. ................... 124
Table 7 13CO chemical shifts in PC:PG:CHOL membrane. ......................... 138

Table 8 Best-fit distances and populations for samples in PC:PG:CHOL. ....... 145
Table 9 13CO chemical shifts in PC:PG membrane. .................................. 158
Table 10 13CO-3‘P internuclear distances and populations for HFPmn_mut. ...165
Table 11 The best-fit A and B population for HFPmn and HFPtr in PC:PG. ....174
Table 12 Summary of the fitting for 13c-3‘P and ‘3C-IQF(C16) experiments. ....174

Table 13 Fitting results for the 13C-‘9F(CS) experiments. ............................ 182

viii

to I
port
the

Fig;
time
figui
was
Fan.
with
this

Figu
AVG
are

IeaC
IeaC'

LIST OF FIGURES

Figure 1. Model (left) and Electron Microscopy (right) of the HIV virus (a) binding
to host cell (b) fusion of viral and host cell membranes (c, d) formation of large
pore and infection of host cell. The triangle represents the viral RNA that enters
the host cells. ....................................................................................... 2

Figure 2. (a) Model of HIV infection. “F” indicates the HIV fusion peptide and the
time sequence is left to right. (b) Model for HIV/host cell fusion. In the left-most
figure, a gp120/gp41 trimer is displayed with the balls representing gp120 and
rods representing gp41. “F” represents the fusion peptide and “A” represents the
transmembrane anchorage of gp41. Fusion proceeds temporally from left to right
with (i) initial state, (ii) receptor binding and fusion peptide membrane insertion,
(iii) gp41 conformational change, and (iv) membrane fusion. .......................... 3

Figure 3. Synthesis scheme for HFPmn and FP represents the sequence
AVGIGALFLGFLGAAGSTMGARS. A black circle represents a resin bead, lines
are drawn to clarify chemical functionalities, an arrow signifies a chemical
reaction, and two arrows signify multiple sequential chemical reactions. All
reactions were carried out at ambient temperature. Reaction a: Fmoc
deprotection in 3 mL of 20% piperidine/DMF (WV), 15 minutes/cycle, 2 cycles.
Reaction b: Peptide synthesis with Fmoc chemistry. 2-hour single couplings were
used for each amino acid with the following exceptions: 4-hour single couplings
for Trp, Ser and Arg residues; 6-hour single couplings for the Leu-12 to Leu-7
residues. Reaction c: Cleavage from the resin using a 4 mL solution containing
TFA/thioanisoIe/ethanedithiol/anisole in 90:5:3:2 volume ratio. After 2.5 hours
reaction time, TFA was removed with nitrogen gas and peptide was precipitated
with cold methyl t-butyl ether. ................................................................. 23

Figure 4. (a) The HPLC chromatograms for the purification of HFPmn. (b) The
MALDI-TOF MS spectrum for the identification of HFPmn. The HPLC fraction
marked with asterisk in (a) was analyzed and the corresponding mass was
labeled using asterisk in (b). .................................................................. 24

Figure 5. Synthesis schemes for (a) HFPdm and (b) HFPdm(Cys). The black
circles, lines and arrows had the same meaning as in Fig. 3. The reactions a and
c were the same as in Fig. 3. In reaction b, the synthesis of HF Pmn with Cys in
(a) followed the coupling time in Fig. 3. For the HFPdm(Cys) synthesis in (b), 4-
hour single couplings were used for each amino acid with the following

ix

er
Ct
12

Li
Hi
Mt

Fig
HF
HF

time
35:8

aste

Flgu
MAL
marl
lab-e

Figu
752

mag
by 0
Shift

I.

Fin

uIC

exceptions: 8-hour single couplings for Trp, Ser, and Arg residues; double
couplings with 4-hours per coupling for the 13CO labeled residue and for the Leu-
12 to Leu-7 residues. Reaction d: Coupling using PyAOP and DIPEA (1:2 molar
ratio) in 4 mL DMF with 6 hour reaction times for Cys and 2 hour reaction time for
Lys. Reaction 9: Cross-linking in 5 mM DMAP, pH = 8.4, open to the air. 1 umol
HFPmn(Cys) in 400 pL solution overnight. Reaction f: Selective deprotection of
Mtt in 3 mL of 1% TFA/DCM (WV), 6 minutes/cycle, 6 cycles. ........................ 25

Figure 6. (a) and (c) The HPLC chromatograms for the purification of HFPdm and
HFPdm(Cys). (b) and (d) The MALDI-TOF MS spectrum for the identification of
HFPdm and HFPdm(Cys). The HPLC fraction marked with asterisk in (a) and (c)
were analyzed and the corresponding peaks were labeled using asterisk in (b)
and (d) respectively. The mass spectra were discussed in the main text. ......... 26

Figure 7. Synthesis schemes for HFPtr. The black circles, lines and arrows had
the same meaning as in Fig. 3. All reaction conditions were the same as shown
in Fig. 3 and Fig. 5 except for reaction 0: Cross-linking in 5 mM DMAP, pH = 8.4,
open to the air. 1 pmol HFPmn(Cys) and 1.5 umol HFPdm(Cys) in 400 uL
solution for 2.5 hours. ........................................................................... 27

Figure 8 (a) The HPLC chromatograms for the purification of HFPtr. (b) The
MALDl-TOF MS spectrum for the identification of HFPtr. The top, middle, and
bottom chromatograms in panel a are for syntheses with HFPtr cross-linking
times of 0.5, 1.5, and 2.5 hours, respectively. The HPLC fraction marked with
asterisk in (a) was analyzed and the corresponding peaks were labeled using
asterisk in (b). The mass spectra were discussed in the main text. ................. 28

Figure 9 (a) The HPLC chromatograms for the purification of HFPte. (b) The
MALDl-TOF MS spectrum for the identification of HFPte. The HPLC fraction
marked with asterisk in (a) was analyzed and the corresponding mass was
labeled using asterisk in (b). .................................................................. 29

Figure 10 1D 13C-15N REDOR pulse sequence. The open columns represent the
72/2 and ”pulses. CP transfers 1H transverse magnetization to 13C. 13C
magnetization is dephased (i.e., reduced) by 13C-‘5N dipolar coupling mediated
by one 15N 71' pulse per rotor period. The 13C It pulses refocus the 13C chemical
Shift. ................................................................................................. 36

Figure 11 Double Cross Polarization (DCP) pulse sequence. CP1 and CP2
indicate the 1H-+15N and 15N—->“°’C cross polarization respectively. There is a

Strar
dime

Fligur
Matt
labels
with F
HFPd
file b
59%,:
bioadé

short delay 1' between the first and second CP process. TPPM decoupling was
applied during the r, CP2 and acquisition periods. ...................................... 45

Figure 12 (AS/So)” (error bars) and best-fit (AS/So)” (lines with or without
diamonds) vs dephasing time (2) for (a) the 13C-3‘P setup and (b) the 13C-‘QF
setup. In panel (a) the experimental data was fit to a two-spin system. In panel
(b), the experimental data was fit to either a two-spin system (dash line) or a
three-spin system (solid line). ................................................................. 49

Figure 13 (a) 13CO region of D-‘3CO-NAL spectrum acquired using the CP pulse
sequence. (b) 13CO region of D-‘3CO-NAL spectrum acquired using the DCP
pulse sequence. The vertical scales in (a) and (b) are the same so that the
relative intensity reflects the DCP transfer efficiency. Both spectra were
processed with 100Hz Gaussian line broadening and baseline correction. (c) A
negative control experiments without 15N CP2 amplitude (of. Fig. 11). The
number of scans used in (a), (b) and (c) was 32. Panel (d) displays the
optimization of 15N CP2 amplitude as given in Fig.11. The 5N rf field was
scanned from 11.8 kHz to 17.8 kHz with the increment of 0.16 kHz. Panel (e)
displays the optimization of 1“5N CP2 offset frequency. The offset was scanned
from -10 kHz to 10 kHz with the increment of 0.5 kHz. ................................. 53

Figure 14 The aliphatic region of the ZD 13C-13C correlation spectra for (a)
HFPmn in PC:PG, (b) HFPdm in PC:PG, (c) HFPtr in PC:PG, (d) HFPmn in
PC:PG:CHOL, (e) HFPdm in PC:PG:CHOL and (f) HFPtr in PC:PG:CHOL. All
spectra were processed with 100 Hz Gaussian line broadening in both
dimensions. The individual peaks were assigned and given in the spectra. For
example, the peak assigned to A6 C6/Ca(|3) represent the cross peak between
CB(f1 dimension) and Ca(f2 dimension) for Ala-6 in B-strand conformation. The
spectra (9) through (I) display the representative 1D slice of the PDSD spectra (a)
through (f) respectively. For the spectra (a), (c), (d), (e) and (f), the 1D slice is
along ~23 ppm in the f1 dimension which corresponds to the CB of Ala-6 in B-
strand conformation. For the spectrum (b), the 1D slice is along ~18 ppm in the f2
dimension which corresponds to the CB of Ala-6 in ct-helical conformation. ......61

Figure 15 13CO spectra of (a) Ala1, (b) lle4, (c) Ala6, (d) Leu9, (e) Leu12 and (f)
Ala14 in HFPmn_mut associated with PC:PG and PC:PG:CHOL. For each
labeled residue, the spectrum with PC:PG is shown In the left and the spectrum
with PC:PG:CHOL is shown in the right. 13co peaks of Ala15 in (g) HFPmn, (h)
HFPdm and (i) HFPtr associated with PC:PG in the top row and PC:PG:CHOL in
the bottom row. All spectra are obtained with the 13C-31P REDOR pulse
sequence with 2 ms dephasing time, and processed with 200 Hz Gaussian line
broadening and baseline correction. All PC:PG spectra are acquired with 3000

xi

scan:
dash

Flgur
HFPt
vent

cos
I30.
let p
Stan

FQUI
rows
tscc
Flgm
d) H:
obtal
32lr
bioa
4132
8185

FIQUI
58m;

The

FQU:
Cons
and

used
Eacr
ai'i‘iia
I938
lad:

HFP.
hidrt
dish

'Id

scans and all PC:PG:CHOL spectra are acquired with 1500 scans. The vertical
dashed lines in (g)-(i) indicate the chemical shift of B—strand 13CO. ................. 72

Figure 16 (a) 1D slices along the chemical shift of Cy of Leu12 for HFPmn,
HFPdm and HFPtr in the top, middle and bottom spectrum respectively. The
vertical dashed lines labeled 1-3 are assigned to the chemical shifts for CO/Cy,
Cor/Cy and CB/Cy cross peaks in helical conformation respectively, and 4-6 are
00/07, Cor/Cy and CB/Cy cross peaks in strand conformation respectively. (b)
13C-3‘P REDOR So spectra for Ala6 and Ala15 samples. In each spectrum, the
left peak corresponds to tat-helical structure and the right peak corresponds to I}
strand structure. .................................................................................. 73

Figure 17 Flow chart of derivation of (AS/So)” for REDOR of HFP-B. The four
rows in each box are in sequence: the site description, its relative population, and
its contributions to So and S1. ................................................................. 84
Figure 18 REDOR So and $1 spectra for membrane—associated (a, b) HFP-A, (c,
d) HFP-B, (e, f) HFP-C, (g, h) HFP-D and (i, j) HFP-E. Spectra a, c, e, g, i were
obtained with 24 ms dephasing time and spectra b, d, f, h, j were obtained with
32 ms dephasing time. Each spectrum was processed with 200 Hz Gaussian line
broadening and baseline correction. Each 80 or $1 spectrum was the sum of (a)
41328, (b) 56448, (c) 45920, (d) 81460, (e) 55936, (f) 79744, (9) 30898, (h)
81856, (i) 45920 or (j) 71040 scans. ........................................................ 90

Figure 19 Plots of (AS/So)” vs dephasing time for membrane-associated HFP
samples prepared with [HFP]inmar of (a) 400 pM or (b) 25 uM. The symbol legend
is: diamonds, HFP-B; triangles, HFP-C; circles, HFP-D; and squares, HFP—E.
The 0“” were ~0.04. ........................................................................... 91

Figure 20 (a) Two antiparallel registries of residues 1-16 of HFP that were
consistent with the REDOR data shown in Fig. 20. The registries are denoted A
and B and the 13CO labeled Ala-14 residue is highlighted in blue. (b) Models
used to calculate (AS/So)” and spin geometries specific for the HFP-C sample.
Each model includes nuclei from three adjacent strands with the Ala-14 13CO
always in the middle strand and 15N in the top and/or bottom strands. The first
letter in the labeling of each model refers to the middle strand/top strand registry
and the second letter refers to the middle strand/bottom strand registry. Registry
X is any registry for which the interpeptide 13CO-‘5N distance was large in the
HFP-H, HFP-l, HFP-J, or HFP-K samples so that d z 0. The Ala-14 13CO is
hydrogen bonded to an amide proton in the top strand. Relevant labeled 13C-15N
distances and ‘5N-13C-‘5N angles are: r1 = 4.063 A; ,1. = 5.890 A; r2 = 5.455 A; r2'
= 6.431 A; 671 = 161.1’; 02 = 131.9°; 03= 130.2'; and 04= 1170’. Each

xii

parameter value was the average of 10 specific values taken from the crystal
structure of outer membrane protein G. .................................................... 92

Figure 21 Contour plots of 1’ vs strand fitting parameters for (a) unconstrained;
(b) partially constrained; and (0) fully constrained fittings. The a, b, 32, and b2
parameters refer to probabilities for different adjacent strand arrangements. In
plot a, the black, green, blue, red, and white regions respectively correspond to
f<19,19<,;2<21,21 <12<23,23<f<25,andf>25.lnplotb,the
regions respectively correspond to 22< 18, 18 < f < 20, 20 < f < 22, 22 < f <
24, and f > 24, and in plot c, the regions respectively correspond to f< 15, 15
< f < 17, 17 < 12 < 19, 19 < f < 21, and 12 > 21. Best-fit parameters were: plot
a, a = 0.22, b = 0.31, f = 16.5; plot b, a = 0.31, b = 0.42, f = 15.1; and plot 0,
a2 = 0.26, b2 = 0.33, f = 12.7. In plot a, the a and b parameters are the fractional
probabilities of adjacent strands having A or B registries, respectively. In plot 0,
the a2 and b2 parameters are the fractional probabilities of domains of A or B
registries, respectively. ......................................................................... 97

Figure 22 Peptide sequences of the basic constructs of HFPmn_mut, HFPmn
and HF Ptr. The specific labeling sites were described in the main text. ......... 110

Figure 23 Synthetic scheme of 5-F-palmitic acid. ...................................... 115

Figure 24 (a) Plot of (AS/So)” vs mol fraction of 5-19F-DPPC at 1' = 16 ms. All
samples contained HFPmn-L9. (b) static 31P spectra for PC:PG and
PC:PG:CHOL bilayer with and without 9 mol fraction 19F-DPPC. Each spectrum
was acquired with 1024 scans and processed with 300 Hz Gaussian line
broadening. The spectra were acquired at 35 °C. ...................................... 118

Figure 25 Panel a displays stopped-flow monitored changes in lipid fluorescence
induced by addition of different HFP constructs to an aqueous solution containing
membrane vesicles. Increased fluorescence is a result of mixing of lipids
between different vesicles and this mixing is one consequence of vesicle fusion.
The lines are color coded: HFPmn (black); HFPdm (red); HFPtr (blue); and
HFPte (green). The total lipid concentration was 150 (M and the HFPmn,
HFPdm, HFPtr, and HFPte concentrations were 1.50, 0.75, 0.50, and 0.37 pM,
respectively, so that peptide strand:lipid = 0.01. The data were collected at 25 °C.
the vesicle composition was 4:1 POPC:POPG, and the initial vesicle diameter
was ~100 nm. Additional data (not shown) were obtained for HFPdm, HFPtr, and
HFPte at 30, 35, and 40 °C. Each data set for each construct was analyzed as
the sum of two exponential buildup functions. Panel b displays Arrhenius plots for
the rate constants of the fast buildup function with legend: HFPdm (open square);

xiii

HFPtr (open circle); and HFPte (open triangle). The best-fit lines are also
displayed and result in the respective activation energies 41 i 3, 26 :l: 1, and 20 :t
1 kJ/mol. .......................................................................................... 122

Figure 26 So spectra for membrane-associated HFP with peptide:lipid ~0.04. The
dotted lines are at 175 ppm. All spectra were obtained with r = 2 ms and were
processed with 200 Hz Gaussian line broadening and baseline correction. The
membrane composition for samples a-d was PC:PG and the membrane
composition for samples e-h was PC:PG:CHOL. The peptides were: a, e,
HFPmn-5GAL; b, f, HFPmn-BFLG; c, g, HFPmn-"FLG; and d, h HFPmn-“AAG.
The numbers of scans summed to obtain spectra a-h were 4823, 3867, 4823,
8500, 3259, 1001, 4320 and 6992, respectively. ....................................... 127

Figure 27 13CO-31P REDOR spectra of membrane-associated HFP with
peptide:lipid ~0.04. Each letter corresponds to a single sample which contained
(a-d) PC:PG or (e-h) PC:PG:CHOL and (a, e) HFPmn-5GAL, (D, f) HFPmn-8FLG,
(c, g) HFPmn-"FLG, or (d, h) HFPmn-“AAG. For each letter/sample, so (left), 31
(right), z'= 16 ms (top), and r= 24 ms (bottom) spectra are displayed. The dotted
lines are drawn for visual comparison of So and 81 peak intensities. Each
spectrum was processed with 300 Hz Gaussian line broadening and baseline
correction. The numbers of So or 81 scans summed to obtain the top and bottom
spectra were respectively: a, 30000, 56000; b, 27509, 29463; c, 20000, 40000; d,
44129, 48296; e, 8448, 52384; f, 5488, 21664; 9, 28032, 52384; and h, 22576,
50240. ............................................................................................. 131

Figure 28 (AS/So) vs dephasing time for membrane-associated HFP in (a,c)
PC:PG or (b,d) PC:PG:CHOL. For panels a and b, the points correspond to
(AS/Sg)°"" and the symbol legend is: squares, HFP2-5GAL; circles, HFP3-8FLG;
triangles, HFPZ-"FLG; and diamonds, HFPZ-“AAG. The vertical dimensions of
each symbol approximately correspond to the i1 0 uncertainty limits. Lines are
drawn between (AS/80f” values with adjacent values of 1. Each (AS/So)” value
was determined by integration of 10 ppm regions of the So and S1 spectra.
Panels 0 and d respectively correspond to the HFPmn-“AAG/PCzPG and the
HFPmn-“AAG/PC:PG:CHOL samples and the points correspond to (AS/SOY“
(vertical lines with error bars) and best-fit (AS/So)” (diamonds). Lines are drawn
between points with adjacent rvalues. For plot c, the best-fit d = 91 i 8 Hz with
corresponding r = 5.12 :l: 0.16 A, r= 0.45 a: 0.02, and 12m... = 5.0. For plot d, the
best-fit d = 85 i 6 Hz with corresponding r = 5.24 :l: 0.13 A, f = 0.32 i 0.02, and

{mm = 3.8. ....................................................................................... 132

Figure 29 (a) Partial membrane insertion (PMI) and (b, 0) full membrane insertion
(FMI) models for antiparallel B strand HFP. The red arrows represent the A1 to

xiv

line

ex;

Fig
dis;
bot
fort

FigL
sarr
earn

the ,

G16 residues in strand conformation and the black lines represent the $17 to
823 residues in random coil conformations. For clarity, black lines are not
displayed in c. Lipids are represented in blue and grey and cholesterol is not
displayed. Three antiparallel strands are displayed in a, b and twelve strands are
displayed in c but the actual number of strands in the oligomer/aggregate is not
known. The curvature and angle of the strands with respect to the bilayer normal
are not known but the models consider that A1-G16 has ~55 A length and that
the transbilayer distance is ~48 A. The experiments do not provide information
about the membrane locations of residues $17 to $23. Relative to FMI model (b),
the FMI [3 barrel variant (c) could have reduced energy because all of the
residues in the membrane interior have backbone hydrogen bonds. ............. 134

Figure 30 REDOR 13C So and S1 NMR spectra for different labeled residues and
different HFP constructs. The dephasing time for each 13CO-31P spectrum was
32 ms and for each 13CO-‘QF spectrum was 24 ms. The membranes contained 9
mol% 16-F-DPPC lipid. Each spectrum was processed with 200 Hz Gaussian
line broadening3 and olynomial baseline correction and was the sum of ~30000
scans for CO-3 P experiments and ~20000 scans for 13CO-‘S’F
experiments. ..................................................................................... 139

Figure 31 Summary of experimental REDOR dephasing (AS/Sg)e"p for the spectra
displayed in Fig. 31. The top box in each panel is the 13CO—3‘P data and the
bottom box is the 13CO-“’F(C16) data. The (AS/80f” values are shown as bars
for different labeled samples and a typical uncertainty is $0.04. ................... 141

Figure 32 ”co-“P and ”co-19F REDOR dephasing curves for different HFP
samples labeled at (a, d) Ala1, (b, e) Ala6 or (c, f) Ala14. The membranes
contained 9 mol% 16-F-DPPC lipid. The color coding of the constructs is given in
the legend of panel b. For 2 ms dephasing time, the typical uncertainty in
(AS/80f” is $0.02 and for the other dephasing times, the typical uncertainty is
$0.04. .............................................................................................. 146

Figure 33 13C So and 81 NMR spectra from 1i’CO-‘QF REDOR experiments of
samples made with 9 mol% 5-F-DPPC lipid. For panels a, b, c, d, and e the
samples respectively contained HFPmn-A1, HFPmn-A6, HF Pmn_mut-A6, HFPtr-
A6, and HFPmn-L9. Each spectrum was processed with 200 Hz Gaussian line
broadening and polynomial baseline correction. Each So and S1 spectrum was
the sum of ~ 20000 scans. 13CO-“’F(C5) REDOR dephasing curves for different
HF Ps are plotted in panel fand the constructs are coded as shown in the legend.
For 2 ms dephasing time, the typical uncertainty in (AS/SOY” is $0.02 and for the
other dephasing times, the typical uncertainty is $0.04. In panel 9 the 13C-‘QF (C5)
and 13C-19F(C16) spectra of HFPmn and HFPtr at 16 or 24 ms dephasing time

XV

item
ther

Fgu
Lhfid
and

ohgo
The

near
96p“
hunt
hnhn
the h

ngr
Star
3P e
hght

resDe
and b
3000C
scans

Figure
dephe
was:
COUeg
relite:
the(3

HQUK
dIiIEIE
hhl4!
circleS
depha
The Q

UIICet-t

FIgtlre
depha
eXQEin

were shown in the stack form with So spectra on the left side and 81 spectra on
the right side. The dash lines reflect the intensities of the So spectra. ............ 147

Figure 34 Insertion models of [3 sheet (a) HF Pmn_mut, (b) HFPmn, and (c) HFPtr.
Lipid headgroups are drawn as blue balls, lipid alkyl chains are drawn in grey,
and peptides are drawn in red. In all models, peptides are represented as
Oligomers with either six (HFPmn and HFPmn_mut) or two (HFPtr) molecules.
The strands are in antiparallel Bsheet structure with adjacent strand crossing
near Phe8 and Leu9. This is the known structure for a large fraction of HFPmn
peptides. The number of strands in a sheet is not known but is likely a small
number. The lines at the C-terminus of HFPtr represent the chemical cross-
linking of the HFPtr construct. For clarity, not all lipid molecules are shown near
the HFP. .......................................................................................... 149

Figure 35 (a) 13c-3‘P REDOR so spectra with 2 ms dephasing time. (b) REDOR
So and 81 spectra for HFPmn_mut with long dephasing time (t = 32 ms for ‘30-
31P experiments and 2' = 24 ms for 13C-‘QF experiments). The left, middle and
right columns in (b) are 1‘3C-3‘P, 13C-19F(CS) and 13C-“’F(C16) experiments
respectively. All spectra were processed with 200 Hz Gaussian line broadening
and baseline correction. ln panel (b) each of the 13C-31P spectra was acquired for
30000 scans and each of the 13C-19F spectra was acquired for ~ 20000
scans. ............................................................................................. 159

Figure 36 Plots of (AS/80V” for HFPmn_mut 13C-31P experiments at 32 ms
dephasing time and 3C-19F experiments at 24 ms. The experimental dephasing
was obtained by integrating over a 1 ppm interval around the 13CO peaks in the
corresponding 80 and 8; spectra shown in Fig. 35. For Leu12, the black bar
represents the (AS/80f” for a-helical conformation and the red bar represents
the (AS/80f” for B—strand conformation. The typical uncertainty is :t 0.02. ..... 160

Figure 37 (a) Plots of 13C-3‘P (AS/80f“ vs. dephasing time for HFPmn_mut with
different labeled positions. The symbols are open diamonds for A1, open squares
for I4, open triangles for A6, crosses for L9, stars for the a-helical L12 and open
circles for A14. (b) Plots of 13C-3‘P (AS/SOY” (open squares) and (AS/So)“ vs.
dephasing time for different residues (as labeled in the figures) of HFPmn_mut.
The typical experimental uncertainties are 10.02003 and the typical corrected
uncertainties are :l:0.03-0.04. ................................................................ 163

Figure 38 REDOR So and 8; spectra for (a) HFPmn and (b) HFPtr at long
dephasing time (T = 32 ms for 13C-3‘P experiments and r = 24 ms for l3C-‘QF
experiments). The So and 8; spectra were shown in black and red respectively.

xvi

AH spi
conec
spech

Fgure
and (i
stand
heop
of (as
symbc
interva
Stspe

Fgum
meaSL
meaSL
UBPgh
dthahl
ol3'P.
Gide l
legion
Leu12
longhu
1tics

Figure
the de
and B-
and at
(AS/s:
JCJEF
Won:
QIEEn
gheni

FigUre
HFP”
ftil 13C
“33980
and b6
9:11 th.
;C,.;F

All spectra were processed with 200 Hz Gaussian line broadening and baseline
correction. The 13C--3“P spectra were acquired for ~ 30000 scans and the ”C-31P
spectra were acquired for ~ 20000 scans. ............................................... 166

Figure 39 Plots of (AS/Sg)°"" vs. dephasing time for the (a) ”C-31P, (b) ”C-”F(C5)
and (c) 13C-”F(C16) experiments for HFPmn and HFPtr in the ot-helical and (3-
strand conformations. The residues Ala6, Leu9 and Leu12 are represented with
the open squares, open circles and open triangles respectively. The uncertainties
of (AS/So)” are typically $0.02 ~ 0.03 and are approximately the size of the
symbols. The (AS/So)” values were determined by integrating over a 1 ppm
interval around the a-helical or B—strand ”CO peaks in the corresponding 80 and
S1 spectra. ....................................................................................... 168

Figure 40 (a) Geometry Model for the consideration of 1i’C-i’IP and 13C-”F
measurement limit. The two circles with 11 A and 14 A radii indicate the
measurement limits of 13C-3‘P and 13C-‘S’F (C16) REDOR respectively. The yellow
triangle shows the geometry of the case where a 13CD has the maximum vertical
distance (3.5 A) relative to the lipid alkyl chain. (b) and (c) Longitudinal positions
of 3‘P, ”F(C5) and ”F(C16) in the membrane bilayer. In panel (b), the dotted
circle has the radius of ~ 10 A and the solid circle has the radius of ~ 14 A. The
region marked in red indicates the possible location of the ”COs of Ala6 and
Leu12 in the B-strand HFPmn as described in the main text. In panel (c), the
longitudinal distance between 31P and 19F(C5) is 10 A and the distance between
”F(05) and ”F(C16) is 12 A. ............................................................... 173

Figure 41 (a)-(d) Plot of (AS/SOY” (open squares) and (AS/30F“ (solid lines) vs.
the dephasing time for Ala6, Leu9 and Leu12 in HFPmn and HFPtr with cit—helical
and B—strand conformations. The data with (AS/SOY” < 0.1 at 32 ms for 13C-3‘P
and at 24 ms for ”C-”F(C16) were not fit and were labeled “N/A”. (e) Plots of
(AS/SOY“ (open squares) and (AS/So)” (solid lines) vs. the dephasing time for
3C-”F(C5) experiments. (f) Contour plots for the data fittings shown in (e). The
regions with x2 values xzmrn +1, xzmrn +2 and mem +3 were shown in red, blue and
green respectively, where xzmin is the best-fit root-mean-squared deviation as
given in Table 13 of the main text. ......................................................... 177

Figure 42 REDOR So and S1 spectra for Ala1, lle4 and Ala14 in HFPmn and
HFPtr at long dephasing time (1- = 32 ms for ”C-31P experiments and z' = 24 ms
for 13’C-‘S’F experiments). The So and 8; spectra are shown in black and red
respectively. All spectra were Processed with 200 Hz Gaussian line broadening
and baseline correction. The 3C-31P spectra were acquired for ~ 30000 scans
and the 13C-31P spectra were acquired for ~ 20000 scans. The arrows indicate
13c-‘9F(C5) dephasing for I|e4 and Ala14. ............................................... 185

xvii

Figure 43 F
(open circle:
calculated 0
St spectra. .

Figure 44 ll
secondary s
chains were
were shown
were displaj
terminus.

Figure 43 Plots of (AS/So)°"” vs. dephasing time for Ala1(open squares), lle4
(open circles) and Ala14 (open triangles) labeled samples. The dephasing was
calculated by integrating over the entire 1 CO peaks in the corresponding 80 and
8; spectra. ........................................................................................ 186

Figure 44 Insertion models for HFPmn_mut, HFPmn and HFPtr and different
secondary structures. The lipid headgroups were shown in blue and the alkyl
chains were displayed in gray. For the peptides, the residues from Ala6 to Leu12
were shown in red with definitive secondary structures and the other residues
were displayed in black. The arrows indicated the direction from N to C
terminus. .......................................................................................... 188

xviii

AF

AII

CH
CH
CP
CS
D: |

d:[

DIP
Dill.
Dirt;
DP(
153t
DTF
DTP
ESR
FMl;

Fmdr

LIST OF SYMBOLS AND ABBREVIATIONS

AHT: Average Hamiltonian Theory

AIDS: Acquired Immune Deficiency Syndrome

AUC: Analytical Ultracentrifugation

CD: Circular Dichroism

CHOL: cholesterol

CHR: C-terminal heptad repeat

CP: Cross Polarization

CS: Chemical Shift

D: Dipolar Coupling

d: Dipolar coupling frequency

DCP: Double Cross Polarization

DIPEA: N, N-diisopropylethylamine

DMAP: dimethylaminopyridine

DMPC: 1, 2-dimyristoyl-sn-glycerol-3-phosphocholine

DPC: dodecylphosphocholine

1-”C-DPPC: [1-13C]-1 , 2-dipalmitoyl-sn-glycero-3-phosphocholine
DTPC: 1, 2-di-O-tetradecyl-sn-glycerol-3-phosphocholine

DTPG: 1, 2-di-O-tetradecyl-sn-glyceroI-3-[phosphor-rac-(1-glycerol)]
ESR: Electron Spin Resonance

FMI: Full Membrane Insertion

Fmoc: 9-fluorenylmethoxycarbonyl

xix

HBTU: O
HEPES: r
HFP; HIV
HFPdm: I
HFPmn: l
HFPmn_r
HFPte: HI
HFPtr: HF
HIV: Hum;
HOBI. 1-h1
HPLC: Hig
IR: Infraret
LUVs: Larg
MALDI: M;
MAS. Mag
MD: MOIec
NAL? N-acl
II-NBD-PE
NHR? N-IEr
NMR:Nucl
NC'E-NUclt
lI-thE: It

OitiPG' 0th

HBTU: O-benzotriazoIe-N,N,N’,N’-tetramethyI-uronium-hexafluoro-phosphate
HEPES: N-(2-hydroxyethyl)piperazine-N-Z-ethanesulfonic acid

HFP: HIV Fusion Peptide

HFPdm: HFP dimer

HFPmn: HFP monomer

HF Pmn_mut: HFP monomer with V2E mutation

HFPte: HFP tetramer

HFPtr: HFP trimer

HIV: Human Immunodeficiency Virus

HOBt: 1-hydroxybenzotriazole

HPLC: High-performance liquid cheomotography

IR: Infrared

LUVs: Large Unilamellar Vesicles

MALDI: Matrix-assisted laser desorption/ionization

MAS: Magic Angle Spinning

MD: Molecular Dynamics

NAL: N-acetyl-Leucine

N-NBD-PE: N-(7-nitro-2,1,3-benzoxadiazol-4-yl)-phosphatiylethanolamine
NHR: N-terminal heptad repeat

NMR: Nuclear Magnetic Resonance

NOE: Nuclear Overhauser Effect

N-Rh-PE: N-(Iissamine Rhodamine B sulfonyI)-phosphatiy|ethanolamine

OMPG: outer membrane protein G

XX

 

Riv!
r. ir
rfl

SD

TFr
TOI

TPF

PAS: Principle Axis System
PDB: Protein Data Bank

PyAOP: 7-azabenzotriazol-1-yloxy-tris-(pyrrolidino)-phosphonium
Hexafluorophosphate

PMI: partial membrane insertion

PDSD: Proton-driven Spin Diffusion

POPC: 1, 2—dimyristoyI-sn-glycerol-3-phosphocholine
POPG: 1, 2-dimyristoyl-sn—glyceroI-3-[phosphor-rac-(1-glycerol)]
REDOR: Rotational-echo Double Resonance

RMSD: root-mean-asquared deviation

r. internuclear distance

rf. radiofrequency

SDS: sodium dodecyl sulfate

SUL: “scatter-uniform” labeling

TFA: trifluoroacetic acid

TOF: time-of-flight

TPPM: Two-pulse phase modulation

xxi

 

 

 

ar

Sy

lie

he

hit
\

ch

CHAPTER I

INTRODUCTION

BACKGROUND

Membrane fusion is an important step in viral infection for widespread
and serious diseases such as influenza and acquired immune deficiency
syndrome (AlDS).(1-2) Understanding of viral fusion is thus important both as a
key step in the viral life cycle and as a target for anti—viral therapeutics.(3-5)
Fusion between two membrane-bound bodies such as cells, viruses or vesicles
is a protein-mediated process and is generally separated into three sequential
steps: (1) binding of the two bodies; (2) mixing of their membrane lipids; (3)
formation of a large fusion pore through which the contents of virus and cell can
mix.(6) Figure 1 illustrates a series of freeze-fracture electron micrographs that
follows the time evolution of human immunodeficiency virus (HIV) infection of a
host cell.(7)

In AIDS, fusion and infection are initialized by strong interactions of two
highly glycosylated viral envelope proteins gp120 and gp41 with the CD4 and
chemokine (e.g. CXCR4) receptors of human T and macrophage cells.(1) The
glycosylation of gp120 is extensive with glycans representing 50% of the
molecular mass of the mature protein.(8) The glycosylation of gp41 is less
abundant relative to gp120, and there are only four or five potential glycosylation
sites.(9-12) However, these potential sites have been thought to be a

requirement for the fusion activity of gp41 .(13) The proteins gp120 and gp41 are

"IN

“l/\\

‘“//\\

 

 

 

 

 

 

 

 

 

 

 

Figure 1 Model (left) and Electron Microscopy (right) of the HIV virus (a) binding
to host cell (b) fusion of viral and host cell membranes (c, d) formation of large
pore and infection of host cell. The triangle represents the viral RNA that enters
the host cells. (Adapted from Reference 7)

Figure
ftom F
99720.
”W984
ITaDSm
With (1";
(iii) 9;)-
REIGI er

T-cell . .. _ .

- ‘ ' ' ' ;_ t

i ”in.” MMA".
II’M ‘

 

 

Figure 2 (a) Model of HIV infection. The time sequence is left to right. (Adapted
from Reference 16) (b) Model for HIV/host cell fusion. In the left-most figure, a
gp120/gp41 trimer is displayed with the balls representing gp120 and rods
representing gp41. “F” represents the fusion peptide and “A" represents the
transmembrane anchorage of gp41. Fusion proceeds temporally from left to right
with (i) initial state, (ii) receptor binding and fusion peptide membrane insertion,
(iii) gp41 conformational change, and (iv) membrane fusion. (Adapted from
Reference 17)

non-
lhr01
Figu

foll0i

non-covalently bound to form a complex and the complex is attached to the HIV
through a transmembrane C-terminal segment in gp41.(14,15) As shown in
Figure 2, a proposed HIV/cell membrane fusion mechanism includes the
following steps: (1) binding of the conserved region in gp120 to CD4 and CXCR4;
(2) conformational changes which finally lead to the exposure of a conserved
segment of about twenty amino acids fusion peptide at the N-terminus of gp41;
(3) interaction between the fusion peptide and target cell membranes which
anchors gp41 in the cell membrane; (4) conformational change of the ecto-
domain of gp41 which locates outside of the viral and cell membranes helps to
bring the HIV and cell membranes close together; (5) mixing of lipids and
formation of large fusion pores.(16,17) There were various experimental
evidences which supported the different steps of the mechanism proposed above.
First of all, it was known that conserved regions in the gp120 subunit are
responsible for the binding of the virus to the CD4 and CXCR4.(14,15) Second,
although the conformational change of gp120 after binding to the receptor was
not well understood, the extended conformation of gp41 after the releasing of
gp120 was characterized and there have been antibodies which can bond to the
extended and exposed N-termlnal heptad repeat (NHR) and C-terminal heptad
repeat (CHR) segments.(2,18-23) In addition, the conformational change from
the extended gp41 was supported by the fact that gp41 folded back on itself and
associated as very stable six-helical-bundle structure, which was observed at the
fusion sites.(24-26) However, the interaction between the N-terminal fusion

peptide and the cell membranes was not clear at this point.

amii
See
the

grea
anal
SIVi
peph
Thel
rest I
vesh

conel

using

suhah
andte
effect (
sUQDes
Charger
IIIR s
dISOIdel

Hip as;

In our studies, the HIV Fusion peptide (HFP) sequence contains 23 native
amino acids: Ala-VaI—Gly-lle-Gly-Ala-Leu-Phe-Leu-GIy-Phe-Leu-Gly-Ala-Ala-Gly-
Ser-Thr-Met-Gly—AIa-Arg-Ser. These residues are located at the N-terminus of
the HIV gp41. The hydrophobic amino acids such as Val, Phe and Leu are in
great abundance in the HFP sequence and are also highly conserved in the
analogs of HFP such as the fusion peptides of the glycoproteins in HIV-ll and
SIV.(27, 28) In the following chapters, HFP and its derivatives will serve as model
peptides of the gp41 in the investigation of structure and membrane insertion.
The HFPs are reasonable substitution for gp41 because (1) in the absence of the
rest part of gp41, the HFP itself can cause rapid fusion and/or leakage of lipid
vesicles or erythrocytes, and (2) several mutational studies have shown strong
correlations between FP-induced fusion and viral/host cell fusion.(29-32)

The secondary structure of the micelle-associated HF P has been probed
using solution Nuclear Magnetic Resonance (NMR) in both sodium dodecyl
sulfate (SDS) or dodecylphosphocholine (DPC) micelles.(33-38) The secondary
and tertiary structures are obtained from chemical shifts and nuclear Overhauser
effect (NOE) crosspeaks.(39,40) For HFP:detergent ~0.01, solution NMR results
suggested the presence of helical structure from Ile4 to Leu12 in negatively
charged SDS or neutral DPC micelles.(33,36-38) However, different solution
NMR studies suggested there may be additional helical, B—turn or more
disordered and dynamic structures in the C-terminus of HFP.(33-38) Structure of
HFP associated with membranes has been studied using circular dichroism (CD)

or infrared (IR) spectroscopy. A previous CD measurement showed that the HF P

adop
nega
signil

vesm

fador
melo
sduhc
Sihuh
HFP |
SUDDO

HFP(

fU”Isl
IIIIErfaC
HFRFE

WIIII Ch;

adopted a significant helical character in SDS detergent or in an environment of
negatively charged vesicles with a 1:200 peptide:lipid molar ratio,(33,41-43) but a
significant B—strand character in 1:30 peptide:lipid ratio,(41) With neutral lipid
vesicles at peptide:lipid molar ratios of ~ 1:200, there are two infrared reports of
predominantly helical structure,(43, 44) three reports of predominantly B
structure,(41,45,46) and one report of mixed helical and B structure.(47) Two
investigators report that the peptide conformation changes from helical to B as
the peptide:lipid ratio is increased from 1:200 to 1:30 while two others report that
the B conformation does not change with these ratios.(41,43-45) There
differences in structure may have to do with differences in peptide sequence,
lipid compositions, sample preparation, or hydration Ievel.(48)

The membrane location of HFP has been suggested as an important
factor to understand the peptide/membrane interaction. Previous studies about
the location of HFP in micelles and/or membranes have been performed using
solution NMR, fluorescence spectroscopy, electron spin resonance (ESR) and
simulation. Unfortunately, there is not yet a consensus for the micelle location of
HFP based on the solution NMR results. There are distinct models which
supported either predominant micelle surface location or micelle traversal by
HFP.(33, 37) In one solution NMR study, residues I4 to A15 were found to be
fully shielded from solvent and residues G3 and G16 were at the micelle/solvent
interface.(37) HFP location in membranes has been primarily probed using a
HFP-F8W mutant and by variation of the tryptophan fluorescence of this mutant

with changes in environment.(49,50) Key results have included: (1) fluorescence

was
salt!
for i
simi
o‘osi
bror
bror
solv
ass:
trypt
lipid
Dhos

local

was higher for membrane-associated HFP-F8W than for HFP-F8W in buffered
saline solution; (2) greater fluorescence quenching by acrylamide was observed
for a soluble tryptophan analog than for membrane-associated HF P-F8W; and (3)
similar fluorescence quenching of membrane-associated HFP-F8W was
observed in samples containing either 1-palmitoyl-2-stearoyI-phosphocholine
brominated at the 6, 7 carbons of the stearoyl chain or the corresponding lipid
brominated at the 11, 12 carbons of the chain. The first two results indicated that
solvent exposure of the HFP-F8W tryptophan is reduced with membrane
association and the third result indicated that the membrane location of the
tryptophan indole group is centered near the carbon 9 position of the brominated
lipid stearoyl chain; i.e. ~8.5 A from the bilayer center and ~11 A from the lipid
phosphorus. In a different set of experiments, ESR spectra showed that 8 M19
location close to the aqueous interface of the membrane and an A1 location
away from this interface.(43) Models for HFP location in membranes have also
been developed from simulations of a single HFP molecule in membranes and
have shown either partial insertion or traversal of the membrane. The HFP
always adopted predominant 0t helical conformation and in one simulation was
generally near the membrane surface with the F8 backbone and sidechain nuclei
respectively 4 A and 6 A deeper than the phosphorus longitude.(51) For a
different simulation, HFP traversed the membrane and the backbone and
sidechain F8 nuclei were at the bilayer center, i.e. ~19 A from the phosphorus

longitude.(52)

stru
crys
mer
and
strul

385C

Show
hydro

aDUp;

dignn

StttigiL
COlttrit
cmSew

The current studies mainly utilized solid-state NMR to investigate the
structures as well as the membrane location of HFPs. Without the need for
crystallization or solvation, solid—state NMR is a useful method to probe
membrane associated peptide/protein systems. Measurements of chemical shifts
and internuclear dipolar couplings provide information about the secondary
structure, tertiary structure, insertion depths and insertion angles of HFPs
associated with membranes. Some of the important results obtained previously in
our group include: (1) Helical, B strand and random coil structure were observed
at specific residues in HFP and the distribution of the conformations at specific
residues was shown to depend on the lipid headgroup and cholesterol
composition of the membrane.(53-57) HFP was also shown to fuse vesicles with
different compositions, which suggests that more than one conformation is
fusogenic.(58) (2) Measurements of dipolar couplings between different HFPs
showed that the B strand structure is oligomeric and contains interpeptide
hydrogen bonding. There may be approximately equal populations of parallel and
antiparallel strand alignment. The adjacent strand in the parallel alignment may
be two residues out-of-registry and the adjacent strands in the antiparallel
alignment are crossing between F8 and L9.(59, 60)

The present work will mainly focus on the solid-state NMR studies of the
structure and insertion depth of membrane-associated HFP oligomers, which will
contribute to understand the question about which characters of HFP may be
closely associated with fusion activities. In Chapter II the synthesis of biological

relevant HIV fusion peptide constructs will be described. In particular, the

experiment.
proposed a
introduce ll
HFP syslel
(REDOR).

(DCP) exp
Theory (AH
and the set:
studies of '
esoecially it
will report I
membrane,
Sheet for at
with the cue
will describe
C0”Silucts:

membranes
disrulition of
HFP CODSlrL

co”CIUSIOn is

experimental conditions for the synthesis of HFP trimer (HF Ptr), which has been
proposed as a HFP construct at the fusion site, will be discussed. Chapter III will
introduce the solid-state NMR methods used to study the membrane-associated
HFP system, including ”C-31P and 1E’C-‘S’F rotational-echo double resonance
(REDOR), proton-driven spin diffusion (PDSD), and double cross polarization
(DCP) experiments. The theoretical approach named Average Hamiltonian
Theory (AHT) for understanding these pulse sequences will be briefly introduced
and the setup procedure will be described in details. Chapter IV will focus on the
studies of the secondary structure of membrane-associated HFP constructs;
especially the dependence on lipid compositions and HFP constructs. Chapter V
will report the tertiary structure of HFP monomer (HFPmn) in a host-cell-like
membrane. The results in this chapter revealed the existence of anti-parallel [3-
sheet for a membrane-associated HFPmn construct and two preferred registries
with the overlapping of most N-terminal hydrophobic residues. Finally, chapter Vl
will describe the studies about the membrane insertion of three different HFP
constructs: V2E mutated HFPmn (HFPmn_mut), HFPmn and HFPtr in
membranes both with and without cholesterol. It has been concluded that both
disruption of membrane caused by HFP insertion and fusion activities of these
HFP constructs followed the trend HFPmn_mut < HFPmn < HFPtr, and the

conclusion is independent on the secondary structure of the peptide.

REFERENCE

1. Eckert, D.M.; Kim, P.S., Mechanisms of Viral Membrane Fusion and its
Inhibition. Annu Rev Biochem 2001, 70, 777-810.

2. Piel, J., The Science of AIDS. 1998.

3. Wild, C. T.; Shugars, D. C.; Greenwell, T. K.; McDanal, C. B.; Matthews, T.
J., Peptides corresponding to a predictive alpha-helical domain of human
immunodeficiency virus type 1 gp41 are potent inhibitors of virus infection. Proc
Natl Acad Sci U S A 1994, 91, (21 ), 9770-4.

4. Kilby, J. M.; Hopkins, 8.; Venetta, T. M.; DiMassimo, 8; Cloud, G. A.; Lee,
J. Y.; Alldredge, L.; Hunter, E.; Lambert, D.; Bolognesi, D.; Matthews, T.;
Johnson, M. R.; Nowak, M. A.; Shaw, G. M.; Saag, M. S., Potent suppression of
HIV-1 replication in humans by T-20, a peptide inhibitor of gp41-mediated virus
entry. Nat Med 1998, 4, (11), 1302-7.

5. Kilby, J. M., Lalezari, J.P., Eron, J.J., Carlson, M., Cohen, 0, Arduino,
R.C., Goodgame, J.C., Gallant, J.E., Volberding, P., Murphy, R.L., Valentine, F.,
Saag, M.S., Nelson, E.L., Sista, PR, and Dusek, A., The safety, plasma
pharmacokinetics, and antiviral activity of subcutaneous enfuvirtide (T-20), a
peptide inhibitor of gp41-mediated virus fusion, in HIV-infected adults. Aids Res.
Hum. Retrovir. 2002, 18, 685-693.

6. Hernandez, L. D.; Hoffman, L. R.; Wolfsberg, T. G.; White, J. M., Virus-cell
and cell-cell fusion. Annu. Rev. Cell. Dev. Biol. 1996, 12, 627-661.

7. Grewe, C.; Beck, A.; Gelderblom, H. R., HIV: early virus-cell interactions. J
Acquir Immune Defic Syndr 1990, 3, (10), 965-74.

8. Geyer, H.; Hoischbach, C.; Hunsmann, G.; Schneider, J., Carbohydrates
of human immunodeficiency virus. J. Biol. Chem. 1988, 263, 11760-11767.

9. Chakrabarti, L.; Guyader, M; Sonigo, P. Sequence of simian
immunodeficiency virus from macaque and its relationship to other human and
simian retroviruses. Nature 1987, 328, 543-547.

10. Dedera, D.A.; Gu, R.; Ratner, L. Role of asparagine linked glycosylation in
human immunodeficiency virus type I transmembrane envelop function. Virology
1992, 187, 377-382.

11. Fenouillet, E.; Jones, I.M.; Gluckman, J.C., Functional role of the glycan

cluster of the human immunodeficiency virus type 1 transmembrane glycoprotein
(gp41) ectodomain. J. Virol., 1993, 67, 150-160.

10

12.

out:
1795

13.
that

gym

14.
and
LES”

15.

Aden
Loeh
Igan
Net:

16.
ton
4264

12. Lee, W.R.; Yu, X.F.; Lee, T.H., Mutational analysis of conserved N-linked
glycosylation sites of human immunodeficiency type 1 gp41. J. Virol. 1992, 66,
1799-1803.

13. Perrin, C.; Fenouillet, E.; Jones, I. M., Role of gp41 glycosylation sites in
the biological activity of Human Immunodeficiency virus type 1 envelope
glycoprotein. Virology 1998, 242, 338-345.

14. Bleul, C. C., Farzan, M., Choe, H., parolin, C., ClarkLewis, l., Sodroski, J.,
and Springer, T.A., The lymphocyte chemoattractant SDF-1 is a ligand for
LESTR/fusin and blocks HIV—1 entry. Nature 1996, 382, 829-833.

15. Oberlin, E., Amara, A., Bachelerie, F ., Bessia, C., Virelizier, J.L.,
ArenzanaSeisdedos, F., Schwartz, 0., Heard, J.M., ClarkLewis, I., Legler, D.F.,
Loetscher, M., Baggiolini, M., and Moser, B., The CXC chemokine SDF-1 is the
ligand for LESTR/fusin and prevents infection by T-cell-line-adapted HIV-1.
Nature 1996, 382, 833-835.

16. Weissenhorn, W.; Dessen, A.; Harrison, S. C.; Skehel, J. J.; Wiley, D. 0,
Atomic structure of the ectodomain from HIV-1 gp41. Nature 1997, 387, (6631),
426-430.

17. Murata, M.; Sugahara, Y.; Takahashi, S.; Ohnishi, S., Ph-Dependent

Membrane-Fusion Activity of a Synthetic 20 Amino-Acid Peptide With the Same
Sequence As That of the Hydrophobic Segment of Influenza-Virus Hemagglutinin.
Journal of Biochemistry 1987, 102, (4), 957-962.

18. Wild, C., and Des, T., A synthetic peptide inhibitor of human
immunodeficiency virus replication: correlation between solution structure and
viral inhibition. Proc. Natl. Acad. Sci. USA 1992, 89, (21), 10537-10541.

19. Munoz-Barroso, l.; Durell, S.; Sakaguchi, K.; Appella, E.; Blumenthal, R.,
Dilation of the human immunodeficiency virus-1 envelope glycoprotein fusion
pore revealed by the inhibitory action of a synthetic peptide from gp41. J. Cell
Biol. 1998, 140, (2), 315-323.

20. Furata, R. A., and Wild, C.T., Capture of an early fusion-active
conformation of HIV-1 gp41. Nat. Struct. Biol. 1998, 5, (4), 276-279.

21. Jiang, S., and Lin, K., HIV-1 inhibition by a peptide. Nature 1993, 365,
(6442), 113.

22. Chen, C. H., and Matthews, T.J., A molecular clasp in the human

immunodeficiency virus (HIV) type 1 TM protein determines the anti-HIV activity
of gp41 derivatives: implication for viral fusion. J. Virol. 1995, 69, (6), 3771-3777.

11

wit.

Pro

26.
pr01
Bat

27.
fusc
imrr

28.
ill-ll.
Dept

29.
stud
men

30
Men
Pun

3i
8mg

32.
fish

33.

dome
SODA
the n-

'm8m
CORR
Biol -

23. Bewley, C. A., and Louis, J.M., Design of a novel peptide inhibitor of HIV
fusion that disrupts the internal trimeric coiled-coil of gp41. J. Biol. Chem. 2002,
277, (16), 14238-14245.

24. Chan, D. C., and Kim, P.S., HIV entry and its inhibition. Cell 1998, 93, (5),
681-684.

25. Weissenhorn, W.; Dessen, A.; Calder, L. J.; Harrison, S. C.; Skehel, J. J.;
Wiley, D. C., Structural basis for membrane fusion by enveloped viruses [In
Process Citation]. Mol Membr Biol 1999, 16, (1 ), 3-9.

26. Schibli, D. J., and Weissenhorn, W., Class I and Class II viral fusion
protein structures reveals similar principles in membrane fusion. Mol. Membr.
Biol. 2004, 21, (6), 361-371.

27. Steffy, K.R.; Kraus, G., Looney, D.J., Wong-Staal, F ., Role of the
fusogenic peptide sequence in syncytium induction and lnfectivity of human
immunodeficiency virus type 2. J. Virol. 1992, 66(7), 4532-4535.

28. Horth, M.; Lambrecht, B.; Khim, M.C.; Bex, F .; Thiriart, C.; Ruysschaert,
J.M., Burny, A., Brasseur, R. Theoretical and functional analysis of the SIV fusion
peptide. EMBO J. 1991, 10(10), 2747-2755.

29. Durell, S. R.; Martin, l.; Ruysschaert, J. M.; Shai, Y.; Blumenthal, R., What
studies of fusion peptides tell us about viral envelope glycoprotein-mediated
membrane fusion. Mol. Membr. Biol. 1997, 14, (3), 97-112.

30. Pecheur, E.; Sainte-Marie, J.; Bienvenue, A.; Hoekstra, D., Peptides and
Membrane Fusion: Towards an Understanding of the Molecular Mechanism of
Protein-Induced Fusion. J Membr Biol 1999, 167, (1), 1-17.

31. Epand, R. M., Fusion peptides and the mechanism of viral fusion. Biochim.
Biophys. Acta-Biomembr. 2003, 1614, (1 ), 116-121.

32. Nieva. J. L.; Agirre, A., Are fusion peptides a good model to study viral cell
fusion? Biochim. Biophys. Acta-Biomembr. 2003, 1614, (1), 104-115.

33. Chang, D. K.; Cheng, S. F.; Chien, W. J., The amino-terminal fusion
domain peptide of human immunodeficiency virus type 1 gp41 inserts into the
sodium dodecyl sulfate micelle primarily as a helix with a conserved glycine at
the micelle-water interface. J. Virol. 1997, 71, (9), 6593-6602.

34. Vidal, P.; Chaloin, L.; Heitz, A.; Van Mau, N.; Mery, J.; Divita, G.; Heitz, F.,
Interactions of primary amphipathic vector peptides with membranes.
Conformational consequences and influence on cellular localization. J. Membr.
Biol. 1998, 162, (3), 259-264.

12

35.
inse

W

IBSI

36.
fusi
spe

37.

Win
inn
16l

Synd

35. Chang, D. K.; Cheng, S. F., Determination of the equilibrium micelle-
inserting position of the fusion peptide of gp41 of human immunodeficiency virus
type 1 at amino acid resolution by exchange broadening of amide proton
resonances. J. Biomol. NMR1998, 12, (4), 549-552.

36. Morris, K. F.; Gao, X. P.; Wong, T. C., The interactions of the HIV gp41
fusion peptides with zwitterionic membrane mimics determined by NMR
spectroscopy. Biochim. Biophys. Acta-Biomembr. 2004, 1667, (1), 67-81.

37. Jaroniec, C. P.; Kaufman, J. D.; Stahl, S. J.; Viard, M.; Blumenthal, R.;
Wingfield, P. T.; Bax, A., Structure and dynamics of micelle-associated human
immunodeficiency virus gp41 fusion domain. Biochemistry 2005, 44, (49), 16167-
16180.

38. Gabrys, C. M., and Weliky, D.P., Chemical Shift Assignment and
Structural Plasticity of 8 HIV Fusion Peptide Derivative in
Dodecylphosphocholine Micelles. BBA-Biomembranes 2007, 1768, 3225-3234.

39. Zhang, H. Y.; Neal, S.; Wishart, D. S., RefDB: A database of uniformly
referenced protein chemical shifts. J. Biomol. NMR 2003, 25, (3), 173-195.

40. Overhauser, A.W., Polarization of Nuclei in Metals. Phys. Rev. 1953, 92(2)
411-415.

41. Rafalski, M.; Lear, J. D.; DeGrado, W. F., Phospholipid interactions of
synthetic peptides representing the N-terminus of HIV gp41. Biochemistry 1990,
29, (34), 7917-7922.

42. Kliger, Y.; Aharoni, A.; Rapaport, D.; Jones, P.; Blumenthal, R.; Shai, Y.,
Fusion peptides derived from the HIV type 1 glycoprotein 41 associate within
phospholipid membranes and inhibit cell-cell Fusion. Structure- function study. J.
Biol. Chem. 1997, 272, (21), 13496-13505.

43. Gordon, L. M.; Curtain, C. C.; Zhong, Y. C.; Kirkpatrick, A.; Mobley, P. W.;
Waring, A. J., The amino-terminal peptide of HIV-1 glycoprotein 41 interacts with
human erythrocyte membranes: peptide conformation, orientation and
aggregation. Biochim. Biophys. Acta 1992, 1139, (4), 257-274.

44. Martin, |.; Schaal, H.; Scheid, A.; Ruysschaert, J. M., Lipid membrane
fusion induced by the human immunodeficiency virus type 1 gp41 N-terminal
extremity is determined by its orientation in the lipid bilayer. J. Virol. 1996, 70, (1),
298-304.

45. Pereira, F. B.; Goni, F. M.; Muga, A.; Nieva, J. L., Permeabilization and
fusion of uncharged lipid vesicles induced by the HIV-1 fusion peptide adopting

I3

an e
1977

46.
pcia
fusm

47.

pept
h).l

48.

Schv
deuw
nicrc

49.

J;lh
vesh:
Sept

50.
shuct
“Duh:

SI

lfihnu
Eldpf,

52.
Debit
J.20(
53.

ESOn
bCUn(

ADDIic
back;
l59,(

55.
Strum
mono

an extended conformation: dose and sequence effects. Biophys. J. 1997, 73, (4),
1977-1986.

46. Peisajovich, S. G.; Epand, R. F.; Pritsker, M.; Shai, Y.; Epand, R. M., The
polar region consecutive to the HIV fusion peptide participates in membrane
fusion. Biochemistry 2000, 39, (7), 1826-33.

47. Martin, l.; Defrise-Quertain, F .; Decroly, E.; Vandenbranden, M.; Brasseur,
R.; Ruysschaert, J. M., Orientation and structure of the NHz-terminal HIV-1 gp41

peptide in fused and aggregated liposomes. Biochim. Biophys. Acta 1993, 1145,

(1), 124-133.

48. Taylor, S. E.; Desbat, B.; Blaudez, D.; Jacobi, 8.; Chi, L. F.; Fuchs, H.;
Schwarz, 6., Structure of a fusion peptide analogue at the air-water interface,
determined from surface activity, infrared spectroscopy and scanning force
microscopy. Biophys. Chem. 2000, 87, (1), 63-72.

49. Agirre, A.; Flach, C.; Goni, F. M.; Mendelsohn, R.; Valpuesta, J. M.; Wu, F.
J.; Nieva, J. L., Interactions of the HIV-1 fusion peptide with large unilamellar
vesicles and monolayers. A cryo-TEM and spectroscopic study. Biochimica Et
Biophysica Acta-Biomembranes 2000, 1467, (1), 153-164.

50. Haque, M. E.; Koppaka, V.; Axelsen, P. H.; Lentz, B. R., Properties and
structures of the influenza and HIV fusion peptides on lipid membranes:
Implications for a role in fusion. Biophys. J. 2005, 89, (5), 3183-3194.

51. Kamath, S.; Wong, T. C., Membrane structure of the human
immunodeficiency virus gp41 fusion domain by molecular dynamics simulation.
Biophys. J. 2002, 83, (1), 135-143.

52. Maddox, M. W.; Longo, M. L., Conformational partitioning of the fusion
peptide of HIV-1 gp41 and its structural analogs in bilayer membranes. Biophys.
J. 2002, 83, (6), 3088-3096.

53. Yang, J.; Gabrys, C. M.; Weliky, D. P., Solid-state nuclear magnetic
resonance evidence for an extended beta strand conformation of the membrane-
bound HIV-1 fusion peptide. Biochemistry 2001, 40, (27), 8126-8137.

54. Yang, J.; Parkanzky, P. D.; Bodner, M. L.; Duskin, C. G.; Weliky, D. P.,
Application of REDOR subtraction for filtered MAS observation of labeled
backbone carbons of membrane-bound fusion peptides. J. Magn. Reson. 2002,
159, (2), 101-110.

55. Yang, J.; Prorok, M.; Castellino, F. J.; Weliky, D. P., Oligomeric beta

structure of the membrane-bound HIV-1 fusion peptide formed from soluble
monomers. Biophys. J. 2004, 87, 1951-1963.

14

56.
Vhflh

sped
mem.

57.

shfle
onenl
Phys.

58.

fusor
whhfli
I45). '

59.
FHV-1
Eastl

60.

flenbi
liner
12975

56. Bodner, M. L.; Gabrys, C. M.; Parkanzky, P. D.; Yang, J.; Duskin, C. A.;
Weliky, D. P., Temperature dependence and resonance assignment of ”C NMR
spectra of selectively and uniformly labeled fusion peptides associated with
membranes. Magn. Reson. Chem. 2004,42, 187-194.

57. Wasniewski, C. M.; Parkanzky, P. D.; Bodner, M. L.; Weliky, D. P., Solid-
state nuclear magnetic resonance studies of HIV and influenza fusion peptide
orientations in membrane bilayers using stacked glass plate samples. Chem.
Phys. Lipids 2004, 132, (1), 89-100.

58. Yang, R.; Prorok, M.; Castellino, F. J.; Weliky, D. P., A trimeric HIV-1
fusion peptide construct which does not self-associate in aqueous solution and
which has 15-fold higher membrane fusion rate. J. Am. Chem. Soc. 2004, 126,
(45), 14722-14723.

59. Yang, J. Solid-state nuclear magnetic resonance structural studies of the
HIV-1 fusion peptide in the membrane environment. Michigan State University,
East Lansing, MI, 2003.

60. Zheng, Z., Yang, R., Bodner, ML, and Weliky, D.P., Conformational
flexibility and strand arrantments of the membrane-associated HIV fusion peptide
trimer probed by solid-state NMR spectroscopy. Biochemistry 2006, 45, 12960-
12975.

15

BAC

test
was
nflev
(HOSE
HFPi
hasl
the F
a pe(
fistl
sec01
Ufing
grown
W881
5%.“

Was

CHAPTER II

OPTIMIZATION OF THE SYNTHESIS OF FUSION PEPTIDE OLIGOMERS

BACKGROUND

Chemically synthesized HFP has been considered to be a useful model
to study the fusion peptide/membrane interaction. However, a practical challenge
was the synthesizing of large quantities of pure peptides, especially biological-
relevant HFP oligomers such as HFPtr. One synthetic route of HFPtr was the
cross-linking between HFPmn with one non-native C-terminal cysteine and
HFPmn with two non-native C-terminal cysteines. However, the major product
has been proved to be HFP dimer (HFPdm) formed from cross-linking between
the HFPmn with one cysteine.(1) An alternative approach was initial formation of
a peptide scaffold with three lysines and amide bonds between the E-NHz of the
first lysine and the COOH of the second lysine and between the e-NHz of the
second lysine and the COOH of the third lysine. HFPtr was then synthesized
using standard 9-fluorenylmethoxycarbonyl (Fmoc) chemistry and synchronous
growth of the peptide chains from the three main chain a-NHzfi This approach
was therefore direct synthesis of a ~90-mer and the consequent yield was at best
5%.(2) In addition, the final product contained significant impurity HFPtr which
was non-separable and which contained one additional lysine in one of the
chains.(2) This impurity was a consequence of undesired intramolecular removal
of the Fmoc protecting group on a Lys 0t-NH2 by a free e-NH2.(3) In this chapter,

the optimization of HFP oligomer synthesis will be described. Firstly, the

16

m0

disc

MA

acic
Trp;
(L0i
uror
wert
lDlF
hex;
Sorr
Pierr
som.
Fostr
Drovi
6th;
mm

InfiLJC

TFA!

monitoring of coupling time for individual residues in HFP sequence will be
discussed, and then, the efficient synthetic schemes for HFPdm and HFPtr will

be provided.(4)

MATERIALS

The Gly or Ala-preloaded Wang Resins and N-Fmoc-protected amino
acids Gly, Ala, Ile, Leu, Val, Phe, Ser(tBu), Thr(tBu), Met, Cys(Trt), Arg(be),
Trp(Boc), Trp(Mtt) and Lys(Boc) were purchased from Peptides International
(Louisville, KY). The coupling reagents O-benzotriazole-N,N,N',N-tetramethyl-
uronium-hexafluoro-phosphate (HBTU) and 1-hydroxybenzotriazole (HOBT)
were purchased from Novabiochem (San Diego, CA). N, N-diisopropylethylamine
(DIPEA) and 7-azabenzotriazol-1-yloxy-tris-(pyrrolidino)phosphonium
hexafluorophosphate (PyAOP) were obtained from Sigma-Aldrich (St. Louis, MO).
Some of the synthesis was done manually in 5 mL polypropylene columns from
Pierce (Rockford, IL) and mixing was accomplished with a rotation stage, and
some of the synthesis was done on an automated peptide synthesizer (ABI 431A,
Foster City, CA). The detailed procedures for the manual peptide synthesis were
provided in appendix 3. All peptides were purified using high-performance liquid
chromatography (HPLC) (Dionex, Sunnyvale, CA) equipped with a 10 mm * 250
mm C18 column (Vydac, Hesperia, CA). “Buffer A” was water with 0.1%
trifluoroacetic acid (TFA), “buffer B” was 90% acetonitrile, 10% water, and 0.1%

TFA, and the gradient was 40% to 80% buffer B over 30 minutes. Peptide

l7

HFI

A Ilne
and the
\

masses were measured with matrix-assisted laser desorption/ionization-time of

flight (MALDI-TOF) mass spectrometry using a Voyager-DE STR

Table 1. Names and sequences of the HIV fusion peptides

 

Name Sequence 8

 

HFPmn AVGIGALFLGFLGAAGSTMGARSWKKKKKKAl3

HFPmn(Cys) AVGIGALFLGFLGAAGSTMGARSWKKKKKCA”

HFPmn(Cys/Gly) AVGIGALFLGFLGAAGSTMGARSWKKKKKCG

AVGIGALFLGFLGAAGSTMGARSWKKKKKKA’3

HFPdm(Cys)
AVGIGALFLGFLGAAGSTMGARSWKKKKKC
AVGlGALFLGFLGAAGSTMGARSWKKKKKCA’
HFPdm
AVGIGALFLGFLGAAGSTMGARSWKKKKKCA"
AVGIGALFLGFLGAAGSTMGARSWKKKKKKA”
HFP" AVGIGALFLGFLGAAGSTMGARSWKKKKK
AVGIGALFLGFLGAAGSTMGARSWKKKKKCG
AVGIGALFLGFLGAAGSTMGARSWKKKKKKA”
AVGIGALFLGFLGAAGSTMGARSWKKKKKC
HFPte

AVGIGALFLGFLGAAGSTMGARSWKKKKKC

AVGIGALFLGFLGAAGSTMGARSWKKKKKKA’

 

a'A line between K and C denotes a peptide bond between the Cys CO
and the Lys s-NH and a line between two Cs denotes a disulfide bond.

biospectrometry workstation (Applied Biosystems, Foster City, CA) and a-cyano-
4-hydroxy cinnamic acid matrix. Peptide synthetic yields were quantified using

280 nm absorbance and the extinction coefficients were 5700, 11600 and 17300

18

monor
wee

seque
that (1

reachc

RESU

HFPn
forfiw

Ofeac

cm" M" for HFPmn, HFPdm and HFPtr respectively. The sequences of HFP
monomers, dimers, trimer and HFP tetramer (HFPte) described in this chapter
were summarized in Table 1. The N-terrninal 23 residues come from the
sequence of gp41. The C-terminal tags contain Lys, Cys and Trp for the reasons
that (1) Lys increases the solubility of HFPs, (2) Cys enables the cross-linking

reactions and (3) Trp helps to quantify the synthesis.

RESULTS AND DISCUSSION

Optimization of coupling time. A manual synthesis was carried out on
HFPmn according to the scheme provided in Figure 3. The detailed procedures
for the manual synthesis of HFPmn are provided in the appendix 3. The coupling
of each residue was optimized by ninhydrin monitoring every two hours to detect
free a-HNz groups.(5) This information provided the basis for longer coupling
times and double couplings at particular residues. It was observed that two-hour
single coupling was sufficient for the residues along the sequence except for
Ser(tBu), Arg(be), and Trp(Boc) residues as well as the residues between Leu-
12 and Leu-7. A complete coupling for these residues were detected for a
coupling time 4~6 hours which means either longer coupling times or double
coupling should be used. Figure 4 shows the HPLC and mass spectrum for the
purified HFPmn using the optimized coupling time. The mass spectrum of the
dominant fraction had an intense peak with m/z = 3149 which was within the
instrumental accuracy of :t 0.1% relative to the expected m/z = 3151. The product

of synthesis was also subjected to the amino acid analysis by the

19

Macrc

Bloch.

showe

1.720

Macromolecular Structure Facility of Michigan State University (Department of
Biochemistry and Molecular Biology, Michigan State University). The results
showed that the sequence contained 6 Gly, 5.922 Ala, 6.257 Lys, 2.711 Leu,
1.720 Phe, 1.880 Ser, 1.080 Met, 1.003 Trp, 1.298 Arg, 1.115 Thr, 0.700 Val and
0.795 Ile, which were consistent with the sequence of HFPmn given in Table 1.
The difficulties for the coupling of Ser(tBu), Arg(be) and Trp(Boc) may be due to
the great size of side chains or side chain protection groups which cause steric
problem and block the active amino group from the newly added amino acid. The
difficulties for the coupling of the residues from Leu7 to Leu12 may come from
the continuous increase of hydrophobicity during the coupling of these apolar
residues, which will further cause the aggregation of peptide chains and prevent
the further coupling.(6)

Synthesis of HFPdm. Figure 5 shows the synthetic schemes of two types
of HFP dimers: HFPdm and HFPdm(Cys). HFPdm was synthesized using the
Cys cross-linking reaction of Cys contained HFPmn (named HFPmn(Cys)) and
HFPdm(Cys) was synthesized using the dimeric scaffold with 9-
fluorenylmethoxycarbonyl (Fmoc) chemistry.(7) HFPdm(Cys) will serve as one of
the building blocks in the synthesis of HFPtr which will be described in the next
session. Figure 6 displays the HPLC and mass spectra for the purification and
identification of HFPdm and HFPdm(Cys). The HPLC retention time of HFPdm
was very well-separated from that of unreacted HFPmn(Cys). The HFPdm
fraction had a mass spectral peak with m/z = 6248 which was within the

instrumental accuracy of :l: 0.1% relative to the expected m/z = 6253, cf. Figure

20

60.1
HFPr
specl
(nevi
dnuh
incwh
fluent
then:
the o
hwern
31 Li
finenl
Spect
wasv
A Dre
imDUr
Fmoc
SUbSE

SCaIic

6b. There was also a peak at m/z = 3124 which corresponded to either the
HFPmn(Cys) fragment formed from cleavage of the disulfide bond in the mass
spectrometer or to doubly charged HFPdm.(8) It has been demonstrated
previously that there was prompt fragmentation where the single intermolecularly
disulfide—bound peptides can be specifically fragmented at the S-S bond by
increasing the laser fluence in MALDl-MS.(8) Since we always used high laser
fluence in the identification of the HFP oligomers to obtain reasonable signal
intensities, it was possible that the interstrand S-S bond was fragmented due to
the over-threshold laser fluence. The peak at m/z = 5733 corresponded to the
internal standard of insulin from bovine pancreas (purchased from Sigma-Aldrich,
St. Louis, MO). Figure 6c displays the chromatogram of the synthesis and the
retention time of HFPdm(Cys) was very close to that of HFPdm. The mass
spectrum of the HF Pdm(Cys) fraction had an intense peak with m/z = 6190 which
was within the :l: 0.1% uncertainty relative to the expected m/z = 6188, cf. Fig. 6d.
A previously published synthesis showed significant higher molecular weight
impurities which was the result of: (1) premature removal of the scaffold lysine
Fmoc group by nucleophilic attack of the scaffold lysine s-NHz; and (2)
subsequent coupling of the next amino acid with both the e- and the ot-NH2 of the
scaffold Iysine.(2) The mass spectrum of HFPdm(Cys) did not show the impurity
corresponding to a peptide with an extra Cys. Minimization of the time between
steps fand d in Figure 5 was critical to eliminating this impurity.

Synthesis of HFPtr. The synthesis scheme for HFPtr was displayed in

Figure 7. HFPtr was formed from cross-linking HFPmn(Cys) and HFPdm(Cys) in

21

a 10:1

amount

The n01

showed

18pm fit
83 shov
chromat
the th
HFPdnu
chromatr
respectiv
HFPmn((
sliectrun
Olit unl
peak at i
6194 We
deavage
5733 cm
the HFPl
Clogs,“m
Sepalaho
Showed I

12461 an

a 10:15 mol ratio. The cross-linking reaction was done at pH = 8.4 with trace
amount of 4-dimethylaminopyridine (DMAP) and with the system open to the air.
The non-stoichiometric ratio was based on initial small-scale syntheses which
showed that cross-linking of HFPmn(Cys) with itself to form HFPdm was more
rapid than cross-linking of HFPmn(Cys) with HFPdm(Cys) to form HFPtr. Figure
8a showed the monitoring of the cross-linking reaction using HPLC. The top
chromatogram in Figure 8a was obtained after 0.5 hour cross-linking time and
the three prominent peaks from left-to-right were HFPmn(Cys),
HFPdm/HFPdm(Cys), and HFPtr, respectively. The middle and bottom
chromatograms were obtained with cross-linking times of 1.5 and 2.5 hours,
respectively, and showed a relative increase in HFPtr and relative decreases in
HFPmn(Cys) and HFPdm(Cys) with longer cross-linking time. The mass
spectrum of the HF Ptr fraction had a peak with m/z = 9307 which was within :1:
0.1% uncertainty relative to the expected HFPtr m/z = 9312, cf. Figure 8b. The
peak at 4653 was assigned to doubly charged HFPtr and the peaks at 3112 and
6194 were assigned to HFPmn(Cys) and HFPdm(Cys) fragments formed from
cleavage of the disulfide bond in the mass spectrometer.(8) The peak at m/z =
5733 corresponded to the internal standard of bovine insulin. One side product in
the HFPtr synthesis was the HFP tetramer (HFPte) which was formed by the
cross-linking between two HFPdm(Cys) molecules. Figure 9a displayed the
separation of HFPte from HF Ptr at the retention time ~ 22 minutes and Figure 9b
showed the mass spectrum of HFPte. There were broad signals peaked at m/z =

12461 and 6235 which were respectively comparable to the expected HF Pte m/z

22

= 123'.
formet
to the
the exl

lower 5

a 3101
scaffolt
been rt
lollowin
reactior
SInthes
reactior
Cross-Iii
HPLC I
and HF
SIDIhES
TIDIBOC
DlDloco
HFPdm
deavag
This mc

Q'NHz b

= 12374 and the m/z of the doubly charged species or the HFPdm(Cys) fragment
formed from cleavage of the disulfide bond in the mass spectrometer.(8) Relative
to the HFPmn, HFPdm, and HFPtr spectra, there were greater uncertainties of
the experimental m/z in the HF Pte spectra because of both broader signals and a
lower signal-to-noise ratio.

The overall yield for the HF Ptr synthesis was about 15 % which indicated
a 3-fold increase compared with the previous coupling scheme from the trimeric
scaffold.(2) In addition, the side product with an extra Lys in the C-terminus has
been removed. The present study increased the HFPtr yield and purity using the
following modifications: (1) HFPtr was formed from a cysteine cross-linking
reaction between HFPmn(Cys) and HFPdm(Cys). Because HFPdm(Cys) was
synthesized using a dimeric scaffold, a successful synthesis required 1/3 fewer
reactions than the earlier HFPtr synthesis. In addition, the purification of the
cross-linking reaction was fairly straightfonlvard because of the separation of the
HPLC peaks corresponding to HFPmn(Cys), HFPdm and HFPdm(Cys), HFPtr,
and HFPte, cf. Figure 8. (2) The monitoring of coupling times in the manual
syntheses of HFPmn showed that longer coupling times were required for the
Trp(Boc), Ser(tBu), Arg(be), and Leu-7 to Leu-12 residues. The new synthetic
protocol used longer coupling times or double coupling at these residues. (3) The
HFPdm(Cys) synthetic protocol was modified to minimize the time between the
cleavage of the Mtt group of the Lys 8-NH2 and the subsequent coupling to Cys.
This modification reduced undesired deprotection of the Fmoc group of the Lys

a-NHz by the e-NHz.

23

CIDDI

peptic
hydro
using
residt
betwe
of the
produ
provid
and rg

and hi

CONCLUSION

This chapter reported efforts on the optimization of the conditions of
peptide synthesis, especially HFPtr. It can be concluded that (1) the
hydrophobicity of HFP sequence increases the challenge of the synthesis and
using either longer coupling times or double coupling for the hydrophobic
residues helps to improve the synthesis, (2) it is important to shorten the time
between the cleavage of Lys e-NHz side chain protection group and the coupling
of the following residue to the unprotected S-NHz in order to eliminate the
producing of by products with extra Lysines and (3) the Cys cross-linking strategy
provides straightfonlvard HPLC purification for the hydrophobic HFP oligomers
and reasonable yield, and should be applicable to the synthesis of other homo

and hetero trimeric peptides.

24

Fm0«

Figure 3
AVGIGAI
are draw
reaction.
reactions
deprotect
Reaction
used IOII
for Trp, E
residues.
TFAlthioa
reactiont
with cold

 

a, b c
—. ———-> FPW Ala
Ala ——> Kg

 

Fmoc

Figure 3 Synthesis scheme for HFPmn and FF represents the sequence
AVGIGALFLGFLGAAGSTMGARS. A black circle represents a resin bead, lines
are drawn to clarify chemical functionalities, an arrow signifies a chemical
reaction, and two arrows signify multiple sequential chemical reactions. All
reactions were carried out at ambient temperature. Reaction a: Fmoc
deprotection in 3 mL of 20% piperidine/DMF (v/v), 15 minutes/cycle, 2 cycles.
Reaction b: Peptide synthesis with Fmoc chemistry. 2-hour single couplings were
used for each amino acid with the following exceptions: 4-hour single couplings
for Trp, Ser and Arg residues; 6-hour single couplings for the Leu-12 to Leu-7
residues. Reaction c: Cleavage from the resin using a 4 mL solution containing
TFA/thioanisole/ethanedithioI/anisole in 90:5:3:2 volume ratio.(6) After 2.5 hours
reaction time, TFA was removed with nitrogen gas and peptide was precipitated
with cold methyl t-butyl ether.

25

 

 

(a) (b)

 

 

Absorption (a.u.)
Intensrty (a u )

~~MW~JJANM~~ ”2......

 

 

 

 

 

 

0 30 2500 4000

Elution Time (min) m/z

Figure 4 (a) The HPLC chromatograms for the purification of HFPmn. The
horizontal axis was the elution time in unit of minute and the vertical axis was the
absorption at 214 nm with arbitrary unit. (0) The MALDI-TOF MS spectrum for
the identification of HFPmn. The HPLC fraction marked with asterisk in (a) was
analyzed and the corresponding mass was labeled using asterisk in (b).

26

d,a

 

a.
(a) Fmoc—AIa—. ——> HZN Cys—Ala—.
Trt
b c
—> ——-> HzN—FPWKs—Cys—Ala
I
SH

HzN—FPWKg—Cys—Ala HzN-FPWKg—Cys—Ala

 

 

I + I

SH SH
HzN-FPWKg—Cys—Ala

e ‘
HzN-FPWK5—Cys—Ala
,f
(b) Fmoc—AIa—C a > Fmoc—Tys—Ala—O
Mtt
9 d, a

 

 

> Fmoc--Lys-AIa—. =

NH2

8 c '
HzN—Lys—Ala—. —> ——> HZN FPWK5 Lys—Ala
H2N—FPWK5—Cys

l
SH

HzN—Cys

Figure 5 Synthesis schemes for (a) HFPdm and (b) HFPdm(Cys). The black
circles, lines and arrows had the same meaning as in Fig. 3. The reactions a and
c were the same as in Fig. 3. In reaction b, the synthesis of HFPmn with Cys(Trt)
in (a) followed the coupling time in Fig. 3. For the HFPdm(Cys) synthesis in (b),
4-hour single couplings were used for each amino acid with the following
exceptions: 8-hour single couplings for Trp(Boc), Ser(tBu), and Arg(be) residues;
double couplings with 4-hours per coupling for the 1300 labeled residue and for
the Leu-12 to Leu-7 residues. Reaction d: Coupling using PyAOP and DIPEA
(1:2 molar ratio) in 4 mL DMF with 6 hour reaction times for Cys and 2 hour
reaction time for Lys. Reaction 8: Cross-linking in 5 mM DMAP, pH = 8.4, open to
the air. 1 umol HFPmn(Cys) in 400 tiL solution overnight. Reaction f: Selective
deprotection of Mtt in 3 mL of 1% TFA/DCM (v/v), 6 minutes/cycle, 6 cycles.

27

Absorption (n u.)

Absorption (a u.)

Absorption (a.u.)

Absorption (a.u.)

Figure 6 (a) and (c) The HPLC chromatograms for the purification of HFPdm and
HFPdm(Cys). The horizontal axes were elution time in unit of minute and the
vertical axis were absorption at 214 nm with arbitrary unit. (b) and (d) The
MALDI-TOF MS spectrum for the identification of HFPdm and HFPdm(Cys). The
HPLC fraction marked with asterisk in (a) and (0) were analyzed and the
corresponding peaks were labeled using asterisk in (b) and (d) respectively. The

 

Mb...— _

i
I l.‘ A J] \\
W“ \_ ,f ’

\-__ ~.- ewe-.4

 

 

 

 

 

 

0 Elution Time (min) 30
*
(C) ‘
r“ "Ji \f,
\Ji , \L-) “\W 4
0 30

Elution Time (min)

 

Intensity (a.u.)

Intensity (a.u.)

 

 

 

 

 

 

 

 

 

 

2000 m/z 15000
*
(d)
2500 15000

m/z

mass spectra were discussed in the main text.

28

 

Fm

le

HZN
HzN

Figui
Iiie 5
as st
DH =
Ill. 81

 

 

 

 

 

 

 

 

a, f 9
Fmoc Ala—. = Fmoc Lys—Ala—. 4’
Mtt
d, a
Fmoc Lys—AIa—. = HzN liys—AIa-O
NH; H2" Ty:
Tl't
H N—FPWK —L s—Ala
o c 2 5 IV
—’ HzN—FPWKg—Cys
I
SH

HzN—FPWKg—Lys—Ala
HzN—FPWKg—Cys—Ala

I

HzN—FPWKg—Cys +
I srr
SI-I

HzN—FPWKs —Ly8—'— Ala
I
HzN—FPWKg—Cys

HzN—FPWKs—Cys—Ala

Figure 7 Synthesis schemes for HFPtr. The black circles, lines and arrows had
the same meaning as in Fig. 3 and Fig. 5. All reaction conditions were the same
as shown in Fig. 3 and Fig. 5 except for reaction 8: Cross-linking in 5 mM DMAP,
pH = 8.4, open to the air. 1 (mol HFPmn(Cys) and 1.5 pmol HFPdm(Cys) in 400

piL solution for 2.5 hours.

29

(3)

Absorption (a.u.)

Fl9ure l
hohzont
$50er
the iden
a are 10
"331160111
indlCaIer
”9m res
and the
speCIIa I

 

 

 

 

 

 

(a) t g
‘7' E 5 \2
' r5 3 5 7M
//§
:2 E E S
to E I. E
v : :4 3‘
w 3 ,1: 1
E- E '5 E
o ff; 3 s c
(I) /: i 1
..Q -- ’ : : :
< : s;
E 1"" A
2 y \
:/ \‘\‘~.
0 30

Elution Time (min)

Figure 8 (a) The HPLC chromatograms for the purification of HFPtr. The
horizontal axes were elution time in units of minutes and the vertical axes were
absorption at 214 nm with arbitrary units. (b) The MALDI-TOF MS spectrum for
the identification of HFPtr. The top, middle, and bottom chromatograms in panel
a are for syntheses with HFPtr cross-linking times of 0.5, 1.5, and 2.5 hours,
respectively. In each chromatogram, the positions of the vertical dotted lines
indicated the elution times of monomer, dimer and trimer for the left, middle and
right respectively. The HPLC fraction marked with asterisk in (a) was analyzed
and the corresponding peaks were labeled using asterisk in (b). The mass

spectra were discussed in the main text.

30

Intensity (a.u.)

(b)

 

 

* *

LJL _

 

 

 

1500

15000

m/z

 

 

 

(a) (bl*

 

 

 

 

 

 

 

31‘ -.~
g 3.
c ,, 3
g .. 3:

i i a
g- ‘2 2's r. 8 l
.8 " '._\ g *
< /// ‘\ H

l
, 30 4500 15000
Elution Time (min) m/z

Figure 9 (a) The HPLC chromatograms for the purification of HFPte. The
horizontal axis was the elution time and the vertical axis represented the 214 nm
absorption in arbitrary unit. (b) The MALDl-TOF MS spectrum for the
identification of HFPte. The HPLC fraction marked with asterisk in (a) was
analyzed and the corresponding mass was labeled using asterisk in (b).

31

REF

1
sold
tfiocl

2.

tumor
whhl
145).

3.
Fnux
Lede.

4.

0190
Cone
IoLip

5
ddec
thh

6.
Fhacn

7

2,2,2.
IIUOre
Damn

8.
dunng
Chem

REFERENCE

1. Yang, R.; Yang, J.; Weliky, D. P., Synthesis, enhanced fusogenicity, and
solid state NMR measurements of cross-linked HIV-1 fusion peptides.
Biochemistry 2003, 42, (12), 3527-3535.

2. Yang, R.; Prorok, M.; Castellino, F. J.; Weliky, D. P., A trimeric HIV-1
fusion peptide construct which does not self-associate in aqueous solution and
which has 15-fold higher membrane fusion rate. J. Am. Chem. Soc. 2004, 126,
(45), 14722-14723.

3. Farrera-Sinfreu, J.; Royo, M.; Albericio, F., Undesired removal of the
Fmoc group by the free epsilon-amino function of a lysine residue. Tetrahedron
Letters 2002, 43, (43), 7813-7815.

4. Qiang, W., and Weilky, D.P., HIV Fusion Peptide and Its Cross-Linked
Oligomers: Efficient Syntheses, Significance of the Trimer in Fusion Activity,
Correlation of 8 Strand Conformation with Membrane Cholesterol, and Proximity
to Lipid Headgroups. Biochemistry 2009,48, (2), 289-301.

5. Kaiser, E., Colescott, R.L., Bossinger, CD, and Cook, P.l., Color test for
detection of free terminal amino groups in solid-phase synthesis of peptides. Anal.
Biochem. 1970, 34, 595.

6. Chan, W. C., and White, P.D., Fmoc Solid Phase Peptide Synthesis: A
Practical Approach 2000, 94-109.

7. Lapatsanis, L.; Milias, G.; Froussios, K.; Kolovos, M., Synthesis of N-
2,2,2-(trichloroethoxycarbonyl)-L-amino acids and N-(9-
fluorenylmethoxycarbonyl)-L-amino acids involving succinimidoxy anion as a
leaving group in amino-acid protection. Synthesis-Stuttgart 1983, (8), 671-673.

8. Patterson, S.; Katta, V., Prompt fragmentation of disulfide-linked peptides

during Matrix-assisted Laser Desorption Ionization Mass Spectrometry. Anal.
Chem. 1994, 66, 3727-3732.

32

Till

info
FtEl

che

sect
Cha
thus
to 11
con:
Spe.
inlet
”)8
VVth

Intrc

i”CIL
CYell

meg

CHAPTER III

OPTIMIZATION OF SOLID-STATE NMR EXPERIMENTS

THERORETICAL BACKGROUND

Solid-state NMR is a useful tool to provide high-resolution structural
information for the membrane-associated peptides.(1-3) For example, for the
REDOR pulse sequence which will be used in the following chapters, the
chemical shift of one specifically-labeled ”C or the dipolar coupling between a
13C and a 15N, 31P or 19F provide valuable structural properties such as
secondary structure, tertiary structure and internuclear distances.(4-7) One
characteristics of the solid samples is the absence of free molecular tumbling and
thus the co-existence of multiple orientations of chemical shielding tensor relative
to the static external magnetic field for a single spin such as a ”C. This will
consequently cause the broadening of the resonance and complicate the
spectrum. In addition, one always wants to extract a specific intra-system
interaction such as isotropic chemical shift or heteronuclear dipolar coupling from
the many other intra-system interactions. In this first part of this chapter, a theory
which has been widely used to analyze solid-state NMR pulse sequences will be
introduced and the analysis of some pulse sequences will be given as examples.

Average Hamiltonian Theory (AHT). Many pulsed NMR techniques,
including the methods used in the following work, apply some combination of the
cyclic radiofrequency (rf) irradiation (also known as pulses) and the cyclic

mechanical manipulation to the nuclear spin systems. This will consequently

33

make
time
Aver
the s
systt
mult‘i
cons

can I

I56H

the

make the intra-system interaction Hamiltonians time-dependent, and make the
time averaging of such Hamiltonians important to the final spectrum. The
Average Hamiltonian Theory (AHT) is helpful to understand the performance of
the spin system under such time-dependent interactions. It states that when a
system is subjected to a cyclic external force, and the inspection is restricted to
multiple integers of the cycle time, it looks like the system is subjected to a
constant Hamiltonian.(8) Three conditions should be satisfied before the AHT
can be applied to analyze a pulse sequence:
(1) The cycle time of rf pulses should be a multiple of the cycle time of
mechanical manipulation;
(2) The net rotation after the block of If pulses should be zero;
(3) The cycle time of n‘ pulses should be small so that the internal Hamiltonian
will not cause significant changes to the spin system during one cycle.
These conditions can be satisfied by carefully choose experimental parameters
such as magic angle spinning (MAS) frequency, rf pulse length, time interval
between pulses and phase cycling strategy.

The lowest-order approximation of the constant Hamiltonian in AHT can

be expressed by Eq. 1

-__1_ C .‘int
H-1ngrH (t) (1)

to represents the cycle time of rf pulses and Him is the interaction Hamiltonian in

the toggling frame (or rf frame).

19"“(0-U;‘(t)-H‘"‘(t)-U,ftr) (2)

34

Fora
hanm
Hani

Until

The t
he p

DUDCE

a); (r)

Phase

IESpe

The e
USed
Obhwr
Hamil
aid It

and D

0“1 CC

QEHEH

For all the discussions below, l-l‘"‘(t) has always been expressed in the rotating
frame under the condition of MAS and after the truncation of the Zeeman
Hamiltonian. The time-dependence of H"‘(t) is generally introduced through MAS.

U,r(t) is a unitary operator defined by the rf pulses.

U’f(t) = fexpr—i (3 dt '- 11"!” (r )1 (3)

The operator fis the Dyson time operator which specifies the time ordering of
the pulse sequence. H”(t) is also expressed in the rotating frame and with the

truncation of Zeeman Hamiltonian, and has the general form
H'f (r) at 73, (r) -[1x 005 W) + Iy sin ¢(t)] = a), (r) - [1, cos (00) + Iy sin ¢(t)] (4)
co,(t) is usually named Rabi frequency and Ix and ly are spin operators. ¢(t) is the

phase of a if pulse and generally 0', 90°, 180° and 270° for x, y, -x, -y pulses
respectively.

The expression of H"‘(t) is dependent on the specific internal interaction.
The expressions for chemical shift (CS) and dipolar coupling (D), which will be
used in the following chapters, will be introduced here. A general procedure to
obtain the expression for any internal Hamiltonian is to initially write the
Hamiltonian in the principal axis system (PAS) as him which is time-independent,
and then to convert into the rotating frame. l-lim has the following form for both CS

and D
. . I . .
Hlflt = Cult .; z (_l)m _ R1113," . I)"; (5)
m=—l ’ i

0"“ contains constants which are different for CS or D. The “R” and “7" terms are

generally expressed as the second-rank tensor in spherical coordinates, and

35

repri

and

In th

CS;

The

alw;
rotat

angl

The

represent the spatial and spin parts of the Hamiltonian respectively. I = 0, 1 or 2
and m values are integers from —I to l.

Ci'“ 5 y‘(int a CS) and Cint a —27"yfh(int a D) (6)
In the PAS, the “R’ terms have their simplest form and the non-zero “R” terms for

CS and D are

R57 = 7% of (7..)

R63” =%'(§Il +622 +533)

R2?” = (IV—21633 -%-(5n +522 +533)] (7b)

R2282, =%'(522 -5”)
The rij in Eq. 7a represents the internuclear distance between two spins iand j. In
Eq. 7b, 611, 622 and (2,3 are three principal values of spin iwith the sequence (in <
(22 < (2,3. In most solid-state NMR techniques with MAS, the Hamiltonians are
always expressed in the rotating frame. The “R” terms can be converted into the
rotating frame from the PAS with the Euler transformation and two sets of Euler

angles.

Riff”, (rotating _ frame) = Z Ding,” (a ")2 Din-,m.(Q )RjgxPAS) (8)
m' m'

The “D” terms are elements in the V\figner rotation matrix. The first set of Euler

angles Q's(a',,6',y') makes a connection between the PAS and some crystal

frame axes system and for powder samples; these angles will finally be

integrated over a spherical surface. The second set of Euler angle always has

the value Q".-:(O,,B”,th) under the condition of fast MAS, where ,B"=54.7° or

36

“magic angle” and wR is the spinning frequency. The second Euler transformation
introduces the time dependence to the “R” terms which is related to the MAS

rotor cycles.

A full expression of “R" terms in the rotating frame will contain 25 terms
due to the Wigner expansion. However, the “7" terms expressed in the rotating
frame are truncated by the Zeeman Hamiltonian which help to simplify the form

of the total Hamiltonian.

T mt(rotatmg frame): U 2'3"" (lab__ frame)U Z (9)
=exp[iwt1 ]T,mt (lab frame)exp[—ia)tlz ]

w is the Larmor frequency and the non-zero “7" terms in the rotating frame are:

To?" =1§B0 and ngt" = yrg’ao (10a)

—. - ~

T013” =1'-11andT2{’0‘f =(i/f)[31;1;’ —i"-if] (100)
72,00” =1’S-z" and T2135” = yrgsg' (10c)

Eqs. 10a-c give the operators of chemical shift, homonuclear dipolar coupling
and heteronuclear dipolar coupling respectively. One typically incorporates Eqs.
6-8 and 10 into Eq. 2 which will further bring time dependence into both “R” and
“7” terms. Many solid-state NMR pulse sequences make use of the combination
of these two time dependences to achieve selective averaging when applying Eq.

1. Two pulse sequences used in the following chapters are analyzed with AHT.

37

 

‘
s, 3
_I_
Rotor 0

Figure 10
.112 and
magnetize
by one 15I
shift.

an

‘H II cpl TPPM decoupllng

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

..: élllll'JLIHIIIIIIITIIIII

Figure 10 1D ”C-”N REDOR pulse sequence. The open columns represent the
m’2 and ”pulses. CP transfers 1H transverse magnetization to ”C. ”C
magnetization is dephased (i.e., reduced) by ”C-‘5N dipolar coupling mediated
by one 15N 7r pulse per rotor period. The ”C It pulses refocus the ”C chemical
shift.

38

hhAE
REE
reso
prim
data
Figu
peric
were
colle
but I

both

tram

The
le0
The
96cc

SEqL

2610

Theory of REDOR experiments. REDOR is one of the most widely used
MAS NMR techniques for analysis of molecular structure in the solid-state.(9-14)
REDOR has found application in detection of formation of chemical bonds and in
resonance assignment of small peptides.(15-17) The wide use of REDOR is
primarily due to the relative simplicity and robustness of its pulse sequence and
data analysis. An example of a ”C-”N REDOR pulse sequence is sketched in
Figure 10. It contains: (1) 1H to ”C cross polarization (CP), (2) a dephasing
period (1') during which a series of ”C and 15N 7: pulses and 1H decoupling field
were applied and (3) ”C acquisition. For each 1, two separate spectra are
collected. The spectrum named “80” is It pulses applied only on the ”C channel
but not on the ”N channel. The spectrum called “81” is the pulses applied on
both channels. For a sample with a single ”CO label on one residue and a single
”N label on another residue, the major interaction Hamiltonians in the rotating
frame for the labeled ”C are

HW’ = HCS + ch + H€N (11)

The terms on the right side of Eq. 11 indicate ”C chemical shift, homonuclear
dipolar coupling between ”Cs and heteronuclear coupling between ”C and ”N.
The heteronuclear dipolar coupling between ”C and 1H can be removed by 1H
decoupling and will not be included in the further discussion. The goal of this
sequence is to selectively detect the ”C and ”N dipolar coupling as well as the
”C isotropic chemical shift, while averaging all the other interactions. The non-

zero term in the general expression of the dipolar coupling Hamiltonian was

H5? = [A - ez’wk’ + B . e’a’R’ + C - (“”11" + D - e‘Z’wR‘ 1 . 1:5; (12a)

39

Eqs.

respe
the E
matri:

9111

iniwh
settj
shit i
0f Mi

0001

x(

While

23'
(a

Fors(
the ar
first a

IhEnS

HUI? =[A-e2‘0’R‘ +3.60” +C-e"“’R’ +D.e‘2’“’R’]-[31;1g —i" if] (120)

Eqs. 12a and b are for heteronuclear and homonuclear dipolar coupling
respectively. The time-independent A, B, C and D are functions of Cint in Eq. 5,

the Euler angle set Q'E(a',,6',y') defined by corresponding Wigner rotation

matrix elements and the internuclear distance r;,-. The non-zero term for chemical

shift was

11,-CS = [at-so + (A - e210”?! + B - eia’R' + C - e—in' + D ~19—21w”) Jam-so]- 1:80 (13)
in which the terms A, B, C and D are functions of 0"“ in Eq. 5 and the Euler angle
set Q'a(a',,6',y'). Eq. 13 indicates time-independence for the isotropic chemical
shift and time-dependence for the anisotropic chemical shift. Under the condition

of MAS and no If pulses, The time integral of H5.) in Eq.12a over the first half of

one rotor cycle is

f d(th) - Hymn
0c {A[e2”‘ - I]+ late"r — 1] — C(e‘”r - I]— D[e_2i" — 1]} - 11's,! = (—B+ C) - Igsj

(14a)

while over the second half of the cycle is

E” d(er).H,j?(mRr)

x{A[e4ifl' _82ifl]+B[82ifl _eifi]_C[e-Zilr _e-ifl]_D[e—4III_e-21'7!]}.I:Szj =(B-C)'I;Sé’
(14b)

For similar reasons, the homonuclear dipolaricoupling Hamiltonian in Eq.12b and
the anisotropic part of chemical shift Hamiltonian in Eq.13 will be reversed for the
first and second half of one rotor period. According to Eq.1, the spin system is

then subjected to a lowest-order averaged Hamiltonian

4o

How
infor

char

A001
and

and

(SO 2

_ _ i C/Z . total C . total j
H_TC(£ dt H (r)+fC/2dt H (t))0CO'- 130 (15)

However, the ”C-”N heteronuclear dipolar coupling which includes the distance
information has to be recovered with the It pulses applied on ”C and/or ”N
channels. The Hamiltonian expressed in Eqs. 12-13 can be generally written as
HDJrelero = (1(0'1252 (163)
Halo... = din-131:1." ..rr 7’] (16b)
HCS,anl'so =d(t)'lz (16C)
According to Eq. 14, the sign of the integral of d(t) will be flipped during the first
and second halves of a rotor cycle. In addition, the sign of the spin operators l2
and 82 will be flipped by the It pulses applied on the corresponding channels.
Table 2 summarizes the sign of Ho and Hogan-so in So and S1 spectra over eight
rotor cycles based on the phase cycling scheme. It is clear that the heteronuclear

dipolar coupling has a non-zero time averaging over the entire period, thus the

difference spectrum will only reflect the heteronuclear dipolar coupling.

Table 2. Hamiltonians in REDOR So and 81 spectra

 

 

half rotor
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

cycle
HD,hetero + - - + + - - + + - - + + - - +
$0 HD,homo + - + - + - + - + - + - + - + -
HCS,aniso + - - + + - - + + - — + + - - +

HD,hetero++++++++++++++++
31HD,homo+'+'+'+'+'+'+'+-

Hcs'am-so + - - + + - - + + - - + + - - +

 

41

con
Tat
cou
ave

3V8

The
It is
corr
the
peri

p10

Whe

Usin

f0r tl

in Er

A more detailed analysis of the evolution of initial density operator can be
completed using the averaged Hamiltonian over the entire phase cycling period.
Table 2 indicates that all the interactions except for the heteronuclear dipolar
coupling have been averaged to zero over 16 rotor periods. Thus, the time
averaging of the total Hamiltonian Ho can be expressed by the zero-order time

averaging of heteronuclear dipolar coupling over 16 rotor periods.

H0 = (HD,hetero>0 = d'IzSz (17)

The term dis a function of B and C according to Eq. 14 and is time-independent.
It is related to the heteronuclear dipolar coupling frequency and can be further
correlated with the internuclear distance between two spins. According to Fig. 10,
the initial direction of the spin operator will be in the transverse plane after the CP
period. Assume the initial magnetization is along x axis in the rotating frame, i.e.

p(0) oc Ix. The time evolution of the magnetization can be calculated as

. . 2 . 3
It (12.122.

p(t)=p(0)-l-!-rt+ 2, 2 3! + ------ (18)

where

"i =IH0aP(0)]
’2 =[Hotri]=IHo,IH0,P(0)I] (19)
"3 =IH0J2]=IH0aIHotiH09P(0)]]I

Using the expression in Eq. 17 for the averaged Hamiltonian and the assumption
for the initial density operator, it can be derived straightforwardly that the r terms

in Eq. 19 for REDOR pulse sequence are

42

Thus

00ml

The

The n

The I
23.

rl acid-IySz
r2 ocdz-Ix (20)

exmtgg
Thus, the time evolution of spin operator for REDOR can be obtained from the

combination of Eqs. 18 and 20.

- - -2. 2 .3, 3
p(t)oc1 -MJ 57+“.1 _M.1 S.+ ......
x 1' y ~ 2' X 3' y z
. . . (21)

(dt)3

(c102 -
+ ------ ]+IySz -[(dt)—T+ ------ ]=Ix cos(dt)+IySZ srn(dt)

2!

 

=]x.[]_

The first term in the right-most expression of Eq. 21 is detectable while the
second term is undetectable. This can be understood with the density matrix form
of these spin operators. In two-spin systems such as an l-S spin pair, the density

matrix of IX and lySz can be written as

F0010‘ loo—10'
0001 000i
Ioc andISoc 22
"1000 yzi000()
_0100_ _O—i00d

 

 

 

 

The detection operator for a two-spin system can be expressed by l”, and has

the matrix form

I+oc

(23)

COCO
COCO
COO-—
OO—‘C

 

 

The FID is proportional to the trace of the product of the matrices in Eqs. 22 and

23.

43

ann
ofsel
sgna

tequl

he p
hang
hang
sewe
hans
mfiar
”N c
inua-

Hahn

The:

"0000‘ F0000“
0000 0000
Ix-IToc andISz-IToc . (24)
0010 y 0010
_0001_ Looo—r‘

 

 

 

 

From Eq. 24 one can see that the trace of first product is non-zero while the trace
of second product is zero. Consequently, Eq. 21 represents that the detectable
signal after the REDOR pulse sequence is related to the l-S dipolar coupling
frequency as well as the evolution time.

Double CP (DCP) experiments. In the DCP experiments (cf. Figure 11),
the polarization transfer follows the route 1H —> ”N -> ”C, and the 15N —i ”C
transfer can be selective for 13C0 and ”Cat by setting the spectrometer
transmitter close to the 13C0 and ”Cot frequencies.(18) The DCP scheme can
serve as building blocks to polarize the labeled ”Cs. The spectra selectivity
transfer is achieved by rendering the conventional Hartmann-Hahn cross
polarization technique frequency dependent. Typically, the rf field applied on the
”N channel is comparable to the ”N spectrometer frequency offset so that the
intra-system Hamiltonians can be analyzed in a tilted rotating frame. The major
Hamiltonians during the ”N -» ”C transfer step is

H total = H CS ,z'so + H CS ,am’so + H D (25)

The terms on the right side of Eq.25 represent the isotropic chemical shift, the
anisotropic chemical shift and the ”C-”N heteronuclear dipolar coupling
respectively. These terms have the similar form as shown in Eqs. 12-13 in a
rotating frame and could be expressed as

HCSJSO = nIIz +9552 (263)

44

HCS’am-so = a, (t)!z +O'S(t)Sz (26b)
H D = d(t)]zSz (26c)
The l and S spins represented ”C and ”N. Here we use a similar strategy to
calculate the zero-order average Hamiltonian during the ”N-”C CP step as used
previously in the REDOR example. However, since the frequency offset Q is
comparable to the Rabi frequency (01, it will be more convenient to apply the AHT
in a tilted togging frame by considering both Q and our Firstly we can transfer the
H0 in Eq. 260 into a tilted rotating frame defined by a and (01. The z axis in the
tilted rotating frame is along the direction of the vector sum of Q and (01. The

Euler angle set between the rotating frame and the tilted rotating frame is (0, 6, 0)

where 656, =arctan(0% ) or (9565 =arctan(0% ) for I or S and a)” and ans
1] IS

are Rabi frequencies applied on I and 8 channels respectively. The dipolar
coupling Hamiltonian in Eq. 26c will consequently be expressed in the tilted

rotating frame as

H Shed = exp[—i6,1y]exp[-i05Sy]HD exp[il93Sy ]exp[i6,ly] =
d(t)cosl9, cosfis -212Sz +d(t)sin (9, sin 63 -(I+S_ +I_S+) / 2+d(t) 00563 sin 05 -SZ(I+ +I_)
+d(t)cos€s sin 9, -IZ(S+ +S_)+d(t)sin 6, sin 63 -(I+S+ +I_S_)/2

(27)
The five terms on the right side of Eq. 27 indicate the Z component, zero-
quantum (ZQ), single-quantum in l spin space (SQI), single-quantum in 8 spin
space (808) and double-quantum (DQ) transfer respectively. The time-

dependent term d(t) has the same representation as in Eq. 16a. In the tilted

45

rotat

W09

The

ntat

efiec

togg!

rotating frame, the effect of rf and chemical shift can be written as a new

propagator which is along the z axis.
U = exp[—i(wll,efllz + wlS,eflSz )1] (28)

The effective Rabi frequency equals the vector sum of chemical shift offset and

rotating frame Rabi frequency for individual spins, and the magnitude of the

effective Rabi frequency TOIIOWSlafiflfl =(IQZ +042 . Based on Eq. 2, the H0 in the

 

toggling frame can be evaluated by

Higggle = UHg/redU—l (29)
Each of the five terms in the right side of Eq.27 can be calculated separately.
flg’fg’e oc 1,3, (30a)

firDongée at (A {Alwyn-wispy] Haiti 11 + B e-l'llwli,efl-wis,eyf>-wrill +Ce-iltwl1,e/f-wis,e17 twirl!

+ De"T(a’Il,efl’-a’IS,efl”)+20’R]‘ )1 S + ( A (Biliary-wisgflflwklt + Beiltwir,eyf-wis,eyf)+wrilt (30b)
+ _
+Ceil(wir,ejj-wis,eyf )‘a’RI’ + Deiltwlr,eyf-ans.ejf Ham It )1. 5+

~ . _2 I _. 2 I . _ . _.
HIDogSgéel «{[Aewwll’efl (DR) +Ae “ml/£174" (DR) ]+[Be’(a’”.ejf 02R)! +83 ‘(a’Il,efl+wR)’]

+[Cei(a’”,efl+wk )t + Ce—i(mll,efl-wR)l ] + [Dei(a’|1,efl+20’R)t + De-“mllflfl'zwl? )l]}Isz

(30c)
+{[ A e'(0’ll,efl"20’R )I _ A e-rtwir.ejf+20R )1 ] + [ B gimme/ram )1 _ B e“(a7ll,eff+a’R It]
+[Cei(a’u,efi+wrt)t _ Ce-i(wir,efl-wii )1] +[ D eitwl1,ejf+2wii )1 _ D e‘i(wll,ejf‘2wR >11} 1y 52
173%? 0c {[Aei(aqs,q01—2wk )r + Ae-"(“’IS.e/f+2“’R )t ] + [Bei(ms’€fl—wR )r + Be-i(a’ls’efl+wml]
+[Ceims’eflmk )t + Ce-iwls’e/f—QR” 1+ [Dei(0IS,ejf+za’R)’ + De—“mls'efl—MR )t ] }Isz (30d)

+{[Aei(a’IS,ej]'-20R )1 _ Ae—i(wIS,efl+2wR )t]+[Bei(a’IS,efl-0R)t _ BeTKa’ISflfl'Ta’R )t]

+[Ceiwl5’e/fwk )t _ Ce-i(a)|g’efl-¢DR )l ] + [Dei(a"s'efl+2wk )t - De—r-(MS,efl—2wk )t ”125),

46

figgglge 0c (Ag—iltwupflmtspjf Have II + Be-iltwil,ejf “145,917 )‘a’RI’ + Ce-iltwil,e/f+wis,ejf )‘W’RI’

+ D e‘iIIOJII,efl+¢’IS,ejf 820R II M 5+ + ( A eiIIG’llxfl’t'a’lSpjf Maui I! + B eiliwir,e/f+wis,efl Harri II (309)

+Ceil(wir,eyy+wls,efl)-wiilt + Deiltwir,e)f+wis,efl)-2wri It )1. S_

 

 

 

 

 

 

air/2
i
1H H CP1 I TPPM decoupling
13C . . Acquisition
: [ CP2
15N

 

 

CP1 7 T [—— cpz

 

Figure 11 Double Cross Polarization (DCP) pulse sequence. CP1 and CP2
indicate the 1H—»”N and ”N—s”C cross polarization respectively. There is a

short delay 1' between the first and second CP process. TPPM decoupling was
applied during the r, CP2 and acquisition periods.

47

The

reco
and
cont

inter
RF 1

2810

All I!
simil
USlng
000C
mag
Harr

Writt.

The
Sam.

dens

The Eqs. 30 suggests the 20, SQI, SDS and DO components can be selectively

recovered when the conditions ain’efl-wlsfifl =nruR, aim,” =an, 0,3,8], =an
and 021 Lejf +w15£fl =an (n = 1 or 2) are satisfied respectively. In practice, the
conditions aim/f =an and wise/7 =an are always avoided because the CSA

interaction will also contribute to the zero-order average Hamiltonian under these
RF rotational resonance conditions.(19,20) Consequently, the general forms of

zero-order average Hamiltonian are

<fi3881€>0 0C (1+S_ +I_S+) for (omefl- -sz,eff = an (313)

(@8800 ac (1,5+ +1_s_ ) for am], mm]? = an (310)

All the time-oscillated terms will be averaged to zero after one rotor period. A
similar derivation of the time evolution of the initial magnetization can be obtained

using Eq. 18 with the assumption that the initial condition p(0)oc Sx. The initial

condition is different from the REDOR sequence because in DCP the 1H
magnetization is firstly transferred to ”N instead of ”C. Vtfith the averaged
Hamiltonian in Eq. 31a, the zero-order averaged Hamiltonian in DCP can be
written as

H0 at: a -IZSZ + b-(I+S_ +I__S+) (32)
The parameters a and b can be derived from Eqs. 30 and are generally not the
same. Based on the zero-order Hamiltonian in Eq. 32, the evolution of initial

density matrix should be

48

’0"c
r]:

7'1:

EXF

to a
Iwi.

Oflt

USE
Iipic
20

san
The
Vae
Witt
thrc
weE

98p

7'0me

r, = [H0, p0] = [a1,s, + bI,S, + b1,S,,S,] .-.- a[1,S,, S,]+b{1ySy, 5,] = in . 1,5,, 451,5,

r2 = [110,4] = [a1,S, + 51,5, + nySyJa-IzSy -ib-IyS,] = tra1,S, +b1ys,,a1,sy —nyS,]

= t[(a2 + b2)(S,Sy — 5,5,)+ ab(IyI, - 1,1,)] = (02 + b2)S, - abr,
(33)

It can be seen from Eq. 33 that the term r2 includes the spin operator Ix, which
means the ”C magnetization can be generated from the coupled ”N

magnetization.

EXPERIMENTAL OPTIMIZATION

In the following chapters, multiple solid-state NMR methods will be used
to achieve measurements of different characters for the HFP/membrane system.
I will give an introduction to the experimental setup in this section and will focus
on the results and interpretations of these experiments in the following chapters.

”C—3’P REDOR experiment. ”C-31P REDOR experiments are going to be
used to measure the distances between ”CO-labeled HFP backbone and the
lipid phosphate 31Ps. The sample used for setup contains 0.8 umol HFPmn and
20 umol [1-”C]-1,2-dipalmitoyl-sn-glycero-3-phosphocholine (1-”C-DPPC). The
sample preparation started with dissolution in chloroform of 20 umol 1-”C-DPPC.
The chloroform was removed under a stream of nitrogen followed by overnight
vacuum pumping. The lipid film was suspended in 2 mL buffer and homogenized
with ten freeze-thaw cycles. Large unilamellar vesicles were formed by extrusion
through a 100 nm diameter polycarbonate filter (Avestin, Ottawa, ON). HFPmn
was dissolved in 2 mL buffer and the HFPmn and vesicle solutions were then

gently vortexed together. The mixture was refrigerated overnight and

49

what

bour

reph
dept
ofet

eacl

(ASP
the

Deh

whe:

whh

the

v-o
II

DID;

a 5,1

ultracentrifuged at ~1500009 for five hours. The membrane pellet with associated
bound HF Pmn was transferred to a 4 mm diameter MAS NMR rotor. (4)

The REDOR pulse sequence is the same as sketched in Figure 10 by
replacing ”N with 31P channel. As described in the previous section, the
dephasing period during the “S1” acquisition contained a ”C n pulse at the end
of each rotor cycle except for the last cycle and a 31P 17 pulse in the middle of
each cycle. In principle, the difference spectrum will contain information about the
13C-3‘P dipolar coupling (d). In practice, the experimental REDOR dephasing
(AS/So)°"” was defined as (So-Si)ng where So and 81 represent the intensities of
the interested ”CO resonance in the So and 8; spectra, respectively.
Determination of d was based on fitting (AS/Sg)°"” to (AS/So)”:

[35.] = l— [Job/221)? + {2 Z” 4771)] —————} (34)

SO k=l 16k kz—l

where A. = dr and Jr, is the kth order Bessel function of the first kind.(21)

During the setup process, the 1H and ”C If fields were initially calibrated
with adamantane and the ”C cross-polarization field was then adjusted to give
the maximum ”CO signal of the sample containing HFPmn and 1-”C-DPPC.
The 31P 11 pulse length was set by minimization of the 81 signal in this sample for
r = 8 ms and the 1H TPPM pulse length was set to give the maximum Sg signal.
Figure 12a displays the plot of (AS/So)MD and (AS/So)“ vs. 1' for HFPmn/1-”C-
DPPC sample. The best-fit d = 68 Hz according to Eq.34 which corresponded to
a 5.6 A ”C-3‘P distance (r) based on the relation d (Hz) = 12250 / P(A). The

50

la)

4 \ S/ S(_)

(al

138/80

 

 

 

 

 

 

l l
0 10 20
Dephasing time (ms)

 

 

 

 

0'0 r r l l
o 10 20 30 40

Dephasing time (ms)

Figure 12 (AS/Sg)"‘” (error bars) and best-fit (AS/Sg)"’" (lines with or without
diamonds) vs dephasing time (1') for (a) the ”C-31P setup and (b) the ”C-”F
setup. In panel (a) the experimental data was fit to a two-spin system. In panel
(b), the experimental data was fit to either a two-spin system (dash line) or a

three-spin system (solid line).

5]

best
crys
p110:
glyCI

51m

00m
inse
with
mod

amir

best-fit NMR value of r is comparable to the 5-6 A values of r observed in the
crystal structures of the related lipids, 1,2-dimyristoyl-sn-glycero-3-
phosphocholine (DMPC) and 1,2-dipalmitoyI-sn-glycero-[phospho-rac-(1-
glycerol)] (DPPC) (which had both been dehydrated) and in molecular dynamics
simulations of gel-phase DPPC.(22-24) The differences between (AS/So)MD and
(AS/So)“ are likely due to: (1) contributions to (AS/So)” from intra- and
intermolecular 31P with comparable values of r which contrasts with the single
”CO-MP spin pair model used to calculate (AS/So)“; (2) two structurally distinct
”C05 in each headgroup with different intra- and intermolecular rvalues; and (3)
structural disorder within the headgroups.(25) Overall, the 1-”C-DPPC fitting
yielded good agreement between the NMR r value and the expected range of r
values in the lipid.

”C-”F REDOR experiment. ”C-”F experiments can provide
complementary distance information for ”C-3‘P experiments in the membrane
insertion studies.(26) These measurements used the same pulse sequence but
with ”F as the third channel. The ”F compound to setup the ”F 7: pulses was a
modified helical peptide F (EQLLKALEFLLKELLEKL) with Phe9 substituted by 2-
amino—3-(4-fluorophenyl) propanoic acid (Sigma-Aldrich, St. Louis, MO) and the
adjacent Leu10 labeled with ”CO. Figure 12b showed the experimental and
simulated dephasing curves for the fluorinated peptide F. It was observed that
the experimental data fit better to a “”F-”C-”F” three-spin system than a “”C-
”F” two spin system, which may indicate that the helical peptide F could form

oligomers and the labeled ”CO could locate close to more than one 19Fs. The

52

REC
6.8
com

000-

with

bonl

prev
cou;
Hert
W88
WErl
inter
Iabe
Fig(
177
test
The

Iian

and

REDOR-determined ”CO-”F distances for the three spin system were 7.6 A and
6.8 A with 162° between the two ”CO-”F vectors. The distances were
comparable with the ~ 7.1 A distance measured in the crystal structure of the
non-fluorinated compound.(27)

DCP experiments. The DCP building block is expected to be incorporated
with other pulse sequences and served to polarize ”C nucleus that are directly
bonded to ”N nucleus and not natural abundance ”Cs. The optimization for ”N
—» ”C transfer during the DCP will be described in this section. According to the
previous discussion about the theory, both the Z0 and DO parts of the dipolar
coupling Hamiltonian can contribute to the transfer under different conditions.
Here the 20 transfer condition with n = 1 as described in the previous section
was chosen and the key parameters such as (1)13, am and .05 (I E ”C and S .=. ”N)
were optimized. The ”C transmitter was always set close to the resonance of
interest so that {11 ~ 0 and the MAS frequency was fixed at 8 kHz. A double ”CD-
Iabeled N-acetyI-Leucine (D-”CO-NAL) was used as the setup compound.
Figure 138 showed a regular 1H-”C CP spectrum of D-”CO-NAL in which the
177.2 and 175.4 ppm can be assigned to the carboxyl and the acetyl ”COs
respectively.(28) In Figure 13b where the DCP pulse sequence was applied, only
the ”CO peak at 175.4 ppm was observed because it was directly bonded to 15N.
Fig. 130 showed that ”C signal came from the ”N—»”C transfer instead of
1H—>”C transfer by removing the CP2 on ”N channel in Figure 11. The ”C
transmitter was set to 160.0 ppm. The optimization of ans is shown in Figure 13d

and the ”N—+”C transfer efficiency was greatly affected by the ”N rf field. Figure

53

13e shows the arrayed spectra for calibrating the ”N frequency offset. The
optimized ”N rf field was 14.8 kHz and the ”N frequency offset was 5.5 kHz.
The center frequency of ”C ramp during the ”N—>”C transfer step was 24.0 kHz
so that the condition of 20 n = 1 transfer was satisfied. In addition, ~ 10 % ”C
ramp and ~ 8 ms contact time was required for a efficient ”N—+”C transfer and

the optimized transfer efficiency was 40-50 % compared with direct 1H-”C CP

experiment.

54

 

WW

I I I I I I
190 170 190 170 190 170

130 Chemical Shirt (ppm)

 

(d)

(0)

Figure 13 (a) 13C0 region of D-”CO-NAL spectrum acquired using the CP pulse
sequence. (b) ”CO region of D-”CO-NAL spectrum acquired using the DCP
pulse sequence. The vertical scales in (a) and (b) are the same so that the
relative intensity reflects the DCP transfer efficiency. Both spectra were
processed with 100Hz Gaussian line broadening and baseline correction. (c) A
negative control experiments without ”N CP2 amplitude (of. Figure 11). The
number of scans used in (a), (b) and (c) was 32. Panel (d) displays the
optimization of ”N CP2 amplitude as given in Fig.11. The ”N rf field was
scanned from 11.8 kHz to 17.8 kHz with the increment of 0.16 kHz. Panel (e)
displays the optimization of ”N CP2 offset frequency. The offset was scanned
from -10 kHz to 10 kHz with the increment of 0.5 kHz.

55

REFERENCE

1. Tycko, R., Biomolecular solid state NMR: Advances in structural
methodology and applications to peptide and protein fibrils. Annual Review of
Physical Chemistry 2001, 52, 575-606.

2. Drobny, G. P.; Long, J. R.; Karlsson, T.; Shaw, W.; Popham, J.; Oyler, N.;
Bower, P.; Stringer, J.; Gregory, D.; Mehta, M.; Stayton, P. 8., Structural studies
of biomaterials using double-quantum solid-state NMR spectroscopy. Annual
Review of Physical Chemistry 2003, 54, 531 -571 .

3. Hong, M., Oligomeric structure, dynamics, and orientation of membrane
proteins from solid-state NMR. Structure 2006, 14, (12), 1731-1740.

4. Qiang, W., Yang, J., and Weliky, D.P., Solid-state nuclear magnetic
resonance measurements of HIV fusion peptide to lipid distances reveal the
intimate contact of beta strand peptide with membranes and the proximity of the
Ala-14-GIy-16 region with lipid headgroups. Biochemistry 2007, 46, (17), 4997-
5008.

5. Qiang, W., Bodner, M.L., and Weilky, D.P., Solid-state NMR Spectroscopy
of HIV Fusion Peptides Associated with Host-Cell-Like Membranes: 2D
Correlation Spectra and Distance Measurements Support a Fully Extended
Conformation and Models for Specific Antiparallel Strand Registries. J. Am.
Chem. Soc. 2008, 130, 5459-5471.

6. Qiang, W., and Weilky, D.P., HIV Fusion Peptide and Its Cross-Linked
Oligomers: Efficient Syntheses, Significance of the Trimer in Fusion Activity,
Correlation of 8 Strand Conformation with Membrane Cholesterol, and Proximity
to Lipid Headgroups. Biochemistry 2009,48, (2), 289-301.

7. Qiang, W., Sun, Y., and Weliky, D.P., A strong correlation between
fusogenicity and membrane insertion depth of the HIV fusion peptide. submitted
to Proc. Natl. Acad. Sci. USA.

8. Haeberlen, U., High Resolution NMR in solids Selective Averaging. 1976.

9. Gullion, T.; Schaefer, J., Rotational-echo double-resonance NMR. J. Magn.
Reson. 1989, 81, (1), 196-200.

10. Hing, A. W.; Tjandra, N.; Cottam, P. F.; Schaefer, J.; Ho, C., An
investigation of the ligand-binding site of the glutamine-binding protein of
Escherichia coli using rotational-echo double-resonance NMR. Biochemistry
1994, 33, (29), 8651-61.

56

11. Gullion, T., Introduction to rotational-echo, double-resonance NMR.
Concepts Magn. Reson. 1998, 10, (5), 277-289.

12. Middleton, D. A.; Ahmed, Z.; Glaubitz, 0.; Watts, A., REDOR NMR on a
hydrophobic peptide in oriented membranes. Joumal of Magnetic Resonance
2000, 147, (2), 366-370.

13. Jaroniec, C. P.; Tounge, B. A.; Herzfeld, J.; Griffin, R. 6., Frequency
selective heteronuclear dipolar recoupling in rotating solids: Accurate ”C- 5N
distance measurements in uniformly 3C, 15N-labeled peptides. Journal of the
American Chemical Society 2001, 123, (15), 3507-3519.

14. Murphy, 0. J., 3rd; Kovacs, F. A.; Sicard, E. L.; Thompson, L. K., Site-
directed solid-state NMR measurement of a ligand-induced conformational
change in the serine bacterial chemoreceptor. Biochemistry 2001, 40, (5), 1358-
1366.

15. Forrest, T. M.; Wilson, G. E.; Pan, Y.; Schaefer, J., Characterization of
Cross-Linking of Cell-Walls of Bacillus- Subtilis By a Combination of Magic-Angle
Spinning Nmr and Gas- Chromatography Mass-Spectrometry of Both Intact and
Hydrolyzed C-13-Labeled and N-15-Labeled Cell-Wall Peptidoglycan. Joumal of
Biological Chemistry 1991, 266, (36), 24485-24491.

16. Merritt, M. E.; Christensen, A. M.; Kramer, K. J.; Hopkins, T. L.; Schaefer,
J., Detection of intercatechol cross-links in insect cuticle by solid-state carbon-13
and nitrogen-15 NMR. Joumal of the American Chemical Society 1996, 118, (45),
11278-11282.

17. Hong, M.; Griffin, R. G., Resonance assignments for solid peptides by
dipolar-mediated C- 13/N-15 correlation solid-state NMR. Journal of the
American Chemical Society 1998, 120, (28), 7113-7114.

18. Baldus, M.; Petkova, A. T.; Herzfeld, J.; Griffin, R. 6., Cross polarization in
the tilted frame: assignment and spectral simplification in heteronuclear spin
systems. Molecular Physics 1998, 95, (6), 1197-1207.

19. Osa, T. G., Griffin, R.G., and Levitt, M.H., Rotary resonance recoupling of
dipolar interactions in solid state nuclear magnetic resonance spectroscopy. J.
Chem. Phys 1988, 89, 692-695.

20. Gan, Z. H., and Grant, D.M., Rotational resonance in a spin-locked field
for solid-state NMR. Chem. Phys. Lett. 1990, 168, 304-308.

21. Mueller, K. T., Analytical Solutions for the Time Evolution of Dipolar-

dephasing NMR Signals. Joumal of Magnetic Resonance Series A 1995, 113, (1),
81-93.

57

22.
shnu
suna

23.
pack
dimy

24.
Nani

25.

Shay
sbnh
Cher

26.
shuc
REE)

27.
nove

28.
With
Statt

22. Venable, R. M., Brooks, BR, and Pastor, R.W., Molecular dynamics

simulations of gel (L8,) phase lipid bilayers in constant pressure and constant
surface area ensembles. J. Chem. Phys 2000, 112, 4822-4832.

23. Pascher, l., Sundell, S., Harlos, K., and Eibl, H., Conformation and
packing properties of membrane lipids: The crystal structure of sodium
dimyristoylphosphatidylglycerol. Biochim. Biophys. Acta. 1987, 896, 77-88.

24. Pearson, R. H., and Pascher, |., Molecular structure of lecithin dihydrate.
Nature 1979, 281, 499-501.

25. Raghunathan, V.; Gibson, J. M.; Goobes, G.; Popham, J. M.; Louie, E. A.;
Stayton, P. S.; Drobny, G. P., Homonuclear and heteronuclear NMR studies of a
statherin fragment bound to hydroxyapatite crystals. Journal Of Physical
Chemistry B 2006, 110, (18), 9324-9332.

26. Toke, O.; Maloy, W. L.; Kim, S. J.; Blazyk, J.; Schaefer, J., Secondary
structure and lipid contact of a peptide antibiotic in phospholipid Bilayers by
REDOR. Biophys. J. 2004, 87, (1), 662-674.

27. Taylor, K. 8., Lou, M.Z., Chin, T.M., Yang, NC, and Garavito, R.M., A
novel, multilayer structure of a helical peptide. Protein sci. 1996, 5, 414-421.

28. Bodner, M. L., Solid state nuclear magnetic resonance of the HIV-1 and

influenza fusion peptides associated with membranes. Ph. D. thesis, Michigan
State University: East Lansing, 2006; p 122.

58

CHAPTER IV

SECONDARY STRUCTURES OF MEMBRANE-ASSOCIATED HIV FUSION
PEPTIDE OLIGOMERS

BACKGROUND

The ~ 20-residue HIV fusion peptide (HFP) (with the sequence
AVGIGALFLGFLGAAGSTMGARS) has been considered to be a good model
peptide to study the properties of HIV gp41 induced membrane fusion process, at
least to the lipid mixing stage. The free HFP causes fusion of liposomes and
erythrocytes, and numerous mutational studies have shown strong correlations
between fusion peptide-induced Iiposome fusion and viral/host cell fusion.(1-4) A
variety of experimental methods have shown that the HF P can assume helical or
non-helical structures when associated with micelles or membranes with different
components or with different peptide to phospholipids molar ratios.(5-16) For
HFP associated with negatively charged sodium dodecyl sulfate (SDS) micelles,
one liquid-state NMR study showed that there was uninterrupted ot-helical
conformation from 14 to M19,(15) while another study showed a helix from l4 to
A14 followed by a Bturn.(11) For HFP associated with neutral
dodecylphosphocholine (DPC) micelles, helical structure was detected from 14 to
L12.(16, 35) Cholesterol is an important membrane component because the
cholesterolzphospholipid molar ratios are ~0.5 and 0.8 for HIV host cell and HIV
membranes, respectively.(i 7) Solid-state NMR provided residue-specific
conformational information about HFP associated with membranes whose lipid

headgroup and cholesterol composition were comparable to that of host cells of

59

the virus. A B strand conformation was observed for residues A1-G16, while A21
appears to be unstructured.(18,19) In addition, recent solid-state NMR studies
showed that the secondary structure of the membrane-associated HFP can be
affected by the existence of pre-HFP domains, i.e. the hairpin formed by CHR
and NHR domains. It has been observed that in the cholesterol-containing
membranes with comparable peptide to lipid molar ratio, more fraction of a-helix
was presented for the HFP samples with the pre-HFP domain than those without
the pre-HFP domain.(20)

This chapter will summarize the residue specific chemical shifts obtained
with 2D solid-state NMR and compare the experimental ”C chemical shifts with
the RefDB database to determine the secondary structures of the labeled
residues.(23) This database correlates the 1H, ”C and ”N chemical shifts of
previously assigned proteins and the secondary structures determined from X-
ray coordinate data of these proteins. In the database, the secondary structures
of residues were classified as helix, beta strand and coil based on the (it and I]!
dihedral angles. A residue was defined as helix if -120 < ¢ < -34 and -80 < r/I< 6.
A residue was defined as beta strand if -180 < d < -40 or 160 < (i < 180 and 70 <
W< 180 or -180 < l,u< -170. A residue with dihedral angles in other regions was
defined as coil.(23) With the information of secondary structure, we will discuss
(1) the effect of the existence of cholesterol on the secondary structure, and (2)
the effect of HFP constructs on the secondary structure. For the third part, the

chemical shift information from four different HFP constructs, HFPmn, HFPdm,

60

HFI

will

MA

1.2
the
The
phi
ell
Sig
ten
“Pt
Th.

the

Chi

Spl

an
Ter

lip)

HF Ptr and HF Pmn_mut, will be summarized and the similaritiesy and differences

will be discussed.

MATERIALS AND METHODS

NMR Sample Preparation. Solid-state NMR samples were made with
ether-linked lipids 1,2-di-O-tetradecyl-sn-glyceroI-3-phosphocholine (DTPC) and
1,2-di-O-tetradecyl-sn-glycerol-3-[phosphor-rac-(1-glycerol)] (DTPG) because
these are commercially available lipids which do not contain carbonyl groups.
The typical ester-linked phospholipids such as 1,2—dimyristoyl-sn-glyceroI-3-
phosphocholine (POPC) and 1,2-dimyristoyI-sn-gcherol-3-[phosphor-rac-(1-
glycerol)] (POPG) were not chosen their large natural abundance lipid ”CO
signals would overlap with the peptide ”CO signals. For the current chapter, the
term “PC:PG” denotes the membrane with 4:1 DTPC:DTPG ratio and the term
“PC:PG:CHOL” denotes the membrane with 8:225 DTPC:DTPGzcholesterol ratio.
These ratios reflect the approximate fraction of phospholipids and cholesterol in
the HIV-infected host cell.(17) The typical peptide strand1lipid molar ratios are
1:50 for PC:PG and 1:25 for PC:PG:CHOL, and the total amount of lipid and
cholesterol are 20 umol for PC:PG and 30 umol for PC:PG:CHOL unless
specifically mentioned.

Each sample preparation began with dissolution in chloroform of the total
amount of lipid and/or cholesterol as described above. The chloroform was
removed under a stream of nitrogen followed by overnight vacuum pumping. The

lipid film was suspended in 2 mL of buffer and homogenized with 10 freeze-thaw

6|

cyc
dia
HF
eth
wis
Pei
PC
ves
PC
ultr
bor

the

ant
Le
(Vi
lat
ms
the
119
lat
Sb

40

cycles. Large unilamellar vesicles were formed by extrusion through a 100 nm
diameter polycarbonate filter. For PC:PG samples, the corresponding amount of
HFPs were dissolved in 30 mL of N—(2-hydroxyethyl)piperazine-N’-2-
ethanesulfonic acid (HEPES) buffer, and the peptide solution was added drop
wise into the vesicle solution with gentle vortex to prevent the aggregation of the
peptides and to increase the population of tit-helical conformation. For
PC:PG:CHOL samples, the HFPs were dissolved in 2 mL HEPES buffer and the
vesicle and HFP solutions were mixed together. Both mixtures with PC:PG and
PC:PG:CHOL were vortexed at ambient temperature for overnight and
ultracentrifuged at ~1500009 for 5 hours. The membrane pellet with associated
bound HFP was transferred to a 4 mm diameter MAS NMR rotor. The majority of
the HFP binds to membranes under these conditions.(21,22)

Solid-state NMR Experiments. All HF Pmn_muts were singly-”CO labeled
and the HFPmn, HFPdm and HFPtr were uniformly-”C-Iabeled at lle4, Ala6 and
Leu12. All solid-state NMR experiments were conducted on a 9.4T spectrometer
(Varian Infinity Plus, Palo Alto, CA). The ”CO chemical shifts for different ”CO-
Iabeled HF Pmn_mut were determined from the ”C-a‘P REDOR So spectra with 2
ms dephasing time and the secondary structures were obtained by comparing
the experimental ”CO chemical shifts to the RefDB databases.(23) A more
detailed assignment was done for HFPmn. HFPdm and HFPtr with uniformly ”C
labeled residues using Proton-driven Spin-diffusion (PDSD) methods. All ”C
shifts were externally referenced to the methylene resonance of adamantane at

40.5 ppm. The REDOR pulse sequences is similar to the sequence sketched in

62

 

(a)

(wdd) wills lesiwalro act

   

r. ‘1

0 L12 Ca!Cy(01)

 

 

 

 

     
 

 

    
 

 

 

 

T 50
60 36 13
13‘0 Chemical Shift (ppm)
13
A6 CB/Ca(a)
b
I l , o a
7
A6 CBICoUi ..
8
° 9
S
‘9
L12 cprcwcl’ a
‘ L12 celcwp :3.
L12 CaICy(B)\ 3 6
L12 Cale(a) o
y o \9 E
, 0
° 9
0 o
I | 60
60 36 13

130 Chemical Shift (ppm)

Figure 14 PDSD spectra for (a) HFPmn in PC:PG and (b) HFPdm in PC:PG.

63

Fig

 

   
  
  
 

 

 
 

 

 

 

   
   

 

 

 

 

(c) m A6 cit/com) '
o o//Ac CWCaIBlg . ,
first ? - f;
c Q '
6E? ‘ - ' ~ ° 6
'1 o 5.3
° L12 cplcylc) 3.
/ . i'36 g
9 L12 08/0703) 0 9
‘Q P 5‘
'0
L12 corc .
a A“? 3
L12 CaIC or 0
¢>
° ”I
T r 6"
60 36 13
13C Chemical Shift (ppm)
((1) '—V 13
or A6 cplcc(p)
, /
/ 8
L12 CyICB(B) V :L is?
ucwcyorcsas 36:33:
\ ‘2.
0 ¢ '4
6 a
S
U
@ ‘3'
l4 Catle or CNBK
o n ‘9
7 j 60
60 36 13

130 Chemical Shift (ppm)

Figure 14 PDSD spectra for (c) HFPtr in PC:PG and (d) HFPmn in PC:PG:CHOL.

64

 

(e)

(wddl slurs IBGIWGtIO act

 

 

 

 

(f)

lwdd) rims realweqo act

 

0 l4 06.le or 68(13)\

 

 

I
60 36

I 60
. 13
130 Chemical Shift (ppm)

Figure 14 PDSD spectra for (e) HFPdm in PC:PG:CHOL and (f) HFPtr in
PC:PG:CHOL.

65

 

 

 

 

 

O-

2'00

130 Chemical Shift (ppm)

Figure 14 (f) HFPtr in PC:PG:CHOL. All spectra were processed with 100 Hz
Gaussian line broadening in both dimensions. The individual peaks were
assigned and given in the spectra. For example, the peak assigned to A6
CB/Cor(8) represent the cross peak between CB(f1 dimension) and 00162
dimension) for Ala-6 in B-strand conformation. The spectra (9) through (I) display
the representative 1D slice of the PDSD spectra (a) through (f) respectively. For
the spectra (a), (c), (d), (e) and (f), the 1D slice is along ~23 ppm in the f1
dimension which corresponds to the 08 of Ala-6 in B-strand conformation. For
the spectrum (b), the 1D slice is along ~18 ppm in the f2 dimension which
corresponds to the C8 of Ala-6 in tit-helical conformation.

66

Fgun
paran
used

kHz r
50 kt
and ‘
NAL

kien

REEi

kw
PCT:
F0rl
H94
for)
rest
conc
San]
Stral
PCll
Adag

Whic

Figure 10 with 31P as the third channel instead of ”N. The experimental
parameters for ”C-3‘P REDOR will be discussed in chapter VI. The parameters
used in PDSD experiments (cf. Figure 14a) are: 10 kHz MAS frequency; 44-64
kHz ramp on the ”C CP n‘ field; 62.5 kHz 1H CP rf field; 2 ms CP contact time;
50 kHz 13c ”/2 pulse rffield; 25 us t1 dwell time; 200 it values; 20 ps t2 dwell time;
and 1 s recycle delay. The parameters were optimized using uniformly labeled
NAL (cf. Figure 14b). Both REDOR and PDSD experiments were done at -50 °C

to enhance the ”C signal.

RESULTS AND DISCUSSION

Residue Specific Chemical Shifts. Figure 14 displays the PDSD spectra
for HFPmn, HFPdm and HFPtr in PC:PG and PC:PG:CHOL.(24) For
PC:PG:CHOL spectra, the assignments were achieved for lle4, Ala6 and Leu12.
For PC:PG spectra, the assignments were achieved for Ala6 and Leu12, but the
Ile4 cross peaks were not clearly identified. Figure 15a-f shows the ”CO peaks
for HFPmn_mut in PC:PG and PC:PG:CHOL. The ”C chemical shifts of these
residues were summarized in the Table 3a through Table 3h. The following
conclusions can be obtained by directly comparing the chemical shifts for these

samples: (1) The HFPmn, HFPdm and HFPtr can adopt both cit—helical and B-

strand structures in PC:PG and adopt only B—strand conformation in
PC:PG:CHOL. (2) There are not obvious chemical shift differences for residue
Ala6 and Leu12 between the three oligomers in either PC:PG or PC:PG:CHOL,

which probably means the oligomerization will not affect the secondary structure

67

of the region around Ala6 and Leu12. (3) There is a mixture of ot-helix and B-
strand conformations for the residues Ala6, Leu9 and Leu12 of HFPmn_mut in
both PC:PG and PC:PG:CHOL. The fraction of a-helix is higher in PC:PG
compared with in PC:PG:CHOL.

Table 3a. Assignments of residues in HF Pmn with PC:PG”

 

 

Residues Cot CB Cy CO
51.0 23.3 175.3
Nae (55.8)” (18.1) (180.3)
Leu12 52.8 46.6 26.1 173.7
(57.7) (42.4) (25.9) (178.8)

 

a'All chemical shifts are in unit of ppm.
b The chemical shifts shown in parentheses indicate ot-helical conformation.

Table 3b. Assignments of residues in HFPdm with PC:PG

 

 

Residues Cot CB OJ CO
50.5 23.7 175.1
“a” (55.8) (18.3) (180.2)
Leu12 53.4 46.0 26.2 173.2
(57.8) Lil-7) (26.0) (179.1)

 

Table 30. Assignments of residues in HFPtr with PC:PG

 

 

Residues Cot C8 Cy CO
51.1 23.3 1754
Na” (55.8) (17.8) (180.3)
Leu12 53.5 45.8 27.0 173.6
(57.8) (41.9y (26.9) (179.2)

 

Table 3d. Assignments of residues in HFPmn with PC:PG:CHOL

 

 

Residues Ca Cfi Cy CO
lle4 59.2 41.8 27.2 174.2
Ala6 50.6 22.8 175.3
Leu12 53.5 46.5 27.0 173.7

 

Table 3e. Assignments of residues in HFPdm with PC:PG:CHOL

 

 

Residues Cot CB 0‘! CO
lle4 58.9 42.0 27.5 173.9
Ala6 50.3 22.7 175.5
Leu12 53.3 45.8 26.5 173.5

 

68

Table 3f. Assignments of residues in HF Ptr with PC:PG:CHOL

 

 

Residues Ca C B Cy CO
lle4 58.2 41.2 27.4 174.2
Ala6 50.4 23.3 175.1
Leu12 52.9 46.3 26.5 174.1

 

Table 39. 13co chemical shift of residues in HFPmn_mut with PC:PG

 

 

Residues Ala1 lle4 Ala6 Leu9 Leu12 Ala14
13
CO
. 175.8 175.5 175.4 176.8
chemical 175.6 176.5
shift (ppm) (180.3) (179.3) (178.9) (179.3)

 

Table 3h. 13CO chemical shift of residues in HFPmn_mut with PC:PG:CHOL

 

 

Residues Ala1 lle4 Ala6 Leu9 Leu12 Ala14
13
co
. 176.1 175.8 175.8
sfi'i‘g'zggfg) "5'5 "4'9 (179.8) (179.2) (179.0) "6'9

 

The results shown above suggested the dependence of global secondary
structure of HFP constructs on the lipid membrane composition, especially the
presence of cholesterol. These data also indicated there was not obvious HFP
constructs-secondary structure correlation for HFPmn_mut, HFPmn, HFPdm and
HFPtr, which might suggested that the secondary structure of HFP was not a
crucial factor to affect the HFP fusion activities because these HFP constructs
were known to have very different fusogenities.(25-27) In the following two
sections, the effect of cholesterol and constructs on the HF P conformation will be
discussed.

Cholesterol-dependence of the HFP Conformation. The presence of
cholesterol in membranes is known to increase the lateral molecular packing

density and membrane tensile strength, decrease the permeability of water

69

through the membrane, and promote formation of the “liquid-ordered phase”.(28-
33) This phase is characterized by a rapid lateral molecular translational diffusion
coefficient similar to that of the “liquid-disordered” phase at high temperature
without cholesterol and high configurational order of the lipid acyl chains similar
to that of the “solid-ordered” phase at low temperature without cholesterol.(33, 34)
The chemical shifts in Table 3a through 3f suggested that Ala6 and Leu12 in
HFPmn, HFPdm and HFPtr will adopt a mixture of a-helical and B—strand
conformation in PC:PG and only B—strand conformation in PC:PG:CHOL. in
another set of experiments shown in Figure 159-i, the 13CO chemical shift of
Ala15 in HFPmn, HFPdm and HFPtr were measured when the peptides were
associated with PC:PG or PC:PG:CHOL. It was observed that the predominant
13CO chemical shift in PC:PG was ~178 ppm with a shoulder at ~176 ppm, while
in PC:PG:CHOL there were single peaks at ~176 ppm. The 178 ppm signal was
assigned to helical conformation and the 176 ppm signal was assigned to [3-
stand conformation for Ala15.(26) The combination of these two pieces of
information suggested there was partial a-helical conformation formed in PC:PG
for the hydrophobic region of HFP constructs, i.e. Ala6 through Ala15. The helical
conformation of the residues Ala6 through Ala15 detected in membranes is in
general agreement with the observed conformation for this region in micelles in
previous solution NMR experiments. There was a general agreement about the
formation of an d—helix from residues lle4 to Leu12 in both SDS and DPC
micelles, while some experiments supported the fact that the helix can be

extended to Met19.(11, 15) A recent solution NMR structure for the N-terminal 23

70

residue HFP also indicated the formation of an a-helix from lle4 to Ala14 in
micelles.(35) In addition, a similar correlation between the presence of
membrane cholesterol and the preference of B—strand conformation has been
observed for the influenza virus fusion peptide so that the correlation may be a
general property of fusion peptides.(36,37) Although the reasons for the
structural effect of membrane cholesterol are poorly understood, it is useful to
consider the increased lateral molecular packing density in cholesterol-containing
membranes and the possibility to form large [3 sheet aggregates for the B—strand
HFPs.(31) Relative to the [3 aggregates, the small monomeric a-helix might
experience a more positive increase in free energy of membrane insertion with
higher packing density.

Construct-dependence of the HFP Conformation. Different HF P
constructs were known to induce lipid mixing and vesicle fusion with different
rates. One previous study concluded that the lipid mixing rate induced by HFPtr
was at least 15 times faster compared with that induced by HFPmn, and HFPdm
has a fusion activity between HFPmn and HFPtr.(25) In addition, there have
been studies both in vivo and in vitro which showed that the V2E mutation has a
transdominant effect on the fusion activity of HFP.(27) One possible explanation
is that there is a required secondary structure for HFP to induced vesicle fusion
and different HFP constructs would adopt different fractions of such fusion active
conformation. In that case one would expect to observe a continuous increase of
a certain secondary structure following the trend HFPmn_mut < HFPmn <

HFPdm < HFPtr. In figure 14a-c, the relative intensities of Cor/CB cross peaks for

71

Ala6 in a-helical and B-strand conformations are different for different HFP
constructs. For HFPmn and HFPtr, the cross peaks for B-strand conformation is
more intense compared with the peaks for a-helix. For HFPdm, on the other
hand, the cross peak for helical structure is more intense compared with that for
strand structure. Figure 16a displays the 1D slice along the chemical shift of Cy in
f1 dimension for HFPmn, HFPdm and HFPtr. This position was chosen because
the chemical shifts of Cy of Leu12 in helical and strand conformations were very
close according to the Table 3a-c, which would make it convenient to compare
the relative intensity of cross peaks for the two conformations directly. Figure 16a
indicates that for Leu12, there is more B-strand than a-helix in HF Pmn and HFPtr,
and more helix than strand in HFPdm. The trend that the relative helical fraction
in HFPdm is larger compared with that in HFPmn and HFPtr was also observed
in other sets of 1:"C-3‘P REDOR experiments for Ala6 and Ala15.(26) As shown in
Figure 16b, the 13CO intensity for a-helical conformation is greater than that for B-
strand conformation in HFPdm, but not in HFPmn or HFPtr. In addition, the 13CO
chemical shifts for HFPmn_mut summarized in Table 3h reveals that for a non-
fusogenic HFP construct, the residues Ala6, Leu9, Leu12 and Ala14 can adopt
both helical and strand structures. The combination of these information
suggested (1) the loss of fusion activity does not associate with the
disappearance of a particular secondary structure because there are mixed
helical and strand conformations in HFPmn_mut, and (2) the increase of fusion
activity does not associate with the increase of a particular secondary structure,

6.9., from HFPmn to HFPdm, there is an increase in the relative population of 01-

72

helix, while from HFPdm to HFPtr, there is an increase in the relative population
of B—strand. This further means (1) both the helical and the 8 strand
conformations are fusion active; or (2) fusion is induced by unstructured HFP.
This transient HFP state would not be apparent in the NMR samples which
reflect the long-time end-state HF P structure.(38, 39) Experimental support for the
first interpretation is an HF Pmn study which showed that the rates of membrane
binding and secondary structure formation were faster than the rate of lipid

mixing.(40)

73

(a) Ala-1 (b) Ila-4

%
%
g

(c) Ala-6 (d) Lou-9

F
%

(e) Lou-12 if) Ala-14

%

 

F
1%

l l I l
190 170 190 170 190 170 190 170

13C Chemical Shift (ppm)

(9) (h) \M‘

 

 

 

   

 

 

 

 

130 Chemical Shift (ppm)

Figure 15 13CO spectra of (a) Ala1, (b) Ile4, (c) Ala6, (d) Leu9, (e) Leu12 and (f)
Ala14 in HFPmn_mut associated with PC:PG and PC:PG:CHOL. For each
labeled residue, the spectrum with PC:PG is shown in the left and the spectrum
with PC:PG:CHOL is shown in the right. 13CO peaks of Ala15 in (g) HFPmn, (h)
HFPdm and (i) HFPtr associated with PC:PG in the top row and PC:PG:CHOL in
the bottom row. All spectra are obtained with the 13C-3‘P REDOR pulse
sequence with 2 ms dephasing time, and processed with 200 Hz Gaussian line
broadening and baseline correction. All PC:PG spectra are acquired with 3000
scans and all PC:PG:CHOL spectra are acquired with 1500 scans. The vertical
dashed lines in (g)—(i) indicate the chemical shift of B-strand 13CO.

74

(a)

 

 

 

 

 

 

 

 

 

 

 

l T
200 100 0

13C Chemical Shift (ppm)

 

——1

(b)

 

 

 

 

 

 

HF Pmn-AB HF Pdm-A6 HF Ptr-A6
1TQO 1 70 170 170
HFPmn A15 HFPdm-A15 HF Ptr A15
1 7'0 190 1 70

13C Chemical Shift (ppm)

Figure 16 (a) 10 slices along the chemical shift of Cy of Leu12 for HFPmn,
HFPdm and HFPtr in the top, middle and bottom spectrum respectively. The
vertical dashed lines labeled 1-3 are assigned to the chemical shifts for CO/Cy,
001le and CB/Cy cross peaks in helical conformation respectively, and 4-6 are
CO/Cy, Col/Cy and Cp/Cy cross peaks in strand conformation respectively. (b)
13C- 31P REDOR So spectra for AlaG and Ala15 samples. In each spectrum, the
left peak corresponds to iii-helical structure and the right peak corresponds to 6-
strand structure.

75

CONCLUSION

The secondary structure studies on a variety of HFP constructs
concluded: (1) ln PC:PG:CHOL, there is a predominant B—strand conformation for
HFPmn, HFPdm and HFPtr, while HFPmn_mut can adopt partial helical
conformation; (2) In membranes without cholesterol, there is a mixture of a-helix
and B—strand secondary structures, and the a—helix is located from Ala6 to Ala15;
(3) There is not an obvious correlation between the secondary structure and
fusion activities for different HFP constructs, which suggested both a-helical and

B—strand conformations can be fusion active.

76

REF

stud
melt

Men
Proll

810;:
fllsic

synt
29. |

fusli
extrl
298

F03
pho
810/

qui.
an Q
197
(Brig
inl‘e
199
10.

hUn
agg

REFERENCE

1. Durell, S. R.; Martin, I.; Ruysschaert, J. M.; Shai, Y.; Blumenthal, R., What
studies of fusion peptides tell us about viral envelope glycoprotein-mediated
membrane fusion. Mol. Membr. Biol. 1997, 14, (3), 97-112.

2. Pecheur, E.; Sainte-Marie, J.; Bienvenue, A.; Hoekstra, D., Peptides and
Membrane Fusion: Towards an Understanding of the Molecular Mechanism of
Protein-Induced Fusion. J Membr Biol 1999, 167, (1), 1-17.

3. Epand, R. M., Fusion peptides and the mechanism of viral fusion. Biochim.
Biophys. Acta-Biomembr. 2003, 1614, (1 ), 1 16-121 .

4. Nieva, J. L.; Agirre, A., Are fusion peptides a good model to study viral cell
fusion? Biochim. Biophys. Acta-Biomembr. 2003, 1614, (1), 104-115.

5. Rafalski, M.; Lear, J. D.; DeGrado, W. F ., Phospholipid interactions of
synthetic peptides representing the N-terminus of HIV gp41. Biochemistry 1990,
29, (34), 7917-7922.

6. Martin, I.; Schaal, H.; Scheid, A.; Ruysschaert, J. M., Lipid membrane
fusion induced by the human immunodeficiency virus type 1 gp41 N-terminal
extremity is determined by its orientation in the lipid bilayer. J. Virol. 1996, 70, (1),
298-304.

7. Kliger, Y.; Aharoni, A.; Rapaport, D.; Jones, P.; Blumenthal, R.; Shai, Y.,
Fusion peptides derived from the HIV type 1 glycoprotein 41 associate within
phospholipid membranes and inhibit cell-cell Fusion. Structure- function study. J.
Biol. Chem. 1997, 272, (21), 13496-13505.

8. Pereira, F. B.; Goni, F. M.; Muga, A.; Nieva, J. L., Permeabilization and
fusion of uncharged lipid vesicles induced by the HIV-1 fusion peptide adopting
an extended conformation: dose and sequence effects. Biophys. J. 1997, 73, (4),
1977-1986.

9. Slepushkin, V. A.; Andreev, S. M.; Sidorova, M. V.; Melikyan, G. 8.;
Grigoriev, V. B.; Chumakov, V. M.; Grinfeldt, A. E.; Manukyan, R. A.; Karamov, E.
V., Investigation of human immunodeficiency virus .fusion peptides. Analysis of
interrelations between their structure and function. AIDS Res Hum Retroviruses
1992, 8, (1), 9-18.

10. Gordon, L. M.; Curtain, C. C.; Zhong, Y. C.; Kirkpatrick, A.; Mobley, P. W.;
Waring, A. J., The amino-terminal peptide of HIV-1 glycoprotein 41 interacts with
human erythrocyte membranes: peptide conformation, orientation and
aggregation. Biochim. Biophys. Acta 1992, 1139, (4), 257-274.

77

11. Chang, D. K.; Cheng, S. F.; Chien, W. J., The amino-terminal fusion
domain peptide of human immunodeficiency virus type 1 gp41 inserts into the
sodium dodecyl sulfate micelle primarily as a helix with a conserved glycine at
the micelle-water interface. J. Virol. 1997, 71, (9), 6593-6602.

12. Waring, A. J.; Mobley, P. W.; Gordon, L. M., Conformational mapping of a
viral fusion peptide in structure-promoting solvents using circular dichroism and
electrospray mass spectrometry. Proteins 1998, Suppl, (2), 38-49.

13. Peisajovich, S. G.; Epand, R. F.; Pritsker, M.; Shai, Y.; Epand, R. M., The
polar region consecutive to the HIV fusion peptide participates in membrane
fusion. Biochemistry 2000, 39, (7), 1826-33.

14. Taylor, S. E.; Desbat, B.; Blaudez, D.; Jacobi, 8.; Chi, L. F.; Fuchs, H.;
Schwarz, G., Structure of a fusion peptide analogue at the air-water interface,
determined from surface activity, infrared spectroscopy and scanning force
microscopy. Biophys. Chem. 2000, 87, (1), 63-72.

15. Jaroniec, C. P.; Kaufman, J. D.; Stahl, S. J.; Viard, M.; Blumenthal, R.;
Wlngfield, P. T.; Bax, A., Structure and dynamics of micelle-associated human
immunodeficiency virus gp41 fusion domain. Biochemistry 2005, 44, (49), 16167-
16180.

16. Gabrys, C. M., and Weliky, D.P., Chemical Shift Assignment and
Structural Plasticity of a HIV Fusion Peptide Derivative in
Dodecylphosphocholine Micelles. BBA-Biomembranes 2007, 1768, 3225-3234.

17. Aloia, R. C.; Tian, H.; Jensen, F. C., Lipid composition and fluidity of the
human immunodeficiency virus envelope and host cell plasma membranes. Proc.
Natl. Acad. Sci. USA. 1993, 90, (11), 5181-5185.

18. Bodner, M. L., Solid state nuclear magnetic resonance of the HIV-1 and
influenza fusion peptides associated with membranes. Ph. D. thesis, Michigan
State University: East Lansing, 2006; p 122.

19. Qiang, W., Bodner, M.L., and Weilky, D.P., Solid-state NMR Spectroscopy
of HIV Fusion Peptides Associated with l-Iost-CeII-Like Membranes: 2D
Correlation Spectra and Distance Measurements Support a Fully Extended
Conformation and Models for Specific Antiparallel Strand Registries. J. Am.
Chem. Soc. 2008, 130, 5459-5471.

20. Sackett, K., and Weliky, D.P., Unpublished experiments.
21. Yang, J.; Gabrys, C. M.; Weliky, D. P., Solid-state nuclear magnetic

resonance evidence for an extended beta strand conformation of the membrane-
bound HIV-1 fusion peptide. Biochemistry 2001,40, (27), 8126-8137.

78

22.
for p
HIV-

23.
refel

24.

25.

fusic
whic
(45)
26.

Ollg=
Corr
t0 Li

27.

909
A081

28.
lipid.
199'

29.
Pho:
elas

30.

Sphi

31.
med

32.
men

33.
Ann.

22. Yang, J.; Weliky, D. P., Solid state nuclear magnetic resonance evidence
for parallel and antiparallel strand arrangements in the membrane-associated
HIV-1 fusion peptide. Biochemistry 2003, 42, 11879-11890.

23. Zhang, H. Y.; Neal, 8.; Wishart, D. S., RefDB: A database of uniformly
referenced protein chemical shifts. J. Biomol. NMR 2003, 25, (3), 173-195.

24. Qiang, W., and Weilky, D.P., unpublished experiments.

25. Yang, R.; Prorok, M.; Castellino, F. J.; Weliky, D. P., A trimeric HIV-1
fusion peptide construct which does not self-associate in aqueous solution and
which has 15-fold higher membrane fusion rate. J. Am. Chem. Soc. 2004, 126,
(45), 14722-14723.

26. Qiang, W., and Weilky, D.P., HIV Fusion Peptide and Its Cross-Linked
Oligomers: Efficient Syntheses, Significance of the Trimer in Fusion Activity,
Correlation of (3 Strand Conformation with Membrane Cholesterol, and Proximity
to Lipid Headgroups. Biochemistry 2009, 48, (2), 289-301.

27. Freed, E. O.; Delwart, E. L.; Buchschacher, G. L., Jr.; Panganiban, A. T.,
A mutation in the human immunodeficiency virus type 1 transmembrane
glycoprotein gp41 dominantly interferes with fusion and lnfectivity. Proc. Natl.
Acad. Sci. U. SA. 1992, 89, (1), 70-74.

28. Bloom, M.; Evans, E.; Mouritsen, O. G., Physical properties of the fluid
lipid-bilayer component of cell membranes: a perspective. Quat. Rev. Biophys.
1991,24, (3), 293-397.

29. Smaby, J. M.; Momsen, M. M.; Brockman, H. L.; Brown, R. E.,
Phosphatidylcholine acyl unsaturation modulates the decrease in interfacial
elasticity induced by cholesterol. Biophys. J. 1997, 73, (3), 1492-1505.

30. Li, X. M.; Momsen, M. M.; Smaby, J. M.; Brockman, H. L.; Brown, R. E.,
Cholesterol decreases the interfacial elasticity and detergent solubility of
sphingomyelins. Biochemistry 2001,40, (20), 5954-5963.

31. Silvius, J. R., Role of cholesterol in lipid raft formation: lessons from lipid
model systems. Biochim. Biophys. Acta-Biomembr. 2003, 1610, (2), 174-183.

32. Binder, W. H.; Barragan, V.; Menger, F. M., Domains and rafts in lipid
membranes. Ange. Chem-Int. Ed. Engl. 2003,42, (47), 5802-5827.

33. Simons, K.; Vaz, W. L. 0., Model systems, lipid rafts, and cell membranes.
Ann. Rev. Biophys. Biomol. Struct. 2004, 33, 269-295.

79

phc
biol

35.
lmr
Bio,

36.

con
Gra

37.
stal
orie
Ph)

38.
Hul
des
fusi

39.
A.S
20C

40.
99‘
strl

34. McMullen, T. P. W.; Lewis, R.; McElhaney, R. N., Cholesterol-
phospholipid interactions, the liquid-ordered phase and lipid rafts in model and
biological membranes. Curr. Opin. Colloid Interface Sci. 2004, 8, (6), 459—468.

35. Li, Y. L., AND Tamm, L.K., Structure and Plasticity of the Human
Immunodeficiency Virus gp41 Fusion Domain in Lipid Micelles and Bilayers.
Biophys. J. 2007, 93, (3), 876-885.

36. Yang, J.; Parkanzky, P. D.; Khunte, B. A.; Canlas, C. G.; Yang, R.; Gabrys,
C. M.; Weliky, D. P., Solid state NMR measurements of conformation and
conformational distributions in the membrane-bound HIV-1 fusion peptide. J. Mol.
Graph. Model. 2001, 19, (1), 129—135.

37. Wasniewski, C. M.; Parkanzky, P. D.; Bodner, M. L.; Weliky, D. P., Solid-
state nuclear magnetic resonance studies of HIV and influenza fusion peptide
orientations in membrane bilayers using stacked glass plate samples. Chem.
Phys. Lipids 2004, 132, (1 ), 89-100.

38. Hofmann, M. W.; Weise, K.; Ollesch, J.; Agrawal, P.; Stalz, H.; Stelzer, W.;
Hulsbergen, F.; de Groot, H.; Gerwert, K.; Reed, J.; Langosch, D., De novo
design of conformationally flexible transmembrane peptides driving membrane
fusion. Proc. Natl. Acad. Sci. U. SA. 2004, 101, (41), 14776-14781.

39. Reichert, J., Grasnick, D., Afonin, S., Buerck, J., Wadhwani, P., and Ulrich,
AS, A critical evaluation of the conformational requirements of fusogenic
peptides in membranes. European Biophysics Journal with Biophysics Letters
2007, 36, (4-5), 405-413.

40. Buzon, V.; Padros, E.; Cladera, J., Interaction of fusion peptides from HIV

gp41 with membranes: A time-resolved membrane binding, lipid mixing, and
structural study. Biochemistry 2005,44, (40), 13354-13364.

80

CHAPTER V

TERTIARY STRUCTURES OF MEMBRANE-ASSOCIATED HIV FUSION
PEPTIDE

BACKGROUND

Previous studies on V2E HIV mutants suggest that HFP oligomerization
is a structural requirement for fusion. Such oligomers could be formed in the (3
strand HFP conformation through inter-peptide hydrogen bonding.(1) Solid-state
NMR REDOR experiments were performed on HFPmn associated with
cholesterol-containing membranes samples in which half of the peptides
contained specific 13C carbonyl backbone labels and the other half of the
peptides contained specific 15N backbone labels. Strong evidence for oligomeric
8 structures was provided by observation of a significant reduction of the ‘30
signals by the 15N nuclei.(2) Previous analytical Ultracentrifugation (AUC) studies
concluded that HFPmn, which contained a number of non-nature charged
residues in the C-terminus, was monomeric in solution and was converted to
oligomers as a result of membrane association.(3) In addition, samples prepared
by different methods had very similar NMR spectra, which indicated that the
oligomeric 6 structure was an equilibrium rather than a kinetically trapped
structure.(4) Since both the host cell and virus membranes contain a large
fraction of . cholesterol.(5) the formation of oligomeric 8 structures may be
biologically relevant when the HIV fusion peptide interacts with the membranes.

The possible biological significance of fusion peptide oligomerization is

also suggested by the following evidence: (1) The atomic-resolution structures of

81

the soluble ectodomain of gp41 showed stable gp41 trimers.(6) These structures
ended ~ 10 residues C-terminal of the fusion peptide domain, and their three N-
termini were close together at the ends of an in-register helical coiled-coil.(7)
Thus, it appears that at least three fusion peptides are in close proximity when
they interact with the target membrane. (2) Experiments and modeling studies
further suggested that the fusion site contained multiple envelope protein
trimers.(8, 9) (3) The functionally disruptive Val-2 to Glu-2 mutation in the gp41
fusion peptide is trans-dominant, i.e., cells expressing 10% mutant proteins and
90% wild-type protein exhibit only 40% of the fusion activity of cells with 100%
wild-type protein. One interpretation of the mutation study was that the mutant
peptide disrupts the correction assembly of a functionally important fusion
peptide oligomer.( 1)

The combination of previous works suggested the importance to study
the tertiary structure of the membrane-associated (3 strand HFP. The specific
aims include: (1) is there a predominant parallel or anti-parallel [3-sheet structure;
and (2) is there a preferred or several preferred registries for the B—sheet
arrangement? In the results and discussion part of this chapter, the evidence for
the existence of anti-parallel B-sheet structure will be provided, and the indirect
evidence will be described for the predominance of anti-parallel structure versus
parallel structure by comparing the present result with previous studies.(2, 10) In
addition, two anti-parallel B-sheet registries have been observed with the overlap
of the N-terminal 16 or 17 residues. Finally, a procedure to quantify the

percentage of certain anti-parallel B—sheet registries will be provided, and the

82

IBSI

HFF

MA'

with
cysl
iuds
HFF
lnte
cha)
MAI

Tabli
Nal

HFF
HFF
HFF
HF;
HFF

a The
the b

DIOC

aSSC

results showed that ~ 25 % of HFPmn has a 16-residue overlap and ~ 30 % of

HF Pmn has a 17-residue overlap.( 10)

MATERIALS AND METHODS

Peptide. All peptides used in this chapter were summarized in Table 4

with their sequences and 13CO and 15N labeled residues. Although there was a

cysteine in the C-terminus, the peptides were predominantly non-cross-Iinked as

judged

by monomeric molecular weight in AUC data. (3,11) The synthesis of

HFP monomers was completed using a 15 mL manual reaction vessel (Peptides

International, Louisville, KY) and followed the same procedure as described in

chapter II for HFPmn and in appendix 3 with FMOC chemistry.(12) HPLC and

MALDI-TOF-MS were used to purify and identify the peptides.

Table 4 peptide sequences and labeling schemes.a

 

 

Name Sequence

HFP-A AVGIGALFLGFLGAAGSTMGARSWKKKKKKAB
HFP-B AVGIGALFLGFLGAAGSTMGARSWKKKKKKAB
HFP-C AVGIGALFLGFLGAAGSTMGARSWKKKKKKA”
HFP-D AVGIGALFLGFLGAAGSTMGARSWKKKKKKAB
HFP-E AVGlGALFLGFLGAAGSTMGARSWKKKKKKAB

 

' The red symbol indicates the 15N-labeled amino acids and
the blue color indicates the 13CO-labeled amino acids.

 

NMR Sample Preparation. The sample preparation followed the

procedure provided in chapter IV and all peptides described in Table 4 were

associated with PC:PG:CHOL.

83

’30-’5N REDOR Experiments and Simulations. The triple resonance MAS
probe was used and tuned to ”C, 1H, and 15N frequencies of 100.8 MHz, 400.8
MHz, and 40.6 MHz, respectively and the 130 transmitter was at 152.4 ppm. The
REDOR sequence was shown in Fig. 10 and the following parameters were used
in the present set of experiments: (1) a 44 kHz 1H 17/2 pulse; (2) 2.2 ms cross-
polarization with 63 kHz 1H field and 76-84 kHz ramped 13c field; and (3) a
dephasing period of duration 1 for which the “So” and “Si” acquisitions contained
62 kHz ‘30 TI' pulses at the end of each rotor cycle except the last cycle and for
which theSi acquisition contained 27 kHz 15N 17 pulses in the middle of rotor
cycles; and (4) 13C detection.(2,13-16) XY-8 phase cycling was applied to the 13C
and 15N pulses during the dephasing period, TPPM 1H decoupling of ~95 kHz
was applied during the dephasing and detection periods, the recycle delay was 1
s, and the MAS frequency was 8000 :I: 2 Hz. REDOR experiments were
calibrated using a Iyophilized “l4” peptide with sequence
AcAEAAAKEAAAKEAAAKA-NHz and a 13co label at Ala-9 and a 15N label at
Ala-13. For the predominant 01 helical conformation of l4, the labeled 13CO-‘5N
distance is ~4.1 A.(13,17)

The So REDOR spectrum contained all ‘30 signals while the Si spectrum
had reduced signals from 130 with proximal 15N and therefore appreciable 13C-
15N dipolar coupling (d). The equation (1 = 3100/r3 expresses the relation between
din Hz and 13C-“"N distance (r) in A. The data analysis focused on integrated So
and 81 intensities in the labeled 13CO region that were denoted as “So” and “81”,

respectively, and an experimental fractional dephasing (AS/So)” = (80’1” -

84

81"”)/So°"" was calculated for each 1'. The (AS/So)” provided the experimental
basis for determination of d and r. The 0- °""’ uncertainty in (AS/So)” was

calculated by

 

“502 x 0512) + (S12 x as:

I
SO, (35)

exp

 

0'

where ”so and 031 are the experimental root-mean-squared noise of the So and

81 spectra, respectively.(18)

Calculations of (AS/So) as a function of spin geometry were denoted
(AS/So)“ and were made using the SIMPSON program.(19) The calculations
were based on two or three spins where one of the spins was the Ala-14 13CO in
a central (3 strand and the other one or two spins were labeled 15N on adjacent
strands. In order to make meaningful comparison between the (AS/So)“ which
were based only on labeled nuclei and (AS/So)” which included contributions
from both labeled and natural abundance nuclei, (AS/So)°°’ were calculated from
the (AS/So)°"” and reflected removal of the natural abundance contribution.

The following parameters/approximations are used:

A1. There is 99% labeling of the Ala-14 13CO and Val-2, Gly-3, Ile-4 or
Gly-5 15N sites. 31 = so for a labeled Ala-9 13co in a molecule with a Val-2, Gly-3,
lie-4 or Gly-5 1“N.

A2. Effects of natural abundance 15N on 13CO 81 signals are evaluated
using the following criteria: (1) 81 = 0 for a labeled Ala-14 13CO separated by one

or two bonds from a natural abundance 15N at Ala-15 and Ala-14. Ala-14 81 is not

85

affected by other natural abundance 15N. (2) 81 = 0 for natural abundance
backbone 13COs at Ala-1 and Val-2, Val-2 and Gly-3, Gly-3 and lie-4, or IIe-4 and
Gly-5 which are separated by one or two bonds from the labeled Val-2, Gly-3, lle-
4 or Gly-5 ‘5N, respectively. 81 = So for other natural abundance backbone 13CO
sites. Criteria (1) and (2) are based on the close distance (5 2.5 A) and
consequent strong (2 200 Hz) dipolar coupling of 13CO and 15N nuclei separated

by one or two bonds.

Figure 17 displays a flow chart for the determination of (AS/So)” for HFP-
B with 13co labeled Ala-14 and 15N labeled Val-2. (AS/So)°°’ for the other HFP
samples were derived based on the same flow chart but only with different 15N
labeling.

A complete derivation of (AS/So)” follows:

exp exp_ exp
[AS] = SO__;SJ__ (35)

30— S3”
55"” is expressed as the sum of contributions from labeled 13CO nuclei (SC’,“”) and
from natural abundance 13CO nuclei (S;“"):
53"!” = 53"” + 55*“- = I - UC + 71 AC (37)

where 1 — Uc is the fractional Ala-14 13CO labeling, Ac is the fractional 13C

natural abundance, and n is the total number of unlabeled peptide backbone CO

sites in an HF P molecule. Sf” is also expressed as the sum of contributions from
labeled 13CO nuclei (Sl’ab) and from natural abundance 13CO nuclei (S{"“' ):

S,” = S,” + s,”- (38)

86

 

 

 

Site oEZhBfion”
Relative Populatlon
[Contrlbutlon to 80]
{Contribution to S1)
9 l '
l l
' HFP labeled carbonyls l HFP nia‘can‘dhyls
1 ' l nAC ,
II I II
(I - _ . ._.-_.__.I.L- -,__
I
, _. _ _ .7“? 'f" ”7‘77"”7".PM}-.- . . _-_.1’_---'._ .1; i"; ...L. ._ H. .-
(1300.414 and 15N v2 13COA14 only . 15N v2 only . IAiavz he carbonyls iothorna calionyls
r 1-UC-UN UN . uc ' I 2530 .f (0112])Ac
= [I [1] i [0] r ' ; 3
l _,i°i_ - 11}---

_ --.Q- ‘ W. ' .10}

I

._,__V_.+,.. -_.,V_.. ._-_.. +_,,-

wlo nearby n.a.15N'. WI") nearby n.a.15N ‘
(t-UC-UN)(1-2AN)1 , (1-UC-UN)2AN l
[‘1 ‘ g

j

l'
l
I
l
l
l
l
I

, 111
if) ' 1_ 10}

l

 

Figure 17 Flow chart of derivation of (AS/So)” for REDOR of HFP-B. The four
rows in each box are in sequence: the site description, its relative population, and
its contributions to So and Sr.

87

with:
S1!“ =(1— UC— UN)(I — 2AN)f + UN (39)
and:
S,“- = (n—2)AC (40)
where 1 — UN is the fractional 15N labeling of the Val-2, Gly-3, lle-4 or Gly-5

residue for HFP-H, HFP-I, HFP-J and HFP-K respectively, AN is the fractional

15N natural abundance and the parameter 1‘.

cor cor_ car car
f=l =l——9———1——=I—-A—S (41)
3601' 5607‘ S0

 

Incorporate Eq. 41 into Eq. 39:

SlIab =(l—UC-UN)(l—2AN)[l—[%S—] ]+UN
0

(42)

= (l - uc- U~)(1-2A~)-(I- uc- UN)(1 -2.,)[§S_)“"+ UN

UC, UN, and ZAN are much less than 1 so that:
(1- UC- UN)(I — 2AN)51- UC— UN - MN (43)
and:
Sllab E l—UC—ZAN— (I—UC—UN—ZAN)(;'£] (44)
0
Incorporate Eqs. 40 and 44 in Eq. 38:
AS cor
SlexP= ]_UC—2AN_(l—UC-UN_2AN)(S—] + ("—2)AC (45)
0

Combine Eqs. 37, 38, 39, and 45:

88

53W —SlexP =[1—UC+nAC]— l—UC "ZAN — (l—UC —UN —2AN)[%S'] '1' ("—2)AC]
0
(46)
and simplify:
ex ex AS COI'
0
Combine Eqs. 37 and 47:

24¢ + 2.4,, + (I — UC — UN — 2A,.)(i‘EJ
(48)

Aim- 50
SO l-UC+nAC

and rewrite:

 

g 6“: l-UC+nAC g ”_ 2AC+2AN (49)
S0 (l-UC-UN-ZAN) S0 (l-UC—UN-ZAN)

Expressions in Eq. 49 were numerically evaluated using Ac = 0.011, AN = 0.0037,
n = 29, and Up = UN = 0.01 which were based on 0.99 fractional labeling of the
Ala-14 13‘CO sites and 0.99 fractional labeling of the Val-2, Gly-3, lie-4 or Gly-5

15N sites:
cor exp
[E] = 1.360(3‘5] - 0.030 (50)

Eq. 50 resulted in:
060’ = 1.360 cap (51)
Input parameters to the SIMPSON program included the 13CO-"SN dipolar

couplings, the Ala-14 13CO chemical shift and CSA principal values, and sets of

89

Euler angles which reflected the orientations of 13CO-‘5N dipolar coupling and
13CO CSA PASs in the fixed crystal frame. The 13CO chemical shift was 175 ppm
and CSA principal values were set to 241, 179, and 93 ppm, respectively.(20)
Determination of Euler angles was based on atomic coordinates of the labeled
nuclei and these coordinates were taken from crystal structure coordinates of
outer membrane protein G (OMPG) (PDB file 2|WW).(21, 22) OMPG was chosen
because the REDOR experiments probed anti-parallel B strand structure in HFP ‘
and this was the predominant OMPG structural motif. After the 13CO coordinates
were obtained from a specific residue in OMPG, 15N coordinates were obtained
from nearby residues in the two adjacent strands. The Results section includes
more detail about the specific choices of these nearby residues. For the two-spin
simulations, the (or, (3, y) Euler angles of the dipolar coupling PAS were (0, 0, 0)
and for the three-spin simulations, the angles for one dipolar PAS was (0, 0, 0)
and for the other PAS were (0, 0.0) where 9 was the angle between two 13CO-“"N
vectors. The Euler angles for the 13CO CSA PAS were calculated using the
known orientation of the PAS relative to the 13CO chemical bonds and the
OMPG-derived orientation of these chemical bonds relative to the crystal

frame.(23)

RESULTS AND DISCUSSION
Anti-parallel ,B-sheet Registries for HFPmn. 13C-‘5N REDOR experiments
were first carried out for HFP-A in Table 4 which contained 13CO labeling at

Ala14, Ala15 and Gly16 and 15N labeling at Ala1, Val2 and Gly3. This labeling

9O

scheme was chosen because: (1) if there were adjacent strand crossing between
Phe-8 and Lou-9, the 13CO-labeled residues on one strand would be hydrogen
bonded to the 15N labeled residues on an adjacent strand with concomitant 13C-
15N dipolar couplings of ~ 45 Hz; and (2) intramolecular 13CO-‘E’N couplings are
negligible. Representative spectra are displayed in Fig. 18a, b and the respective
(AS/So)”p were ~ 0.41 and ~ 0.50 for REDOR dephasing time 1: = 24 and 32 ms.
These values suggested that there was a large population of HFP with the
putative anti-parallel strand registry.(10) Unambiguous analysis of these data
was challenging because there were contributions of three distinct 1300s and
because different combinations of anti-parallel strand registries could fit the data.

Ambiguity was reduced by studying samples for which HFPmn had only a
single 13CO and a single 15N label. Four HFPs were prepared and all had a 13CO
label at Ala-14. This residue had been 13’00 labeled in HFP-A and had been
previously observed to give a fairly sharp signal.(24) The 15N label was at Val-2
(HFP-B), Gly-3 (HFP-C), Ile-4 (HFP-D), or Gly-5 (HFP-E). The variation of the
REDOR data among the different HFPs was striking (of. Figure 186-j and Figure
19a). For r= 32 ms, the (AS/So)” were ~0.3 for the HFP-C and HFP-D samples
and ~0 for the HFP-B and HFP-E samples. These data suggested that there
were two anti-parallel registries which could be classified: (1) Ala-14 on one
strand opposite Gly-3 on the adjacent strand; and (2) Ala-14 on one strand
opposite lie-4 on the adjacent strand. These two registries were denoted A and B

and are displayed in Fig. 20a.

9]

Fig
d) l
obt

brc
411
811

 

 

meadow
WW 9.1.91.0)...
mammalian.)

88040081...
mol... WM...

250 150 150
13C Chemical Shift (ppm)

 

 

 

 

 

 

 

 

Figure 18 REDOR So and Sr spectra for membrane-associated (a, b) HFP-A, (c,
d) HFP-B, (e, f) HFP-C, (g, h) HFP-D and (i, j) HFP-E. Spectra a, c, e, g, iwere
obtained with 24 ms dephasing time and spectra b, d, f, h, j were obtained with
32 ms dephasing time. Each spectrum was processed with 200 Hz Gaussian line
broadening and baseline correction. Each So or 81 spectrum was the sum of (a)
41328, (b) 56448, (c) 45920, (d) 81460, (e) 55936, (f) 79744, (9) 30898, (h)
81856, (i) 45920 or G) 71040 scans.

92

 

 

 

 

 

0 10 20 30
Dephasing time (ms)

 

(b)

8
0A
0
(Q
U)
3

0.4‘

(10-

 

 

 

l l l
0 10 20 30

Dephasing time (ms)

Figure 19 Plots of (AS/So)” vs dephasing time for membrane-associated HFP

samples prepared with [HFPlinitial of (a) 400 (M or (b) 25 pM. The symbol legend
is: diamonds, HFP-B; triangles, HFP-C; circles, HFP-D; and squares, HFP-E.

The a°°'were ~0.04.

93

(bl

(3) AVG IGALFLGFLGAAG AVG | GALFLGFLGAAG
GAAGLFGLFLAGI GVA GAAGLFGLFLAG | GVA

"1 Q o o Q till 0 ® 0 0
I11 \11’ 1l1 \11’ I11 l \11' l I
<1ch 91®®® G 96
112112 /12’ /12’ 1 112 1 /12’ 1
GD 09 ® 09 0 09 009 0

Figure 20 (a) Two antiparallel registries of residues 1-16 of HFP that were
consistent with the REDOR data shown in Fig. 20. The registries are denoted A
and B and the 13CO labeled Ala-14 residue is highlighted in blue. (b) Models
used to calculate (AS/So)” and spin geometries specific for the HFP-C sample.
Each model includes nuclei from three adjacent strands with the Ala-14 13CO
always in the middle strand and 15N in the top and/or bottom strands. The first
letter in the labeling of each model refers to the middle strand/top strand registry
and the second letter refers to the middle strand/bottom strand registry. Registry
X is any registry for which the interpeptide 13CO-‘5N distance was large in the
HFP-H, HFP-l, HFP-J, or HFP-K samples so that d as 0. The Ala-14 13CO is
hydrogen bonded to an amide proton in the top strand. Relevant labeled 13C-‘E’N
distances and 15N-‘3C-‘5N angles are: r1 = 4.063 A; N = 5.890 A; r2 = 5.455 A; ['2'
= 6.431 A; 611 = 161.1’; 02 = 131.9'; i93= 130.2°; and HF 117.0°. Each
parameter value was the average of 10 specific values taken from the crystal
structure of outer membrane protein G.

94

The
[HF
ass
and
con
F 191
of

”181

ofl

stur

cor
at ;

the

The samples used to obtain data for Figures 18 and 193 were made with
[HFPmnlinitial e 400 uM. In order to check for possible effects of HFP self-
association in aqueous solution prior to membrane binding, two additional HFP-C
and HFP-E samples were made with [HFPmnlinitiai z 25 pM which is a
concentration for which HF P is known to be monomeric in the HEPES buffer.(4)
Figures 19a,b illustrates that very similar (AS/So)” were obtained for both values
of [HFPmnjrnmar and the apparent strand registries appear to be due to
membrane-association.

The registries proposed in Figure 20a were consistent with a previous set
of PDSD experiments in which “scatter-uniform” labeled (SUL) HFPmn were
studied and crosspeaks were observed between Ala6 and Gly10 and between
lle4 and Ala13.(25) In addition, it was worthwhile comparing the REDOR
dephasing obtained at 1- = 24 ms for HFP-A to a previous set of triply-labeled
HFP monomer samples where 50 % triply 1"’CO-Iabeled peptide and 50 % triply
15N-Iabeled peptide had been used. In those cases the combination HFPmn-
(A14A15G16)c/HFPmn-(A14A15616)N provided ~ 4 °/o dephasing and the
combination HFPmn-(A14A15G16)c/HFPmn-(GSA6L7)N gave ~ 16 % dephasing
at 24 ms, while Figure 183 showed the HFP-A sample gave ~ 55 % dephasing at
the same dephasing period.(2) Although the probability for a 13C-Iabeled HFP
aligned with a 1‘r’N-Iabeled HFP was only half of this probability in the present
study, there was still a clear trend that the dephasing will increase when the

labeling pattern satisfy an anti-parallel instead of a parallel B-sheet arrangement.

95

an.

the

Sir

re

re

th

Quantitative Anti-parallel ,B-sheet Registry Models. More quantitative
analysis of the (AS/So)” of the samples was done using calculations of
(AS/So)” based on different models for registries of three adjacent strands with
the overall goal of quantization of the populations of the different registries. The
strands were denoted, “top”, “middle”, and “bottom”. Figure 20b displays the
models as well as spin geometries specific to the HFP-C sample. The models
were focused on registries at the middle strand Ala14 whose 13CO group was
hydrogen bonded to an amide proton in the top strand. Each model was labeled
by two letters which were either A, B, or X. The first letter described the registry
relating the middle strand and the top strand and the second letter described the
registry relating the middle strand and the bottom strand. For registry A, Ala14 in
the middle strand was across from Gly3 in the adjacent strand and for registry B,
Ala14 in the middle strand was across from Ile4 in the adjacent strand (of. Figure
20a). Registry X was defined as any registry for which the inter-peptide 13CO-‘5N
distance was large in the HFP-B, HFP-C, HFP-D, and HFP-E samples so that d is
0. Registry X could include the in-register parallel strand arrangement. Such
registry has been proposed for membrane-associated gp41 constructs which
contain the HFP.(26)

Model )0( had (AS/So)“ = 0 for all dephasing times while models AX, XA,
BX, and X8 resulted in two-spin systems for which (AS/SOP“ were primarily
dependent on the 13CO-“"N distance. Models AA, BA, AB, and BB were three-
spin systems for which (AS/So)” depended both on the two 13CO-‘5N distances

and on the angle between the two 13CO-"5N vectors.(27) For all samples and all

96

m0r

timt

mm
081
the

601

10-

fil

fi.

models, (AS/So)“ were calculated for each of the five experimental dephasing

times.

The fractional populations of each of the models were calculated with
fitting of the (AS/So)” and the (AS/So)”. The fitting was primarily based on the
data from the HFP-C and HFP-D samples because many of the (AS/So)” for
these samples were appreciably positive. Fitting was accomplished with the

equafions:

 

{11%. if: 13%.):1

2
‘ 6'0"
(“1'1 1

12=Z

j=l k

(52)

Mo.

ll
~

calc 9 sim
(1%.)... = 2;, 1 . (1%.)... 1531

for which j was the index of the sample, k was the index of the dephasing time, /

was the index of the model, and fr was the fractional population of model I.

Three types of fitting were done and differed in the choice of which 1', were
fitted and which were set to zero. For all fittings, 2 f, = 1. For “unconstrained”
fitting, there was no correlation between the registry of the middle and top
strands and the registry of the middle and bottom strands. All f were therefore
fitted and each f was a function of “a” and “b” which were defined as the
fractional probabilities of two adjacent strands having A or B registries,
respectively. The fractional probability of the X registries was then 1 — a — b.
Each f was the product of the fractional probabilities of the middle strand/top

strand registries and middle strand/bottom strand registries with resulting fAA = 32,

97

f3A= ab, fA3= ab, f33= b2, fAX= a(1-a—b),fx,q = a(1—a-b),fax= b(1—a—b),
fxa = b(1 - a - b), and fxx = (1 — a - b)2. “Partially constrained” fitting was done
based on the idea that there were domains of antiparallel strand registry and
domains ofX registry so that no. = a2, rs. = ab, rag = ab, r35 = 52, fax = 0, 5.. = 0,
fax = 0, fxa = 0, and fxx = 1 — (a + b)2. For partially constrained fitting, physically
meaningful expressions of a and b included: (1) a/b which was the ratio of
probability that two adjacent strands had A registry to the probability that they
had B registry; and (2) (a + b)2/[(1 -(a + b)]2 which was the ratio of the total
population of the A and B antiparallel structures to the population of the X
structures. For “fully constrained” fitting, it was assumed that [3 strand domains
would form with only A or only B or only X registries so that fat = 62, fBA = 0, fag =
0, 1135 = b2, rAx = 0, 1).. = 0, fax = 0, 5.3 = 0, and rxx s 1 — a2 - b2. In this fitting, the
fractional populations of the A, B, and X strand arrangements were 82, b2, and
1 - a2 - b2, respectively.

The results of unconstrained fitting are displayed in Figure 21a as a 20
contour plot of )1 vs a and b. The best-fit a = 0.22 and b = 0.31 with {mm = 16.5
and good-fit a and b represented in the black region.(18, 20) The good-fit regions
of the plot showed negative correlation between a and b as might be expected
from the positive correlation between (AS/Sn)°‘"‘ and either a or b for both the
HFP-l and HFP-J samples. The (AS/Soft"c were also computed for the HFP-H
and HFP-K samples using the best-fit a and b. At 1= 32 ms, maximum (AS/So)”c
of 0.08 and 0.09 were obtained for the HFP-H and HFP-K samples, respectively,

and can be compared to the maximum (AS/So)” = 0.05 d: 0.04 for these samples.

98

Figure 21 b displays the ZD contour plot of partially constrained fitting with best-fit
a = 0.31 and b = 0.42 with [min = 15.1. At r= 32 ms, these a and b values led to
(AS/So)“’° = 0.11 and 0.13 for the HFP-H and HFP-K samples, respectively.
Figure 21c displays the 2D contour plot of fully constrained fitting with best-fit 32
= 0.26 and b2 = 0.33 with 22min = 12.7 and (AS/So)“’° = 0.09 and 0.12 at r= 32
ms for the HFP-H and HFP-K samples, respectively. For all three fittings, the
23mm are reasonable, as evidenced by being within a factor of 2 of 8', the. number
of degree of freedom of the fitting. This suggests that each model is plausible.
The limits of the good-fit black regions have been generously set and include all
parameter space with f 2-3 units higher than 33mm.

The best-fit f of the three fittings were used to calculate PA, P3, and PX
which were fractional populations of the A, B, and X registries, respectively: PA =
fAA "‘ (113.4 + 018 1' fo '1' fXA)/2 ; P8 = fee ‘1 (fBA 1' 1'48 + fax '1' fXB)/2 : and Px = f)o< 1' (fo
+ fxa + fax + fx3)/2 with PA + P3 + Px = 1 . The resulting fractional populations
were: (1) unconstrained fitting, PA = 0.22, P5 = 0.31, and Px = 0.47; (2) partially
constrained fitting, PA = 0.23, P3 = 0.31, and Px = 0.46; and (3) fully constrained
fitting, PA = 0.26, P3 = 0.33, and Px = 0.41 . An overall result of the three fittings
was therefore PA is 0.25, Pa z 0.30, and PX z 0.45. In addition, examination of the
values in the black regions of the three plots showed that the approximate range
of reasonable values for the sum PA + P3 was 0.5-0.6 and the corresponding

range of PX was 04-05.

99

(a)

(b)

(C)

 

1.0

0.8—1

0.6-1

0.4“

0.2—1

 

 

 

 

0.2

0.4

0.6

0.8 1.0

 

1.0

0.8-1

0.6-

0.4—

0.2..

 

 

 

0.2

0.4

0.6

0.8 1.0

 

1.0

0.8-'1

0.6"

0.4":

0.2-7

 

 

 

0.8 1.0

100

Figure 21 Contour plots of ,1? vs
strand fitting parameters for (a)
unconstrained; (b) partially
constrained; and (6) fully
constrained fittings. The a, b, a,
and b2 parameters refer to
probabilities for different adjacent
strand arrangements. In plot a, the
black, green, blue, red, and white
regions respectively correspond to
f<19,19<f<21,21<f<23,
23 < f < 25, and f > 25. In plot b,
the regions respectively correspond
tof<18,18<12<20,20<f<22,
22 < f < 24, and f > 24, and in
plot c, the regions respectively
correspond to f< 15, 15 < 2’2 <17,
17<f<19,19<j<21,andf>
21. Best-fit parameters were: plot a,
a = 0.22, b = 0.31, f = 16.5; plot b,
a = 0.31, b = 0.42, = 15.1; and
plot c, a2 = 0.26, b = 0.33, ,y’ =
12.7. In plot a, the a and b
parameters are the fractional
probabilities of adjacent strands
having A or B registries, respectively.
In plot c, the a2 and b2 parameters
are the fractional probabilities of
domains of A or B registries,
respectively.

Previous studies in our group using ZD PDSD methods detected inter-
peptide or inter-residue crosspeaks with long mixing times between Ala6 and
Gly10, and between lle4 and Gly13.(25) These observations were consistent with
the A and B registries. Compared with the previous SUL samples, the REDOR
data were more quantitatively analyzed and also significant based on our
knowledge because it provided the first residue-specific structural model for the
B—strand HF Ps. There would be complete (registry A) or nearly complete (registry
B) inter-peptide hydrogen bonding for residues Ala1 to Gly16 which form the
apolar region of the HFP. These hydrogen bonding patterns would be favored if
this region were predominantly located in the membrane interior. The existence
of multiple p-strand structures is also consistent with a recent 13C and 15N
assignment of a membrane-associated HFP with SUL at Ph68, Leu9 and Gly10.
(28) There were two crosspeaks with comparable intensity for the Leu9
13CO/Gly10 15N correlation and two crosspeaks with comparable intensity for the
Gly10 13CO/Gly10 15N correlation. For a given pair, the two 13C shifts differed by
~0.5 ppm and were both consistent with B—strand conformation whereas the
Gly10 15N shifts were 107 and 111 ppm. The two crosspeaks may correlate with
the multiple [El-strand structures inferred from analysis of the REDOR data in the
present study. It was also interesting to compare the current REDOR work with
one anti-parallel registry suggested by a previous REDOR work where samples
contained an equimolar mixture of a HFP with three sequential 13CO labels and a
HF P with three sequential 15N labels were used.(2) Data were only acquired for a

single dephasing time (1' = 24 ms) and the best-guess anti-parallel registry had

101

Ala14 hydrogen bonded with Leu7 which is different than the registries A and B
in the present study. The Ala14/Leu7 could be one of the X registries but it is
noted that there was significant uncertainty in the determination of this registry
because of the multiple 13CO and 15N labels. Because of the single site 13CO and
15N labeling in the present study, there was definitive determination of the A and
B registries and these registries are biologically reasonable.

The model peptide-membrane system was used to mimic the environment
of membrane-associated gp41 protein so it would be worthwhile considering the
registries A and B in the context of the full gp41 proteins. The gp41 ectodomain
structures to-date show a symmetric trimer with an in-register parallel coiled-coil
extending over residues 30-80.(7) The residues N-terminal of Ala30 are
disordered and the soluble ectodomain construct also lacked the N-terrninal HFP.
Although there is no evidence that the oligomeric state of the membrane-
associated HFP of the present study is a trimer, it is interesting to consider the
anti-parallel B-sheet structure in the context of a putative trimeric state of intact
gp41. It is difficult to understand this structure in the context of a single gp41
trimer, but this structure could be understood considering two trimers denoted “C”
and “D” with respective HFP strands Cr, 02 and C3 and D1, Dz and 03. A
010302020301 anti-parallel B-sheet structure could be formed with the Cr, C2 and
Ca strands parallel to one another, and the Dr, D; and 03 strand parallel to one
another, and the C and D strands anti-parallel to one another with 03 hydrogen
bonded to C1 and 02, C2 hydrogen bonded with 03 and D2, etc. There is some

support for this model from internuclear distance measurements on a HFP trimer

102

construct composed of three HFP strands chemically cross-linked at their C-
termini. The ‘30-‘30 and 13C-“"N distances determined for this membrane-
associated trimer were consistent with the A anti-parallel registry deduced from
the present study.(13) The existence of multiple registries for a membrane-
associated HFP could be biological relevant. One factor favoring the formation of
A registry is inter-peptide hydrogen bonding for all the residues between Ala1
and Gly16. This hydrogen bonding would reduce the unfavorable Born energy of
CO and NH dipoles in the low electric environment of the membrane interior. For
the B registry, Ala1 is not part of the hydrogen bonding B—sheet registry and if the
HFP N-terminus is charged, better charge solvation might be achieved relative to
the A registry. Ala1 could adopt a broader range of conformation in the B registry
which might facilitate the location of the charged N-terminus in a solvated
environment. A greater distribution of the conformations for Ala1 is supported by
Iinewidths which were broader than those of residues in the central region of
HF P.(29) Although the ionization state(s) of membrane-associated B—strand HFP
have not yet been experimentally determined, there is evidence for a charged
amino terminus in the related influenza fusion peptide in its helical
conformation.(30) The B—sheet registries may also relate to the fusion activities of
membrane-associated HFPs. Either HIV or HFP with the Val2 —+ Glu2 point
mutation is known to be nonfusogenic.(1,31,32) In the context of our results, this
lack of fusion activity may be related to a change in the strand registries arising
from the charged Glu2 side chain. The V2E mutation is trans-dominant, that is,

mixtures of wild-type and mutant proteins correlated with fusion activities which

103

were reduced by much more than would be expected from the fraction of mutant
proteins. This effect could be explained by registry changes for several strands
near the mutant HFP which might affect HFP oligomerization and/or membrane
Iocations.(32)

Finally, the data in our study restrict the X registry to structures other than
the Ala14NaI2, Gly3, Ile4 or Gly5 anti-parallel registries. There are therefore
many possibilities for the X structures and one of these possibilities, Ala14/Leu7
anti-parallel registry, has been discussed previously. Another reasonable
possibility is parallel B-strand structure either in-register or close to in-register.
This structural model is appealing because therefore most of the residues in the
Ala1 to Gly16 region oculd have inter-peptide hydrogen bonds and this region
could be located in the membrane interior. Previous solid-state NMR 13C-‘E’N
distance measurements were consistent with some population of in-register
parallel strand structure over residues Gly5 to Gly13 in addition to anti-parallel
population over residue Gly5 to Gly16.(2,33) In addition, infrared studies on
constructs containing the first 34 or first 70 residues of gp41 were consistent with
predominant in-register parallel B-sheet structure from residue Ala1 to Gly16.(26)
The interpretation of the infrared data was based on shifts in peak wavenumbers

of ‘30, labeled relative to natural abundance peptides.

CONCLUSION
In the present work, different B-sheet registry models were tested with

REDOR 13CO-‘f’N distance measurements on a few selectively labeled samples.

104

Twc
regi
hyd
000
I002
dev
bec
whi
act

str.

Two of the registries were shown to have significant population and both
registries were consistent with complete or nearly complete inter-peptide
hydrogen bonding for the apolar N-terrninal domain of the HFP. This hydrogen
bonding scheme would be favored if a significant part of this domain were
located in the membrane interior where there is low water content. The
development of a detailed structural model for B—strand HFP is significant
because this is the observed conformation in cholesterol-containing membranes
which reflect the composition of membranes of the host cells of HIV. HFP fusion
activity is also observed for vesicles with this membrane composition and the (3-

strand conformation may therefore be a physiologically relevant HFP structure.

105

REFERENCE

1. Freed, E. 0.; Delwart, E. L.; Buchschacher, G. L., Jr.; Panganiban, A. T.,
A mutation in the human immunodeficiency virus type 1 transmembrane
glycoprotein gp41 dominantly interferes with fusion and infectivity. Proc. Natl.
Acad. Sci. U. SA. 1992, 89, (1), 70-74.

2. Yang, J.; Weliky, D. P., Solid state nuclear magnetic resonance evidence
for parallel and antiparallel strand arrangements in the membrane—associated
HIV-1 fusion peptide. Biochemistry 2003, 42, 11879-11890.

3. Yang, R.; Prorok, M.; Castellino, F. J.; Weliky, D. P., A trimeric HIV-1
fusion peptide construct which does not self-associate in aqueous solution and
which has 15-fold higher membrane fusion rate. J. Am. Chem. Soc. 2004, 126,
(45), 14722-14723.

4. Yang, J.; Prorok, M.; Castellino, F. J.; Weliky, D. P., Oligomeric beta
structure of the membrane-bound HIV-1 fusion peptide formed from soluble
monomers. Biophys. J. 2004, 87, 1951-1963.

5. Aloia, R. C.; Tian, H.; Jensen, F. C., Lipid composition and fluidity of the
human immunodeficiency virus envelope and host cell plasma membranes. Proc.
Natl. Acad. Sci. U. S.A. 1993,90, (11 ), 5181-5185.

6. Lu, M.; Blacklow, S. C.; Kim, P. S., A trimeric structural domain of the HIV-
1 transmembrane glycoprotein. Nat Struct Biol 1995, 2, (12), 1075-82.

7. Caffrey, M.; Cai, M.; Kaufman, J.; Stahl, S. J.; Vlfingfield, P. T.; Covell, D.
G.; Gronenbom, A. M.; Clore, G. M., Three-dimensional solution structure of the
44 kDa ectodomain of SIV gp41. EMBO J. 1998, 17, (16), 4572-4584.

8. Munoz-Barroso, I.; Durell, S.; Sakaguchi, K.; Appella, E.; Blumenthal, R.,
Dilation of the human immunodeficiency virus-1 envelope glycoprotein fusion
pore revealed by the inhibitory action of a synthetic peptide from gp41. J. Cell
Biol. 1998, 140, (2), 315-323.

9. Bentz, J., Minimal aggregate size and minimal fusion unit for the first
fusion pore of influenza hemagglutinin-mediated membrane fusion. Biophys J
2000, 78, (1), 227-45.

10. Qiang, W., Bodner, M.L., and Weilky, D.P., Solid-state NMR Spectroscopy
of HIV Fusion Peptides Associated with Host-CeIl-Like Membranes: 2D
Correlation Spectra and Distance Measurements Support a Fully Extended
Conformation and Models for Specific Antiparallel Strand Registries. J. Am.
Chem. Soc. 2008, 130, 5459-5471.

106

11 .
$0111
8101
12.
Prat

13.

flexi
hmm
129'

14.
hypl
thoc

15.

VVeb
dout
1995

16.
Cont

17.
sohd-
Cher.

18.
UnaF

19.
Shnul
147,

20.

mem.
1010):.
49-55

21
01 the
c01110

22.
911.; E

11. Yang, R.; Yang, J.; Weliky, D. P., Synthesis, enhanced fusogenicity, and
solid state NMR measurements of cross-linked HIV-1 fusion peptides.
Biochemistry 2003, 42, ( 12), 3527-3535.

12. Chan, W. C., and White, P.D., Fmoc Solid Phase Peptide Synthesis: A
Practical Approach 2000, 94-109.

13. Zheng, Z., Yang, R., Bodner, M.L., and Weliky, D.P., Conformational
flexibility and strand arrantments of the membrane-associated HIV fusion peptide
trimer probed by solid-state NMR spectroscopy. Biochemistry 2006,45, 12960-
12975.

14. McDowell, L. M.; Holl, S. M.; Qian, S. J.; Li, B; Schaefer, J., Inter-
tryptophan distances in rat cellular retinol-binding Protein Ii by solid-state NMR.
Biochemistry 1993, 32, (17), 4560-4563.

15. Anderson, R. C.; Gullion, T.; Joers, J. M.; Shapiro, M.; Villhauer, E. B.;
Weber, H. P., Conformation of [1-130,‘5N]acetyl-L-carnitine - rotational- echo,
double-resonance nuclear-magnetic-resonance spectroscopy. J. Am. Chem. Soc.
1995, 1 17, (42), 10546-10550.

16. Gullion, T., Introduction to rotational-echo, double-resonance NMR.
Concepts Magn. Reson. 1998, 10, (5), 277-289.

17. Long, H. W.; Tycko, R., Biopolymer conformational distributions from
solid-state NMR: alpha-helix and 3(10)-helix contents of a helical peptide. J. Am.
Chem. Soc. 1998, 120, (28), 7039-7048.

18. Bevington, P. R.; Robinson, D. K., Data Reduction and Error Analysis for
the Physical Sciences. 2nd ed.; McGraw-Hill: Boston, 1992; p 38-52.

19. Bak, M.; Rasmussen, J. T.; Nielsen, N. C., SIMPSON: A general
simulation program for solid-state NMR spectroscopy. J. Magn. Reson. 2000,
147, (2), 296-330.

20. Gabrys, C. M.; Yang, J.; Weliky, D. P., Analysis of local conformation of
membrane-bound and polycrystalline peptides by two-dimensional slow-spinning
rotor-synchronized MAS exchange spectroscopy. J. Biomol. NMR 2003, 26, (1),
49-68.

21. Yildiz, 0., Vinothkumar, K.R., Goswami, P., and Kuhlbrandt, W., Structure
of the monomeric outer-membrane porin OmpG in the open and closed
conformation. EMBO Journal 2006, 25, (15), 3702-3713.

22. Mehring, M., Principles of high-resolution NMR in solids. 2nd, rev. and enl.
ed.; Springer-Verlag: Berlin; New York, 1983.

107

23. Oas, T. G.; Hartzell, C. J.; McMahon, T. J.; Drobny, G. P.; Dahlquist, F. W.,
The carbonyl 13C chemical-shift tensors of 5 peptides determined from 15N
dipole-coupled chemical shift powder patterns. J. Am. Chem. Soc. 1987, 109,
(20), 5956-5962.

24. Qiang, W., Yang, J., and Weliky, D.P., Solid-state nuclear magnetic
resonance measurements of HIV fusion peptide to lipid distances reveal the
intimate contact of beta strand peptide with membranes and the proximity of the
Ala-14-Gly-16 region with lipid headgroups. Biochemistry 2007,46, (17), 4997-
5008.

25. Bodner, M. L., Solid state nuclear magnetic resonance of the HIV-1 and
influenza fusion peptides associated with membranes. Ph. D. thesis, Michigan
State University: East Lansing, 2006; p 122.

26. Sackett, K.; Shai, Y., The HIV fusion peptide adopts intermolecular parallel
B—sheet structure in membranes when stabilized by the adjacent N-terminal
heptad repeat: A 13C FTIR study. J. Mol. Biol. 2005, 350, (4), 790-805.

27. Vogt, F. G., Gibson, J.M., Mattingly. SM, and Mueller, K.T.,
Determination of molecular geometry in solid-state NMR: Rotational-echo double
resonance of three-spin systems. J. Phys. Chem. B 2003, 107, (5), 1272-1283.

28. Bodner, M. L., Gabrys, C.M., Struppe, J.O., and Weliky, D.P., C-13-C-13
and N-15-C-13 correlation spectroscopy of membrane-associated and uniformly
labeled human immunodeficiency virus and influenza fusion peptides: Amino
acid-type assignments and evidence for multiple conformations. J. Chem. Phys
2008, 128, 052319.

29. Yang, J.; Gabrys, C. M; Weliky, D. P., Solid-state nuclear magnetic
resonance evidence for an extended beta strand conformation of the membrane-
bound HIV-1 fusion peptide. Biochemistry 2001, 40, (27), 8126-8137.

30. Zhou, Z.; Macosko, J. 0.; Hughes, D. W.; Sayer, B. G.; l-iawes, J.; Epand,
R. M., N-15 NMR study of the ionization properties of the influenza virus fusion
peptide in zwitterionic phospholipid dispersions. Biophysical Journal 2000, 78, (5),
2418-2425. '

31. Pereira, F. B.; Goni, F. M.; Muga, A.; Nieva, J. L., Perrneabilizatlon and
fusion of uncharged lipid vesicles induced by the HIV-1 fusion peptide adopting
an extended conformation: close and sequence effects. Biophys. J. 1997, 73, (4),
1977-1986.

32. Kliger, Y.; Aharoni, A.; Rapaport, D.; Jones, P.; Blumenthal, R.; Shai, Y.,
Fusion peptides derived from the HIV type 1 glycoprotein 41 associate within

108

phospholipid membranes and inhibit cell-cell Fusion. Structure- function study. J.
Biol. Chem. 1997, 272, (21), 13496-13505.

33. Zheng, Z., Qiang, W., and Weliky, D.P., Investigation of finite-pulse
radiofrequency-driven recoupling methods for measurement of intercarbonyl
distances in polycrystalline and membrane-associated HIV fusion peptide
samples. Magn. Reson. Chem. 2007, 245, 8247-8260.

109

BA

Pei
bee
cate
mai
ach

moc

fluor
fluor

in bL

CHAPTER VI

MEMBRANE INSERTION OF HIV FUSION PEPTIDES

BACKGROUND

Current models of HIV/host cell infection include interaction of the fusion
peptide with the host cell membrane.(1,2) The membrane location of HFP has
been hypothesized to be significant structural factors for understanding the
catalysis of fusion by HFP. Previous studies about the micelle location of HFP
mainly utilized solution NMR methods where high resolution results had been
achieved.(3-7) However, there is not yet a consensus and there were distinct
models of both micelle surface location and micelle traversal.

HFP location in membranes has been primarily probed with tryptophan
fluorescence of a HFP-F8W mutant.(8-9) Key results have included: (1)
fluorescence was higher for membrane-associated HFP-F8W than for HF P-F8W
in buffered saline solution; (2) greater fluorescence quenching by acrylamide was
observed for a soluble tryptophan analog than: for membrane-associated HFP-
F8W; and (3) similar fluorescence quenching of membrane-associated HFP-F8W
was observed in samples containing either 1-palmitoyl-2-stearoyl-
phosphocholine brominated at the 6, 7 carbons of the stearoyl chain or the
corresponding lipid brominated at the 11, 12 carbons of the chain. The first two
results indicated that solvent exposure of the HFP-F8W tryptophan is reduced
with membrane association and the third result indicated that the membrane
location of the tryptophan indole group is centered near the carbon 9 position of

the brominated lipid stearoyl chain; i.e. ~8.5 A from the bilayer center and ~10 A

110

from the lipid phosphorus. Infrared and solid-state NMR spectra of membrane-
associated HFP suggested that the HFP-F8W had predominant [3 strand
conformation under the conditions of the fluorescence experiments. In a different
set of experiments, electron spin resonance spectra showed that chromium
oxalate in the aqueous phase quenched the signal of membrane-associated HFP
which was spin—labeled at M19 but did not quench HFP spin-labeled at A1.(10)
These data indicated a M19 location close to the aqueous interface of the
membrane and an A1 location away from this interface. Models for the
membrane location of helical HFPs have also been developed by simulations
and there have been distinct models supporting either partial insertion or
traversal of the membrane. In one simulation the peptide was generally near the
membrane surface with the F8 backbone and sidechain nuclei respectively 4 A
and 6 A deeper than the phosphorus longitude.(11) In a different simulation, HFP
traversed the membrane and the backbone and sidechain F8 nuclei were at the
bilayer center, i.e. ~19 A from the phosphorus longitude.( 12)

The present work has been focused on the study of membrane location of the
HFP in a host-celI-Iike environment. The goal of the research was to build up a
high resolution insertion model which would correlate with l-IFP membrane fusion
activities. To achieve the goal, different HF P constructs with very different fusion
activities were considered as model peptides based on the previous and present
studies using fluorescence spectroscopy.(13-15) The HFPmn induces fusion with
moderate rate whereas very little fusion is observed for HFPmn_mut. The high-

resolution structures of the soluble ectodomain of gp41 which lack the HFP are

111

HFPmn AVGlGALFLGFLGAAGSTMGARSWKKKKKKA
HFPmn_mut AEGIGALFLGFLGAAGSTMGARSWKKKKKKA

AVGIGALFLGFLGAAGSTMGARSWKKKKKIFA
HFPtr AVGlGALFLGFLGAAGSTMGARSWKKKKKCIZ
AVGIGALFLGFLGAAGSTMGARSWKKKKKCA

Figure 22 Peptide sequences of the basic constructs of HFPmn_mut, HFPmn
and HFPtr. The specific labeling sites were described in the main text.

112

trimeric and suggest that HFP may interact with the target cell membrane in a
trimer unit.(16) The putative functional significance of trimers is supported by the
15-40 fold higher vesicle fusion rates of the chemically cross-linked HFPtr
relative to HFPmn.( 14) Thus, the fusion rates are ordered HFPmn_mut < HFPmn
< HF Ptr and the present study examines the structures and membrane locations
of these constructs with correlation to their very different fusion activities. Solid-
state NMR is suitable and more advantageous compared with other biophysical
methods because with an appropriate isotope labeling scheme, it can provide
residue-specific membrane insertion depth information with minimum
perturbation to the membrane system. In this chapter, the 13C-31P and 13C-‘S’F
REDOR experiments which were used to probe the membrane location of
different HFP residues will be described, initially with triply 13CO-Iabeled HFPmn
to get a qualitative sense,(17) and finally with singly 13CO-labeled HFPmn_mut,
HFPmn, and HFPtr to achieve systematic, semi-quantitative insertion

models.(18, 19)

MATERIALS AND METHODS

Peptides. The sequences of HFPmn_mut, HFPmn and HFPtr were
summarized in Figure 22. The triply-labeled peptides followed the nomenclature
of HFPmn-BFLG, HFPmn-5GAL, etc, and the singly-labeled peptides followed the
nomenclature of HFPmn_mut-1A, HFPtr-1A, etc. In general, terms before the
dash express HFP sequences with specific C-terminal tags and terms after the

dash indicate the 13CO labeling positions. All peptides were synthesized using

113

the strategies described in chapter II for HFPmn and HFPtr and all syntheses
were completed either automatically on peptide synthesizer (Applied Biosystems
431A, Foster City, CA) or manually with a 15mL reaction vessel (Peptides
International, Louisville, KY).(20) The C-terrninal Lys(Boc), Trp(Boc) and Cys(T rt)
were introduced to increase the solubility, add an A230 chromophore and achieve
the synthesis of higher order HFP oligomer with cross-linking,
respectively.(14,15,21) All peptides were purified using HPLC with a water-
acetonitrile gradient, and identified using MALDI-TOF mass spectrometer.

Lipid mixing induced by HFPs. Mixing of lipids between membrane
vesicles was monitored by a fluorescence assay to show that different HFP
oligomers have very different lipid-mixing activities.(22) Together with the
previous results which indicated the non-fusogenicity of HFPmn_mut, this work
provide the rationale of choosing HFPmn_mut, HFPmn and HFPtr as model
peptides to study the membrane location-fusion activity correlation.(13) Two
types of large unilamellar vesicles (LUVs) were prepared. The “unlabeled LUVs”
contained POPC and POPG in a 4:1 mol ratio. This composition approximately
reflected the ratio of neutral to negatively charged lipids in membranes of host
cells of HIV and correlated with the lipid composition used in previous structural
studies of viral fusion peptides.(23,24) The “labeled LUVs” contained 77 mol%
POPC, 19 mol% POPG, 2 mol% of the fluorescent lipid N-(7-nitro-2,1,3-
benzoxadiazoI-4-yl)-phosphatiylethanolamine (N-NBD-PE), and 2 mol% of the
quenching lipid N-(lissamine Rhodamine B sulfonyl)-phosphatidylethanolamine

(N-Rh-PE). HFP-induced fusion was examined in a solution with an

114

unlabeledzlabeled vesicle ratio of 1:9 so that a labeled vesicle would likely fuse
with an unlabeled vesicle. The resultant lipid mixing would yield a larger average
distance between fluorescent and quenching lipids and increased fluorescence.

LUV preparation began with dissolution of lipid in chloroform followed by
removal of the chloroform with nitrogen gas and overnight vacuum pumping. The
lipid film was suspended in 5 mM HEPES buffer and the lipid dispersion was
homogenized with ten freeze-thaw cycles. LUVs were formed by extrusion
through a filter with 100 nm diameter pores (Avestin, Ottawa, ON).

Fluorescence was recorded on a stopped-flow fluorimeter (Applied
Photophysics SX.18MV-R, Surrey, UK) using excitation and emission
wavelengths of 465 and 530 nm, respectively. For a single run, one syringe in the
fluorimeter contained HFPmn, HFPdm, HFPtr, or HFPte dissolved at a
concentration of 3.00, 1.50, 1.00, or 0.75 (M in HEPES buffer. A second syringe
contained labeled and unlabeled LUVs at 300 pM total lipid concentration. At
time zero, equal volumes of the two solutions were mixed and fluorescence was
recorded every second for 200 s. The HF P concentrations were chosen so that
the HFP strand:lipid ratio was always 0.010.

Most reports of fluorescence based lipid mixing have focused on AFrumn,
the net change in fluorescence after the fusogen is added to the vesicles.
AFmWn is typically compared to Ademm, the change caused by addition of a
detergent which completely solubilizes the vesicles. Because of the very large
average distance between fluorescent and quenching lipids in the solubilized

vesicles, AFdetemm is the maximum observable fluorescence change. The

115

“percent lipid mixing” is typically defined as AFrusogen/Ademm x 100. In order to
provide some comparison between our stopped-flow fluorescence data and the

lipid mixing literature, the raw data at each time point, Fraw(t) were converted to

F(t) = {[Fraw(t)-Eniria%Fmax} X 100 (54)

Fm” was a typical value of fluorescence at t = 0 and 13qu was chosen to

normalized F(t):

provide semi-quantitative comparison between F(t) and earlier studies of percent
lipid mixing induced by HFPs.(21) A single value of Frnma, and a single value of
AFmax were used for all of the data.

At the end of the 200 s collection time, the fluorescence from HFPmn-
induced lipid mixing was still appreciably increasing and it was therefore difficult
to fit these data to a buildup function. The fluorescence of the HFPdm, HFPtr,
and HFPte constructs had leveled off and these data fitted much better to the
sum of two exponential buildup functions than to a single buildup function:

F(t)=F0 +F,(l-e"‘1‘)+F2(l-e"‘2’) (55)
where F0, k1, F1, k2, and F2 were fitting parameters. The best-fit value of F0 was
close to 0 because of the way F(t) was calculated in Eq. 54. A convention was
chosen that k1 > k2 so that k1 and F1 were respectively the rate constant and
overall fluorescence change of the faster lipid mixing process and k2 and F2 were
the rate constant and overall fluorescence change of the slow process.
Data were collected for each construct at 25, 30, 35 and 40 °C and each HFPdm,

HF Ptr, and HFPte data set was fitted with Eq. 55. For each construct,

116

OTs O F O F O

a f
—> —->
W 3 5 3 5

Figure 23 Synthetic scheme of 5-F-palmitic acid.

117

the temperature dependence of k1 was fitted to the Arrhenius Equation In
k1 = In A - Ea/RT where R was the ideal gas constant, T was the absolute
temperature, and A and E. were the pre-exponential factor and activation energy,
respectively. Three independent runs were done for each construct and
temperature.

Synthesis of the precursor of 5-19F-DPPC. In general, the experiments
described in this chapter measured the internuclear distances between the
selectively labeled 13CO in the HFP strands and the 31P or 19F in the lipid bilayer.
Besides the 31P nucleus in the phosphate group, two types of fluorinated lipids
were used in order to cover the entire depth of lipid bilayer. The 1-palmitoyl-2-
(16-fluoropalmitoyl)-sn-glycero-3-phosphocholine (16-‘9F-DPPC) was purchased
from the Avanti Polar Lipids and the precursor of 1-palmitoyI-2-(5-
fluoropalmitoyl)-sn—gchero-3-phosphocholine (5-‘9F-DPPC), named 5-F-palmitic
acid, was synthesized in our lab. The 19F nuclei were located 8.5 A and 20.7 A
away from the layer of 3‘P for the 5-‘9F-DPPC and 16-‘9F-DPPC, respectively.(25)
The synthetic scheme for 5-F-palmitic acid is described in Figure 23. The overall
yield of 5-‘9F-palmitic acid was ~40% and each step was monitored using thin
layer chromatography with iodine and phosphomolybdic acid as visualization
reagents. The intermediate products were purified using silica gel column
chromatography with a mixture of pentane and ethyl acetate as developing
solutions.(26-30) The 5-19F-DPPC was synthesized by Avanti Polar Lipids using

the 5-F-palmitic acid as a precursor.

118

Reaction conditions in Figure 23 included: (a) 68.2 g undecyl bromide
(Sigma-Aldrich, St. Louis, MO) in 350 mL dry diethyl ether was added to 6.94 9
Mg in 100 mL dry diethyl ether. Reflux at 34 °C for 2 hours. (b) The diethyl ether
was removed and 28.0 9 methyl 4-(chloro-formyl) butyrate (Sigma-Aldrich, St.
Louis, MO) in 100 mL dry benzene was added to the Grignard solution from step
a and 27.5 g CdCI2 in 350 mL dry benzene. Reflux at 78 °C for 1 hour. (c) NaBH4,
NaH2P04 and 5-keto-methyl palmitate each at 1 M concentration were dissolved
in dry methanol. The mixture was stirred at 0 °C for 15 minutes and at ambient
temperature for 1 hour. (d) 5-hydroxy-methyl palmitate and 0.5 M tosyl chloride
(Sigma-Aldrich, St. Louis, MO) each at 0.5 M concentration were dissolved in dry
CH2CI2 with 0.025 M 4-(dimethylamino)pyridine. The mixture was cooled and
held at 0 °C. dry pyridine was added dropwise over 40 minutes to reach a final
concentration 0.5 M, and then the mixture was stirred at 0 °C for 2 hours. (e) 0.05
M 5-O-tosyl-methyl palmitate and 0.1 M tetrabutylammonium fluoride (Sigma-
Aldrich, St. Louis, MO) in dry CH3CN were stirred at ambient temperature for 96
hours. (f) 5-F-methyl palmitate and KOH powder were each added into dry
methanol at 0 °C to reach a final concentration of 0.1 M of each reagent. The
mixture was stirred at 0 °C for 2 hours.

Solid-state NMR Sample Preparation and Experiments. The lipid bilayer
samples were prepared using the same procedure as in chapter IV, however,
with different lipid compositions. “PC:PG” in this chapter denotes a lipid bilayer

with DTPC:DTPG in 4:1 molar ratio for triply-“CO-labeled HFPs and

119

(a)

 

1.0

0.2—
o l
0.04 0 08

0

(AS/3016””

 

l

T
0.12 0.16 0.20
mol fraction of 5-19F-DPPC

1....

50 -50 50 -50

 

 

(b)

   

PC:PG
(w/o 19F-DPPC)

PC:PG:CHOL
(w/o ‘9F-DPPC)

  

 

 

 

PC:PG:CHOL PC:PG
(w/ 19F-DPPC) (w/ 19F-DPPC)

 

   

 

50 -50 50 -50
31P Chemical Shift (ppm)
Figure 24 (a) Plot of (AS/So)” vs mol fraction of 5-19F-DPPC at r = 16 ms. All
samples contained HFPmn-L9. (b) Static 3‘12 s ectra for PC:PG and
PC:PG:CHOL bilayer with and without 9 mol fraction 1 F-DPPC. Each spectrum
was acquired with 1024 scans and processed with 300 Hz Gaussian line

broadening. The spectra were acquired at 35 °C.

 

120

DTPC:DTPGz‘9F-DPPC in molar ratio 8:211 for singly-”CO-labeled samples.
“PC:PG:CHOL” denotes a lipid bilayer with DTPC:DTPGzcholesterol in 8:225
molar ratio for triply-‘3CO-labeled samples and DTPC:DTPGzcholesteroI:19F-y
DPPC in molar ratio 8:2:5:1 for singly-‘3CO-Iabeled samples. The molar ratio for
peptide strand to DTPC+DTPG was 1:50 for PC:PG samples and 1:25 for
PC:PG:CHOL samples, respectively. Samples containing 100% 19F-DPPC form
non-bilayer structures.(31) In order to maintain bilayer structure in the NMR
samples, 0.09 lipid mol fraction of 19F—DPPC was initially determined with
measurements on a series of samples which differed in their mol fraction of 5-
‘gF-DPPC, cf. Figure 24a. The choice of 0.09 mol fraction ‘9F-DPPC for
subsequent samples was based on: (1) maximum (AS/So)”; and (2) relatively
constant (AS/So)” over the 0.07-0.14 mol fraction range. Static 31P NMR spectra
were consistent with overall bilayer structure in samples containing 0.09 mol 19F-
DPPC (of. Figure 24b).

The 13C-31P REDOR experiments were conducted on a triple-resonance
MAS probe and the 13C-‘S’F REDOR experiments were conducted on a
quadruple-resonance MAS probe. The 1H, 13C and 31P channels were tuned at
400.8 MHz, 100.8 MHz and 162.2 MHz respectively for 13C-3‘P experiments and
the 1H, 13C and 19F channels were tuned to 398.7 MHz, 100.2 MHz and 375.1
MHz respectively for 13C-‘9F experiments. The following parameters were used in
both 1:‘C-3‘P and 13C-‘QF experiments: 50 kHz 1H 1t/2 pulse, 1 ms CP between 1H
and 13C channels, 50 kHz constant 1H field and 55-66 kHz ramped 130 field

during CP and 50 kHz 13C 1: pulse in the REDOR dephasing period. A 95 kHz ‘H

121

decoupling field during dephasing and acquisition periods was used for 13C-3’1P
experiments and a 75 kHz 1H decoupling field was used for 13C-‘QF experiments.
The 1: pulses alternatively applied either on the 31P (50 kHz) or 19F (33 kHz)
channel were calibrated by maximizing (AS/SOY” in standard samples as
described in the experimental setup in chapter Ill.

Natural Abundance Correction for (AS/So)”. The natural abundance
correction was required for quantitatively analyzing the REDOR experimental
data. The triply and singly-labeled samples followed different natural abundance
correction procedures.

For the triply-labeled samples which were associated with ether-linked
phospholipids, the natural abundance correction started from considering that
both So and 81 came from the labeled 13CO and the natural abundance of

unlabeled residues.

 

 

[Ejerp= Saab+ Sga_ Sllab_ Slrra = 1 _ Sllab — Slna (56)
SO Séab+ Sgt! S60b+ Sga 60b+ S60

lab

A few algebraic manipulations led to the relation between (AS/So) , which was

the contribution to REDOR dephasing from the labeled 13co, and (AS/So)”.

lab exp na
.49: _ 56'“ E _ 56'" 95
111 101111901 I 114-HS.) l

The term S5“/S,§“” was related to the number of unlabeled residues vs. labeled

 

 

residues in the sequence. The term (AS/So)“ considered the contribution to
REDOR dephasing from all the unlabeled residues and was approximately

calculated as the average of all available (AS/So)”. Although the latter

122

calculation is an approximation, uncertainties in (AS/So)" have a relatively small
impact on the value of (AS/SOY”. For example, consider the spectra for the
HFPZ-“AAG in PC:PG samples at 2' = 24 ms. The values of (Sam/80"”),
(AS/So)”, and (AS/So)" are 0.084, 0.419 t 0.014, and 0.134, respectively, and
result in (AS/so)“ = 0.443 :I: 0.015. If (AS/So)” were 0.0 or 0.25, (AS/so)“ would
be 0.454 or 0.433, and are within the experimental uncertainty of the reported
(AS/soy“.

For the singly-labeled samples, 0.09 mol fraction 19F-DPPC with natural
abundance 13006 were incorporated into the lipid bilayer. The comparable

formulas to Eq. 56 and Eq. 57 were

(58)

 

( AS )“P __ 53"” + 5510mm) + 55‘“ (DPPC) — SF” — 51"“ (HFP) — 51"“ (DPPC)

S 56"” +561“ (HFP)+S6’“ (DPPC)

and

 

lab na na exp
(as) =(1+S0 (HFP)+S0 (DPPC)).[:A§J

S0 Séab So (59)

_w.(é§.] (HFP)—m-(i‘é) (DPPC)

Stllab SO Séab SO

The terms (AS/SOY”, (AS/So)" (HFP) and (AS/So)” (DPPC) represented the
contribution to REDOR dephasing from the labeled 13CO, the natural abundance
of unlabeled residues and the natural abundance 13CO from 16F-DPPC,
respectively. The So terms in Eq. 59 had the numerical values

53"” =1

sg“(HFP) = 29x 0.011 = 0.319 (60)
S3“(DPPC) = 2.5x 2x 0.011 = 0.055

123

The values of (AS/Soy“a (HFP) were calculated from the experimentally available
(AS/So)” for a specific HF P construct and the values (AS/So)"a (DPPC) were

Table 5a. The (AS/So) "° (HFP)

 

 

 

”CO-3113 a 13CO-(16-19F) a b 13CO-(5-19F) a b
Peptide HF Pmn_mut HF Pmn HFPtr HF Ptr HFPmn
0.007 0.023 0.010 0.018 0.002
0.118 0.133 0.099 0.004 0.016
r(ms) 16 0.319 0.260 0.188 0.028 0.032
24 0.465 0.353 0.294 0.135 0.158
32 0.444 0.380 0.308 -- --

 

aThe 13CO-31P values were based on the (S1/So)°Xp of samples labeled at Ala1, Ile4, Ala6,
Leu9, Leu12, or Ala14, and for HFPmn_mut and HFPmn, Ala21. The ”co-”How; values
were based on samples labeled at Ala1, Ile4, Ala6, Leu9, Leu12, or Ala14 and the 1 CO-
19F(C5) values were based on samples labeled at Ala1, AlaG, or Leu9.

The maximum rfor 13CO-19F experiments was 24 ms.

 

Table 5b. The (AS/So) "° (DPPC)

 

 

1300-31 P 13CO-(16-19F) a 1300-(5-19F) a
2 0.025 0.000 0.011
8 0.318 0.000 0.118
r(ms) 16 0.695 0.000 0.567
24 0.811 0.000 0.896
32 0.907 .— ...-

 

aThe maximum rfor1300-19F experiments was 24 ms.

 

calculated using a 13C-3‘P or 13C-‘QF two-spin system with internuclear distance

of the setup model compound. Table 5 provides the numerical values.

RESULTS AND DISCUSSION

124

 

100
d

75 -1 //t‘

I

50 -1 -'-'

A
I
v

\

Fluorescence (a.u.)

25 q j. //—fl—M~/‘
0”.

I
e
1e

‘* 7"
———-'

 

_‘_A.
A#
v—v—

 

 

 

 

 

 

 

M
0 1
0 100 200
Time (s)
-1
(b)
A A
. O ' A
-2-1 0 A
‘ 0
SE
5 D
-3-4 0
u
U.
'4 T T 1
3.20 3.28 3.36
(1/T)‘1000 (K")

Figure 25 Panel a displays stopped-flow monitored changes in lipid fluorescence
induced by addition of different HFP constructs to an aqueous solution containing
membrane vesicles. Increased fluorescence is a result of mixing of lipids
between different vesicles and this mixing is one consequence of vesicle fusion.
The lines are color coded: HFPmn (black); HFPdm (red); HFPtr (blue); and
HFPte (green). The total lipid concentration was 150 pM and the HFPmn,
HFPdm, HFPtr, and HFPte concentrations were 1.50, 0.75, 0.50, and 0.37 11M,
respectively, so that peptide strandzlipid = 0.01. The data were collected at 25 °C,
the vesicle composition was 4:1 POPC:POPG, and the initial vesicle diameter
was ~100 nm. Additional data (not shown) were obtained for HFPdm, HFPtr, and
HFPte at 30, 35, and 40 °C. Each data set for each construct was analyzed as
the sum of two exponential buildup functions. Panel b displays Arrhenius plots for
the rate constants of the fast buildup function with legend: HFPdm (open square);
HFPtr (open circle); and HFPte (open triangle). The best-fit lines are also
displayed and result in the respective activation energies 41 :t 3, 26 i 1, and 20 :I:
1 kJ/mol.

125

Fusion Activities of Different HFP Constructs. The lipid-mixing assay
provided the rationale for choosing different HFP constructs to study the
membrane location-fusion activity correlation. Previous study in our group
suggested the fusion activities for HFPmn, HFPdm and HFPtr followed the trend
HFPmn < HFPdm < HFPtr. The present study introduces a higher order HFPte
and suggests that HFPtr may have the highest catalytic efficiency among all the
four HFP constructs. Figure 256 shows stopped-flow fluorescence data which
track lipid mixing induced by HFPs and suggests that the long-time lipid mixing
rates are ordered HFPmn < HFPdm < HFPtr - HFPte. Both HFPtr and HFPte
induced similar rapid lipid mixing while HFPdm induced slower mixing and
HFPmn induced little mixing. For each construct, data were acquired at 25, 30,
35,‘and 40 °C and the HFPdm, HFPtr, and HFPte data could be fit well to a
biexponential buildup function, cf. Eq. 55. The best-fit parameters of the 35 °C
data are listed in Table 6 as an example of data fitting. For each of the three
constructs, k2 is 0.1 k1 and F1 e: F2. In addition, kit’lkrd’" at 2.5 and kr'elkr" z 1.3.
Figure 25b displays Arrhenius plots for the k1 rate constants and the best-fit Eas
and In As are listed in Table 6. The values of E, and InA for HFPdm and HFPtr
are comparable to those reported in a previous study.( 14) The data show that
Ead’" > E.” > E,“ and In Adm > In At, > In A)... An increased number of strands in
the oligomer is therefore correlated with decreased InA and E2, with concomitant
opposite effects on k1. The activation entropies were calculated using the
transition state theory equation AS‘ = R x [In(Ah/kgT) - 2] where R is the ideal

gas constant, h is Planck’s constant, k3 is Boltzmann’s constant, and T is the

126

absolute temperature.(14) The resultant AS‘ were all negative with A§dm > ASE,
> ASE. but we do not understand the sign or trends of the AS’ values. These
data indicate: (1) cross-linking increases the rate and extent of HFP-induced lipid
mixing and decreases activation eneI‘QY; and (2) the increase in Iipid-mixing-per-
strand and decrease in activation energy with cross-linking levels off at HFPtr. It
might be expected that oligomer folding would be more difficult with an increasing
number of monomer units so the putative trimeric oligomerization state of gp41
and other class I viral fusion proteins may be the optimal balance between higher

catalytic efficiency and more difficult folding.

Table 6. Fitting parameters for the lipid mixing kinetics at 35°C 3'”

 

 

to re F0 F1 F: s,‘ As”
9011511” (1635-1) (1033") (a.u.) (a.u.) (a.u.) (kJ/mol) I" A (J/moI-K)

54.9 4.9 0.5 19.6 19.8 40.6 13.2

HFPdm (0.3) (0.2) (0.1) (0.3) (0.1) (3.4) (1.2) '152
139 17.3 0.9 50.1 50.0 25.8 8.0

HFP" (3) (0.3) (0.2) (0.2) (0.2) (1.2) (0.8) '195
185 20.7 0.9 49.8 50.1 20.1 6.2

”me (7) (0.3) (0.3) (0.3) (0.2) (0.7) (0.5) '210

 

' Fitting uncertainties are given in parentheses. The variation of a parameter value from fitting
data of different runs was less than the fitting uncertainty of a single run.

b The k1, k2, F0, F1 and F2 were obtained using Eq. 2 in the main text.

c Ea and In A were calculated using In k1 = InA — Ea/RT and k1 values from temperatures
between 25 and 40 °C.

d As’ was calculated using A31 = R x [In (Ah/kBT ) — 2).

Membrane Insertion of triply-labeled HFPs detected by 13C-3’P REDOR
methods. The long-term goal of the membrane insertion study is a detailed

structure of the membrane location of the HFP in helical and [3 strand

127

conformations. There was relatively little information about the membrane
location of HFP. It was also likely that a large number of 13CO sites had 13CO-31P
distances beyond the REDOR detection limit. In addition, HFP 13C Iinewidths are
fairly broad which leads to overlap of 13CO resonances from different residues
and the need for specific 13CO labeling. In an effort to reduce the numbers of
specifically labeled peptides needed to develop a membrane location model,
samples were first made with four peptides each of which had 13'CO labels at
three sequential residues between G5 and G16. The G5-G16 region was
therefore rapidly scanned for 13CO-3‘P proximity. Although the (AS/So)” data for
each of the samples had contributions from three distinct 1300 sites, the
individual (AS/So) would only be appreciably greater than zero for 13CO-3‘P
distances 5 8 A. The regions of HFP proximate to 31P were defined from the
REDOR data on the triply labeled samples and these regions provided a basis
for choosing sites for single 13CO labeled peptides which will be described later in
this chapter. .

The secondary structure information that can be obtained with the triply-
labeled samples was limited by the overlapping of multiple 13CO labelings,
nonetheless, the local peptide conformation was examined by analysis of the
13CO chemical shift distributions in So spectra of HF Ps obtained with r= 2 ms, cf.
Figure 26. The data supported the following models: (1) the major fraction of
peptides in PC:PG and PC:PG:CHOL adopted a B strand conformation from GS

to G16; and (2) there is a minor fraction of peptides in PC:PG with helical

128

conformation. The detailed experimental support for the models is based on the

 

 

(h)

 

 

N---

 

 

1 1 1 1
170 190 170
13C Chemical Shift (ppm)

1
190

Figure 26 So spectra for membrane-associated HFP with peptide:lipid ~0.04. The
dotted lines are at 175 ppm. All spectra were obtained with r = 2 ms and were
processed with 200 Hz Gaussian line broadening and baseline correction. The
membrane composition for samples a-d was PC:PG and the membrane
composition for samples e-h was PC:PG:CHOL. The peptides were: a, e,
HFPmn-5GAL; b, f, HFPmn-SFLG; c, g, HFPmn-"FLG; and d, h HFPmn-“AAG.
The numbers of scans summed to obtain spectra a-h were 4823, 3867, 4823,
8500, 3259, 1001, 4320 and 6992, respectively.

129

known correlation between larger 13CO chemical shifts and local helical
conformation and smaller 1300 chemical shifts and local 8 strand conformation.
(33) For example, average database values in ppm units of 13CO chemical shifts
of helix (strand) conformations are: Gly, 175.5 (172.6); Ala, 179.4 (176.1); Leu,
178.5 (175.7); and Phe, 177.1 (174.2). For the 5GAL, 8FLG, "FLG, and “AAG
samples, the peak chemical shifts were ~175, 174, 175, and 176 ppm,
respectively, and correlated with B strand conformation for the Ala, Leu, and Phe
residues. For the 8F LG and 14AAG samples associated with PC:PG, there were
shoulders at ~178 and 179 ppm, respectively, which correlated with helical
conformation of Ala, Leu, and Phe residues. These results were consistent with
previous studies of the conformation of membrane-associated HF P with
peptide:lipid ~ 0.04 and with previous observations of greater preference for B
strand conformation in cholesterol-containing membranes.(34-41)

The values of (AS/So)” were extracted from integrating the 13CO signals
of the corresponding 80 and 81 spectra. The experimental values were used to
obtain (AS/So)” and the corrected dephasing was fit to (AS/So)” from Eq. 34 to
achieve the minimum root-mean-squared deviation (RMSD).

AS lab AS sim
{1.1—1 ‘113‘111. 12
H (01.1). (61)
The fitting parameter (1 represented the dipolar coupling frequency which could
be converted into the internuclear distance r and f was the fraction of 1300 that

was close enough to the lipid 31P to provide non-zero dephasing. This fraction

130

lab

parameter was considered because at large rvalues, (AS/So) reached plateau

values of ~1. The corrected uncertainty 0"" = (1 + Sam/80"") x a' ”P while the
experimental uncertainty was calculated using Eq. 35.(42)

Figure 27 displays the z' = 16 and 24 ms REDOR spectra of triply-labeled
membrane-associated HFP samples and Figure 28a,b displays comparative
plots of (AS/So)” for the different samples. The data demonstrated that samples
containing HFPmn-“AAG have qualitatively larger (AS/So)” than do samples
containing HFP labeled at other residues. Using the conformational results from
Figure 26, it appears: (1) a significant fraction of [3 strand HFP are in close
contact with membranes; and (2) the 14AAG (A14 to G16) region is closer to the
lipid 3‘12 than is the 5GALFLGFLG (G5 to G13) region. Figure 28c,d displays plots
of (AS/so)“ and best-fit (AS/so)” for HFPmn-“AAG in PC:PG and
PC:PG:CHOL. The best-fit rwas ~5.2 A in both membrane compositions and the
best-fit f in PC:PG and PC:PG:CHOL were 0.45 and 0.32, respectively. It was not
possible to fit the HFPmn-“AAG data well without inclusion of the f parameter.
Although the (AS/So)” had contributions from three 13CO sites which would each
have a distinct r, the number of data points and signal-to-noise dictated fitting to
a single r value. The best-fit r should therefore be considered as both
approximate and as likely representing the population of 13CO sites with greatest
d and smallest r. Fitting was not done for data from the other samples because of

the small (AS/So)” and because the (AS/So)” do not always reach asymptotic

values at large 1'.

131

As a control experiment, spectra were also obtained for samples made
with HFPmn labeled at 8FLG that did not contain C-terrninal lysines (named
HFPmnl-BFLG). For r= 2, 8, 16, and 24 ms, (AS/So)°"” = -0.02, 0.06, 0.11, and
0.08 for the HFPmn1-8FLGIPC:PG sample and 0.01, 0.03, 0.01, and —0.01 for
the HFPmn1-8FLG/PC:PG:CHOL sample. These values correlated with the
(AS/so)” of the respective HFPmn-BFLG/PCzPG and HFPmn-
8FLG/PC:PG:CHOL samples (circles in Figure 28) and suggested that the
additional C-terminal lysines do not greatly affect the REDOR results.

The position of the HFP in the membrane has been postulated to be a
significant structural factor in its fusion activity and to our knowledge; this study is
the first example of direct distance measurements between the HF P and the lipid
headgroups. Values of r ~ 5-6 A were detected between the 1300s of residues
from A14 to G16 and the lipid 31Ps. These r values support intimate association
of the HF P and membranes containing either only phospholipids or phospholipids
and cholesterol. The average r for 5GALFLGFLG 13COs was likely greater than 8
A (d s 25 Hz) as evidenced by the significantly smaller (AS/So)”, of. Figure 28.
Thus, relative to the 5GALFLGFLG residues, the 14AAG residues are much closer
to the lipid 3‘P. The G5 to G16 13co chemical shift distributions of this study were
consistent with a major population of HFP with B strand conformation for these
residues. This result correlated with previous studies which supported the
following structural features: (1) B strand HFP was fully extended between A1
and G16;(35,43) (2) B strand HFP formed hydrogen bonded oligomers or

aggregates;(39,44) and (3) a

132

171.11.31.11.
,AAAA

XML EILJL
1113,
51./MM.
ALIA

.WAAA
than. JUL

1111 TVTT 1111 ITITI

190 170 190 170 190 170 190 170
'30 Chemical Shift (ppm)

Figure 27 13CO-3‘P REDOR spectra of membrane-associated HFP with
peptide:lipid ~0.04. Each letter corresponds to a single sample which contained
(a-d) PC:PG or (e-h) PC:PG:CHOL and (a, e) HFPmn-5GAL, (b, t) HFPmn-BFLG,
(c, g) HFPmn-"FLG, or (d, h) HFPmn-“AAG. For each letter/sample, so (left), 31
(right), r= 16 ms (top), and 1' = 24 ms (bottom) spectra are displayed. The dotted
lines are drawn for visual comparison of So and 81 peak intensities. Each
spectrum was processed with 300 Hz Gaussian line broadening and baseline
correction. The numbers of So or 81 scans summed to obtain the top and bottom
spectra were respectively: a, 30000, 56000; D, 27509, 29463; c, 20000, 40000; CI,
44129, 48296; e, 8448, 52384; f, 5488, 21664; 9, 28032, 52384; and h, 22576,
50240.

133

 

Ia)

 

 

 

 

I l
0 10 20
Dephasing time (ms)

 

 

 

 

 

I
O 10 20
Dephasing time (ms)

Figure 28 (AS/So) vs dephasing time for membrane-associated HFP in (a,c)
PC:PG or (b,d) PC:PG:CHOL. For panels a and b, the points correspond to
(AS/So)” and the symbol legend is: squares, HFP2-5GAL; circles, HFP3-8FLG;
triangles, HFP2-"FLG; and diamonds, HFP2-“AAG. The vertical dimensions of
each symbol approximately correspond to the :l:1 0' uncertainty limits. Lines are
drawn between (AS/SOY” values with adjacent values of 2'. Each (AS/So)” value
was determined by integration of 10 ppm regions of the So and Sr spectra.
Panels c and d respectively correspond to the HFPmn-“AAGIPC:PG and the
HFPmn-“AAG/PC:PG:CHOL samples and the points correspond to (AS/SOY”
(vertical lines with error bars) and best-fit (AS/So)” (diamonds). Lines are drawn
between points with adjacent rvalues. For plot c, the best-fit d = 91 :I: 8 Hz with
corresponding r = 5.12 :I: 0.16 A, f = 0.45 i 0.02, and 23m = 5.0. For plot d, the
best-fit d = 85 :t 6 Hz with corresponding r = 5.24 :I: 0.13 A, f = 0.32 i 0.02, and

12min = 3..8

134

significant fraction of the oligomers have an antiparallel arrangement with
adjacent strand crossing between F8 and L9.(9,36,45,46) Some of these studies
also supported conformational disorder at A21.(34,35) Although there are some
data supporting a population of parallel strand arrangement,(47) “partial
membrane insertion (PMI)” and “full membrane insertion (FMI)” models are only
presented for the antiparallel arrangement, cf. Figure 29. There have been high-
resolution structures for the ~130-residue “soluble ectodomain” region of gp41
which begins about ten residues C-terminal of the HFP and ends about twenty
residues N-terminal of the gp41 transmembrane domain.(2,16,48-50) These
structures showed trimeric gp41 with the residues closest to the HFPs in a
parallel in-register coiled-coil. Antiparallel HFP strand arrangement in the context
of gp41 would then require at least two gp41 trimers. As demonstrated in chapter
V, Strands from trimer C (Cr, C2, 03) would be parallel to one another and
strands from trimer D (D1, D2, 03) would be parallel to one another and an
antiparallel interleaved strand arrangement could be formed as Cngc2D2chi.
There is solid-state NMR evidence for the antiparallel arrangement of membrane-
associated HFPs which were cross-linked at their C-termini. (36)

For antiparallel strands between A1 and G16, the 14AAG residues in both
the PMI and FMI models are at the ends of the hydrogen-bonded oligomer and
are closer to the lipid headgroups than residues 5GALFLGFLG. The F8 and L9
residues are at the center of the hydrogen-bonded oligomer and are most deeply
membrane-inserted in all models. This result is consistent with the smallest

(AS/So)” values observed for the 8F LG samples and with the large number of

135

 

    

in e '- ' . ‘ as
Figure 29 (a) Partial membrane insertion (PMI) and (b, c) full membrane insertion
(FMI) models for antiparallel B strand HFP. The red arrows represent the A1 to
G16 residues in strand conformation and the black lines represent the S17 to
823 residues in random coil conformations. For clarity, black lines are not
displayed in c. Lipids are represented in blue and grey and cholesterol is not
displayed. Three antiparallel strands are displayed in a, b and twelve strands are
displayed in c but the actual number of strands in the oligomer/aggregate is not
known. The curvature and angle of the strands with respect to the bilayer normal
are not known but the models consider that A1-G16 has ~55 A length and that
the transbilayer distance is ~48 A. The experiments do not provide information
about the membrane locations of residues S17 to $23. Relative to FMI model (b),
the FMI [3 barrel variant (c) could have reduced energy because all of the
residues in the membrane interior have backbone hydrogen bonds.

136

apolar sidechains in the central 7LFLGFL (L7 to L12) region. Relative to the “FLG
samples, the models also predict smaller r and larger (AS/So)” for the 5GAL and
11FLG samples which generally correlates with the experimental data, of. Figure
28a-b. The models suggest small r and significant (AS/So)°"” for HFPs labeled at
the N-terminal residues and future studies could examine samples labeled in this
manner.

The PMI model in Figure 29a would likely perturb the membrane and has
some similarity with: (1) the PMI of extended conformation internal fusion
peptides postulated from structures of dengue, Semliki forest, herpes, and
vesicular stomatitis viral fusion proteins; (2) the PMI of helical influenza fusion
peptide determined from electron spin resonance experiments; and (3) a PMI
model based on the HFP-F8W fluorescence measurements.(9,51-56) However,
the locations of lipids in the perturbed leaflet in the PMI model are not clear. For
the FMI model of Figure 29b, the positions of the lipids are clearer but there are
non-hydrogen bonded CO and NH groups at the sheet edges with large Born
energies. These energies would be reduced for a FMI (3 barrel structure, Figure
29c. There is correlation between the FMI model and the deep insertion of the
Trp sidechain suggested from fluorescence studies of the HFP-F8W mutant.(8, 9)
In the context of gp41, individual HF P trimers would be on the same side of the
membrane in the PMI model but would be on different sides of the membrane in
the FMI model. It is not clear how this FMI trimer topology would relate to the
positions of the viral membrane-anchored gp41 trimers and the host cell

membranes. The free energy difference between the A1 to G16 FMI state and a

137

non-inserted state is ~3.9 kJ/mol as calculated from the sum of individual residue
free energy values derived from transmembrane helices.(57) The calculated
difference for the I4 to G13 sequence is —2.3 kJ/mol and leads to the general
conclusion that the free energy calculations do not strongly distinguish between
the PMI and FMI models. Future studies could discriminate between the PMI and
FMI models using REDOR distance measurements between peptide nuclei and
lipid acyl chain nuclei.(58)

There are similarities between these PMI and FMI models of oligomeric (3
strand HFP and PMI and FMI models which have been developed for a single
HFP in a helical confonnation.(11, 12,59) Much of the experimental data for
helical HFP insertion has been based on detergent rather than membrane
samples and there has been support for both surface location and micelle
traversal by HFP.(3-5,60,61) Our results on oligomeric B strand HFP were
consistent with the previous observations that the A15 and G16 residues of
monomeric helical HFP were close to the water-micelle interface and that the F8
to G10 were furthest from this interface. Thus, there may be common features
shared by helical and fl strand HFP micelle and membrane location.

In summary, 13CO-31P distance measurements suggested that the 1‘AAG
residues have close proximity to the lipid 31P while the 5GALFLGFLG residues
are relatively farther away from the phosphate groups. The ZD 13C chemical
shifts measurements described in chapter IV supported the conclusion that there
may not be a particular secondary structure required for HF P inducing the fusion.

Given the residue-dependence of the proximity to membrane headgroups, it is

138

possible that the membrane location of HFP affects the fusion activities of
different HFP constructs. In the following sections, the membrane location of
HFPs with tit-helical and B-strand secondary structures will be discussed with
singly-labeled HFP samples. The single 13CO labeling was used to avoid the
signal overlapping and to achieve an analysis of the individual conformations.
The selection of secondary structures was also helped by using PC:PG:CHOL or
PC:PG lipid bilayers. Most of the hydrophobic residues in HFPs adopted B-strand
conformation in PC:PG:CHOL while some N-terrninal residues adopted a mixture
of tat-helical and B-strand conformations in PC:PG.

Membrane Insertion of ,B-strand HFPs. In this part of the membrane
insertion studies, a systematic investigation on the membrane location of (3-
strand HFPmn_mut, HFPmn and HFPtr using 13c5113 and ‘3C-19F REDOR
experiments will be reported. For all the samples described in this section,
PC:PG:CHOL membranes were used to make most of the N-terminal residues in
B conformation. The reasons to study the B-strand HFPs are: (1) The previous
triply-labeled HFP studies demonstrated that B-strand HFP is in intimate contact
with membranes and merits serious consideration as a fusogenic
confonnation;(17) (2) The membranes of host cells of HIV have a cholesterolzlipid
ratio of ~0.45, and the membranes of HIV have a cholesterolzlipid ratio of
~0.8.(23) The chemical shifts derived from previous triply-labeled experiments
suggested a predominant population of HFP with B—strand conformation which
further indicated that B-strand could potentially be a biological relevant structure.

In addition, the current membrane location study was also trying to buildup a

139

correlation between the membrane position of HFP and the fusion activities of
different HFP constructs. Thus HFPmn_mut, HFPmn and HFPtr were chosen as
model constructs because of the very different fusion activities of these
constructs. In order to achieve a systematic membrane location study which
could cover the entire range of lipid bilayer, two types of 19F-DPPCs, 16-19F-
DPPC and 5-‘9F-DPPC, were used and the REDOR experiments were denoted
as ”co-“P, 13co-191=(C16) and 13co-‘9F(cs).

Initially, the 1300-311:1 and 13co-19F(c16) REDOR reveal different
membrane locations for different HFP constructs. HFP samples were prepared
with single 13CO labels at Ala1, lle4, Ala6, Leu9, Leu12, Ala14, or Ala21. The
labeled constructs are referred to as HFPmn_mut-1A, HFPmn-1A, HFPtr-1A, etc
in order to be consistent with the nomenclature used in this chapter. Residues 1-
16 are in the apolar region of the HFP sequence while AI321 is in the more polar
c-terminal region. Figure 30 displays ‘36-3‘P and 13C-‘9F(C16) REDOR spectra
for samples containing 9 mol% fraction 16-‘9F-DPPC lipid. The rvalue was 32 or
24 ms for the 13C-3‘P or ‘30-‘9F spectra, respectively. Table 7 summarizes the
peak chemical 13CO chemical shift for each sample as well as characteristic 1"’CO
shifts for helical or (3 strand conformation.(33) All of the peak shifts agree better
with (3 strand than with helical conformation. There may be a ~30% population of
helical HF Pmn_mut as evidenced by a shoulder at ~180 ppm for samples labeled
at Ala6, Leu9, or Leu12. The typical linewidth for samples labeled between Ala1
and Ala14 was 3-5 ppm while the linewidth for samples labeled at Ala21 was ~ 8

ppm (not shown in Figure 30). The chemical shift and linewidth data support

140

predominant B strand conformation for the N-tenninal apolar regions of all three
constructs and more disordered structure in the C-terminal polar regions. The
chemical shifts were consistent with previous triply-labeled HFP experiments as
well as the 20 uniforrnIy-Iabeled PDSD results in Chapter IV.

Table 7. 1300 chemical shifts in PC:PG:CHOL membrane a

 

 

Ala1 Ile4 Ala6 Leu9 Leu12 Ala14

HFPmn_mut 176.5 175.2 175.3 172.7 174.3 176.3
HFPmn 174.3 174.5 175.3 173.0 173.7 176.5
HFPtr 174.7 174.6 175.2 173.5 174.2 176.5
o-helix b 179.4 177.7 179.4 178.5 178.5 179.4
p-strand " 176.1 174.9 176.1 175.6 175.6 176.1

 

1' All chemical shifts are given in unit of ppm.
b The typical error for a standard conformation is 1 1.53m.

For each So and 81 spectrum, a 1 ppm interval around the peak was
integrated to calculate the (AS/So)”. For each pair of So and Sr spectra shown in
Figure 30, and (AS/So)” is graphically displayed in Figure 31. Subsequent
sections of this section describe quantitative distance and population analysis of
the (AS/So)°"" vs 1 data of all of the constructs while in this paragraph, a
qualitative analysis is provided based on Figure 31. For example, all constructs
labeled at Ala1 have 13CO-3‘P (AS/80f” as 0.8 which is interpreted to mean that a
major fraction of Ala1 13COs are 5 — 6 A from a 31F. In some contrast, the
HFPmn-9L and HFPtr-9L samples displayed 13CO-3‘P (AS/So)°"" s 0 which is
interpreted to mean that most of these Leu9 13COs are >8 A from a 31P.
Membrane location analysis based on this general approach assumes that all of
the constructs have the known HFPmn structure in which the Ala1 to Gly16

region is fully extended and HFPs assemble into an antiparallel 8 sheet structure

141

with adjacent strand crossing near Phe8 and Leu9, and this anti-parallel structure

was supported by 13C-‘E’N REDOR experiments described in Chapter V.(43)

 

 

 

 

 

Ala1 lle4 Ala6
13CO-“P REDOR
111211
$411 I

 

 

 

13(30491: REDOR

HFPmn_mut

tkAAAA

HFPmn

 

 

 

 

 

eeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee

 

 

 

 

 

190170 190170 190170 190170 190170 190170
13C Chemical Shift(ppm)

142

Leu9 Leu12 Ala14
13CO-3‘P REDOR

HFPmn_mut

 

 

 

 

1360-191: REDOR

 

   

 

 

 

A 1:} “A” .................

 

 

 

13.1. 1.1.1.1.

. . ——7— -q—U—-1—'—
190170 190170 190170 190170 190170190170

13C Chemical Shift (ppm)

 

Figure 30 REDOR 13C So and Sr NMR spectra for different labeled residues and
different HFP constructs. The dephasing time for each 13CO-31P spectrum was
32 ms and for each 13CO-‘S’F spectrum was 24 ms. The membranes contained 9
mol% 16-F-DPPC lipid. Each spectrum was processed with 200 Hz Gaussian
line broadening and polynomial baseline correction and was the sum of ~30000
scans for 13CO-3’1P experiments and ~20000 scans for 13CO-‘QF experiments.

143

(a) HFPmn_ mut (b) HFPmn (c) HFPtr

 

 

 

 

 

 

 

 

- ~ - 1 - M 1 —-
I 1366 31F .
l i
1 I I I I
e r
o 1
.2.- I'll I ,_ .-.1 _- o._ _...l
<01” ‘ 1 1 1* T -- 111mm“— _ -__.--__..-_.I
311300-191:
0. — — - - - — 0 . —-____-_-.-._.- 0 b “4.12....”1
A1 14 A5 L9 L12A14 A1 14 A6 L9 L12A14 A1 14 A6 L9 L12A14

Residue

Figure 31 Summary of experimental REDOR dephasing 0813/30)” for the spectra
displayed in Fig. 31. The top box in each panel is the 3300-1P data and the
bottom box is the 1300-19F(C16) data. The (AS/So)” values are shown as bars

for different labeled samples and a typical uncertainty is :l:0.04.

144

The results of the analysis analysis include: (1) All HFPmn_mut samples
have 1300-3‘P (AS/So)” > 0.3 and 13co-‘91=(c16) (AS/Sg)”‘" e 0 which indicates
that HFPmn_mut lies on the membrane surface near the phosphate headgroups
and is far from the bilayer center. (2) The HFPmn and HFPtr samples with Ala1
or Ala14 labeling show ”CO-“P (AS/So)” > 0.5 and ‘3CO-‘9F(C16) (AS/sole” e
0. These residues are near the ends of the [3 sheet structure which must
therefore be close to the phosphate headgroups. (3) The HFPtr-A6 and HFPtr-L9
samples have “CO-31P (AS/So)°"” e o and 1300-‘9F(C16) (AS/So)” significantly
greater than 0. This suggests that the interior residues of the HFPtr B sheet are
close to the bilayer center; i.e. a significant fraction of HF Ptr is deeply inserted in
the membrane. (4) HFPmn samples labeled at Ile4, Ala-6, or Leu-9 have 1300-
% (AS/so)” and ‘3CO—‘9F(C16) (AS/so)” s 0 which suggests that these
interior [3 sheet residues are neither close to the headgroups or the bilayer center
and are instead located midway between these two regions.

Figure 32 displays plots of (AS/So)” vs 1 for samples labeled at Ala1,
Ala6 or Ala14. These residues were selected to represent the N-terminal, middle
and C-terminal parts of the apolar region of the HFP sequence. These data
support the qualitative discussion of membrane location of the previous
paragraph. The samples labeled at Ala1, of. Figure 32a, (1, have 13CO-31P
(AS/So)” that increase rapidly with twhile the 13CO-"’F(C16) (AS/So)” z 0 for
all 2'. After removal of natural abundance 13C contributions from the 13CO-31P
(AS/So)” based on the procedure described in the “materials and methods”

section, the remaining (AS/SOY” reach ~1 at large 1. The (AS/SOY“ represent

145

only the Ala1 13CO signals and the value of ~1 indicates that the N-termini of all
constructs are close to the phosphate headgroups. These N-termini are likely
positively charged and are therefore attracted to the negatively charged
phosphates. Samples labeled at Ala14, cf. Figure 32c, f, show similar large 13CO-
3‘P (AS/So)” and ”CO-19F(C16) (AS/So)” e 0 indicating proximity to the
phosphate headgroups. However, at large 1, (AS/SOY“ s: 0.65 which might be
explained by two populations of Ala14 1300s in a HFP [3 sheet structure: (1)
13005 in strands near the edge of the sheet with close contact to lipid 31Ps; and
(2) 13005 in interior strands which are far from any lipid molecule. REDOR
(AS/So)” for samples labeled at Ala6, of, Figure 32b, e, depends on HFP
construct. HF Pmn_mut was the only construct with 13C-3‘P (AS/So)” significantly
greater than zero and HFPtr was the only one with 13C-"’F(C16) REDOR
(AS/So)” significantly greater than zero. These results revealed that the middle
regions of different HFP constructs had different membrane locations. HF Ptr had
the deepest insertion and induced the most rapid vesicle fusion while
HFPmn_mut was located on the membrane surface and was the least fusogenic

construct.

The previous 13C-31P and 13C-19F(C16) results suggested that the interior
[3 sheet region of HFPmn was located midway between the headgroups and the
bilayer center and more direct evidence was provided by experiments on
samples containing 9 mol fraction 5-‘9F DPPC lipid. The position of the CS
carbon along the bilayer longitude of gel phase DPPC bilayers is ~10 A from 31P

and ~12 A from the bilayer center.(25) Although the cholesterol-rich membranes

146

Table 8. Best-fit distance and population parameters for samples in PC:PG:CHOL.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

abc
7 13CO-31P
Ala1 lle4 Leu12 Ala14
r(A) 5.2 (2) 5.5 (2) 5.7 (2) 6.3 (3)
HFPmn_mut
f 0.86 (4) 0.45 (3) 039(2) 0.75 (3)
r A 4.8 2 8.6 2 5.7 2
HFPmn ( ) ( ) n.d. ( ) ( )
f 0.87 (3) 0.70 (3) 0.77 (3)
r A 5.1 2 7.9 2 9.4 1 5.9 2
HFP" ( ) ( ) ( ) ( ) ( )
f 0.83 (3) 0.66 (3) 0.63 (3) 0.72 (3)
13cones-‘91:)
_ Ala1 lle4 Leu12 Ala14
r(A)
HFPmn_mut f n.d. n.d. n.d. n.d.
r(A)
HFPmn f n.d. n.d. n.d. n.d.
HA)
HFPtr f n.d. n.d. n.d.
Ala1 lle4 Leu12 Ala14
rtA)
HFPmn_mut f n.d. -- -- ---
r(A)
HFPmn f n.d. -- -- —-
r(A)
HFPtr f n.d. --- -—- -—-

¥

 

°The best-fit rand f are given with uncertainties in parentheses. The typical 2min < 5 and the

uncertainties of d and f were determined from the region encompassed by x =12mm + 1.

b n.d. means “not determined". Fitting was not done for samples with (AS/80f” < 0.1 at t= 32
ms (1300-31P) or at r= 24 ms (13CO-19F), or for samples with no clear buildup curve.

°A dashed line means the experiment was not done.

147

13CO-3‘P REDOR 13CO-‘W” REDOR

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1.0 a 1.01
(a) ; (d)
g 0.64. A A a 0.6‘
Ala1 ° 0 °
3 04 x 3 0.41
<1 6 g
V 0.24 x o.2<
] X a Q
0.04. 9 0.01 0 ° A 9
rL 1 f 7 T
0 15 30 0 10 20
1.0 —— 1.0
(b) o HFPmn_mut {(9)
0.8* A HFPmn 08]
x HFPtr Q
a 0.61 g; 0-61
° 03’
Ala6 8 (14. a 0.4
s o O O S x
0.24 o 0.24 X
X X
00 a 2 3 2 A 00 8 6 2
O
0 15 30 0 10 20
1.0 ’ ' ‘ ' 1.0
.(c) (f)
0.8 ' 0.8‘
3 Q
25 0.6 fig 06*
Ala14 8 A g ‘ 8
‘3, 04 x 3 0.4
0.2 A 3 0.21
O X A
0.0 9 x 0.01 I e 3 9
I I l j T
0 15 30 0 10 20
Dephasing time (ms) Dephasing time (ms)

Figure 32 13CO-3‘P and 13CO-‘9F REDOR dephasing curves for different HFP
samples labeled at (a, d) Ala1, (b, e) Ala6 or (c, f) Ala14. The membranes
contained 9 mol% 16-F-DPPC lipid. The color coding of the constructs is given in
the legend of panel b. For 2 ms dephasing time, the typical uncertainty in
(AS/So)” is 10.02 and for the other dephasing times, the typical uncertainty is
:l:0.04.

148

 

 

(al

 

 

 

190 1'70 190

../\..J\. €1.11

 

 

 

 

 

 

 

190 170190 170190 170
1901'701901'7

13C Chemical Shifl (ppm)

 

_a
O

 

 

 

(f) o HFPmn_mut-A6
.1 A HFPmn-A6
0-8 x HFPtr-A6
. HFPtr-L9
g 0.6 .
$1"
<1 0 4
V A
A
02 _' l-
A 9
0.0 1 8 a g
0 1'0 2'0

Dephasing time (ms)

Figure 33 13C So and S1 NMR spectra from 13CO- 19F REDOR experiments of
samples made with 9 mol% 5-F-DPPC lipid. For panels a, b, c, d, and e the
samples respectively contained HFPmn-A1, HFPmn-A6, HFPmn_ mut—A6, HFPtr-

A6, and HFPmn- L9.

149

(9)

So 81

 

| 1

:~ l l," b.1l'."".'1.'1
Jlfidm. 1",” .11., .11.)» all “v,

HFPmn-A6-F5-16ms

l

. ,1 .111
'lp‘ur'l.§\l ‘11 {.‘N'u (LIL 911;} ’1")

HFPmn-L9-F5-16ms

l '1' ' ' t' ‘ ' 1 I
51‘ it"! 5." - 1‘1." 12“?!“ ".191 1.1111111 “it

HFPtr-A6-F16-16ms

' ' 1 1 - 1' "1 ' ' ‘ .E
"1'11? 1'4 ‘}%§1‘,flf 1!"..1'J'L'a1'11l1 f. '1‘

HFPtr-L9-F16-16ms

So 81

 

I ‘ I ‘ I
_r".“AH‘-l 1.1.11 '1‘ ‘g'df‘V‘N. iv \ 1'3}

HFPmn-A6-F5-24ms

awn—n. .. .....-c.--u.‘....s... l—‘v: ”swan“.

: 1 .
11.,1‘1 1'1"1."""11,-11'""9"l|"'l"

HFPmn- L9- FS- 24ms

 

fill‘filaJ \m1

HF Ptr-A6-F 1 6-24ms

' 11 ' -. l 'l ' ..1
(View "K #1; tit-hivull:.1t-1'

HFPV-L9~F16-24ms

 

 

1'1‘“W'-~-“~'~‘-.w*iri~iv-t~'Wel ~11

HF Pmn-A6-F1 6-24ms

u-..“

awnvm-a.’ ’w- (~11- --vt ,- V

HF Pmn-L9-F 1 6-24ms

" . . 'l '
..l’ it}; 1'1 «'13',‘ “1
1'1" 1" 1111111 1""111'1‘1 W"

H FPtr-AB-F5-24ms

"118‘“ 1.713.,{1'1'fqm'11t11'2k‘1'9

HFPtr-L9-F5-24ms

Figure 33 130 So and S1 NMR spectra from 13CO-19F REDOR experiments of
samples made with 9 mol% 5-F-DPPC lipid. For panels a, b, c, d, and e the
samples respectively contained HFPmn-A1, HFPmn—A6, HFPmn_mut-A6, HFPtr-
A6, and HFPmn- L9. Each spectrum was processed with 200 Hz Gaussian line
broadening and polynomial baseline correction. Each 80 and S1 spectrum was
the sum of~ 20000 scans. 13CO-19F(C5) REDOR dephasing curves for different
HFPs are plotted in panel fand the constructs are coded as shown in the legend.
For 2 ms dephasing time, the typical uncertainty' in (AS/So)°"" is $0. 02 and for the
other dephasing times, the typical uncertainty is +0. 04. In panel 9 the 13C-19F(C5)
and1C-19F(C16) spectra of HFPmn and HFPtr at 16 or 24 ms dephasing time
were shown in the stack form with So spectra on the left side and S1 spectra on
the right side. The dash lines reflect the intensities of the So spectra.

150

Increasing Fusogenicity ——————-—

(a) (b) (e)
Ala1 Gly16

      

HFPmn_mut

Figure 34 Insertion models of [3 sheet (a) HFPmn_mut, (b) HFPmn, and (c) HFPtr.
Lipid headgroups are drawn as blue balls, lipid alkyl chains are drawn in grey,
and peptides are drawn in red. In all models, peptides are represented as
oligomers with either six (HFPmn and HFPmn_mut) or two (HFPtr) molecules.
The strands are in antiparallel Bsheet structure with adjacent strand crossing
near PheS and Leu9. This is the known structure for a large fraction of HFPmn
peptides. The number of strands in a sheet is not known but is likely a small
number. The lines at the C-terminus of HFPtr represent the chemical cross-
linking of the HFPtr construct. For clarity, not all lipid molecules are shown near
the HFP.

151

of the present study form a liquid-ordered rather a gel phase bilayer, the acyl
chain conformation is ordered in both phases with the C5 carbon approximately
midway between the 31P and the bilayer center.(62) Figure 33 shows the REDOR
spectra at z- = 24 ms for samples labeled at Ala1 and Ala6 as well as plots of
(AS/So)” vs rfor Ala6 samples. As a negative control, HFPmn-1A, cf. Figure 33a,
had (AS/So)” z 0 which was consistent with the proximity of Ala1 and the lipid
phosphate groups deduced from the 13CO-31P REDOR data. Figure 33f shows
that among constructs labeled at Ala6, HFPmn-6A was the only one with non-
zero (AS/So)°"”. Similar results were obtained for HFPmn-9L as shown in Figure
33e for the r = 24 ms spectra and in Figure 33f for the dephasing curve. These
data along with (AS/So)” e o for the 13CO-3‘P and 13CO-‘9F(C16) experiments
support a membrane location for the HFPmn interior which is midway between
the phosphate headgroups, and the bilayer center, i.e. intermediate between
HFPmn_mut and HFPtr. This location correlates with the intermediate fusion rate
of HFPmn. The more clearly supplementary experiments shown in Figure 339
proved that the HFPmn-6A and HFPmn-9L have dephasing in the ‘30-‘9F(cs)
REDOR, but not in the 13c-‘9l=(c1ts) REDOR, while the HFPtr-6A and HFPtr-9L
have dephasing in the 13C-19F(016) but not the ‘30-‘9F(C5) REDOR.

Figure 34 shows experimentally-based membrane insertion models for
HFPmn_mut, HFPmn and HFPtr in antiparallel [3 sheet structure. The B strand
conformation was supported by the 13CO peak chemical shifts for the labeled
residues, cf. Table 7, and the antiparallel [3 sheet structures for HFPmn and

HFPtr were based on previous experiments. (36,43) The depths of insertion

152

follow the trend that HFPmn_mut < HFPmn < HFPtr. Quantitative analysis of the
REDOR data was done by first removing the natural abundance 13CD

"b which represent the signals

contributions and then fitting the remaining (AS/So)
of only the labeled 1300s. The fitting model was two populations of spin pairs (eg.
two 13CO-3‘P or two 13CO-‘QF pairs) with one pair having fractional population f
and magnitude of dipolar coupling d > 0 and the other pair having fractional
population 1 - f and d = 0. For a single spin pair, d is quantitatively related to the
internuclear distance r by a r‘3 dependence. The existence of the 1 - f, d = 0
population was ascribed to 1""CO nuclei in the [3 sheet interior that were far from
any region of the membrane and for the 13CO-19F data, the dilute 19F spin density
because of the 0.09 mol fraction of fluorinated lipid. Because each data set only
contained 4 or 5 points, it was not reasonable to fit the data to more sophisticated
structural models, eg. multiple 31P nuclei. The two spin pair model was at least
reasonable as evidenced by typical best-fit 12 < 5.

Table 8 summarizes the spin pair populations and best-fit internuclear
distances for the f fractional populations. A summary of the quantitative data
analysis includes: (1) For all HFPmn_mut samples, the best-fit 13CO-3‘P
distances are in the 5.0 — 6.3 A range. These data and reasonable values of van
der Waals radii are consistent with close contact of the [3 sheet region of
HFPmn_mut with the phosphate headgroups. (2) For HFPmn and HFPtr samples
labeled at Ala1 or Ala14, the best-fit 1r’CO-:”1P distances are in the 4.8 — 5.9 A

range with best-fit f > 0.7. For more interior [3 sheet residues, the (AS/SOY” are

small, eg. Ala6 or Leu9, and could not be reliably fitted or the fitted distances are

153

in the 8 - 10 A range, eg. lle4 or Leu12. These data suggest membrane insertion
of the lle4 to Leu12 region of HFPmn and HFPtr with the termini of the B sheet,
eg. Ala1 and Ala14, in contact with the lipid headgroups, of. Figure 34. (3) For all
HFPmn_mut samples and most HFPmn and HFPtr samples, the 13CO-‘9F
(AS/SOY“ are small and could not be reliably fitted. The exceptions are the
samples containing 5-‘9F lipid and HFPmn labeled at AlaG or Leu9 and the
samples containing 16-‘9F lipid and HFPtr labeled at AlaG or Leu9. The best-fit
13CO-‘S’F distances in these samples are in the 7 - 8 A range and the best-fit f
are in the 0.34 - 0.39 range. These analyses are consistent with partial
membrane insertion of the interior B sheet residues of HFPmn and deeper
insertion of HF Ptr.

The membrane location of the HFP provides useful information to
understand the perturbation of membranes and the catalysis of membrane fusion.
The present study provides residue-specific membrane locations based on solid-
state NMR experiments for three different HFP constructs with very different
fusogenicities. Insertion models for different B strand HFPs will be discussed in
the context of previous and present work. The previous study on triply-labeled
HFPmn showed that relative to the Gly5 to Gly13 13(305, the Ala1 to Gly3 and
Ala14 to Gly16 13COs were closer to the lipid 3‘Ps.(15,17) Two insertion models
were proposed with either insertion into a single leaflet or membrane traversal of
both leaflets. Another study focused on the secondary and tertiary structure of
HFPmn in membranes with physiologically relevant cholesterol content and

supported the formation of small oligomers in an antiparallel B sheet structure

154

with adjacent strand crossing near Phe8 and Leu9.(36) Therefore, Ala1 to Gly3
and Ala14 to Gly16 were close to one another in adjacent strands of the sheet
and were at the termini of the sheet. It was therefore reasonable that both
regions could be close to the phosphate groups. All of these results are
consistent with the results of the present study and with the HFPmn insertion
model present in Figure 34. Furthermore, the proximity of interior B sheet
residues to 5-‘9F but not 16-‘9F lipid nuclei supports membrane insertion into a
single leaflet rather than membrane traversal by HFPmn. This partial insertion
model is also consistent with an earlier fluorescence study showing proximity of
residue 8 of HF Pmn to the middle region of a single leaflet.(8,9)

In Figure 34, HFPmn_mut and HFPtr are also represented by antiparallel
B sheet structure. Evidence for this structure includes: (1) earlier 13CO-‘5N
REDOR measurements on HFPtr; and (2) peak 13CO chemical shifts in
HFPmn_mut and HFPtr which are typically within 0.5 ppm of the corresponding
shift of HFPmn, of. Table 7; and (3) the trend of best-fit 1"co-313 Is for
HF Pmn_mut are consistent with a B sheet structure where relative to the interior
residues, eg. Ala6. A greater fraction of the terminal residues, eg. Ala1, have
contact with the lipid molecules. The surface location of B sheet HFPmn_mut is
also strongly supported by small 13CO-‘QF (AS/So)” in all HFPmn_mut samples.
It is very interesting that the charged Glu2 residue near the terminus of the B
sheet appears to induce a significant change in HFP membrane location. The
HFPtr antiparallel B sheet is most reasonably described with a minimal unit of two

HFPtr molecules “C” and “D” and adjacent antiparallel B strands arranged in an

155

CDCDCD structure. Because of the close contact of HF Ptr Ala6 and Leu9 1300s
with the 16-‘9F lipid nuclei, it is not possible to discount membrane traversal by
HFPtr, ie. molecules C and D on opposite sides of the membrane. However, the
displayed model in Figure 34c is more consistent with viral fusion in which
multiple gp41 trimers would initially bind to the same outer leaflet of the target
cell membrane. The Figure 34c model also correlates with the definitive
membrane locations of HFPmn and HFPmn_mut. It is also definitive that relative
to HFPmn, HFPtr is more deeply inserted in the membrane. The present study
does not provide data to explain the reason for this deeper insertion but it may be
related to formation of larger and more hydrophobic oligomers by HF Ptr relative
to HF Pmn.

The present study focuses on membranes with biologically relevant
cholesterol content in which all of the constructs have predominant B strand
conformation. For HFPmn and HFPtr associated with membranes without
cholesterol, significant populations of molecules with helical conformation are
detected.(15) Helical conformation is also observed for HFPmn in detergent
micelles at low HFP:detergent where each micelle contains at most one
HF Pmn.(3, 4, 6) There is reasonable correlation between the membrane locations
of B sheet HFPs and the current data on the micelle and membrane locations of
helical HFPs. One point of agreement between all of the data is that in either
helical or B strand conformations, Ala14 and Ala15 residues are near the
membrane or micelle surface. These residues are on the border between the

apolar and polar regions of the HFP sequence which approximately matches the

156

polarity change at the membrane or micelle surface. The location of these
residues appears to be an intrinsic property of the HFP sequence that is
independent of conformation. For one model of helical HFPmn in a micelle, the
lle4 to Ala15 region traverses the micelle interior.(3) This correlates with the
similar membrane location of this region in B sheet HFPmn in the present study,
of. Figure 34b. Molecular dynamics simulations on a single molecule of
HFPmn_mut or HFPmn in a membrane show a surface location or shallow
insertion, respectively, which correlates with the B sheet HFPs of the present
study, of. Figure 34a, b.(11)

It has been proposed that two requirements of virus-cell fusion are
assembly of multiple fusion peptides and destabilization of the target cell
membrane.(63) The present study provides insight into these requirements
including a possible link between them. There is a clear positive correlation
between the depth of HFP membrane insertion and fusogenicity. The correlation
can be understood by a second correlation between depth of membrane insertion
and membrane destabilization. For insertion into a single leaflet, there will be
perturbation in the packing of lipids near HFP which will likely destabilize this
region of the leaflet and reduce the activation energy needed to form membrane
fusion intermediate states such as stalks and fusion pores.(64) It is reasonable
that deeper insertion into a single leaflet will cause greater destabilization and
therefore faster fusion rate which correlates with experimental observations for

HFPmn_mut, HFPmn, and HFPtr.

157

For these three constructs, there may also be a positive correlation
between number of molecules in the B sheet assembly and depth of membrane
insertion. A larger assembly would likely be more hydrophobic and therefore
more stable in the membrane interior. Evidence to support this hypothesis
includes: (1) HFPtr has the deepest insertion and is pre-organized into trimers; (2)
HFPmn_mut has the shallowest insertion and relative to HFPmn and HFPtr,
HF Pmn_mut has the greatest helical population which is likely helical monomers.
Inhibition of HFP oligomeric assembly in HFPmn_mut is reasonable because of
charge repulsion between Glu2 sidechains from different molecules. The HFP
assembly/fusion correlation is also supported by virus-cell and cell-cell fusion
studies with the gp41 V2E mutant. Viruses or cells expressing 91% wild-type
gp41 and 9% gp41 V2E mutant had only ~40% of the fusion activity of the
corresponding system expressing 100% wild-type gp41.(65) Assembly of multiple
HFPs for efficient fusion has been inferred from this “trans-dominant” effect.

In this section, the membrane locations have been determined for three
different HFP constructs in membranes with biologically relevant cholesterol
content. All three constructs adopt predominant B strand conformation for the N-
terminal region and are less structured in the C-terminal region. HFPmn_mut is
the least fusogenic construct and is located on the membrane surface. HFPmn
has intermediate fusion rate and its lle4 to Leu12 region is inserted into one
leaflet of the bilayer. HFPtr has the putative trimeric HF P state of gp41 and is the
most fusogenic construct with the deepest membrane insertion that extends to

the bilayer center. This study therefore correlates membrane insertion depth into

158

a single leaflet and fusion rate and this correlation is reasonably understood in
terms of destabilization of the lipid packing. In addition, the present work
including use of 5-‘9F-DPPC lipid describes a general approach to study the
membrane locations of specifically labeled peptides and proteins, and may also
be applicable to more uniformly labeled systems with appropriately modified
REDOR pulse sequences.(66)

Membrane Insertion of a-he/ical HFPs. In this section, we will report the
studies on the membrane location of HFP which was associated with PC:PG, in
particular the membrane location of the residues in HFP which formed a—helical
conformation. The motivations to study the helical HFPs include: (1) There is not
yet a consensus that B-strand is the only secondary structure which has fusion
activity even though it has been shown that HFP adopted a predominant B-strand
conformation in a biologically relevant cholesterol-containing
membrane.(6, 14,67, 68) In fact, a previous lipid-mixing assay suggested that the
a-helical and B-strand HFPs may have comparable fusion activities.(14) (2) The
helical conformation may also be a biologically relevant structure for HFP present
in the entire gp41. There has been experimental evidence which supported the
formation of partial a-helical structure for HFP in PC:PG:CHOL membranes when
the HFP was combined with the NHR and CHR parts of gp41 .(69) (3) The results
of membrane location of helical HFPs can be directly compared with the HFP
location in micelles studied by solution NMR, and also with the simulation studies
in which only ct-helix structure was considered. (4) It will be interesting to

compare the membrane location results from the helical HFPs with those

159

obtained for the B-strand HFPs, and to learn whether the membrane location of
HFPs could be correlated to the fusion activity of different HFP constructs in a
conformation-independent manner.

There have been residue-specific conformation studies for the HF P
associated with micelles where the HFP adopted helical conformation. It has
been generally accepted that the helical conformation extends from Ile4 to Leu12,
however, one study reported that the helix extended to Met19 while another
study claimed that there was a turn at Ala15 and Gly16.(3,4,6, 7) The residue-
specific secondary structure of HFPs associated with PC:PG membranes have
been studied using solid-state NMR. Previously we showed that Ala6 and Ala15
of HFP in PC:PG membranes can adopt both a-helical and B—strand
conformations.(15) In the present set of experiments, we put a single 1"CO-label
on either Ala1, lle4, Ala6, Leu9, Leu12 or Ala14 which covers the entire helical
region. The HFPmn_mut, HFPmn and HFPtr were chosen as the model
constructs to study the membrane insertion/fusion activity correlation as was
done for the cholesterol-containing membranes.

Table 9 summarizes the 13CO chemical shifts for the labeled residues in
different HFP constructs. The information was extracted from the corresponding
“co-“P REDOR so spectra with r = 2 ms as shown in Figure 35a. The
difference in chemical shifts between different 2' and different REDOR
experiments for a singly labeled residue was always less than 0.3 ppm. The 13CO
chemical shift is correlated to the local conformations of proteins. The empirical

correlation database RefDB has been established by liquid-state NMR

160

assignment of proteins. The 13CO chemical shift ranges for Ala, lie and Len to be
considered in tat-helical conformation are 1781-1807, 1784-1790 and 177.2-
179.8 ppm and the range for these residues to be B-strand are 174.5-177.5,
173.4-176.1 and 174.2-177.0 ppm respectively.(33)

Table 9. 1300 chemical shifts in PC:PG membranes 8

 

 

Ala1 lle4 Ala6 Leu9 Leu12 Ala14
HFPmn_mut 175.6 176.5 177.8" 180.0 179.9 180.3
175.8 175.8 176.9

HFPmn 174.4 178.8 179.8 180.0 179.8 179.8
174.8 175.5 175.7 176.0

HFPtr 175.4 178.3 179.8 180.1 180.0 180.0
174.8 175.3 175.6 178.1

 

’ All chemical shifts are given in unit of ppm.

b The AlaG 13CO peak for HF Pmn_mut is broad and is not correlated with a single secondary
structure (of. Fig. 35)

The following conclusions can be obtained by comparing the
experimental results with the database values: (1) Ala1 is not helical for any of
the three constructs. (2) For all three constructs, there is always some fraction of
HFPs which adopt helical conformation. (3) Table 9 showed that the variation in
the chemical shifts for lle4 between different constructs in or-helical conformation
is larger than those for Ala6, Leu9 and Leu12. Figure 35a showed that lle4 has
broad single peaks in all three constructs while Ala6, Leu9 and Leu12 have either
a predominantly single peak or two separated peaks. These results suggested
that lle4 has a more flexible secondary structure compared with Ala6, Leu9 and

Leu12, which probably mean that lle4 is the N-terminus of a helix.(6)

161

(a)

Ile4 AlaS LeuQ Leu12 Ala 14

Winntnn

mAWiAAA
”#11111

190170 190170 190170 190170 190170 190170
13C Chemical Shift (ppm)

 

 

 

(b)

SO 81 So 81
Ala1 A ‘ A A
S S

0 1

leeches

190170 190170 190170 190170 190170 190170
130 Chemical Shift (ppm)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 35 (a) ‘30-3‘P REDOR so spectra with 2 ms dephasing time. (o) REDOR
So and S1 spectra for HFPmn_mut with long dephasing time (t = 32 ms for 13C-
31P experiments and r = 24 ms for 13C-‘9F experiments). The left, middle and
right columns in (b) are ‘30-3‘P, ‘30-‘9F(C5) and 13c-191=(c18) experiments
respectively. All spectra were processed with 200 Hz Gaussian line broadening
and baseline correction. In panel (b) each of the 13C-31P spectra was acquired for
30000 scans and each of the ‘30- 9F spectra was acquired for ~ 20000 scans.

162

(a) 130319 REDOR (b) 13C-‘f’F(CS) REDOR

 

 

 

 

 

 

 

   

0.0 I;
Ile4 AlaG L909 Leu12 Ala14

Ala1 lle4 Ala6 Leu9 Leu12 Ala14 .
Residue

Residue

(c) 13C-19F(C16) REDOR

 

.0 .0 7‘
a) on o
t l

(AS/30PM
o
a
1

.°
N
I

 

 

00"
N81 "04 A186 Leu9 1.8012 Ala14

Residue

Figure 36 Plots of (AS/180)” for HFPmn _mut 13c 31F experiments at 32 ms
dephasing time and :1-10 9F experiments at 24 ms. The experimental dephasing
was obtained by integrating over a 1 ppm interval around the 300 peaks in the
corresponding So and $1 spectra shown in Fig. 35. For Leu12, the black bar
represents the (AS/So)” for a-helical conformation and the red bar represents
the (AS/So)” for B-strand conformation. The typical uncertainty is :I: 0.02.

163

Figure 35b displays the long time REDOR So and 81 spectra (1 = 32 ms
for 13C-31P REDOR and 1' = 24 ms for 13C-19F REDOR) for HFPmn_mut and
figure 36 summarizes the corresponding experimental dephasing. In figure 36,
the residue Leu12 has two (AS/80V” in each plot which correspond to a-helical
and B—strand conformations. The values were obtained by integrating over a 1
ppm interval around the corresponding peaks in Figure 35b. For the other
residues, only one experimental dephasing was shown either because there is
only one major secondary structure (Ala1 and LeuQ, of. Figure 35b), or because
the peaks corresponding to distinct conformations are not clearly separated (lle4,
Ala6 and Ala14, of. Figure 35b). The general conclusion for HFPmn_mut is that
there are large experimental dephasings for all the residues in 13C-31P REDOR,
e.g. (AS/So)” > 0.2, and there is almost zero dephasing in both 13C-19F(C5) and
13C-11'1F(C16) experiments. The results indicate the peptide backbone of
HFPmn_mut from Ala1 to Ala14 of HFPmn_mut has close contact with the lipid
phosphate groups and little contact with the lipid alkyl chains. Figure 37a shows
the experimental 13C-31P REDOR dephasing curves, and the dephasing curves
for the 130-19F experiments are basically flat lines with (AS/So)” < 0.1 for the
longest dephasing time. A more quantitative analysis can be completed by first
calculating the (AS/So)” from the (AS/So)” using Eq. 59, and then fitting the
dipolar coupling frequency (d) and a fraction parameter (A) of (AS/So)“ by
comparison with the (AS/So)“ which were calculated using Eq. 34. The best-fit
130-31P dipolar coupling frequency can be correlated to the 1300-31P internuclear

distance through r = (12250/d)"3 with r in units of A and d in units of Hz. The

164

population parameter A indicates the fraction of 13C0s that are close to a 31F and

provide non-zero (AS/So)”. The data fitting for the 130-31P experiments are

(a)

 

 

 

 

1.0
o
0.8 —
A
A
g. 0.6— 0
(Q 04
co ' 3
< 0
V alt
0.2 -— D x
8 ' x
x
0.0 d 8 n
I r l
0 10 20 30

Dephasing time (ms)

Figure 37 (a) Plots of 130-31P (AS/SOY” vs. dephasing time for HFPmn_mut with
different labeled positions. The symbols are open diamonds for A1, open squares
for l4, open triangles for A6, crosses for L9, stars for the a-helical L12 and open
circles for A14.

165

(b)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1.0 1.0
Ala-1 lie-4
0.8 "1 0.8 ‘1
0.6 ~ 0.6 '1
o o
a)
2;. 0.4 — a 0.4 ‘1
D
0.2 -i 0.2 ..
0.0 -1 D 0.0 '7
l r l l l l
0 10 20 30 0 10 20 30
Dephasrng time (ms) Dephasing time (ms)
1.0 1.0
Ala-6 Lou-9
0.8— o 0.8 4
o 0.6- o 0.6“
‘0 Q
h D U)
(1 Q4 --1 4 04 ~
0.2 "l D 02 d
i f r T T T
0 10 20 30 0 10 20 30
Dephasing time (ms) Dephasing time (ms)
1.0 1'0
Lou-12 Ala-14
0.8- 0.8 —i
0.6 -
06*
(3’ A a?
a a
V 04_ A 0.4 "
0.2 -1
0.0 -1
I r I T T T
0 10 20 30 0 10 20 30
Dephasing time (ms) Dephasing time (ms)

Figure 37 (b) Plots of 130-31P (AS/SOY” (open squares) and (AS/So)” vs.
dephasing time for different residues (as labeled in the figures) of HFPmn_mut.
The typical experimental uncertainties are 1:0.02-003 and the typical corrected
uncertainties are :l:0.03-0.04.

166

provided in Figure 37b. Data fitting was not carried out for the 130-19F
experiments because the build-up of REDOR dephasings was not observed
experimentally. Table 10 provides the best-fit distances for all the labeled
residues. The labeled 13005 are 6.6-10.1 A away from the 31Ps in the lipid
phosphate groups. Under the experimental conditions, the temperature is below
the transition temperature of the phospholipids of the lipid bilayer which means
the membranes are in the gel phase. The typical membrane longitudinal distance
between 31P and 19F(C5) is 10 A in the gel phase lipid bilayer according to a
simulated gel-phase DPPC bilayer.(25) This suggests the HF Pmn__mut backbone
is located higher than the longitude of the phosphate groups, since there is no
contact between the backbone 1300s and 19F(CS). The simulated gel phase
DPPC membrane bilayer structure indicated that there is a ~ 10 A water layer
above the bilayer surface.(25) Our results suggest that the HFPmn_mut may
bind to the lipid bilayer surface through hydrogen bonds with the water molecules.
Figure 35a showed that the 1300 peaks for Ala1, lle4 and Ala6 in the
HFPmn_mut were broad. This suggested that these residues may not have a
well-defined secondary structure and the carbonyl oxygens and the amide
protons of these residues may not form intra-peptide hydrogen bonds. Thus there
may be hydrogen bonds formed between the peptide backbone of these residues
and the water molecules, or between the backbone of these residues and the
lipid headgroups. All distances provided in Table 10 are greater than 6 A, which
approximately equals the sum of the length of a P=O bond (~1.8 A), a C=O bond

(~1.2 A) and two van der Waals radius of the oxygen nuclei (~3.0 A). This

167

reflects the 1300-31P distance when the lipid phosphate group is in close
proximity to ct-helix backbone. The fitting parameters A in Table 10 are close to 1,
which suggested that the residues Ala1 through Ala14 in almost

(a)

13C-31P REDOR
Ala6 Leu9 Leu12

 

z--.“ __-_.-__ __ --w *_____ _*_.___1

_ l _. . _
200 150 200 150 200 150

13C-19F(CS) REDOR

 

 

AlaB Leu9 Leu12
’T’H'T—"’_.T”—” ~' " _ ‘ " "" _fi"— ‘7 r ‘:
200 150 200 150 200 150
13C-19F(C16) REDOR
Ala6 Leu9 Leu12
j‘“"’—' _____ T— H" ' ' “ F "" “ ""“‘ """_"_"r '” " " " ' '” ' ' WT“ T“
200 150 200 150 200 150

130 Chemical Shift (ppm)

168

(b)
13C-31P REDOR

Ala6 Leu9 Leu12

200 150 200 150 200 150

13C-19F(C5) REDOR
Ala6 Leu9 Leu12

  

 

 

 

200 150 200 150 200 150
13C-19F(C16) REDOR
Ala6 Leu9 Leu12
200 150 200 150 200 150

130 Chemical Shift (ppm)

Figure 38 REDOR So and 81 spectra for (a) HFPmn and (b) HFPtr at long
dephasing time (1- = 32 ms for 13C-31P experiments and r = 24 ms for l3C-111F
experiments). The So and 81 spectra were shown in black and red respectively.
All spectra are processed with 200 Hz Gaussian line broadening and baseline

correction. The
spectra were acquired for ~ 20000 scans.

169

30-31P spectra were acquired for ~ 30000 scans and the 13C-31P

 

 

 

 

 

 

 

 

 

1.0
HFPmn-helical
0.81
0
‘31 o 4 O
.1
g 8><
0.2 — o/
0.0-1
F T l
0 10 20 30
Dephasing time (ms)
1.0
HFPmn-strand
0.8 —
Q
3; o.6~ /
3° /
O
u——--—U-—n
0.0-
l l l
0 10 20 30
Dephasing time (ms)

 

 

 

 

 

 

 

 

1.0
HFPtr-helical
0.89
a, 0.6—
9‘
<3 0
V /
0.2“ f87<2/A
D/CI
o.o- B u/
l T r
0 10 20 30
Dephasing time (ms)
1.0
HFPtr-strand
0.8—
5.: o.6~
8’
0.2-
0.0-1 9 \D‘TD a
A
l l l
0 1O 20 30
Dephasing time (ms)

Figure 39 Plots of (AS/So)” vs. dephasing time for the 13C-31P experiments.

170

 

 

 

 

 

HFPmn-helical

\A

370‘ -

“U7

 

—O

 

 

 

 

 

 

 

r
10

Dephasing time (ms)

 

(AS/$019)“)

 

HFPmn-strand

“3%

 

 

l
10

Dephasing time (ms)

 

 

 

 

 

 

1.0
0.8-1 HFPtr-helical
3* 0.6—
0
63504
a
S
0.2-1 0/0 o
"—L
0.0- 07égf E, ‘D
0
i 7
0 10 20
Dephasingtime(ms)
1.0
HFPtr-strand
0.8-1
3- 0.6-1
9‘
g 04H
51
0.2-1
0.0-1 8" 8 =8
l l
o 10 20
Dephasing time (ms)

Figure 39 Plots of (AS/So)MD vs. dephasing time for the 130-19F(CS) experiments.

171

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1.0 1.0
HFPmn-helical HFPtr-helical
0.8- 0.8-
g. 0.6-1 a 0.61
94 9.2
S 3 0
0.2— A///§ 02.. ————-—-A%fl
0.0-1 é———-6——————0 0.0—1
1 1 1 T
0 10 20 0 1O 20
Dephasing time (ms) Dephasing time (ms)
1.0 1.0
HFPmn-strand HFPtr-strand
0.8- 0.8-1
:1 q
3.; 0.8-a it; 0.6—
O O
a °-“ E049
0.2-j 0/0 0.22 i 875
0-0'1 Kgfifi/ 0.0— Q A A A
l l 1 T
0 1O 20 0 10 20

Dephasing time (ms) Dephasing time (ms)

Figure 39 Plots of (AS/So)” vs. dephasing time for the (a) 130-31P, (b) 13C-111F(C5)
and (c) 13C-19F(C16) experiments for HFPmn and HFPtr in the a-helical and p
strand conformations. The residues Ala6, Leu9 and Leu12 are represented with
the open squares, open circles and open triangles respectively. The uncertainties
of (AS/So)” are typically 10.02 ~ 0.03 and are approximately the size of the
symbols. The (AS/So)11x11 values were determined by integrating over a 1 ppm

interval around the tit-helical or B-strand 13CO peaks in the corresponding So and
So spectra.

172

all HFPmn_mut have close contact with the 31Ps. Together with the fact that
there is no contact with the 19Fs, it suggests that the residues Ala1 through Ala14
in almost all the HFPmn_mut adopted the surface location. Furthermore, the
membrane location of the tit-helical HFPmn_mut in PC:PG is consistent with the
observation of the surface membrane location of the B-strand HFPmn_mut in
PC:PG:CHOL, which may indicate the surface location of HFPmn_mut is an
intrinsic property of the peptide sequence and is independent of the membrane
composition and peptide conformation.(18) A consequence of the surface
location of HFPmn_mut is that the peptide does not penetrate the polar
phospholipid headgroups and disrupt the membrane interior.

Table 10. 13CO-31P internuclear distances and populations for HFPmn_mut

 

 

Ala1 Ile4 Ala6 Leu9 Leu12 " Ala14
1300-31,: 6.9 (2) D 8.7 (2) 6.6 (1) 10.1 (3) 8.4 (2) 7.6 (2)
distance (A) 7.8 (2)
. 0.94 (3)
Population A 0.98 (2) 0.93 (3) 0.88 (1) 0.90 (3) 0 97 (2) 0.95 (3)
2 2.0
x 5.1 1.7 9.7 1.3 1.8 1.4

 

1' For Leutz, the first value corresponds to helical conformation and the second value
corresponds to strand conformation.

1’ The uncertainties are shown in the parentheses and were determined by the x2 = x2m1n+1
criterion.

Figure 38 displays the long time REDOR spectra for the residues Ala6,
Leu9 and Leu12 for HFPmn and HFPtr. These residues were selected because
(1) there is obvious non-zero 13C-19F dephasing in some spectra for these
residues which is qualitatively different from HFPmn_mut; and (2) the or-helical
and B-strand peaks are in general well-separated which facilitates quantitative

analysis of the data as a function of conformation. The information obtained for

173

the residues Ala1, lle4 and Ala14 will be discussed later in this section. Figure
39a and 6 plots the 13C-31P and 1110-19F(C16) experimental dephasing curves
respectively for the a-helical and B-strand conformations for HFPmn and HFPtr. It
can be qualitatively concluded that the HFPmn and HFPtr are deeply inserted
into the region of phospholipid alkyl chains because both constructs show some
non-zero 13C-19F dephasings.

A quantitative analysis was done for Ala6, Leu9 and Leu12 in the a-
helical HFPmn, the or-helical HFPtr, the B-strand HFPmn and the B-strand HFPtr.
The resultant membrane locations may be correlated with the different fusion
activities of HFPmn and HFPtr in PC:PG bilayer. The data fitting was based on
the following considerations:

(1) For each HFP in each conformation, there are two populations which differ in
their membrane location. The population which is closer to the membrane
surface is named “surface-located population” and the population closer to the
membrane interior is named “deeply-inserted population”.

(2) The typical uncertainty for the 130-31P and 13C-19F REDOR (AS/So)” is :I:
002003. Only the 1300s with (AS/So)” > 0.1 at large dephasing times (2' = 32
ms for 13C-31P and r = 24 ms for 13C-19F) are considered to have fittable
dephasing relative to the uncertainties. According to the REDOR universal
dephasing curve,(32) AS/So = 0.1 corresponds to ~ 11 A 13C-31P distance at 32
ms and ~ 14 A 13C-111F distance at 24 ms. Thus, we consider that the
measurement limits for 13C-31P and 13'C-19F REDOR experiments are 11 and 14 A,

respectively. The distance between the 31P and the 19F(C16) in a gel phase

174

    

1' i. .L

, . . ' ,1 1 3" phate-waterlnterface
‘ I a} 1: g 1", ' 4".“‘1 .-- ' 1'
«15.191303 Eli-11a _.'

-. If 3.571
Alkyl C NudGUS’ . Alkyl Chain Region

phate-alkyl chain Interface

 
 

 

v 19F(C16)

A Membrane Bilayer Longitude

  

 

1 I I ‘1 '
”F(CS) ‘ ,- 1

 

‘ . .

'19F(c16)
V

Figure 40 (a) Geometry Model for the consideration of 13C-31P and 13C-11’F
measurement limit. The two circles with 11 A and 14 A radii indicate the
measurement limits of 13C-31P and 13C-11’F(C16) REDOR respectively. The yellow
triangle shows the geometry of the case where a 13CO has the maximum vertical
distance (3.5 A) relative to the lipid alkyl chain. (b) and (c) Longitudinal positions
of 31P, 19F(cs) and 1°F(C16) in the membrane bilayer. In panel (b), the dotted
circle has the radius of ~ 10 A and the solid circle has the radius of ~ 14 A. The
region marked in red indicates the possible location of the 13COs of Ala6 and
Leu12 in the B-strand HFPmn as described in the main text. In panel (c), the
longitudinal distance between 31P and 19F(C5) is 10 A and the distance between
19F(cs) and 19mom is 12 A.

175

DPPC molecule is ~ 24 A.(25) The overlapped region of the two circles in Figure
40a indicates the region where a 1300 nucleus in the peptide backbone can have
detectable 13C-31P and 13C-19F(C16) REDOR dephasings simultaneously. Using
the geometrical parameters in Figure 40a, it can be obtained that the maximum
vertical distance from the points in the overlapped region to the lipid alkyl chain is
3.5 A. This means a 13CO nucleus has to be located within 3.5 A away from a
lipid alkyl carbon nucleus to have detectable 13C-31P and 130-19F(C16)
dephasings simultaneously. However, the shortest distance from a peptide
backbone 13CO to a lipid alkyl carbon nucleus in the lipid alkyl chain is
approximately 5.3 A which is the sum of the bond length of a C-H (1.4 A), the
bond length of a C=O (1.2 A), the van der Waals radius of a hydrogen atom (1.2
A) and the van der Waals radius of an oxygen atom (1.5 A). Consequently, in the
following quantitative analysis we consider that a specific 13CO will not have
detectable 13C-31P and 13C-19F(C16) dephasings at the same time.

(3) A 13co in the surface-located peptide does not contact 13C-19F(C16), and this
does not mean that such a 13CO nucleus must contact 31P. A 1300 in the deeply-
inserted peptide does not contact 130-31P, and this does not mean that such a
1300 nucleus must contact with 19F(C16).

The 130-31P and 13C-111F(C16) (AS/So)” were first corrected using Eq. 59
to obtain (AS/So)”, and then the (AS/So)“ were fitted to the simulated dephasing
(AS/So)” of either a 130-X two-spin system or a X-130-X three-spin system
where X E 31P or 19F(C16). The three-spin fitting was only applied to these

samples with visibly large difference between their (AS/SOY” and the (AS/So)”m of

176

Table 11 The best-fit A and 8 population parameters for HFPmn and HFPtr in PC:PG.

 

 

 

Parameter A Parameter B
Ala-6 Leu-9 Leu-12 Average Ala-6 Lou-9 Leu-12 average
HFPmn(helical) 0.62 0.62 0.65 063(2)" 0.27 —-. 0.31 029(2)
HFPtr(heIical) m a 0.29 0.20 024(4) 0.28 0.37 0.33 033(4)
HFPmn(strand) --- 0.61 0.66 064(2) _- 030 .... 0.30 c
HFPtr(strand) —- 0.41 0.44 042(2) 0.32 0.32 -- 0.32 c

 

1’ The dashed lines indicate that these data have (AS/So)” < 0.1 at 32 ms dephasing time for 13C-
31P data and (AS/So)” < 0.1 at 24 ms dephasing time for 13C-11’F(C16) data.

b The uncertainties are shown in parentheses. The uncertainties in A were determined from the
difference between A for each residue and the average A. The uncertainties in B were similarly
calculated.

1’ The uncertainties were not determined either because there is only one A or 8 value or
because the two values are the same.

Table 12 Summary of the fitting for 13C-31P and 13C-19F(C16) experiments.“

 

 

 

 

 

 

 

 

Surface-located Populatiorzl Deeply-inserted Populatiog
r (A) 0(°) x r (A) 96’) in
A6 4.8(3) 10(2) 0.9 10.4(3) 98(3) 0.4
HF Pmn
, L9 6.3(2) 5.1 11 .8(3) 21(2) 1.5
(helical)
L12 5.2(2) 9.2 7.2(3) 24(2) 1.5
A6 >11 7.4(2) 0.9
HF Ptr
. L9 6.6(3) 2.1 8.1(3) 3.6
(helical)
L12 4.6(2) 2.3 6.1(1) 9.0
A6 >11 >14
HFPmn
L9 7.9(2) 0.6 8.4(3) 0.4
(strand)
L12 6.2(2) 6.9 >14
A6 >11 6.9(2) 21(2) 1.0
HFPtr
L9 6.7(2) 6.2 6.8(2) 0.1
(strand)
L12 6.7(1) 14.8 >14

 

1' The uncertainties are provided in the parentheses and were determined by the X2 = xzmin+1
criterion. If the 6values were provided, the fitted values are from X-1 C-X simulations.

177

the two-spin system. The two-spin-system fitting has two variable parameters: (1)
the 13C-X dipolar coupling frequency d which depends on the 13C-X internuclear
distance r as r'3; and (2) a parameter denoted A or B which respectively reflects
the maximal fractional 13C-31P or 13C-19F(C16) dephasing. The logical basis of the
A or B parameters includes: (1) two HFP populations; (2) some 13C0s may be
shielded from any region of the membrane by other peptides. This is most clear
for B-strand HFPs in the interior of a B-sheet; and (3) the 13C-19F dephasing may
not reach 1 because of the 0.09 mol fraction of fluorinated lipids. The three-spin-
system fitting has three variable parameters: (1) the ”ox dipolar coupling
frequency d; (2) the angle 19 between two 13C-X vectors; and (3) the population
parameters A or B as in the 13C-X fittings. The fitting was done for every
individual samples with non-zero 13C-31P or 13C-19F(C16) dephasing curves to
obtain the best-fit population parameters A or B. Table 11 summarizes the best-
fit A and B parameters for Ala-6, Leu-9 and Leu-12. The averaged A or B
parameters were calculated for each construct and each conformation because
the individual values for these residues are close to each other. This probably
indicates that the fraction parameter is independent of the residues that were
labeled in the region of HFP from Ala-6 to Lou-12.

The following changes can be observed by comparing the averaged A
and B parameters: (1) The sum of A and B was close to 1 for HFPmn but less
than 1 for HFPtr, regardless of the secondary structures. This suggests that all

13COs in HFPmn are close to some regions of the membrane, but this may not

178

be true for HFPtr. For the helical HF Ptr, the three helices may be close together

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(a)
HFPmn-helical
1.0 1.0
0 8 “351130319 0 a, Ala6(13c-19F(c16))
0.6 -1 0.6“
O
U)
Q 04 - £04-
0.2 - 0.2%
00 -‘ 00‘ U/
1 1 T 1 T
0 10 20 30 0 10 20
Dephasing time (ms) Dephasing time (ms)
1.0 1.0
Leu9(13C-31P) Leu9(130-19F(016))
0.8 -1 0.8-1
0.6 - 0.6-
O O
(I)
2 0.4 a a 0.44
0.2 .. 02—.
0.0 — n o.o~ o_9..”u/a
l m r V i
0 10 20 30 0 10 20
Dephasing time (ms) Dephasing time (ms)
1.0 1.0
0 Leu12(13C-31P) 0 8 Leu12(13c-19F(016))
.8 4 .
0.6 d D 0.6m
«8 §
<2 ()4 1 < 0.4a
0.2 .1 012—1 B/El/E/‘D
0.0 A El 0.0—1
1 r l u r
0 10 20 30 0 10 20

Dephasing time (ms)

Dephasing time (ms)

Figure 41 (a) Fitting curves for the helical HFPmn in PC:PG.

179

(b)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

HFPtr-helical
1.0
Ala6(13c-19F(c16))
0.8-1
(30.69
N/A 3
‘1 0.4a
0.2-a
0.0-4
T l
0 10 20
Dephasing time (ms)
1.0 1,0
Leu9(13c-31P) Leu9(‘3C-‘9F(C16))
0.8— 0.8-
0.5“ o 0.6m
3
0.4-4 <1 0.4..
024 0'21 D
0-0‘ D 00~ 13 D
T l l I I
0 10 20 30 0 10 20
Dephasing time (ms) Dephasing time (ms)
1.0 1.0
Leu12(13C-31P) Leu12(‘3C-‘9F(016))
0.8— 0.8-1
0.6- 0.6-1
6’
a
0.4—i <1 0,4—
CI
0.2— //3\/fl 0.2-1
CI
0.0—r 0.0-4 CI
l r T i l
o 10 20 30 0 10 20

Dephasing time (ms)

Dephasing tlme (ms)

Figure 41 (b) Fitting curves for the helical HFPtr in PC:PG.

180

 

 

 

(c)
HFPmn-strand

N/A N/A

 

 

 

 

 

 

 

 

 

 

 

 

1.0 1.0
13 -31
Leu1” C P) o... Leu9<13C-19the»
0.6 a 0.6 d
O
(I)
Q o 4 .. €04 4
02 .. 0.2 d V/
0.0 a 0.0 -« .
j T l l l
0 10 20 30 0 10 20
Dephasing time (ms) Dephasing time (ms)
1.0
Leu12(13C-31 P)
0.8 ..
El
0.6 a
O
a 044 N/ A
0.2 ~
0
0.0—
l I 1
0 10 20 30
Dephasing time (ms)

Figure 41 (c) Fitting curves for the strand HFPmn in PC:PG.

181

(d)

HFPtr-strand

 

1.0
0.8-1

0.6-t

(SS/SO

0.2 -

0.04

Ala6(13c-19F(c16))

 

 

 

i 1
10 20

Dephasing time (ms)

 

 

Leu9(13C-31P)

 

 

T T T
1 0 20 30

Dephasing time (ms)

 

Leu12(13C-31P)

 

 

 

i T
10 20 30
Dephasing time (ms)

N/A

 

Leu9(13C-19F(C16))

 

 

 

1
0 10 20

Dephasing time (ms)

N/A

Figure 41 (d) Fitting curves for the strand HFPtr in PC:PG.

182

(e)

 

 

 

1'0 I” 111-helical HFPmn
o 31Leu12
§ 0. 6 7
G0 4 1:
<1 0 2 D .
0.0 i / _- __ _‘_ ___.
0 10 20
Dephasing time (ms)
1'0 B-strand HFPmn
0 8 - Ala6
00.6“
a 0. 4
<1 .
0.2, M
001 1

0 i 10 2o
Dephasing time (ms)

1.0

 

0.87
o 0.61
0.4~
0.2-
0.01

AS/S

 

W

D

tit-helical HFPtr
Leu9

 

 

0

1.0—

0.8“
O 0.6“
0.4“
0.2“
0.0“

AS/S

 

P.,/V”

1'0

1

20
Dephasing time (ms)

B-strand HFPmn
Leu12

 

 

O

10

T

20

Dephasing time (ms)

Figure 41 (e) Plots of (A18/So)’°" (open squares) and (AS/So)” (solid lines) vs. the

dephasing time for 13C

183

19F(C5) experiments.

3

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

300 300
7; 230,1 a-helicalHFPmn Leu12 17g 230. a-helicalHFPtr Leu9
i” 2601 1130-19F1C5» 3 2609 (130'1QF105D
a 240. § 2404
C . 0 t
3 2204 g 220«
O’ i 1
,g 200- E 200«
g 180« '3 1801
Q 1 :3 1
§ 160- 3 160-
% 14041 i 140: .
1204 -- 1204
a 1 0 « -
0.0 02 0.4 06 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0
Fraction parameter F Fraction parameter F
200 200
g 180: B-strand HFPmn AlaG E 1801 B-strand HFPmn Leu12
; 1602 (13C-19F(CS)) I; 1601 (13C-19F(05))
1 > 1
g; 140. § 1404
8 1201 g 120-
E 100. a 100
U) ‘ c 1
E. 80* a 80-
D. t 3 i
8 6°: 8 6°:
‘11 10‘ ‘0 a ‘1‘"
‘ 1
g 201 - a 20+ -
0.0 . 02 - 04 . 016 . 08 71.0 0.0 Y 012 ' 0.4 ' 06 ' 081 1.0
Fraction parameter F Fraction parameter F

Figure 41 (a)-(d) Plot of (AS/So)“ (open squares) and (AS/So)” (solid lines) vs.
the dephasing time for Ala6, Leu9 and Leu12 in HFPmn and HFPtr with tat-helical
and B-strand conformations. The data with (AS/So)” < 0.1 at 32 ms for 13C-31P
and at 24 ms for 13C-19F(016) were not fit and were labeled “N/A”. (e) Plots of
(AS/So)” (open squares) and (AS/So)” (solid lines) vs. the dephasing time for
3C-19F(CS) experiments. (f) Contour plots for the data fittings shown in (e). The
regions with x2 values xzm +1, Xme +2 and xzmm +3 were shown in red, blue and
green respectively, where xzmio is the best-fit root-mean-squared deviation as
given in Table 13 of the main text.

184

and
awe
largl
far 2
HF F
the

para
inse
be

effe
DPI
PO:
par
p0;
ind
We

wl

ar

(16

El

and form a bundle. The 13003 locate at the inner side of the bundle will be far
away from any part of the membrane. For the strand HFPs, the HFPtr may form
larger B-sheet oligomers in which more 1300s are located in the inner strand and
far away from the membrane. (2) The A parameters decrease from HFPmn to
HFPtr independent of the secondary structures, while the 8 parameters remain
the same. According to the logical basis to introduce the A and 8 parameters, the
parameter A will be affected by (1) the change in surface-located and deeply-
inserted population, and (2) the change in shielding effect. The parameter 8 will
be affected by (1) the change in two populations, (2) the change in shielding
effect, and (3) the dilution effect due to the utilization of 9 mol fraction of 19F-
DPPC. The decrease in A parameter indicates an increase in the deeply-inserted
population, or an increase in shielding effect or both. However, since the B
parameter remains the same, there must be an increase in the deeply-inserted
population. The increase of shielding will lead to the decrease in B parameters,
independent of the dilution effect. In other words, if there is no population change,
we would expect to see a decrease in 8 parameters from HFPmn to HFPtr,
which is not consistent with the data in Table 11.

The averaged fraction parameters A and B are used to refine the 13C-3‘P
and 13C-"’F(C16) distances using either a two-spin or a three-spin system as
described previously. Table 12 summarizes the internuclear distances (r), the
inter-vector angles (9) and the best-fit K values. The 6 values are in general
equal or less than 90° rather than close to 180°, which means these residues are

not located in the same longitudinal position as 19F(C16). Figure 41a-d showed

185

all the fitting for the data with (AS/So)” > 0.1 at long dephasing time for the
residues Ala6, Leu9 and Leu12.

The (AS/So)” for the 13C-‘9F(C5) REDOR experiments are provided in
Figure 39b. There are four samples which have (AS/So)” > 0.1 for the 13C-
19F(CS), a-helical HFPmn-L12, ot-helical HFPtr-L9, B-strand HFPmn-A6 and B-
strand HFPmn-L12. These data were fitted to 13C-‘S’F two-spin systems using Eq.
59. There were two fitting parameters: the 13C-19F dipolar coupling frequency (d)
which depends on the 13C-‘S’F distance (r) and the fraction parameter (F) as
explained previously. Table 13 summarizes these best-fit parameters and Figure
41a shows the (AS/So)“ data and fitting curves. Figure 41f shows the contour
plots which reflect the best-fit regions for the four samples. For the ot-helical
HF Pmn-L12 and tat-helical HFPtr-L9 samples, the best-fit regions were restricted
to one single area. The best-fit F for a-helical HFPmn-L12 was 0.16 and it was
roughly half of the parameter B for the ol-helical HFPmn (of. Table 11). The
relation FIB - 0.5 may be explained by the model that the peptide inserts deeply
into one leaflet of the membrane bilayer so that a 13CO is in close proximity with
19F(C16) in both leaflets but with ‘9F(05) only in a single leaflet. This model will
be discussed later in the section with the data of lle4 and Ala14. The best-fit F for
a-helical HFPtr-L9 was 0.24 which was close to both A and B parameters for the
a-helical HFPtr (cf. Table 11). Figure 41f indicated that there were multiple best-
fit regions for the Ala6 and Ala12 in B-strand HFPmn. Thus it will be difficult to
determine whether the dephasings for these residues were contributed by

surface-located or deeply-inserted population. However, the best-fit d values for

186

respectively. All spectra were P

HFPmn

 

 

13031;: 13C-19F(CS) 13C-‘9F(C16)

Ala1 ls WAQQ A0) Q

\s A
lle4 I \ a
\

Alal4 d ’ \ A
—T—l'— —l—_l'_ T 1
190170 190170 190170

130 Chemical Shift (ppm)
flEEtl:
130-31): 130-19F(C5) 13C-19F(016)
\ P
lle4 J N
\

Ala14

_m— —|_'|_ _l__l__

190170 190170 190170
13C Chemical Shift (ppm)

Figure 42 REDOR So and S1 spectra for Ala1, lle4 and Ala14 in HFPmn and
HF Ptr at long dephasing time (r= 32 ms for 13C—3‘P experiments and t = 24 ms
for l3‘C-19F experiments). The So and 81 spectra are shown in black and red
rocessed with 200 Hz Gaussian line broadening
and baseline correction. The 30-3‘P spectra were acquired for ~ 30000 scans
and the 13C-3‘P spectra were acquired for ~ 20000 scans. The arrows indicate

‘30-‘9l=(cs) dephasing for lle4 and Ala14.

187

 

 

HFPmn 13c-31P HFPtr 13c-31P

 

 

(AS/So)9XP
O o O
(.3 L" T
o
\.
\.
\.
.o .O .C' .
<1: 'T’ ‘i‘
8\l>cr‘l3\

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0 10 20 3o 0 20 30
HFPmn 13c-19F(CS) HFPtr13C-19F(C5)
g. 02- 0.24
Q)
C? /8
3 0,0- 32%8/ 00- 9...;8/
0 10 20 0 1‘0 20
HFPmn 13C-19F(C16) HFPtr13C-19F(C16)
& 0.2« 0.2-—
2L
0
(I)
.3; i.e.—e .
v W
0-07 E k; 0.0_ 6
0 1‘0 20 0 1'0 20
Dephasing time (ms) Dephasing time (ms)

Figure 43 Plots of (AS/So)” vs. dephasing time for Ala1(open squares), lle4
(open circles) and Ala14 (open triangles) labeled samples. The dephasing was
calculated by integrating over the entire CO peaks in the corresponding 80 and
S1 spectra.

188

these two residues were 20-50 Hz which corresponded to 8.5-11.2 A 13C-‘QF
distances. Together with the results that the 13C-19F(C16) distances for these two
residues were greater than 14 A (of. Table 12), it appears that they are located

along the longitude between the 3‘P and 19F(CS) positions in the bilayer (cf.

 

 

Figure 40b).
Table 13 Fitting results for the 13C-19F(C5) experiments.
d (Hz) F r(A) x2
ct-helical HFPmn-L12 202(16) 8 016(1) 5.2(2) 1.5
ct-helical HFPtr-L9 125(8) 024(2) 6.1 (1 ) 1.4
B-strand HFPmn-A6 31(2) 027(2) 9.7(2) 0.9
B-strand HFPmn-L12 39(2) 022(2) 9.0(2) 1.1

 

a The uncertainties are shown in parentheses and were determined by the x2 = xzmin'” criterion.

 

The quantitative distance fitting suggested that (1) for both HFPmn and
HFPtr in both secondary structures, there is some deeply-inserted population and
(2) for the deeply-inserted population, the HFPtr inserts into the region of
phospholipid alkyl chains more deeply compared with the HFPmn. For the ot-
helical conformation, the HFPtr has a closer contact with 19F(C16) compared with
the HFPmn (7.4 A vs. 10.4 A for A6, 8.1 A vs. 11.8 A for L9 and 6.1 A vs. 7.2 A
for L12). For the B-strand conformation, the 13CO-‘9F(C16) distance for Ala6 in
HFPmn is greater than 14 A, while in HFPtr it is 6.9 A. For Leu9 the 13CO-
1"F(C16) distance is 8.4 A for HFPmn vs. 6.8 A for HFPtr.

Figure 42 displays the spectra at long dephasing time for Ala1, lle4 and

Ala14 of HFPmn and HFPtr, and Figure 43 displays the plots of (AS/So)” for

these residues by integrating over the entire 13CO peaks. The a-helical and B-

189

ot-hellcal HFPmn_mut a-holical HFPmn

 

Figure 44 Insertion models for HFPmn_mut, HFPmn and HFPtr and different
secondary structures. The lipid headgroups were shown in blue and the alkyl
chains were displayed in gray. For the peptides, the residues from Ala6 to Leu12
were shown in red with definitive secondary structures and the other residues
were displayed in black. The arrows indicated the direction from N to C terminus.

190

strand 13CO peaks were not well-separated in the spectra for lle4 and Ala14, thus
we roughly consider that the 13C-31P and 13C-‘9F(C5) dephasings shown in
Figure 43 for these two residues came from both conformations. The
experimental build up curves provide qualitative information about whether or not
there is close contact between these residues and 3‘Ps or 19Fs. In general, Ala1
has close contact with 31F and has no contact with either 19F(Cb) or 19F(Cte) in
both constructs, which means the N-terminus of the HFPs is always located
close to the phosphate groups. There may be electrostatic attraction between the
positively charged peptide N-terminus and the negatively charged lipid
phosphate groups, which facilitates this close proximity. For lle4 and Ala14, there
seems to be some contact with 31P and 19F(C5), but not with 19F(C16). These
results suggest that both lle4 and Ala14 in some HFPs are located along the
membrane longitudes between positions of 31P and 19F(CS). This will support the
model that the peptide backbone in ol-helical structure is inserted into the outer
leaflet rather than both leaflets. The length of a a-helix from lle4 to Ala14 is about
16 A. The longitudinal distance between 31F and ‘9F(C16) is ~ 22 A, while the
distance between ”F(CS) and ‘9F(C16) is ~ 13 A (of. Fig. 40c). Thus if the helix
inserted into both leaflets and the 13CD of lle4 is between 31P and 19F(CS) of the
outer leaflet, the 1"’00 of Ala14 will be close to the 19F(C16), which is inconsistent
with the data in Figures 42 and 43.

Figure 44 provides semi-quantitative insertion models for the PC:PG
associated HFPmn_mut, HFPmn and HFPtr in the ol-helical and B—strand

conformations. Comparing with the HFPmn_mut which is non-fusogenic, both the

191

HFPmn and HFPtr have some population of insertion into the membrane interior,
and the membrane insertion is observed in both helical and strand structures.
This suggests a positive between the membrane insertion and the fusion activity.
The membrane insertion may be correlated with the disruption of the membrane
bilayers. Our models suggested that in the hydrophobic region of HFP from Ala1
to Gly16, the N-terminal residue Ala1 and the C-terminal residue Ala14 are closer
to the phosphate groups, while the residues in between such as Ala6, Leu9 and
Leu12 are closer to the phospholipid alkyl chains. It is reasonable to think that
such a “V-type” insertion (cf. Figure 44) will induce disruption to the membrane
outer leaflet. Comparing with the HF Pmn, the HF Ptr has a closer contact with the
19F(C16), and may also have a larger deeply-inserted population. Since the
HFPtr was known to have a higher fusion activity than the HFPmn,(14,15) our
results suggest that there is a positive correlation between the fusion activity and
the insertion depths as well as the population of deeply inserted peptide. In the
previous section focused on the PC:PG:CHOL bound HFPmn_mut, HFPmn and
HFPtr, we have proposed that there was a positive correlation between the
fusion activity and the membrane insertion depths for the B—strand HFPs. Thus,
the fusion activity/membrane insertion depth and population correlation is

independent of the membrane composition.

It is interesting to compare the present insertion models for the ot—helical
HFPs in PC:PG with the previous simulation works with a-helical HFPs in a

similar membrane environment. In one simulation which focused on the 16-N-

terminal-residue model HFP and its V2E’ mutant, the wild-type HF P was found to

192

insert obliquely into the bilayer while the mutant was located on the surface.
According to their models, the residues from Ala6 to Leu12 in the wild-type HFP
were located 0 - 5 A below the longitude of the phosphate groups, while for the
V2E mutant, these residues were located around the same position as the
longitude of the phosphate groups.(11) We generally agree with the statement
that the wild-type monomer inserts into the bilayer while the mutant is located on
the surface. However, the best-fit distances for Ala6, Leu9 and Leu12 suggested
that the HFPmn_mut is located 6 -10 A above the phospholipid groups. The
residues Ala6, Leu9 and Leu12 in the surface-located HFPmn are not likely to be
below the phosphate layer because all of the three residues are > 14 A away
from the 19F(CS). For the deeply-inserted population, the residues from Ala6 to
Leu12 seem to be in a longitudinal position between 19F(C5) and 19F(C16)
instead of between 31P and 19F(C5) because Ala6 and Leu9 have contact with
1°F(C16) and Leu12 has contact with both 19F(05) and 19Holes). In another
simulation with the 23-residue HF Pmn and its V2E mutant, it has been proposed
that the wild-type HFP adopted a predominant fully inserted configuration with
the residues Ala6 through Leu12 locating 10-20 A deeper along the membrane
longitude relative to the phosphate group, and the mutant had a predominant
transmembrane configuration with Ala6 and Leu12 located 20-30 A deeper
longitude relative to the phosphate group.(12) Our results do not support the
transmembrane configuration for the HFPmn_mut. A fraction of the HFPmn has a
deep insertion into the region of membrane alkyl chains. However, our results

seem to agree more with the insertion into the outer leaflet where the peptide

193

only contacts the alkyl chains in one leaflet, rather than a fully inserted model

where the peptide has contact with the alkyl chains in both leaflets.
CONCLUSION

In this chapter, we described the studies of the membrane insertion of
three different HFP constructs in two different biologically relevant membranes,
PC:PG which reflects the relative ratio between the phospholipids with neutral
and negatively charged headgroups in the host cells and PC:PG:CHOL which
reflects the ratio of phospholipidszcholesterol in HIV-infected host cells. The
membrane location of these constructs is correlated with their different abilities to
induce vesicle fusion. These studies provide experimentally-based semi-
quantitative models for the membrane locations of the HFPmn_mut, HFPmn and
HFPtr in both ot-helical and B-strand conformations. These models suggested
that the residues from Ala6 to Leu12 had the greatest tendency to insert into the
bilayer interior, regardless of the secondary structure. For the V2E-mutated
HFPmn, the residues from Ala1 to Ala14 are located 6-10 A higher relative to the
longitude of the phosphate groups. Together with the stopped-flow fluorescence
studies, we propose that there is a positive correlation between the fusion activity
and the membrane insertion tendency and depth for the HFPs, i.e. membrane
insertion is a requirement for a HFP construct to become fusogenic and a HFP
construct with greater fusion activity will have greater inserted population with
greater insertion depth. It is reasonable to correlate the population of insertion

and the insertion depth to the membrane disruption induced by the HFP. Thus,

194

our results may further suggest a positive correlation between the fusion activity

of HFPs and their abilities to disrupt the membrane bilayer.

195

REFERENCE
1. Staffs, N. a. E., New Hope in HIV Disease. Science 1996, 274, 1988-1991.

2. Weissenhorn, W.; Dessen, A.; Harrison, S. C.; Skehel, J. J.; Wlley, D. 0.,
Atomic structure of the ectodomain from HIV-1 gp41. Nature 1997, 387, (6631),
426-430.

3. Jaroniec, C. P.; Kaufman, J. D.; Stahl, S. J.; Viard, M.; Blumenthal, R.;
Vanfield, P. T.; Bax, A., Structure and dynamics of micelle-associated human
immunodeficiency virus gp41 fusion domain. Biochemistry 2005,44, (49), 16167-
16180.

4. Chang, D. K.; Cheng, S. F.; Chien, W. J., The amino-terminal fusion
domain peptide of human immunodeficiency virus type 1 gp41 inserts into the
sodium dodecyl sulfate micelle primarily as a helix with a conserved glycine at
the micelle-water interface. J. Virol. 1997, 71, (9), 6593-6602.

5. Morris, K. F.; Gao, X. F.; Wong, T. C., The interactions of the HIV gp41
fusion peptides with zwitterionic membrane mimics determined by NMR
spectroscopy. Biochim. Biophys. Acta-Biomembr. 2004, 1667, (1), 67-81.

6. Li, Y. L., AND Tamm, L.K., Structure and Plasticity of the Human
Immunodeficiency Virus gp41 Fusion Domain in Lipid Micelles and Bilayers.
Biophys. J. 2007, 93, (3), 876-885.

7. Gabrys, C. M., and Weliky, D.P., Chemical Shift Assignment and
Structural Plasticity of a HIV Fusion Peptide Derivative in
Dodecylphosphocholine Micelles. BBA-Biomembranes 2007, 1768, 3225-3234.

8. Agirre, A.; Flach, C.; Goni, F. M.; Mendelsohn, R.; Valpuesta, J. M.; Wu, F.
J.; Nieva, J. L., Interactions of the HIV-1 fusion peptide with large unilamellar
vesicles and monolayers. A cryo-TEM and spectroscopic study. Biochimica Et
Biophysica Acta-Biomembranes 2000, 1467, (1 ), 153-164.

9. Haque, M. E.; Koppaka, V.; Axelsen, P. H.; Lentz, B. R., Properties and
structures of the influenza and HIV fusion peptides on lipid membranes:
Implications for a role in fusion. Biophys. J. 2005, 89, (5), 3183-3194.

10. Gordon, L. M.; Curtain, C. C.; Zhong, Y. C.; Kirkpatrick, A.; Mobley, P. W.;
Waring, A. J., The amino-terminal peptide of HIV-1 glycoprotein 41 interacts with
human erythrocyte membranes: peptide conformation, orientation and
aggregation. Biochim. Biophys. Acta 1992, 1139, (4), 257-274.

196

11. Kamath, S.; Wong, T. C., Membrane structure of the human
immunodeficiency virus gp41 fusion domain by molecular dynamics simulation.
Biophys. J. 2002,83, (1), 135-143.

12. Maddox, M. W.; Longo, M. L., Conformational partitioning of the fusion
peptide of HIV-1 gp41 and its structural analogs in bilayer membranes. Biophys.
J. 2002, 83, (6), 3088-3096.

13. Kliger, Y.; Aharoni, A.; Rapaport, D.; Jones, P.; Blumenthal, R.; Shai, Y.,
Fusion peptides derived from the HIV type 1 glycoprotein 41 associate within
phospholipid membranes and inhibit cell-cell Fusion. Structure- function study. J.
Biol. Chem. 1997, 272, (21), 13496-13505.

14. Yang, R.; Prorok, M.; Castellino, F. J.; Weliky, D. P., A trimeric HIV-1
fusion peptide construct which does not self-associate in aqueous solution and
which has 15-fold higher membrane fusion rate. J. Am. Chem. Soc. 2004, 126,
(45), 14722-14723.

15. Qiang, W., and Weilky, D.P., HIV Fusion Peptide and Its Cross-Linked
Oligomers: Efficient Syntheses, Significance of the Trimer in Fusion Activity,
Correlation of 6 Strand Conformation with Membrane Cholesterol, and Proximity
to Lipid Headgroups. Biochemistry 2009, 48, (2), 289-301.

16. Caffrey, M.; Cai, M.; Kaufman, J.; Stahl, S. J.; Wlngfield, P. T.; Covell, D.
G.; Gronenbom, A. M.; Clore, G. M., Three-dimensional solution structure of the
44 kDa ectodomain of SIV gp41. EMBO J. 1998, 17, (16), 4572-4584.

17. Qiang, W., Yang, J., and Weliky, D.P., Solid-state nuclear magnetic
resonance measurements of HIV fusion peptide to lipid distances reveal the
intimate contact of beta strand peptide with membranes and the proximity of the
Ala-14-Gly-16 region with lipid headgroups. Biochemistry 2007, 46, (17), 4997-
5008.

18. Qiang, W., Sun, Y., and Weliky, D.P., A strong correlation between
fusogenicity and membrane insertion depth of the HIV fusion peptide. submitted
to Proc. Natl. Acad. Sci. USA.

19. Qiang, W., and Weilky, D.P., unpublished experiments.

20. Chan, W. C., and White, P.D., Fmoc Solid Phase Peptide Synthesis: A
Practical Approach 2000, 94-109.

21. Yang, R.; Yang, J.; Weliky, D. P., Synthesis, enhanced fusogenicity, and

solid state NMR measurements of cross-linked HIV-1 fusion peptides.
Biochemistry 2003, 42, (12), 3527-3535.

197

22. Struck, D. K.; Hoekstra, D.; Pagano, R. E., Use of resonance energy
transfer to monitor membrane fusion. Biochemistry 1981 , 20, (14), 4093-9.

23. Aloia, R. C.; Tian, H.; Jensen, F. C., Lipid composition and fluidity of the
human immunodeficiency virus envelope and host cell plasma membranes. Proc.
Natl. Acad. Sci. USA. 1993, 90, (11), 5181-5185.

24. Brugger, 8.; Glass, 8.; Haberkant, P.; Leibrecht, I.; Vlfieland, F. T.;
Krasslich, H. G., The HIV lipidome: A raft with an unusual composition. Proc. Natl.
Acad. Sci. U. SA. 2006, 103, (8), 2641-2646.

25. Venable, R. M., Brooks, B.R., and Pastor, R.W., Molecular dynamics

simulations of gel (L0,) phase lipid bilayers in constant pressure and constant
surface area ensembles. J. Chem. Phys 2000, 112, 4822-4832.

26. McDonough, B.; Macdonald, P. M.; Sykes, B. D.; McElhaney, R. N., F-19
Nuclear Magnetic-Resonance Studies of Lipid Fatty Acyl Chain Order and
Dynamics in Acholeplasma-Laidlawii B Membranes - a Physical, Biochemical,
and Biological Evaluation of Monofluoropalmitic Acids As Membrane Probes.
Biochemistry 1983, 22, (22), 5097-5103.

27. Birdsall, N. J., Lee, A.G., Levine, Y.K., and Metcalfe, J.C., 19F NMR of
monofluorostearic acids in lecithin vesicles. Biochim. Biophys. Acta. 1971, 241,
693-696.

28. Sibi, M. P., Rutherford, D., and Shanna, R., A new electrophilic alaninol
synthon. A general route to oxazolidinones of D or (R)-2-amino alcohols form L-
serine. J. Chem. Soc. Perkin. Trans. 1994, 1, 1675-1678.

29. Jackson, R. F. W., and Perez-Gonzalez, M., Synthesis of N-(tert-
butoxycarbonyl)-beta-iodoalanine methyl ester: A useful building block in the
synthesis of nonnatural alpha-amino acids via palladium catalyzed cross coupling
reaactions. Org. Syntheses 2005, 81, 77-81.

30. Lang, L. X., Jagoda, E., Ma, Y., Sassaman, MB, and Eckelman, W.C.,
Synthesis and in vivo biodistribution of F-18 labeled 3-cis-, 3-trans-, 4-cis-, and 4-
trans-fluorocyclohexane derivatives of WAY 100635. Bioorganic & Medicinal
Chemistry 2006, 14, 3737-3748.

31. Hirsh, D. J., Lazaro, N., Wright, L.R., Boggs, J.M., McIntosh, T.J.,
Schaefer, J., and Blazyk, J., A new monofluorinated phosphatidylcholine forms
interdigitated bilayers. Biophys. J. 1998, 75, (4), 1858-1868.

32. Mueller, K. T., Analytical Solutions for the Time Evolution of Dipolar-

dephasing NMR Signals. Journal of Magnetic Resonance Series A 1995, 113, (1),
81-93.

198

33. Zhang, H. Y.; Neal, 8.; Wlshart, D. 8., RefDB: A database of uniformly
referenced protein chemical shifts. J. Biomol. NMR 2003, 25, (3), 173-195.

34. Yang, J.; Gabrys, C. M.; Weliky, D. P., Solid-state nuclear magnetic
resonance evidence for an extended beta strand conformation of the membrane-
bound HIV-1 fusion peptide. Biochemistry 2001, 40, (27), 8126-8137.

35. Bodner, M. L., Solid state nuclear magnetic resonance of the HIV-1 and
influenza fusion peptides associated with membranes. Ph. D. thesis, Michigan
State University: East Lansing, 2006; p 122.

36. Zheng, Z., Yang, R., Bodner, M.L., and Weliky, D.P., Conformational
flexibility and strand arrantments of the membrane-associated HIV fusion peptide
trimer probed by solid-state NMR spectroscopy. Biochemistry 2006, 45, 12960-
12975.

37. Yang, J.; Parkanzky, P. D.; Khunte, B. A.; Canlas, C. G.; Yang, R.; Gabrys,
C. M.; Weliky, D. P., Solid state NMR measurements of conformation and
conformational distributions in the membrane-bound HIV-1 fusion peptide. J. Mol.
Graph. Model. 2001, 19, (1), 129-135.

38. Wasniewski, C. M.; Parkanzky, P. D.; Bodner, M. L.; Weliky, D. P., Solid-
state nuclear magnetic resonance studies of HIV and influenza fusion peptide
orientations in membrane bilayers using stacked glass plate samples. Chem.
Phys. Lipids 2004, 132, (1 ), 89-100.

39. Yang, J.; Prorok, M.; Castellino, F. J.; Weliky, D. P., Oligomeric beta
structure of the membrane-bound HIV-1 fusion peptide formed from soluble
monomers. Biophys. J. 2004, 87, 1951-1963.

40. Bodner, M. L.; Gabrys, C. M.; Parkanzky, P. D.; Yang, J.; Duskin, C. A.;
Weliky, D. P., Temperature dependence and resonance assignment of 13'C NMR
spectra of selectively and uniformly labeled fusion peptides associated with
membranes. Magn. Reson. Chem. 2004, 42, 187-194.

41. Zheng, Z., Qiang, W., and Weliky, D.P., Investigation of finite-pulse
radiofrequency-driven recoupling methods for measurement of intercarbonyl
distances in polycrystalline and membrane-associated HIV fusion peptide
samples. Magn. Reson. Chem. 2007, 245, 8247-8260.

42. Bevington, P. R.; Robinson, D. K., Data Reduction and Error Analysis for
the Physical Sciences. 2nd ed.; McGraw-Hill: Boston, 1992; p 38-52.

43. Qiang, W., Bodner, M.L., and Weilky, D.P., Solid-state NMR Spectroscopy

of HIV Fusion Peptides Associated with Host-CeII-Like Membranes: ZD
Correlation Spectra and Distance Measurements Support a Fully Extended

199

Conformation and Models for Specific Antiparallel Strand Registries. J. Am.
Chem. Soc. 2008, 130, 5459-5471.

44. Yang, J.; Weliky, D. P., Solid state nuclear magnetic resonance evidence
for parallel and antiparallel strand arrangements in the membrane-associated
HIV-1 fusion peptide. Biochemistry 2003, 42, 11879-11890.

45. Gordon, L. M.; Mobley, P. W.; Pilpa, R.; Sherman, M. A.; Waring, A. J.,
Conformational mapping of the N-terrninal peptide of HIV-1 gp41 in membrane
environments using 13C-enhanced Fourier transform infrared spectroscopy.
Biochim. Biophys. Acta-Biomembr. 2002, 1559, (2), 96-120.

46. Castano, 8.; Desbat, 8., Structure and orientation study of fusion peptide
FP23 of gp41 from HIV-1 alone or inserted into various lipid membrane models
(mono-, bi- and multibi-layers) by FT-IR spectroscopies and Brewster angle
microscopy. Biochim. Biophys. Acta-Biomembr. 2005, 1715, (2), 81-95.

47. Sackett, K.; Shai, Y., The HIV fusion peptide adopts intermolecular parallel
B-sheet structure in membranes when stabilized by the adjacent N-terrninal
heptad repeat: A 13C FTIR study. J. Mol. Biol. 2005, 350, (4), 790-805.

48. Chan, D. C.; Fass, D.; Berger, J. M.; Kim, P. 8., Core structure of gp41
from the HIV envelope glycoprotein. Cell 1997, 89, (2), 263-273.

49. Tan, K.; Liu, J.; Wang, J.; Shen, 8.; Lu, M., Atomic structure of a
thennostable subdomain of HIV-1 gp41. Proc. Natl. Acad. Sci. USA. 1997, 94,
(23),12303-12308.

50. Yang, Z. N.; Mueser, T. C.; Kaufman, J.; Stahl, S. J.; V\fingfield, P. T.;
Hyde, C. C., The crystal structure of the SN gp41 ectodomain at 1.47 A
resolution. J. Struct. Biol. 1999, 126, (2), 131-144.

51. Macosko, J. C.; Kim, C. H.; Shin, Y. K., The membrane topology of the
fusion peptide region of influenza hemagglutinin determined by'spin-Iabeling
EPR. J. Mol. Biol. 1997, 267, (5), 1139-1148.

52. Han, X.; Bushweller, J. H.; Cafiso, D. S.; Tamm, L. K., Membrane
structure and fusion-triggering conformational change of the fusion domain from
influenza hemagglutinin. Nat. Struct. Biol. 2001, 8, (8), 715-720.

53. Modis, Y.; Ogata, S.; Clements, 0.; Harrison, S. 0., Structure of the
dengue virus envelope protein after membrane fusion. Nature 2004, 427, (6972),
313-319.

54. Gibbons, D. L.; Vaney, M. C.; Roussel, A.; Vigouroux, A.; Reilly, B.;
Lepault, J.; Kielian, M.; Rey, F. A., Conformational change and protein protein

200

interactions of the fusion protein of Semliki Forest virus. Nature 2004, 427,
(6972), 320-325.

55. Roche, S., Bressanelli, S., Rey, F.A., and Gaudin, Y., Crystal Structure of
the low-pH form of the vesicular stomatitis virus glycoprotein G. Science 2006,
313, 187-191.

56. Heldwein, E. E., Lou, H., Bender, F.C., Cohen, G.H., Eisenberg, R.J., and
Harrison, 8.0., Crystal Structure of glycoprotein B from herpes simplex virus 1.
Science 2006, 313, 217-220.

57. Hessa, T., Kim, H., Bihlmaier, K., Lundin, C., Boekel, J., Andersson, H.,
Nilsson, L, White, SH, and Heijne, G., Recognition of tranmembrane helices by
the endoplasmic reticulum translocon. Nature 2005, 433, 377-381.

58. Toke, O.; Maloy, W. L.; Kim, S. J.; Blazyk, J.; Schaefer, J., Secondary
structure and lipid contact of a peptide antibiotic in phospholipid Bilayers by
REDOR. Biophys. J. 2004, 87, (1), 662-674.

59. Charloteaux, B., Lorin, A., Crowet, J.M., Stroobant, V., Lins, L., Thomas,
A., and Brasseur, R., The N-terrninal 12 residue long peptide of HIV gp41 is the
minimal peptide sufficient to induce significant T-cell-like membrane
destablization in vitro. J. Mol. Biol. 2006, 359, 597-609.

60. Chang, D. K.; Cheng, S. F., Determination of the equilibrium micelle-
inserting position of the fusion peptide of gp41 of human immunodeficiency virus
type 1 at amino acid resolution by exchange broadening of amide proton
resonances. J. Biomol. NMR1998, 12, (4), 549-552.

61. Langham, A.; Kaznessis, Y., Simulation of the N-terminus of HIV-1
glycoprotein 41000 fusion peptide in micelles. J. Pept. Sci. 2005, 11, (4), 215-224.

62. Bloom, M.; Evans, E.; Mouritsen, O. 6., Physical properties of the fluid
lipid-bilayer component of cell membranes: a perspective. Quat. Rev. Biophys.
1991, 24, (3), 293-397.

63. Nir, S., and Nieva, J.L., Interaction of peptides with liposomes: pore
formation and fusion. Prog. Lipid Res. 2000, 39, 181-206.

64. White, J. M., Delos, S.E., Brecher, M., and Schornberg, K., Structures and
mechanisms of viral membrane fusion proteins: Multiple variations on a common
theme. Crit. Rev. Biochem. Mol. Biol. 2008,43, 189-219.

65. Freed, E. O.; Delwart, E. L.; Buchschacher, G. L., Jr.; Panganiban, A. T.,
A mutation in the human immunodeficiency virus type 1 transmembrane

201

glycoprotein gp41 dominantly interferes with fusion and infectivity. Proc. Natl.
Acad. Sci. U. SA. 1992, 89, (1), 70-74.

66. Jaroniec, C. P.; Tounge, B. A.; Herzfeld, J.; Griffin, R. G., Frequency
selective heteronuclear dipolar recoupling in rotating solids: Accurate 1 C-‘5N
distance measurements in uniformly 13C, 1 N-Iabeled peptides. J. Am. Chem. Soc.
2001, 123, (15), 3507-3519.

67. Nieva, J. L.; Nir, S.; Muga, A.; Goni, F. M.; IMlschut, J., Interaction of the
HIV-1 fusion peptide with phospholipid vesicles: different structural requirements
for fusion and leakage. Biochemistry 1994, 33, (11), 3201 -3209.

68. Reichert, J., Grasnick, D., Afonin, S., Buerck, J., Wadhwani, P., and Ulrich,
AS, A critical evaluation of the conformational requirements of fusogenic
peptides in membranes. European Biophysics Journal with Biophysics Letters
2007, 36, (4-5), 405-413.

69. Sackett, K., and Weliky, D.P., Unpublished experiments.

202

CHAPTER VII
SUMMARY AND FUTURE WORK

The overall goal of this project is to understand the HIV-induced
membrane fusion mechanism from the atomic-resolution structural point of view.
The HIV fusion peptide was used as a model peptide for the entire fusion protein
gp41 to provide insights into structure and membrane insertion. My specific goals
included the measurements of structure for the membrane associated HIV fusion
peptides and the detection of the HFP-membrane interaction. A summary of the
conclusions for my studies: (1) The HFP trimer has been considered as a
biologically relevant construct for HF Ps based on the structure of the soluble part
of gp41, and my study provided a feasible approach to chemically synthesize the
HFPtr. This improved synthetic scheme utilized the cross-linking of C-terminal
_ lysines and cysteines to buildup a trimeric scaffold. With the optimized
experimental conditions described in Chapter II, the overall yield has been
increased by at least a factor of three and the purity has also been greatly
improved.(1,2) In addition, the method to buildup the trimeric scaffold should be
applicable to the synthesis of other hetero- or homo-trimers. (2) The structures of
membrane-associated HFPs have been studied using solid-state NMR methods,
in particular, the secondary structures of HFPmn, HFPdm and HFPtr in both
cholesterol-containing and non-cholesterol-containing membranes, and the
tertiary structure of HFPmn in cholesterol-containing membrane. The population

of B-strand conformations is correlated with the membrane cholesterol. There

203

was not an obvious dependence of secondary structure on cross—linking. Since
there had been shown a positive correlation between the extent of cross-linking
and the fusion activity of different HFP constructs,( 1) it did not seem that the
fusion activity can be correlated with the secondary structure. The tertiary
structure of HFPmn in cholesterol-containing membrane has been studied using
13C-15N REDOR experiments, and the results supported two specific anti-parallel
B-sheet registries. (3) The membrane location of HFPs has also been studied
mainly with 13C-3‘P and 13C-‘S’F REDOR methods. A systematic distance
measurement method has been developed to detect the distances between the
backbones of a series of singly-”CO labeled peptides and different positions in
lipid bilayers such as 31P and 19F in the fluorinated phospholipids 5-‘9F-DPPC
and 16—19F-DPPC. The studies concluded that HFPs with both ol-helical and B-
strand conformations can insert into the membrane bilayer interior. The N-
terminus was always located close to the lipid phosphate groups while the
hydrophobic middle regions of HFP, e.g. Ala6 to Leu12 are more likely to be
located close to the phospholipid alkyl chains in a single leaflet. The cross-linked
HFPs such as HFPtr inserted more deeply and may have larger deeply-inserted
population compared with the HF Pmn.

It is important to have local membrane disruption to induce membrane
fusion.(3) Our model proposed in chapter Vl suggested that the membrane
disruption can be caused by the insertion of HFPs into the phospholipid alkyl
chains of the membrane outer leaflet. There is a positive correlation between the

membrane insertion depth and population and the ability to fuse the vesicles for

204

different HFP constructs, and the correlation is independent of the secondary
structure of HFPs. From the methodology point of view, the systematic
measurement of membrane location of the peptide backbone 13003 provides
atomic resolution information about membrane insertion of the HIV fusion
peptides. The method should be in principle applicable to other membrane
associated peptides where commonly used structural characterization methods
such as solution NMR and X-ray crystallography are not applicable. However,
this method may be better applied to a small B—sheet oligomeric peptide/protein
system than to a larger B-sheet oligomeric peptide/protein system because in the
latter case most of the labeled 13Cs will be located in the center of the B—sheet
and will be far away from any region of the membrane.

The experimental results provide insights for understanding the HIV-
induced membrane fusion mechanism. It has been proposed previously that both
the assembly of HFPs and the membrane disruption caused by HFPs could be
pre-requisites for membrane fusion.(3) Overall, my research contributes to the
fusion mechanism by studying whether the structure and membrane insertion
could potentially be important to affect the fusion activity of the HFP.

First of all, the secondary structure does not seem to be a crucial factor to
affect the fusion activity because there is not an obvious correlation between the
secondary structures and the fusion activities for HFPs with different extents of
cross-linking.

Second, it had been proposed that the V2E mutant was non-fusogenic

because the substitution of Val2 to Glu2 prevented the formation of fusion active

205

assembly of HFPs.(4) My work for the first time demonstrated that HFPmn in
host-cell like membrane environments will adopt anti-parallel B—sheet structure
with at least two registries. In addition, the work provided a general approach to
detect the registries for B—sheet structures. The tertiary structure may be
important for HFP to induce fusion. It will be interesting to investigate whether
there are certain registries associated with the non-fusogenic HFPmn_mut and
whether there are some other assembly patterns associated with the highly-
fusogenic HFPtr.

Third, the membrane insertion studies clearly showed that there is a
positive correlation between the membrane insertion depths of different HFP
constructs and the fusion activities of these HFPs. There may also be a positive
correlation between the deeply-inserted population and the fusion activity of
these HFPs. VWth the reasonable hypothesis that the insertion would induce
membrane disruption, our results may suggest that there is a positive correlation
between the ability to disrupt the membrane and the ability of fuse the
membranes for different HFP constructs. The distance measurements also
demonstrated that the most hydrophobic region of HFPs insert into the
membrane, i.e. starting from around Ala6 and ending at around Leu12.

The ultimate goal of the project is to understand the fusion mechanism
induced by the gp41 from the structural point of view. Large proteins can be
expressed with amino acid-type 13co and 15N labelings.(5) This will bring up a
problem if one would like to apply the 13c-“”l=> or 13c-‘9l= REDOR distance

measurements to the membrane-associated protein systems. Since the isotope

206

labels are on all residues of a particular amino acid type rather than on a singly
labeled residue, the ‘30 signals from all labeled residues will be generated during
the cross polarization (CP) period. In order to solve the problem, the regular CP
has to be replaced by a double cross polarization (DCP) sequence where the 13C
magnetization is generated through the route 1H —r 15N —> 13C.(6) Consequently,
the only observable ‘30 label at the end of DCP period will be the one with an
adjacent 15N label. This modification will help to filter out the natural abundance
13C and the other isotopically labeled residues. The DCP building block can then
be incorporated into other distance measurement pulse sequences such as 13C-
31P REDOR and the modified pulse sequence can be used for the protein
samples.

Another interesting future direction will be to apply the 13C-15N REDOR
methods to other HFP constructs such as HFPtr or HFPmn_mut to study the
possible tertiary structures such as anti-parallel (Bi-sheet registries or parallel 6-
sheet registries. The predominant tertiary structure for these constructs can then
be compared with the predominant 16 or 17-residue-overlapped anti-parallel [3-
sheet for HFPmn, and the results may provide insight on whether there is a
correlation between the tertiary structure and the fusion activity for different HFP

constructs.

207

REFERENCE

1. Yang, R.; Prorok, M.; Castellino, F. J.; Weliky, D. P., A trimeric HIV-1
fusion peptide construct which does not self-associate in aqueous solution and
which has 15-fold higher membrane fusion rate. J. Am. Chem. Soc. 2004, 126,
(45), 14722-14723.

2. Yang, R.; Yang, J.; Weliky, D. P., Synthesis, enhanced fusogenicity, and
solid state NMR measurements of cross-linked HIV-1 fusion peptides.
Biochemistry 2003, 42, (12), 3527-3535.

3. Nir, S., and Nieva, J.L., Interaction of peptides with liposomes: pore
formation and fusion. Prog. Lipid Res. 2000, 39, 181-206.

4. Freed, E. O.; Delwart, E. L.; Buchschacher, G. L., Jr.; Panganiban, A. T.,
A mutation in the human immunodeficiency virus type 1 transmembrane
glycoprotein gp41 dominantly interferes with fusion and infectivity. Proc. Natl.
Acad. Sci. U. S.A. 1992,89, (1), 70-74.

5. Curtis-Fisk, J., Spencer, RM, and Weliky, D.P., Isotopically Labeled
Expression in E. coli, Purification, and Refolding of the Full Ectodomain of the
Influenza Virus Membrane Fusion Protein. Protein Expression and Purification,
2008, 61, 212-219.

6. Baldus, M.; Petkova, A. T.; Herzfeld, J.; Griffin, R. G., Cross polarization in

the tilted frame: assignment and spectral simplification ln heteronuclear spin
systems. Molecular Physics 1998, 95, (6), 1197-1207.

208

APPENDIX 1. NMR FILES CHECKLIST

Figure 14 2D PDSD experiments @ ..lhome/wei4bldata/2D-PDSD/
HFPmn in PC:PG (HFPmn-PCPG-50ms)
HFPdm in PC:PG (HFPdm-PCPG-50ms)
HFPtr in PC:PG (HF Ptr-PCPG-50ms)
HFPmn in PC:PG:CHOL (HFPmn-PCPGCHOL-50ms)
HFPdm in PC:PG:CHOL (HFPdm-PCPGCHOL-SOms)
HFPtr in PC:PG:CHOL (HFPtr-PCPGCHOL-SOms)

Figure 18 10 ‘30-‘5N REDOR experiments @../homelwei4b/data/13C-15N-
REDOR
A14(‘3C)-V2(15N) HFPmn in PC:PG:CHOL (130-15N-A14-V2)
A14(130)-G3(15N) HFPmn in PC:PG:CHOL (13C-15N-A14-G3-new)
A14(13C)-l4(15N) HFPmn in PC:PG:CHOL (13C-15N-A14-I4)
A14(‘3C)-G5(‘5N) HFPmn in PC:PG:CHOL (13C-15N-A14-G5-new)

10 ‘30-3‘P REDOR experiments @../home/wei4bldata/1SC-31 P
Figure 18 A1V2G3(‘3C) HFPmn in PC:PG (FPmnK3-A1VZG3)
A1VZG3(1SC) HFPmn in PC:PG:CHOL (FPmnK3-A1V2G3-CHOL)
Figures 26 and 27 GSA6L7(‘3C) HFPmn in PC:PG (FPmnK3-G5A6L7-PCPG)
' G5A6L7(130) HFPmn in PC:PG:CHOL (FPmnK3-G5A6L7-chol)
F8L9G10(130) HFPmn in PC:PG (FPmn-F8LQG10-PCPG)
F8L9G10(‘3C) HFPmn in PC:PG:CHOL (FPmn-F8L9G10-chol)

209

F11L12G13(13C) HFPmn in PC:PG (FPmnK3-F11L12G13-PCPG)
F11L12613(‘3C) HFPmn in PC:PG:CHOL (FPmnK3-F11L12G13-chol)
A14A15G16(‘3C) HFPmn in PC:PG (FPmnK3-A14A15G16-PCPG)
A14A15G16(‘3C) HFPmn in PC:PG:CHOL (FPmnK3-A14A15G16-chol)

The following spectra were displayed in Figures 30, 33, 35, 38 and 41 with

detailed

names described in the corresponding figure captions.

A1(”C) HFPmn in PC:PG (HFPmn-A1-PCPG)

A1(‘3C) HFPmn in PC:PG:CHOL (HFPmn-A1-chol)

14(130) HFPmn in PC:PG (HFPmn-I4-PCPG)

l4(‘3C) HFPmn in PC:PG:CHOL (HFPmn-I4-chol)

A6(‘3C) HFPmn in PC:PG (HFPmn-A6-PCPG)

A6(130)HFPmn in PC:PG:CHOL (HFPmn-A6-chol)

L9(13C) HFPmn in PC:PG (HFPmn-L9-PCPG)

L9(‘3C) HFPmn in PC:PG:CHOL (HFPmn-L9-chol)

1.12030) HFPmn in PC:PG (HFPmn-L12-PCPG)

L12(13C) HFPmn in PC:PG:CHOL (HFPmn-L12-chol)

A14(‘3C) HFPmn in PC:PG (HFPmn-A14-PCPG)

A14(13C) HFPmn in PC:PG:CHOL (HFPmn-A14-chol)

A15(13C) HFPmn in PC:PG (HFPmn-A15—PCPG)

A15(13C) HFPmn in PC:PG:CHOL (HFPmn-A15-chol)

A1(13C) HFPmn_mut in PC:PG (HFPmnmut-A1-PCPG)

A1(‘3C) HFPmn_mut in PC:PG:CHOL (HFPmnmut—A1-chol)

2l0

l4(13C) HFPmn_mut in PC:PG (HFPmnmut-l4-PCPG)
l4(‘30) HFPmn_mut in PC:PG:CHOL (HFPmnmut-l4-chol)
A6(13C) HFPmn_mut in PC:PG (HFPmnmut-A6-PCPG)
A6(‘3C) HFPmn_mut in PC:PG:CHOL (HFPmnmut-A6-chol)
L9(‘3C) HFPmn_mut in PC:PG (HFPmnmut-L9-PCPG)
L9(1aC) HFPmn_mut in PC:PG:CHOL (HFPmnmut-L9-chol)
L12(13C) HFPmn_mut in PC:PG (HFPmnmut-L12-PCPG)
L12(13C) HFPmn_mut in PC:PG:CHOL (HFPmnmut-L12-chol)
A14(13C) HFPmn_mut in PC:PG (HFPmnmut—A14-PCPG)
A14(13C) HF Pmn_mut in PC:PG:CHOL (HFPmnmut-A14-chol)
A1(13C) HFPtr in PC:PG (HFPtr-A1-PCPG)

A1(13C) HFPtr in PC:PG:CHOL (HFPtr-A1-chol)

14(130) HFPtr in PC:PG (HFPtr-l4-PCPG)

14(130) HFPtr in PC:PG:CHOL (HFPtr-l4-chol)

A6(130) HFPtr in PC:PG (HFPtr-A6-PCPG)

A6(13C) HFPtr in PC:PG:CHOL (HFPtr-A6-chol)

L9(‘3C) HFPtr in PC:PG (HFPtr-L9-PCPG)

L9(‘3C) HFPtr in PC:PG:CHOL (HFPtr-L9-chol)

L12(‘3C) HFPtr in PC:PG (HFPtr-L12-PCPG)

L12(13C) HFPtr in PC:PG:CHOL (HFPtr-L12-chol)

A14(13C) HFPtr in PC:PG (HFPtr-A14-PCPG)

A14(‘3C) HFPtr in PC:PG:CHOL (HFPtr-A14-chol)

211

1D ‘3C-19F(C16) REDOR experiments @../home/wei4b/data/13C-19F(C16)
A1(‘3C) HFPmn in PC:PG (HFPmn-A1-PCPG)

A1(‘3C) HFPmn in PC:PG:CHOL (HFPmn-A1-chol)

l4(‘30) HFPmn in PC:PG (HFPmn-l4-PCPG)

l4(‘3C) HFPmn in PC:PG:CHOL (HFPmn-l4-chol)

A6(13C) HFPmn in PC:PG (HFPmn-A6-PCPG)

A6(‘3C) HFPmn in PC:PG:CHOL (HFPmn-A6-chol)

L9(‘3C) HFPmn in PC:PG (HFPmn-L9-PCPG)

L9(‘3C) HFPmn in PC:PG:CHOL (HFPmn-L9-chol)
L12(‘3C) HFPmn in PC:PG (HFPmn-L12-PCPG)

L12(‘3C) HFPmn in PC:PG:CHOL (HFPmn-L12-chol)
A14(13C) HFPmn in PC:PG (HFPmn-A14-PCPG)

A14(‘3C) HFPmn in PC:PG:CHOL (HFPmn-A14-chol)
A1030) HFPmn_mut in PC:PG (HFPmnmut-A1-PCPG)
A1(‘3C) HFPmn_mut in PC:PG:CHOL (HFPmnmut-A1-chol)
14(13C) HFPmn_mut in PC:PG (HFPmnmut-l4-PCPG)
I4(‘3C) HFPmn_mut in PC:PG:CHOL (HFPmnmut—l4-chol)
A6(‘3C) HFPmn_mut in PC:PG (HFPmnmut—A6-PCPG)
A6(13C) HFPmn_mut in PC:PG:CHOL (HFPmnmut-A6-chol)
L9(‘3C) HFPmn_mut in PC:PG (HF Pmnmut—L9-PCPG)
L9(‘3C) HFPmn_mut in PC:PG:CHOL (HFPmnmut—L9-chol)
1.12030) HFPmn_mut in PC:PG (HFPmnmut-L12-PCPG)
L12(‘3C) HFPmn_mut in PC:PG:CHOL (HFPmnmut—L12-chol)

212

A14(‘3C) HFPmn_mut in PC:PG (HFPmnmut—A14-PCPG)
A14(‘3C) HFPmn_mut in PC:PG:CHOL (HFPmnmut—A14-chol)
A1(‘3C) HFPtr in PC:PG (HFPtr-A1-PCPG)

A1('3C) HFPtr in PC:PG:CHOL (HFPtr-A1-chol)

l4(‘3C) HFPtr in PC:PG (HFPtr-l4-PCPG)

I4(13C) HFPtr in PC:PG:CHOL (HFPtr-l4-chol)

A6(‘3C) HFPtr in PC:PG (HFPtr-A6-PCPG)

A6030) HFPtr in PC:PG:CHOL (HFPtr-A6-chol)

1.9630) HFPtr in PC:PG (HFPtr-L9-PCPG)

L9(‘3C) HFPtr in PC:PG:CHOL (HFPtr-L9-chol)

L12(‘3C) HFPtr in PC:PG (HFPtr-L12-PCPG)

L12(‘3C) HFPtr in PC:PG:CHOL (HFPtr-L12-chol)
A14(‘3C) HFPtr in PC:PG (HFPtr-A14-PCPG)

A14(‘3C) HFPtr in PC:PG:CHOL (HFPtr-A14-chol)

10 ‘30-‘9F(CS) REDOR experiments @../home/wei4b/data/13c-19F(cs)
A1(‘3C) HFPmn in PC:PG:CHOL (HFPmn-A1-chol)

A6(13C) HFPmn in PC:PG:CHOL (HFPmn-A6-chol)

L9(‘3C) HFPmn in PC:PG:CHOL (HFPmn-L9-chol)

A6(13C) HFPmn_mut in PC:PG:CHOL (HFPmnmut-A6-chol)

L9(‘3C) HFPmn_mut in PC:PG:CHOL (HFPmnmut-L9-chol)

A6(‘3C) HFPtr in PC:PG:CHOL (HFPtr-A6-chol)

L9(‘3C) HFPtr in PC:PG:CHOL (HFPtr-L9-chol)

213

A1030) HFPmn in PC:PG (HFPmn-A1-PCPG)

l4(‘3C) HFPmn in PC:PG (HFPmn-l4-PCPG)

A6(‘3C) HFPmn in PC:PG (HFPmn-A6-PCPG)

L9(‘3C) HFPmn in PC:PG (HFPmn-L9-PCPG)

L12(13C) HFPmn in PC:PG (HFPmn-L12-PCPG)
A14(13C) HFPmn in PC:PG (HFPmn-A14-PCPG)

l4(13C) HFPmn_mut in PC:PG (HFPmnmut-l4-PCPG)
A6(13C) HFPmn_mut in PC:PG (HFPmnmut—AG-PCPG)
L9(‘3C) HFPmn_mut in PC:PG (HFPmnmut-L9-PCPG)
L12(‘3C) HFPmn_mut in PC:PG (HFPmnmut-L12-PCPG)
A14(‘3C) HFPmn_mut in PC:PG (HFPmnmut-A14-PCPG)
I4(‘3C) HFPtr in PC:PG:CHOL (HFPtr-l4-PCPG)

A6(‘3C) HFPtr in PC:PG:CHOL (HFPtr-A6-PCPG)
L9(13C) HFPtr in PC:PG:CHOL (HFPtr-L9-PCPG)
1.12030) HFPtr in PC:PG:CHOL (HFPtr-L12-PCPG)
A14(‘3C) HFPtr in PC:PG:CHOL (HFPtr-A14-PCPG)

Figure 13 1D DCP setup experiments @..Ihome/wei4b/datalDCP

Double 130 labeled NAL DCP spectra (dcp-double13CNAL)

214

APPENDIX 2. TROUBLE SHOOTING FOR MAS PROBES

1. Sample spinning problems

(1) The sample is not spinning at all.

a. The alignment for the pieces in spinning module is not good enough if you can
feel some difficulty when inserting the rotor;

b. The drive and bearing gas have been reversed when installing the spinning
module with the end piece that is connected to the magic angle spinning
adjustment, the groove on the and piece has to be placed on the drive gas side.
(2) The rotor is spinning, but the MAS controller is not reading.

a. Restart the MAS controller;

b. Cut the tip of the optical fiber;

0. The optical fiber is broken if you can see red laser all over the probe instead of
just the spinning module region;

d. The two pinholes on the spinning module front panel are too far from each
other. This is always because you are using a 6 mm spinning module front panel
intead of a 4 mm one.

(3) The rotor is not spinning stably.

a. Remark the rotor;

b. Repack the rotor tightly;

c. Reset the bearing gas parameters in ACC panel. The typical values for a 4 mm

MAS probe are: adjust 2 psi, span 100 Hz and maximum 35 psi.

2. Probe Tuning problems

215

When you change the configuration of a probe, follow the suggested capacitors
on the manual. Check the low power tuning before inserting the probe into the
magnet.

(1) There is no resonance. There is always some contacting problem. Check the
connection of every capacitor and coil, especially the weak soldering spots. The
following figure is for the HFXY probe. The four spots that typically have contact

problems are pointed outwith red arrows.

“
@ Splnnlng Module

 

 

 

 

CHY CHX CHH CHF

Figure A1 Schematic pictures for a quadruple HFXY 4mm MAS probe.

(2) There is resonance, but the position is not moving at all when you move the
tuning rods. This is special for the 1H and 19F channels of the HFXY probe, and
the problem is the screws on the 1H and 19F tuning rods are off. As shown in the
pictures below, these tuning rods use the plastic screws to connect the
adjustable knob outside the rod and the copper piece inside the rod. The screw

will be off if it has been over adjusted.

216

 

 

C J—W mm— --o
k A
Screw

 

 

k

tunnlng rod

Figure A2 Schematic pictures for a 1H or 19F tunning rode on a quadruple HFXY
4mm MAS probe.

(3) There is resonance in the low power tuning, but it’s not at the correct
frequency, and you cannot get the right frequency by adjusting the tune and
match knobs in the probe. In this case you may want to try different sets of
capacitors. The one suggested on the manual is NOT always perfect. For the
HFXY probe, the current configuration is HFPC (X E 3‘P and Y E 13C), and it is
good for routine experiments such as 130-31P REDOR, 13C-‘E’F REDOR and 13C-
13C 20 correlation. If you find the tuning for this probe is not satisfying and want

to change the pieces, here are some suggestions:

2Q— Splnnlno Mom)

 

 

 

 

 

M
=_.- 4 4T
CI-lfl CHF

Figure A3 Schematic pictures for a quadruple HFXY 4mm MAS probe.

217

a. For CH X and CH Y, you can always change the capacitors 1 and 2, and you
can adjust the length of the coil bridged over capacitor 2 gently.

b. For CH H and CH F, the current fonrvard/reflection voltage ratio is 8:1 and 9:1
respectively and it works well for REDOR and PDSD experiments. It is NOT
recommended to adjust any pieces connecting these two channels because they
are very sensitive and the connection soldering points are weak. However, the
coil in the red circle above is adjustable and is the ONLY adjustable part for the
two channels. If you do readjust the coil, double check all the soldering points

before assembling the probe.

218

APPENDIX 3 FMOC MANUAL PEPTIDE SYNTHESIS

The following solutions are commonly used in the FMOC peptide synthesis, and
one can always make them before the synthesis:

Deprotection Solution: piperidine/DMF solution with 1:3 volume ratio.

Capping Solution: acetic anhydride/pyridine/DMF solution with 2:1 :3 volume ratio.
Coupling Solution: 0.45M HBTU and 0.45M HOBt in DMF.

These solutions should be placed in brown bottles, sealed with parafilms, and
stored in the fume hood.

The Cleavage Solution should be fresh and should be prepared right before
utilization.

Cleavage Solution: TFA/thioanisoIe/EDT/anisole solution with 90:5:3:2 volume

ratio.

Procedures for regular peptide synthesis. (The procedures were applicable for
the synthesis of HFPmn, HFPmn(Cys), HFPmn_mut and HFPdm(Cys) used in
this thesis):

1. We generally use the Ala or Gly-preloaded Wang Resin for the manual
synthesis. A typical scale of synthesis was 0.1 mmol. For the synthesis of HIV
fusion peptide monomer, we use the preloaded Wang Resins with substitution
of 0.6 mmoI/g, and for the synthesis of Cys-containing dimer as described in
chapter II, we use preloaded Wang resins with substitution of 0.2 mmng. The
synthesis starts from weighing the appropriate amount of Wang Resin and

placing the resin into a 5 mL polypropylene column from Pierce (Rockford, IL).

219

. Wash the resin with DCM, drain the DCM, and repeat these for two or three
times.

. Soak the resin in DCM for 2 hours, and then drain the DCM.

. Wash the resin with DMF, and drain the DMF.

. Soak the resin with 3 mL Deprotection solution, place the column on a
rotation stage and mix for 5 minutes, and then drain the deprotection solution.
. Soak the resin with 3 mL Deprotection solution, mix on a rotation stage for 20
minutes, and drain the deprotection solution.

. At the same time with step 6, dissolve the Fmoc-protected amino acid into the
coupling solution. The amino acid: resin ratio used in our lab is 5:1 for the
unlabeled residues and 3:1 for the ‘30 or 15N labeled residues. The amino
acid:HBTU:HOBt molar ratio should be 9:9:10. (For example, for an unlabeled
residue, the amount of amino acid is 0.5 mmol if 0.1mmol resin was used,
and the amino acid was dissolved into 1mL 0.45M Coupling Solution.)

. Wash the resin with DMF, drain the DMF, and repeat these for 6 - 8 times in
order to totally remove the piperidine.

. Add DIPEA into the amino acid solution. The DIPEAzamino acid molar ratio is

2:1.

10.Add DMF into the amino acid solution until the total volume reaches ~ 4 mL,

and transfer the solution into the column.

11.Place the column onto the rotating stage, and mix for 2 hours. (The coupling

time may vary for different amino acids and peptide sequences as described

220

in chapter ll, cf. the figure captions of Figures 3 and 5) One can check the
completeness of the coupling step using the Ninhydrin test.*

12.After the coupling step, drain the coupling solution.

13. Repeat the steps 4 through 12 for the following residues.

14.After the coupling step for the last residue, repeat steps 4 though 6 to
deprotect the last Fmoc group.

15.Wash the resin 3 ~ 5 times with DMF and drain, and then 2 ~ 3 times with
DCM and drain.

16. Place the column into a dessicator with high vacuum pump for at least 6
hours to remove the residual DCM.

17.Soak the dry resin with 3 mL fresh cleavage solution, seal the column with
parafilm, and mix it for 2 hours. (The Arg(be) requires more cleavage time,
and add 0.5 hour for every extra Arg(be) residue.)

18.After the cleavage, remove the parafilm, then remove the cap of the column,
and then collect the solution part with a 50 mL centrifuge tube and discard the
residual resin.‘""Ir

19.Remove the TFA from the solution using N2 gas flow in the fume hood, and
add cold methyl t-butyl ether drop wise into the residual solution (~ 1 mL).
One should usually see white cotton-like precipitate while adding the ether.

20.Place the centrifuge tube into the freezer for overnight, and separate the
supernatant and precipitate through centrifugation. Collect the precipitate and

discard the supernatant.

221

21.Remove the residual ether in the precipitate using N2 gas flow, dissolve the
peptide into 5 mL DI. water, and lyophilize to get crude product.
22.HPLC purification with a H2O/Acetonitrile gradient and MS identification as

described in chapter ll.

*Ninhydrin Test:

Solution A: 5% Ninhydrin in ethanol (w/v); Solution B: 80% Phenol in ethanol
(w/v); Solution C: KCN in pyridine (2 mL 0.001M KCN in 98 mL pyridine)

Sample a few resin beads and wash with DMF and DCM several times. Transfer
the resin to a small glass tube and add 2 drops of the above solutions. Mix well
and heat to 100 °C for 5 minutes. The presence of resin-bound free amine is
indicated by blue resin beads.

"Cleavage

The reagents thioanisole, EDT and TFA have pungent odor. There will be
pressure during cleavage, thus please remove the parafilm and cap very

carefully before collecting the cleavage solution.

Procedures for the Cys crosslinking between HFPmn(Cys) to form HFPdm.

1. Dissolve the purified HFPmn(Cys) into D.I. H2O in a 1.5 mL eppendorf tube to
reach the peptide concentration ~ 5 mM.

2. Add DMAP powder into the peptide solution to reach the concentration of 10

mM. and adjust the pH of the solution to ~ 8 using 1M NaOH.

222

3. Gently vortex the eppendorf tube for 2 hours with the solution exposed to the
air.

4. Dilute the peptide solution with 10-time volume of D.l.water, freeze the diluted
solution and lyophilize it for overnight to obtain the crude product.

5. HPLC purification with a H2O/Acetonitrile gradient as described in chapter ll.

Procedures for the Cys crosslinking between HFPmn(Cys) and HFPdm(Cys) to
form HFPtr.

1. Dissolve the purified HFPmn(Cys) and HFPdm(Cys) into D.l. H2O in an 1.5 mL
eppendorf tube to reach the concentrations of 5 mM and 7.5 mM respectively.

2. Add DMAP powder into the peptide solution to reach the concentration of 10
mM, and adjust the pH of the solution to ~ 8 using 1M NaOH.

3. Gently vortex the eppendorf tube for 2.5 hours with the solution exposed to
the air.

4. Dilute the peptide solution with 10-time volume of D.l.water, freeze the diluted
solution and lyophilize it for overnight to obtain the crude product.

5. HPLC purification with a H2O/Acetonitrile gradient as described in chapter Il.

Procedures for the Cys crosslinking between HFPmn(Cys) to form HFPte

1. Dissolve the purified HFPdm(Cys) into D.I. H2O in an 1.5 mL eppendorf tube to
reach the concentration of 5 mM.

2. Add DMAP powder into the peptide solution to reach the concentration of 10

mM, and adjust the pH of the solution to ~ 8 using 1M NaOH.

223

3. Gently vortex the eppendorf tube overnight with the solution exposed to the
air.

4. Dilute the peptide solution with 10-time volume of D.l.water, freeze the diluted
solution and Iyophilize it for overnight to obtain the crude product.

5. HPLC purification with a H2O/Acetonitrile gradient as described in chapter II.

224

APPENDIX 4 INTRODUCTION TO NMR
This appendix helps to explain the description of Average Hamiltonian Theory in

chapter lll.

Spin Behavior in a magnetic field. Nuclei with spin of 1/2 such as 1H, 13C, 15N, 31P
and 19F have two possible spin states with magnetic quantum number m = 1/2 or

-1/2. If the nucleus is placed in a magnetic field, there will be an interaction

between the nuclear magnetic moment Z and the external magnetic field
—B_0°=Boé . The energy associated with It in E; is given by

E = "2'30 = #1230 = “thBo (A1)

A

m =-1/2

  
 
 

AE = E-1/2- E1/2

Energy

 

m = +1/2
’

 

 

Magnetic Field

Figure A4 Energy Splitting for a spin 1/2 nucleus in the external magnetic field.

In Eq. A1 the term #2 represents the 2 component of the nuclear magnetic
moment because the external magnetic field is always defined as being along the

z direction in the laboratory frame. 7 is the gyromagnetic ratio and his Planck’s

225

constant divided by 21:. Thus, the splitting between the two energy levels with m
= 1/2 and -1/2 is shown in Figure A4 for a nucleus with spin 1/2, and the energy
difference between the two levels are

AE=yhB0 :I:-co (A2)

where Va is also known as the Larmor frequency.

Magnetization and time evolution of magnetization. The concept of magnetization
in the field of NMR was defined as the vector sum of individual magnetic

moments fl. It is usually convenient to describe the system using magnetization
because it provides a macroscopic picture for the spin system.

A7! = 2;: (A3)

The direction of the magnetization for an equilibrium spin system is along the z
axis because the Boltzmann population of individual ii that are aligned along the
z axis is larger than the Boltzmann population of the it aligned along the -z axis.

To understand the time evolution of magnetization, it will be helpful to first

consider the time evolution of an individual magnetic moment. An individual
magnetic moment [1 in a field B is subject to a torque A7 , which is defined as

A7 = [ME (A4)

Analogous to the Newton’s Law where force is the time derivative of momentum,

in the rotational motion, torque is the time derivative of angular momentum Z.

| a.

117: Z (A5)

a.

t

226

In the quantum mechanics, the angular momentum and magnetic moment are
correlated through the constant of gyromagnetic ratio.

[7 =75 (A6)

The combination of Eqs. A4, A5 and A6 gives the time evolution of an individual

magnetic moment.

d- _. .. .. ..
EP=YPX3=PXYB (A7)

The discussion for an individual magnetic moment can be extended to the

macroscopic magnetization, thus

d - .. s
—M=M B A8
dt X7 ( )

Time evolution of the spin system. In quantum mechanics, a spin system can be
expressed as the linear combination of time-independent eigenfunctions and

their corresponding time-dependent coefficients.

‘1’ =Zcml/Im =26": Im) (A9)

The ensemble average of these coefficients forms a matrix which is generally
named density matrix. The density matrix is the matrix representation of an

operator known as density operatora.

E=(n|o|m)=a,,m (A10)

The system can be represented as the density matrix and density operator. The

time evolution of the density operator reflects the time evolution of a spin system.

227

According to the Eq. A9, the time—dependent Schrodinger Equation can be

written as

idzd—f-L— (’)| n): 2c, (t)H| n)(A11)

n

Usmg the orthogonal condition between different eigenfunctions and some

algebra, the time evolution of the coefficient ck(t) is
rick—(€Z=ZC,(1)(/.|H|n) (A12)

dt n
Thus the time evolution of an arbitrary element in the density matrix under the

Hamiltonian is given by

 

d kam d ,2 d; d . —_e
(LII >= 0;; “Wei—U -—c—"c;, =iZc—_kc; (n njylm)_l§c,c,,(k|y|n)(A13)
=i21kI0I'IIX'1IHI m") “iZXkIHI ")1" IUI "'I= (kIIG-HII'")

In the operator form, or for the entire density matrix, the time evolution can be

expressed as
d0") = -i[H,a'(t)] (A14a)

dt
This differential equation has a solution when the Hamiltonlan is time-

 

independent

d(t) = cxp[-r'Ht] - 0(0) - exp[th] (A14b)

Sprn interaction Hamiltonians. In quantum mechanics, the Hamiltonian operator
is associated with the total energy of the system. In NMR, the nuclear spin

Hamiltonians are associated with the energies of different types of coupllngs

228

between the nuclear spins with different types of magnetic fields. For example,

the Zeeman Hamiltonian (H 2) corresponds to the coupling of nuclear spins with
the external static magnetic field. The chemical shift Hamiltonian (HCS)

corresponds to the coupling with the induced magnetic fields originating from

orbital motions of electrons. The dipolar coupling Hamiltonian (H D) corresponds

to the coupling with the magnetic field from the magnetic dipole moment of

another spin. The expression for H 2 can be derived from Eq. A1 and the relation
Zi=yh7,where751x2+1yy+122.

Elma, = 42 - 30 => HZ = —yn‘i . 30 (A15)

In the lab frame where the z-axis is set along the direction of the external
magnetic field, we have 30 = BO}.

”2 = —yhBOIZ (A16)

For HCS and H D, it is always convenient to first express the Hamiltonian in the

principal axis system, and then to transform them into the lab frame through
Vlfigner Rotation introduced in Chapter III of the main text.

For HCS , we could write an analogous expression as Eq. A15.
ECS = —;-§s => HCS = —yri‘I'-'és = (—7h)'i.§.iéo (A17)
In Eq. A17, the term BS represents the induced magnetic field due to the motion

of electrons and the term 3‘: is a tensor which reflects the shielding effect. As

shown in Fig. A5, the fact that E is a tensor rather than a numerical value

229

indicates that the induced field and the original field can be along different

directions, and the effective field is the vector sum of Bo and BS.

30 r
8s

Beff

 

Figure A5. The vector representation of the fact that the induced shielding field is

not necessarily parallel to the original field.

To evaluate HCS in the principal axis system, one always expresses the
Hamiltonian as the product of the elements of two second-rank Cartesian

tensors or spherical tensors.

i-EEFZEQoeiyj (A18)
w

In Eq. 18, the term 50 817 is a dyadic product of vectorsfio and 7 which forms a
tensor.
For HD, the field produced by a magnetic dipole 7r; at the origin of the

coordinate system can be expressed as

79', = [30" "Zr-”1] (A19a)
r

in which F: w and r = (Ixz + y2 + 22 (A19b)
r

230

The vector F is the internuclear vector between the two coupled dipoles. The
expression of the dipole-dipole interaction and the associated Hamiltonian thus

can be written analogous to Eqs. A15 and A17 as

- - 13F fix" -?)—" 7’]
ED=-#s'31=- #1 #53 #1 .Us

2 r (A20)
=>H —-——h 7’75 [3(I.F)(S-?)—1.S]
D— 3
r

In most cases, the magnitude of Zeeman Hamiltonian is much larger than the
magnitude of chemical shift and dipolar coupling Hamiltonians. Therefore, there

is always a truncation effect for the HC5 and H D, and the terms containing spin
operators which do not commute with Hz (or 12 according to Eq. A16) will be
averaged to zero. Thus, the truncated HCS is proportional to 12, the truncated

heteronuclear H D is proportional to IZSZ, and the truncated homonuclear H D is

proportional to IZSZ +i-(I+S_ + I_S+) where 1:, = I,r :t i] y and Si = S, i iSy.

Rotating frame. Another type of important nuclear spin interaction is the
interaction between spins and the external rf magnetic field because the If pulses
are widely utilized in NMR. In the case of magic angle spinning probes with

solenoid coils, the If field is oscillating with the carrier frequency a)” along the

MAS rotor axis which is tilted by 9 relative to the external magnetic field. Thus,
the rf field can be decomposed into two parts that are either parallel or

perpendicular to the external magnetic field.

if = 8, cos 6004th + a]; + Bl sin 000%th + 0)]; (A21)

231

In Eq. A21, B, and (care the amplitude and phase of the rf field, and can be time-

dependent

The 2 component in Eq. A21 is an oscillating field with frequency w,f. However,

this field can be neglected because the time scale of the oscillation (nanosec) is
much shorter than the experimental time scales such as dwell time and If pulses
(microsec), and the time average of the oscillations field is zero.

The x component in Eq. A21 can be decomposed as two rotating fields with

angular frequencies ”If and —w,, since

cos[a),ft + p] = écxp[i(a),ft + 02)] + %exp[—i(ro,ft + (0)] (A22)
With the condition warmth, the rotating field with frequency —ro,f can be

neglected. In a classical picture, the nucleus spin [1 is rotating around the z-axis

in the lab frame under the effect of Bo with angular frequency wL. The rotating

field with frequency a),f zwL will provide efficient torque to the nucleus spin

which will induce the transition of magnetization over time.
—- d —~ _. ~ . .. -
N = a}: = 7px B,f,x = ”13,.” sm < ,u,B,f,x > (A23)

As shown in Eq. A23, an efficienct torque requires an approximately time-

independent angle < AB,” > which can be satisfied by the rotating field with

frequency dog. The field with frequency —w,., will have a time-dependent angle

< .173ng > and the interaction according to Eq. A23 will be reduced over time.

232

It is usually more convenient to make the rf field appear static by transforming the

system to a rotating frame with angular frequency (0,1. In the rotating frame, for
the simplest case that 0),f = 0L, the field acting on the magnetization is just the

static 81 field along the x axis. If the initial magnetization is along the z axis, e.g.
equilibrium state, the effect of applying a If field is to tilt the magnetization into

the xy plane (also known as the transverse plane) according to Eq. A8.

Magic Angle Spinning (MAS) and spinning sidebands. Manipulation of the spin
interaction such as chemical shift and dipolar coupling can be achieved by

rotating the sample about an axis which is tilted by an angle 6 with respect to the

external magnetic field Bo. The angle 6, known as magic angle, satisfies the

condition 300s2 6—1 = 0 and I95 54.7°. Here we will analyze the effect of the MAS

on the chemical shift Hamiltonian.

According to Eq. 5 in chapter III, a general expression to the chemical shift is
1
Hrs = CCS Z. 21(‘1)m'R1C:m']I.C:l (A24)

where the values of I are 0, 1 or 2. In the condition of high field where Hz >> H”,
the non-secular terms (I = 1 and 2) in Eq. A24 will become time-dependent due to
the manipulation of Zeeman Hamiltonian, and their time averages will be zero.
Thus, eq. A24 can be simplified as

Hes = cs (1?me +111on +R20T20) (A25)

On the right side of Eq. A25, the middle term represents the anti-symmetric part

of the Hamiltonians. This part is not detectable for chemical shift and therefore

233

can be neglected. The last term on the right side of Eq. A25 represents the

anisotropic part of the chemical shift Hamiltonian (or Hm) in the principal axis

system. When transforming into the lab frame, this term will the periodic with the

period —l-where wk is the sample spinning frequency. The rotational angular
0R

frequency of the transverse magnetization under the periodical Hm will also be

periodical and will have the form

a)(t) = A+B§(t) (A26)

The terms A and B are time-independent and are related to the Euler angle set
(0t, 0, y) which correlates the principal axis system and the rotor frame, and the

magic angle 0. The term §(t) is periodic and related to the sample spinning

frequency. The F ID due to this precession can be expressed as
S(t) cc exp[r'a)(t)] = exp(iAt)exp[iB £§(t')dr'] (A27)

The second term on the right side of Eq. 27 will be unity for multiples of the
sample spinning period. The FID will consequently contain a sequence of echoes

separated by lle. Fourier Tansformation of the FID with echoes provides
spinning sidebands at multiples of wk away from the isotropic chemical shift in

the frequency domain.
Evolution of transverse magnetization and its Fourier relationship to the NMR

spectrum. Consider a simple NMR experiments where a 90° pulse was applied

followed by detection. Initially the magnetization is along 2 axis and the 90° pulse

234

will flip the magnetization to the xy plane. Consider the situation that at the
beginning of the acquisition period, the magnetization is along the x axis. During
the acquisition period, the spin system is subject to only the Zeeman Hamiltonian.
The time evolution of the magnetization is controlled by Eq. A14a with the density
operator 0(0) cc 1, and H2 ac (001,, where IX and Iz are the x and 2 components of
the spin operator and 00 = yBo . Since the Zeeman Hamiltonian is time-
independent, according to Eq. A14b, the density operator at time t is

a(t) cc cxp[—iroolzt]1x cxp[ia)01 zt] = I x cos(a)ot) + [y sin((oot) (A28)

This means the magnetization will rotate in the transverse plane. The detected
time-domain signal is the transverse magnetization. It is convenient to write the

transverse magnetization in Its complex form ,u+ = ,er +ipy oc I+ . Wlthout
considering relaxation, the expectation value will be
5(1) “=- (r‘) at: Tr{a'(t)1+} -_- Tr{ [1,, cos(w0t)+1y sin(m0r)]-[1, +i1y]} (A29)
It is easier to evaluate the time-domain signal in the matrix form where
l 0 l 1 O —i
I =—- and I — A30
" 2Il o] y 2L 0I( )

Using Eq. A30, one can evaluate the time-domain signal in Eq. A29 as

S (t) at exp[iar0t] (A31)
Eq. A31 indicates an oscillation in the time-domain signal, and after the Fourier
transformation the frequency—domain spectrum showed a single resonance line

at frequency 020 . Considering relaxation which makes the tranverse

235

magnetization return to its thermoequilibrium state, eg. zero, there will be a
exponential decay term adding to the time-domain signal expressed in Eq. A31.
S(t)ocexp[iroot—AJ] (A32)

The parameter Ais the decay constant. The Fourier transformation of Eq. 32
results in a Lorentzian Iineshape in the frequency domain, and Iinewidth is

proportional to the decay constant ,1.

Effect of the powder average on evolution of the transverse magnetization. This
part provides a qualitative picture about how the recovered heteronuclear dipolar
coupling affects the 13C signal in the REDOR 81 spectrum. Considering that the
zero-order average heteronuclear dipolar coupling Hamiltonian under Magic
Angle Spinning is the only interaction in the 81 spectrum.

< H mm,” >Ooc 0,5 sin(2,6’) sin y-IZSZ (A33)

In Eq. 33 the (0,5 represents the magnitude of the dipolar coupling frequency
which is related to the gyromagnetic ratios and the internuclear distance. The
angles ,6 and 7 are Euler angles between the principal axis system of the

dipolar coupling frame and the lab frame at a single time in the rotor cycle. For an
individual l-S internuclear vector, the value of these angles is related to the
orientation of the l-S vector relative to external magnetic field. The time evolution
of initial density operator, which is proportional to I)( can be expressed as

d(t) = I x cos(é'rt) + 21 ySz sin(a")t) and c?) at 0,3 sin(2,B) sin 7 (A34)

236

The term with the spin operator IySz is undetectable based on Eq. 22 in the main
text. Thus, the detected time-domain signal is
S(t) cc Tr{a(t)I+} = Tr{IJr cos((2'rt)-(Ix + in)} oc cos(a31)[1+iTr{Iny}] (A35)

With the matrix form of the spin operators IX and IV

“0010“ Too-i0“
0 0 O l O 0 0 —'
Ixoc and Iyoc . l , one can obtain from Eq. A35 that
1 0 0 0 r 0 0 0
_0 1 O 0‘ _0 i O 0_

 

 

 

 

Tr{Ix1y} = 0 and

S(t)occos(a")t) (A36)

Eq. A36 indicates that the time-domain signal of spin l due to the I-S dipolar
coupling with a particular orientation relative to the magnetic field is a cosine
oscillation with its own frequency a”). In a powder sample where an individual I-S
spin pair can adopt any orientation relative to the external field, the net
transverse magnetization will decay since it will be the sum of the time-domain
signals of individual spin pairs with same initial phase but different oscillation
frequencies (Figure A6). Thus, it can be qualitatively understood. that in the
REDOR Si spectrum there is an attenuation in the detected signal relative to the
So spectrum. In REDOR experiments, one acquires Si spectra with different
dephasing times, which correspond to different evolution times of the transverse
magnetization under the heteronuclear dipolar coupling. Thus, it is
understandable that there will be a buildup curve because the transverse

magnetization is not entirely reduced at short dephasing times.

237

   

 

.‘:
8
..C. 8
'5
9' s
U) '4-0
—' _, a
“ti or ' ‘ "
:3 (U
jg 2 Time
.fi
5

Figure A6 Decay of the transverse magnetization due to the powder averaging.

238

Bibliography

1. Cavanagh, C., Fairbrother, W.J., Palmer, AR, and Skelton, NJ. (1996).
Protein NMR Spectroscopy: Principles and Practice. San Diego: Academic Press.

2. Schmidt-Rohr, K., and Spiess, H.W. (1996). Multidimentional Solid-state NMR
and Polymers. London: Academic Press.

3. Stejskal, E.O., and Memory, JD. (1994). High Resolution NMR in the Solid
State. New York: Oxford University Press.

4. Tycko, R. (1994). Nuclear Magnetic Resonance Probes of Molecular
Dynamics. Dordrecht: Kluwer Academic Publishers.

5. Haeberlen, U. (1976). High Resolution NMR in Solids Selective Averaging.
New York: Academic Press.

6. Mehring M. (1983). Principles of High Resolution NMR in Solids. New York:
Springer-Verlag.

7. Jaroniec CM. (2008). “Heteronuclear Decoupling and Recoupling” in US.-

Canada Winter School on Biomolecular Solid State NMR. Stowe, Vermont,
January 20-25, 2008.

239

         

   

   

ATE

III 1IIIljljjljljljjjllll“

1293

   

MlCi-litsflxltl|
3