STRUCTURE AND FUNCTION STUDY OF HIV AND INFLUENZA FUSION PROTEINS
By
Shuang Liang

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Chemistry—Doctor of Philosophy
2017

ABSTRACT
STRUCTURE AND FUNCTION STUDY OF HIV AND INFLUENZA FUSION PROTEINS
By
Shuang Liang
Human immunodeficiency virus (HIV) and influenza virus are membrane-enveloped
viruses causing acquired immunodeficiency syndrome (AIDS) and flu. The initial step of HIV and
influenza virus infection is fusion between viral and host cell membrane catalyzed by the viral
fusion protein gp41 and hemagglutinin (HA) respectively. However, the structure of gp41 and HA
as well as the infection mechanism are still not fully understood. This work addresses (1) full
length gp41 ectodomain and TM domain structure and function and (2) IFP membrane location
and IFP-membrane interaction. My studies of gp41 protein and IFP can provide better
understanding of the membrane fusion mechanism and may aid development of anti-viral
therapeutics and vaccine.
The full length ectodomain and transmembrane domain of gp41 and shorter constructs
were expressed, purified and solubilized at physiology conditions. The constructs adopt overall 
helical structure in SDS and DPC detergents, and showed hyperthermostability with Tm > 90 °C.
The oligomeric states of these proteins vary in different detergent buffer: predominant trimer for
all constructs and some hexamer fraction for HM and HM_TM protein in SDS at pH 7.4; and
mixtures of monomer, trimer, and higher-order oligomer protein in DPC at pH 4.0 and 7.4.
Substantial protein-induced vesicle fusion was observed, including fusion of neutral vesicles at
neutral pH, which are the conditions similar HIV/cell fusion. Vesicle fusion by a gp41 ectodomain
construct has rarely been observed under these conditions, and is aided by inclusion of both the FP

and TM, and by protein which is predominantly trimer rather than monomer. Current data was
integrated with existing data, and a structural model was proposed.
Secondary structure and conformation of IFP is a helix-turn-helix structure in membrane.
However, there has been arguments about the IFP membrane location. 13C-2H REDOR solid-state
NMR is used to solve this problem. The IFP adopts major  helical, minor  strand secondary
structure in PC/PG membrane. The  helical IFP’s with respectively 13CO labeled Leu-2, Ala-7
and Gly-16 all show close contacts with the lipid acyl chain tail, suggesting IFP has strong
interaction with the membrane. By screening the current IFP topology models, it either has a
membrane-spanning confirmation, or it promotes lipid trail protrusion.
IFP bounded lipid membrane structure was studied by paramagnetic relaxation
enhancement (PRE) solid-state NMR to provide more information about the detailed IFP
membrane location model. The T2 relaxation time and rate were measured for membrane with or
without IFP and with or without Mn2+. Based on the results, it is concluded that IFP does not
promote lipid protrusion at both gel phase and liquid phase, which is evidenced by that the R2
difference with and without Mn2+ is smaller for IFP free membrane than IFP bounded membrane,
meaning IFP does not induce a smaller average distance between lipid acyl chain and aqueous
layer. By integrating these results, a IFP membrane spanning model was proposed, in which IFP
N-terminal helix adopts a 45° angle with respect to membrane normal.

Dedicated to mom, dad and Luyao

iv

ACKNOWLEDGEMENTS

I would like to take this chance to thank my advisor Dr. David P Weliky for giving me a
chance to join this wonderful group, teaching me incredible things not only about science but also
about life, and being patient and always helpful. I want to thank him for all the help, suggestion,
support and courage he gave me when I faced any difficulties. He taught me NMR and biological
knowledge, techniques and, most importantly, learning and problem-solving abilities in the past
five years. I also want to thank my committee members: Prof. John McCracken, Prof. Heedeok
Hong and Prof. Daniel Jones, for their guidance, support and valuable suggestions.
I am appreciative that I can join the Weliky group, thank Dr. Li Xie, Dr. Koyeli Banerjee,
Dr. Punsisi Ratnayake, Dr. Ujjayini Ghosh, Lihui Jia, Robert Wolfe and Ahinsa Ranaweera for not
only being colleagues that help and support me in my research, but also as my good friends that
we share happiness and sorrow. I am also appreciative to have the Max T Rogers NMR facility
staffs Dr. Dan Holmes and Dr. Li Xie. Thank them for all the help and suggestions whenever I
have any NMR issues or questions throughout my graduate school.
I am grateful to for everybody in Hong’s lab and Geiger’s lab. Thank them for all the help,
valuable suggestions, and for sharing instrument. I also got to know so many good friends in
Michigan State University. I will never forget all the joy we had together and thank them for
accompanying to be a better and mature person.
At the end, I want to express my gratefulness for my family. Thank my mom and dad for
all the care, support and encouragement to let me study in the United States. Special thanks for my
husband Luyao for all his love and the wonderful moments that we spent together. Word cannot
express my love for my family and this love will last forever.

v

TABLE OF CONTENTS

LIST OF TABLES ....................................................................................................................... viii
LIST OF FIGURES ........................................................................................................................ x
KEY TO ABBREVIATIONS .................................................................................................... xviii
Chapter 1 Introduction to HIV gp41 and influenza hemagglutinin fusion proteins .............. 1
1.1 Introduction to HIV gp41 fusion protein............................................................................... 2
1.1.1 HIV virus structure and its infection pathway ................................................................ 2
1.1.2 Structure of gp41 ............................................................................................................ 5
1.1.3 Possible membrane fusion mechanism ......................................................................... 15
1.2 Introduction to influenza hemagglutinin fusion peptide ..................................................... 21
1.2.1 Influenza virus structure and its infection pathway ...................................................... 21
1.2.2 Structure of HA2 .......................................................................................................... 26
1.2.3 Proposed membrane fusion mechanism ....................................................................... 32
1.3 Introduction solid-state NMR techniques for membrane proteins ...................................... 34
1.3.1 Magic-Angle Spinning (MAS) ..................................................................................... 34
1.3.2 13C-2H Rotational-Echo Double-Resonance NMR (REDOR) ..................................... 36
1.3.3 Paramagnetic Relaxation Enhancement NMR (PRE) .................................................. 42
REFERENCES ............................................................................................................................. 45
Chapter 2 Materials and Methods............................................................................................. 55
2.1 Materials .............................................................................................................................. 56
2.2 Gp41 constructs expression and purification ...................................................................... 57
2.3 Circular Dichroism (CD) ..................................................................................................... 61
2.4 Size-Exclusion Chromatography (SEC).............................................................................. 62
2.5 Protein induced vesicle fusion............................................................................................. 62
2.6 Solid phase peptide synthesis, purification and characterization ........................................ 63
2.7 solid state NMR sample preparation ................................................................................... 65
2.8 solid state NMR................................................................................................................... 65
REFERENCES ............................................................................................................................. 69
Chapter 3 Expression, purification and functional characterization of HIV gp41
ectodomain and transmembrane domain ................................................................................. 72
3.1 Introduction ......................................................................................................................... 73
3.2 Result and discussion .......................................................................................................... 75
3.2.1 Protein solubilization and purification ......................................................................... 75
3.2.2 Influence of FP and TM on hyperthermal Îą helical hairpin structure .......................... 82
3.2.3 Oligomeric states vary in different detergent ............................................................... 88
3.2.4 Comparison of gp41 oligomeric states in SEC and AUC ............................................ 95
3.2.5 Hairpin protein-induced vesicle fusion at both physiologic and low pH ..................... 97
3.2.6 FP_HM_TM structural model .................................................................................... 100

vi

3.2.7 Relationship between vesicle fusion and HIV/cell fusion .......................................... 102
3.3 Conclusion......................................................................................................................... 106
REFERENCES ........................................................................................................................... 107
Chapter 4 IFP membrane location studied by 13C-2H REDOR NMR ............................... 113
4.1 Introduction ....................................................................................................................... 114
4.2 Result and discussion ........................................................................................................ 118
4.2.1 13CO-2H REDOR spectra, buildups, and fittings ........................................................ 118
4.2.2 Effects of sample preparation method and lipid charge ............................................. 126
4.2.3 Close contact of helical IFP and 2H in DPPC_D10................................................ 130
4.3 Conclusion......................................................................................................................... 132
REFERENCES ........................................................................................................................... 133
Chapter 5 IFP effects on membrane studied by 2H paramagnetic relaxation enhancement
(PRE) solid-state NMR ............................................................................................................. 137
5.1 Introduction ....................................................................................................................... 138
5.2 Result and discussion ........................................................................................................ 143
5.2.1 Sample preparation methods ...................................................................................... 143
5.2.2 2H FID, spectra and T2 relaxation time of pure lipid membrane ................................ 145
5.2.3 The effect of Mn2+ concentration on lipid T2 ............................................................. 153
5.2.4 FID, spectra and T2 of IFP-bounded membrane with 5% Mn2+ ................................. 158
5.2.5 T2 of IFP-bounded membrane with 20% Mn2+ .......................................................... 172
5.2.6 IFP bounded membrane structure ............................................................................... 176
5.3 Conclusion......................................................................................................................... 179
REFERENCES ........................................................................................................................... 180
Chapter 6 Summary and future directions ............................................................................ 184
APPENDICES ............................................................................................................................ 189
APPENDIX A Location of NMR data .................................................................................... 190
APPENDIX B Solid-state NMR raw data .............................................................................. 195
APPENDIX C Additional PRE data ....................................................................................... 219
REFERENCES ........................................................................................................................... 227

vii

LIST OF TABLES

Table 3.1 Analysis of CD spectra in 0.25% DPC at pH 4…...…………………………………..86
Table 3.2 Analysis of SEC traces………………………………………………………………..90
Table 4.1 Best-fit exponential buildup parameters for 13CO–2H REDOR of IFP in DPPC and
DPPG…………………………………………………………………………………………...122
Table 4.2 FWHM of membrane associated IFP……………………………………………..….125
Table 4.3 Best-fit exponential buildup parameters for 13CO–2H REDOR of IFP in DPPC…..…128
Table 5.1 2H NMR spectra peak splitting (kHz)…………………………………………….…..151
Table 5.2 Best-fit 2H T2 (s) of D10 and D8 lipids measured by quecho experiment…………...152
Table 5.3 D10 2H NMR spectra peak splitting affected by Mn2+ concentration……………..…155
Table 5.4 Best-fit 2H T2 (s) of D10 with various Mn2+ concentration……………………….…155
Table 5.5 IFP-bounded D10 membrane 2H NMR spectra peak splitting with 5% Mn2+……….164
Table 5.6 IFP-bounded D8 membrane 2H NMR spectra peak splitting with 5% Mn2+………...164
Table 5.7 Best-fit 2H T2 (s) of IFP-bounded membrane with 5% Mn2+………………………169
Table 5.8 Best-fit 2H T2 relaxation rate (R2, kHz) of IFP-bounded membrane with 5% Mn2+….169
Table 5.9 IFP-bounded D10 membrane 2H NMR spectra peak splitting (kHz) with 20%
Mn2+…………………………………………………………………………………………….175
Table 5.10 Best-fit 2H T2 (s) of IFP-bounded membrane with 20% Mn2+……………………175
Table 5.11 Best-fit 2H T2 relaxation rate (R2, in unit kHz) of IFP-bounded membrane with 20%
Mn2+…………………………………………………………………………………………….175
Table B1 IFP L7c in DPPC-D4 studied by REDRO raw data (Figure 4.3)……………………195
Table B2 IFP L7c in DPPC-D8 studied by REDRO raw data (Figure 4.3)……………………196
Table B3 IFP L7c in DPPC-D10 studied by REDRO raw data (Figure 4.3)………………..…197
Table B4 IFP A7c in DPPC-D4 studied by REDRO raw data (Figure 4.4)...…………...……..198
Table B5 IFP A7c in DPPC-D8 studied by REDRO raw data (Figure 4.4)……………………199

viii

Table B6 IFP A7c in DPPC-D10 studied by REDRO raw data (Figure 4.4)…………………...200
Table B7 IFP G16c in DPPC-D4, DPPC-D8 and DPPC-D10 studied by REDRO raw data (Figure
4.5)……………………………………………………………………………………………...201
Table B8 IFP L2c in pure DPPC-D8 and DPPC-D10 studied by REDRO raw data (Figure
4.6)……………………………………………………………………………………………...203
Table B9 D10 with various concentration of Mn2+ at 25 °C studied by PRE (Figure 5.10)……204
Table B10 D10 with various concentration of Mn2+ at 50 °C studied by PRE (Figure 5.11)……206
Table B11 D10 at 25 °C studied by PRE (Figure 5.16)…………………………………………208
Table B12 D10 at 50 °C studied by PRE (Figure 5.17)………………………………………….210
Table B13 D8 at 25 °C studied by PRE (Figure 5.18)…………………………………………..212
Table B14 D8 at 50 °C studied by PRE (Figure 5.19)…………….…………………..………..214
Table B15 D8 reproducibility at 25 °C studied by PRE (Figure 5.20)………………………....216
Table B16 D10 with 20% Mn2+ studied by PRE (Figure 5.22)……………………………….…217
Table C1 D10 samples at 25 °C ……………………………………………………………….219
Table C2 D10 samples at 50 °C …………………………………………………………….…221
Table C3 D8 samples at 25 °C ……………………………………………………………...…223
Table C4 D8 samples at 50 °C ……………………………………………………………...…225

ix

LIST OF FIGURES

Figure 1.1 Cutaway schematic of the structure of an HIV virion.....................................................3
Figure 1.2 (A)Stages of the HIV‑1 life cycle. The yellow boxes in the diagram show the possible
antiretroviral therapy mechanism. (B) Schematic infection model of HIV. The spikes on virion are
gp120/gp41 complex. (a) Binding of HIV on host cell, (b) Hemifusion of viral and host cell
membrane, (c) Pore formation, and (d) complete fusion and entry of viral genetic material into the
host cell. (C) Electron microscopy pictures of HIV infection process correspond to panel B. (D)
Schematic infection model of an alternative HIV entry pathway via clathrin-mediated
endocytosis………………………………………………………………………………………...4
Figure 1.3 (A) Schematic diagrams of full-length HIV gp41 and corresponding colors: FP  fusion
peptide, red; NHR  N-helix region, blue; Loop, grey; CHR  C-helix region, green; MPER 
membrane-proximal external-region, pink; TM  transmembrane domain, orange; and endo 
endodomain, white. (B) Amino acid sequences with colors matching segments in panel A except
for the endo domain in black. The sequence is from the HXB2 laboratory strain of HIV and have
the gp160 precursor residue numbering, 1-511 for gp120 and 512-856 for gp41…………………..7
Figure 1.4 (A)Sequence of the N36/C34 complex. (B) The end-on view of the N36/C34 complex
looking down the three-fold axis of the trimer. The N36 residues are in blue and the C34 residues
are in red. (C)The side view of the N36/C34 complex. The amino termini of the N36 helices point
toward the top of the page, while those of the C34 helices point toward the bottom………………..8
Figure 1.5 (A) Schematic of the HIV gp140 construct studied in comparison to full-length gp160.
N-linked glycans are shown and numbered on their respective Asn residues. The FP, NHR, CHR,
MPER, transmembrane (TM), and endodomain elements in gp41 are indicated. The mutations are
shown in red, as well as the added N332 glycan site. The color coding is preserved in (B) and (C).
(B) Side view of the gp140 trimer. The main domains are labeled correspond to panel A and
glycans are shown as spheres. (C) gp41 in gp140 complex. Secondary structure determination was
ambiguous at the dashed line areas……………………………………………………………….10
Figure 1.6 Compare of NHR coiled-coil trimer at pre- and post-fusion states. (A) Top view of the
locations of the three N-terminal helices at pre-fusion state derived from cryo-electron microscopy.
(B) Top view of the locations of the same helices in the post-fusion state derived by X-ray
crystallography. (C) Superposition of the arrangement of the three N-terminal helices in pre-(cyan)
and the post-fusion (magenta) conformation……………………………………………………..11
Figure 1.7 Models of FP in membrane. (A) Antiparallel β sheet FP with residue 16→1/1→16 or
17→1/1→17 registries for adjacent strands binding to membrane. The FP is shown as red lines
and the lipids are the blue sphere with gray lines.(B) Two membrane locations of FP studied by
13 2
C- H REDOR: majorly deep and minorly shallow insertion. The FP is 13CO labeled on Gly_5,
Leu_12 and Gly_16 residues. The labels on both side of the diagram are the approximate
membrane locations of the 2H’s and 31P’s where P  phosphorous in lipid; Chol_d6, Chol_d7,
PC_d4, PC_d8 and PC_d10 are deuterium in cholesterol or phosphocholine lipids……………..13

x

Figure 1.8 Models of MPER and TM. (A) MPER665-683 in DPC micelle at pH 3.5 showing a
straight -helix; (B) MPER662-683 in DPC micelles at pH 6.6 showing a L-shape kink between two
-helices; (C) Crystallography result of NHR547–575-CHR630-662-MPER663-675 monomer. The CHR
and MPER forms a continuous helix and the MPER in this structure ends at I675. (D) Molecular
dynamics (MD) simulation result of TM681-707 in membrane with Arg694 snorkeling toward the
membrane surface. TM peptide is shown in grey, Arg residues in blue, lipid molecules in yellow
and water molecules in red. (E) Liquid-state NMR result of MPER675-683-TM684-704 peptide. The
MPER and TM form a continuous -helix with a turn at 690GGLV693 residues………………….14
Figure 1.9 Stages of membrane fusion…………………………………………………………...17
Figure 1.10 Schematic diagram of Model 1. (a) Trimer gp120 and gp41 at pre-fusion state; (b)
displacement of gp120 and PHI formation; (c) SHB formation; (d) gp41 at the SHB post-fusion
state. ‘A’ represents the transmembrane domain and ‘F’ represent the FP domain……………….17
Figure 1.11 Schematic diagram of Model 2. (a) Trimer gp120 and gp41 at pre-fusion state; (b)
PHI formation and IFP inserted in host cell membrane; (c) Viral and host cell membrane
hemifusion; (D)gp41 at the SHB post-fusion state. FP is shown in red, NHR in blue, CHR in green
and MPER in white……………………………………………………………………………….18
Figure 1.12 Membrane fusion Model 3 of gp41 ectodomain monomer and hexamer. The different
domains of gp41 are color coded the same as Figure 1.3 and the TM and endodomain are not
shown. One of the monomers is not displayed in steps 3−5. The initial gp41 structure of step 1 and
the final SHB structure of step 7 are based on high-resolution structures………………………..19
Figure 1.13 Schematic representation of the structure of influenza virus. Viral envelope is shown
as gray sphere and the membrane proteins (HA, NA and M2) are shown as spikes on the
membrane. Inside the membrane is M1 that surrounds the viral ribonucleoproteins (vRNP). There
are eight single-stranded, negative-sense RNA segments and each encoded one or two
proteins…………………………………………………………………………………………...24
Figure 1.14 Influenza viral life cycle…………………………………………………………….25
Figure 1.15 (A) Pre-fusion structures of HA2 ectodomain (Protein Data bank entries 1RD8). This
structure includes residues 38 - 170 of HA2 at pH 7.5; (B) Post-fusion structures of HA2 (Protein
Data bank entries 1QU1). This structure includes residues 38 - 175 of HA2 at pH 5. (C) The
structure of pre-fusion (left) and post-fusion (right). The HA2 is divided into A – H regions and
the residue numbering is labeled. The -strand of HA1 is also shown as “1” and the disulfide bond
between 141 and 1372 is indicated…………………………………………………………….......28
Figure 1.16 NMR structures of IFP. (A) (B) H3_20 in DPC micelle at pH5 and pH 7.4
respectively. Amino acids are labeled in the diagram. Backbone of the peptide is shown as ribbon
representations and side chains shown as stick models. (C) (D) Longitudinal view and lateral view
of H1_23 in DPC micelle at pH 7 and pH 4, with the hydrophobic side chains in yellow, polar side
chains in green, acidic side chains in red, and Gly residues shown in white van der Waals
representation; (E) (F) H1_23 and H3_20 in lipid membrane at pH 5 or pH 7. Carbon, nitrogen

xi

and oxygen atoms in green, blue and red vertices. The dashed lines are between F9 N and G16 CO
with distances 3.9 Å and 5.5 Å respectively……………………………………………………..30
Figure 1.17 Mechanism of membrane fusion promoted by HA2. Host cell membrane is the blue
bilayer on top in each diagram and viral membrane on bottom. The IFP is in green and N-terminal
part of ectodomain of HA2 is in red while the C-terminal domains are in navy. The HA1 is not
shown in this figure. (A) In the prefusion state, the protein is anchored to the viral membrane by a
C-terminal transmembrane domain. (B) pH decreasing to pH 5 in endosome triggers a
conformational change resulting in an extended intermediate and exposing IFP to the target
membrane. (C) The intermediate collapses. (D) Hemifusion stalk. (E) Fusion pore formation. As
the hemifused bilayers open into a fusion pore, the final zipping up of the C-terminal ectodomain
segments snaps the refolded trimer into its post-fusion conformation, preventing the pore from
resealing………………………………………………………………………………………….33
Figure 1.18 Geometry of the geometry of the 13C – 2H vector in solid state NMR sample under
MAS. The sample is spun rapidly in a cylindrical rotor about a spinning axis oriented at the magic
angle ( = 54.7) with respect to external magnetic field B0…………………………………….35
Figure 1.19 13C-2H REDOR NMR pulse sequence. The columns represent the /2 or  pulses. CP
= cross polarization that transfers 1H transverse magnetization to 13C and can enhance the 13C
signal. The CP is followed by a 13C-2H dipolar evolution for a period of time which is called
dephasing time (). Adjacent 13C  pulses are separated by one rotor period as are adjacent 2H 
pulses. 13C is the detecting channel……………………………………………………………….37
Figure 1.20 Evolution of dipolar coupling as a function of rotor period in REDOR experiments.
The dipolar coupling is averaged out over each rotor period by MAS. In S0 experiment, rotor
synchronized 13C  pulses do not interfere with the MAS averaging of the heteronuclear dipolar
interaction. In S1 experiment, rotor-synchronized 13C and 2H  pulses re-introduce the 13C-2H
dipolar interaction………………………………………………………………………………..39
Figure 1.21 13C-2H REDOR spectra with a 40 ms dephasing time for the I4 peptide as well as
S/S0 vs τ. Black squares are the experimental dephasing of I4 peptide. Blue triangles are the bestfit calculated by the SIMPSON program (without 2H relaxation) with a 22 Hz 13C-2H dipolar
coupling. The red line is the best-fit exponential buildup………………………………………..40
Figure 1.22 Quadrupolar echo (quecho) pulse sequence………………………………………..44
Figure 2.1 (A) Structure of DPPC-D4, DPPC-D8 and DPPC-D10. (B) Approximate membrane
locations of the 2H’s in the membrane bilayer without protein. The lipid 2H are for the membrane
gel-phase…………………………………………………………………………………………57
Figure 2.2 (A) Schematic diagrams of full-length HIV gp41 and the four gp41 constructs being
studied in this work with domains and corresponding colors: FP  fusion peptide, red; N-helix,
blue; Loop, grey; C-helix, green; MPER  membrane-proximal external-region, pink; TM 
transmembrane domain, orange; and endo  endodomain, white. The four constructs have nonnative SGGRGG replacing native residues 582-627. (B) Amino acid sequences with the same color

xii

coding as panel A and the non-native C-terminal G6H6 or G8H6 in black. The H6 is for Co2+-affinity
chromatography and the G6/G8 are necessary spacers for exposure of the H6 tag……………….58
Figure 2.3 DNA sequences of gp41 inserts. Each line is 75 nucleotides…………………………60
Figure 2.4 (A) 13C - 2H REDOR pulse sequence and (B) quecho pulse sequence……………….67
Figure 3.1 (A) Full-length HIV gp41 and the four gp41 constructs. (B) Amino acid sequences with
the same color coding as panel A and the non-native C-terminal G6H6 or G8H6 in black…….….77
Figure 3.2 Static 13C NMR spectra of lysate pellets labeled with 1-13C Gly. Each spectrum was
the sum of 1000 scans…………………………………………………………………………….78
Figure 3.3 SDS-PAGE after Co2+-affinity chromatography of the solubilized pellet enriched in
inclusion body protein. The protein is FP_HM and different solubilization conditions for the pellet
are noted. Only the MW marker lane and relevant elution lane(s) are displayed………………...81
Figure 3.4 SDS-PAGE of the purified HM (MW = 13.7 kDa), HM_TM (MW = 16.7 kDa), FP_HM
(MW = 16.5 kDa), and FP_HM_TM (MW = 18.9 kDa) using solubilization condition: PBS at pH
7.4 with 8M Urea, 0.5% SDS, and 0.8% Sarkosyl…………………………………………….….81
Figure 3.5 Sequence coverage of FP_HM, HM_TM, and FP_HM_TM after trypsin digestion….82
Figure 3.6 Circular dichroism spectra of samples containing ~10 ÂľM protein concentration in
different buffer + detergent solutions: (A) 10 mM Tris at pH 7.4 and 0.2% SDS; (C) 20 mM
phosphate at pH 7.4 and 0.25% DPC; and (D) 20 mM acetate at pH 4.0 and 0.25% DPC. The
spectra in panels A, C, and D spectra were obtained at ambient temperature. (B) the 222 values
derived from spectra in 0.2% SDS at pH 7.4 and temperatures between 25 and 90 °C. The panel A
vs. B differences between the ambient-temperature 222 values of the same construct may reflect
measurement uncertainties in protein concentrations and use of different CD instruments……...84
Figure 3.7 Circular dichroism spectra of four gp41 constructs with ~10 ÂľM protein concentration
in 10 mM Tris at pH 7.4 and 0.2% SDS at 25, 60 and 90 °C. The panels are noted corresponds to
each construct………………………………………………………………………………….…85
Figure 3.8 SEC of gp41 constructs under the following conditions: (A) 10 mM Tris at pH 7.4, 150
mM NaCl, 0.2% SDS, and ambient temperature; (B) 20 mM phosphate at pH 7.4, 150 mM NaCl,
and 0.25% DPC at 4 °C; and (C) 20 mM acetate at pH 4.0, 150 mM NaCl, and 0.25% DPC at 4
°C. SEC was obtained with a Superdex 200-increase column, 1 mg/mL protein loading with ~10fold dilution in the column, and A280 detection. The arrows in the plots are at the elution volumes
of the MW standards, and some of the peaks are identified with dashed lines and with MW’s
calculated from interpolation between MW standards…………………………………………....89
Figure 3.9 SEC traces of replicate samples in: (A) 10 mM Tris buffer at pH 7.4 with 0.2% SDS
and 150 mM NaCl; and (B, C) 20 mM phosphate buffer at pH 7.4 with 0.25% DPC and 150 mM
NaCl. The (C) replicates differ in the presence vs. absence of 2 mM DTT reducing agent.
Constructs and their replicate traces are labeled and dashed lines were used to show peak
alignment………………………………………………………………………………………....94
xiii

Figure 3.10 Vesicle fusion assays of gp41 proteins. Fusion was initiated by addition of an aliquot
of protein stock solution at 0 s, and subsequent fusion was monitored by increased fluorescence
associated with inter-vesicle lipid mixing. The stock contained 40 M protein in buffer at pH 7.4
with 0.2% SDS, and the protein + vesicle mixture contained [protein] = 0.5 M, [POPC+POPG]
= 150 M, and vesicle molar compositions and pH’s: (A) POPC:Chol = 2:1 at pH 3.2; (B)
POPC:POPG:Chol = 8:2:5 at pH 3.2; (C) POPC:Chol = 2:1 at pH 7.4; and (D) POPC:POPG:Chol
= 8:2:5 at pH 7.4. Fusion wasn’t observed for any vesicle composition after addition of an aliquot
of buffer + 0.2% SDS without protein. The assay dead time is ~5 s……………………………..98
Figure 3.11 Structural model of FP_HM_TM based on circular dichroism spectra of the four
constructs as well as other data. A monomer is shown for clarity but the model should be valid for
trimers and hexamers. Approximate residue numbers are displayed………………………..….101
Figure 3.12 Schematic illustrating (A) trimer and (B) monomer respectively favored in the
absence and presence of peptide inhibitor. Panel B displays “C34” inhibitor which contains Chelix residues 628-661. The sequence color coding matches Figure 3.1 and 3.11, and loops
between structured regions are not displayed for clarity. The FP’s from different trimers or
monomers adopt antiparallel  sheet structure. Fusion is enhanced in panel A vs. B because of
greater clustering of membrane-perturbing protein regions in the trimer vs. monomer. This
enhancement exists for the displayed hemifusion state as well as membrane states that precede
hemifusion…………………………………………………………………………………...….105
Figure 4.1 (A) The 20 lowest energy conformers of the H3_20 IFP in lipid bilayers by the EPR
data at pH 5 (red) and 7.4 (yellow). A phospholipid is shown for reference and nitrogen, oxygen
and phosphorus atoms are colored in yellow, red and green, respectively. The white line represents
the level of the lipid phosphate groups. The hydrophobic hydrocarbon is in dark gray and interface
is in light gray. (B) Membrane locations of closed structure H1_23 IFP. Dashed lines are the
hydrocarbon core. (C) Lipid tail protrusion induced by IFP. IFP, lipid tails and phosphates are in
green, gray and orange, respectively. A thin gray plane shows the average phosphorus position in
the upper leaflet. One lipid is shown in sticks, with an acyl tail protruding into the polar layer. (D)
IFP deeply inserted in one membrane leaflet at low pH. Hydrophobic residues are highlighted in
yellow and hydrophilic residues in red. Phosphorus and nitrogen atoms are in tan and blue,
respectively. (E) Membrane-spanning conformation model of IFP in membrane at low pH. IFP,
phosphate group, amine group and acyl chain are in purple, orange, blue and green, respectively.
(F) IFP binds on membrane surface at neutral pH. N-terminal helix, C-terminal helix and
hydrophobic side chains are in red, blue and white, respectively……………………………….116
Figure 4.2 H2_20 IFP sequence, closed structure model and semi-closed model. Labeled amino
acids are in red in sequence and orange in the cartoon model………………………………….118
Figure 4.3 13CO–2H REDOR data of samples that contain either IFP_L2c in membrane with
DPPC:DPPG 4:1 ratio. Samples were prepared with DPPC_D4, DPPC_D8 and DPPC_D10, and
the corresponding data are displayed in purple, green and red, respectively. (A) S0 (black) and S1
(colored) REDOR spectra for  = 40 ms. (B) and (C) Plots of (S/S0) vs  are displayed for the 
helical and  strand peaks. The solid lines are best-fits to A(1 – e-) and the fitting parameters are
in Table 4.1……………………………………………………………………………………...119

xiv

Figure 4.4 13CO–2H REDOR data of samples that contain either IFP_A7c in membrane with
DPPC:DPPG 4:1 ratio. Samples were prepared with DPPC_D4, DPPC_D8 and DPPC_D10, and
the corresponding data are displayed in purple, green and red, respectively. (A) S0 (black) and S1
(colored) REDOR spectra for  = 40 ms. (B) and (C) Plots of (S/S0) vs  are displayed for the 
helical and  strand peaks. The solid lines are best-fits to A(1 – e-) and the fitting parameters are
in Table 4.1………………………………………………………………………………….…..120
Figure 4.5 13CO–2H REDOR data of samples that contain either IFP_G16c in membrane with
DPPC:DPPG 4:1 ratio. Samples were prepared with DPPC_D4, DPPC_D8 and DPPC_D10, and
the corresponding data are displayed in purple, green and red, respectively. (A) S0 (black) and S1
(colored) REDOR spectra for  = 40 ms. (B) Plot of (S/S0) vs  are displayed for the  helical
peak. The solid lines are best-fits to A(1 – e-) and the fitting parameters are in Table 4.1…….121
Figure 4.6 13CO–2H REDOR data of samples that contain either IFP_L2c in membrane with DPPC
prepared by organic co-solubilization method. Samples were prepared with DPPC_D8 and
DPPC_D10, and the corresponding data are displayed in green and red, respectively. (A) S0 (black)
and S1 (colored) REDOR spectra for  = 40 ms. (B) Plot of (S/S0) vs  are displayed for the 
strand peak. The solid lines are best-fits to A(1 – e-) and the fitting parameters are in Table
4.3……………………………………………………………………………………………….127
Figure 4.7 Possible models for IFP topology in membrane. The lipid molecules are shown in blue,
D10 2H atoms in lipid acyl chain shown as red dots, IFP in green and labeled residues in orange.
(A) Membrane spanning model and (B) lipid tail protrusion model………………………..….131
Figure 5.1 Labeling scheme of PRE experiments. Paramagnetic species are Mn2+ ions, and they
bind to the surface of lipid membrane. Lipid composition used was DPPC:DPPG 4:1 mole ratio.
DPPC-D8 and DPPC-D10 were used. The hydrophobic core is 34 Å, which is a comparable length
with the PRE detection range. The Mn2+ ions are labeled in red, lipid molecules in blue and
deuterium atoms shown as red dots. (A) Lipid acyl chain without protrusion. (B) Lipid acyl chain
protrudes towards aqueous surface……………………………………………………………...142
Figure 5.2 Quecho pulse sequence. 2H spectra were acquired with different 1 and 2 and a fixed
(1 -2) value. Typically, the pulse length was set ≈1.5 s; 1, 2 were set between 10 and 1000 s;
the recycle delay was set to 1 s. Samples are detected at 25 °C (gel phase lipids) and 50 °C (fluid
phase lipids)………………………………………………………………………………….…142
Figure 5.3 D10 at 25 °C. Top: FID. Middle: FID stacked plots with increasing 1 and 2. Bottom:
spectra stacked plots. All spectra were acquired with 5000 scans, processed with 300 Hz
exponential line broadening, data shift = -11, and baseline correction………………………...146
Figure 5.4 D10 at 50 °C. Top: FID. Middle: FID stacked plots with increasing 1 and 2. Bottom:
spectra stacked plots. All spectra were acquired with 5000 scans, processed with 300 Hz
exponential line broadening, data shift = -11, and baseline correction…………………………147
Figure 5.5 D8 at 25 °C. Top: FID. Middle: FID stacked plots with increasing 1 and 2. Bottom:
spectra stacked plots. All spectra were acquired with 5000 scans, processed with 1000 Hz
exponential line broadening, data shift = -11, and baseline correction………………………….148
xv

Figure 5.6 D8 at 50 °C. Top: FID. Middle: FID stacked plots with increasing 1 and 2. Bottom:
spectra stacked plots. All spectra were acquired with 5000 scans, processed with 1000 Hz
exponential line broadening, data shift = -11, and baseline correction…………………………149
Figure 5.7 2H NMR spectra D10 or D8 at 25 or 50 °C. Spectra were acquired with 5000 scans,
data shift = -11, and baseline correction. D10 spectra were processed with 300 Hz exponential line
broadening, and D8 with 1000 Hz exponential line broadening……………………………..…150
Figure 5.8 Quecho experimental (black squares) and best fit (red lines) plots of D10 and D8 at 25
and 50 °C………………………………………………………………………………………..151
Figure 5.9 2H NMR spectra of D10 with different mole percentage of Mn2+………………….154
Figure 5.10 Quecho experimental (black squares) and best fit (red lines) plots of D10 with various
concentration of Mn2+ at 25 °C………………………………………………………………….156
Figure 5.11 Quecho experimental (black squares) and best fit (red lines) plots of D10 with various
concentration of Mn2+ at 50 °C………………………………………………………………….157
Figure 5.12 IFP_D10 at 25 °C. Top: FID. Middle: FID stacked plots with increasing 1 and 2.
Bottom: spectra stacked plots. All spectra were acquired with 5000 scans, processed with 300 Hz
exponential line broadening, data shift = -11, and baseline correction…………………………160
Figure 5.13 IFP_D10 at 50 °C. Top: FID. Middle: FID stacked plots with increasing 1 and 2.
Bottom: spectra stacked plots. All spectra were acquired with 5000 scans, processed with 300 Hz
exponential line broadening, data shift = -11, and baseline correction…………………………161
Figure 5.14 2H NMR spectra of D10 samples at 25 and 50 °C with 5% Mn2+…………………162
Figure 5.15 2H NMR spectra of D8 samples at 25 and 50 °C with 5% Mn2+……………………163
Figure 5.16 Quecho experimental (black squares) and best fit (red lines) plots of D10, D10_Mn,
IFP_D10 and IFP_D10_Mn with 5% Mn2+ at 25 °C……………………………………………164
Figure 5.17 Quecho experimental (black squares) and best fit (red lines) plots of D10, D10_Mn,
IFP_D10 and IFP_D10_Mn with 5% Mn2+ at 50 °C…………………………………………...166
Figure 5.18 Quecho experimental (black squares) and best fit (red lines) plots of D8, D8_Mn,
IFP_D8 and IFP_D8_Mn with 5% Mn2+ at 25 °C………………………………………………167
Figure 5.19 Quecho experimental (black squares) and best fit (red lines) plots of D8, D8_Mn,
IFP_D8 and IFP_D8_Mn with 5% Mn2+ at 50 °C……………………………………………...168
Figure 5.20 Reproducibility of D8 at 25°C. The quecho experimental (black squares) and best fit
(red lines) plots as well as the fitted T2 and R2 values are shown………………………………172
Figure 5.21 2H NMR spectra of D10 samples at 25 and 50 °C with 20% Mn2+………………..173

xvi

Figure 5.22 Quecho experimental (black squares) and best fit (red lines) plots of D10_Mn’ and
IFP_D10_Mn’ with 20% Mn2+ at 25 °C and 50 °C……………………………………………..174
Figure 5.23 Model for IFP-bounded membrane. The lipid molecules are shown as blue sphere
with two lines, 2H atoms in lipid acyl chain shown as red dots, IFP in green, labeled amino acids
in orange and Mn2+ binds on membrane surface………………………………………………...178

xvii

KEY TO ABBREVIATIONS

AA

Amino acid

AB

Aqueous binding

AIDS

Acquired immunodeficiency syndrome

AUC

Analytical ultracentrifugation

CD

Circular dichroism

CDC

Centers of Disease Control and Prevention

Chol

Cholesterol

CHR

C-heptad repeat

CP

Cross polarization

DCM

Dichloromethane

DEPBT

3-(Diethylphosphoryloxy)-1,2,3-benzotriazin-4(3H)-one

DIEA

N,N-Diisopropylethylamine

DM

n-Decyl- β-D-Maltoside

DMF

Dimethylformamide

DNA

Deoxyribonucleic acid

DPC

n-Dodecylphosphocholine

DPPG

1,2-Dipalmitoyl-sn-glycero-3-phosphocholine

DTPC

1,2-di-O-tetradecyl-sn-glycero-3-phosphocholine

E.coli

Escherichia coli

Endo

Endodomain

EPR

Electron paramagnetic resonance

xviii

FID

Free induction decay

Fmoc

Fluorenylmethyloxycarbonyl

FP

Fuison peptide

FWHM

Full width at half maximum

GuHCl

Guanidinium chloride

HA

Hemagglutinin

HBTU

O-(benzotriazol-1-yl)-N,N,N’,N’-tetramethyluronium hexafluorophosphate

HEPES

4-(2-Hydroxyethyl)piperazine-1-ethanesulfonic acid

HIV

Human immunodeficiency virus

HOBt

1-Hydroxybenzotriazole

HPLC

High-performance liquid chromatography

IFP

Influenza fusion peptide

IPTG

Isopropyl β-D-thiogalactopyranoside

LB

Luria-Bertani broth

MAS

Magic-angle spinning

MD

Molecular dynamics

MES

2-(N-morpholino)ethanesulfonic acid

MPAA

4-Mercaptophenylacetic acid

MPER

Membrane-proximal external region

NA

Neuraminidase

NDB-PE

N-(7-Nitrobenz-2-oxa-1,3-diazol-4-yl)-1,2-dihexadecanoyl-sn-glycero-3phosphoethanolamine

NHR

N-heptad repeat

NMR

Nuclear magnetic resonance

xix

NOE

Nuclear Overhauser effect

OC

Organic co-solubilization

PBS

Phosphate-buffered saline

POPC

1-Palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine

POPG

1-Palmitoyl-2-oleoyl-sn-glycero-3-[phospho-rac-(1-glycerol)]

PRE

Paramagnetic relaxation enhancement

Quecho

Quadrupolar echo

REDOR

Rotational-echo double-resonance

Rh-PE

N-(Lissamine rhodamine b sulfonyl)-1,2-dihexadecanoyl-sn-glycero-3phosphoethanolamine

RMSD

Root-mean-square deviation

RNA

Ribonucleic acid

RP

Recombinant protein

Sarkosyl

Sodium lauryl sarcosinate

SDS

Sodium dodecyl sulfate

SDS-PAGE

Sodium dodecyl sulfate polyacrylamide gel electrophoresis

SEC

Size-exclusion chromatography

SHB

Six helical bundle

SPR

Surface plasmon resonance

t-Boc

tert-Butyloxycarbonyl

TCEP

Tris(2-carboxyethyl)phosphine

TFA

Trifluoroacetic acid

THF

Tetrahydrofuran

TM

Transmembrane

xx

vRNP

Viral ribonucleoproteins

WHO

World Health Organization

DTPG

1,2-di-O-tetradecyl-sn-glycero-3-phospho-(1'-rac-glycerol)

xxi

Chapter 1
Introduction to HIV gp41 and influenza hemagglutinin
fusion proteins

1

1.1 Introduction to HIV gp41 fusion protein
1.1.1 HIV virus structure and its infection pathway
The Acquired Immunodeficiency Syndrome (AIDS) is a disease that causes human
immune system failure and it is caused by the Human Immunodeficiency Virus (HIV).1 HIV/AIDS
is one of the most significant infectious disease challenges in the world. According to World Health
Organization (WHO), there were approximately 36.7 million people living with HIV and 1.1
million people died from HIV related diseases in 2015. There is no cure or vaccine for HIV. The
HIV antiretroviral drugs can suppress the viral replication in human and allow human immune
system to recover its capability to defend infection. However, only 46% of AIDS patients are using
these antiretroviral drugs and the cost of the drug is around $20,000 per year.
HIV virus has two types: HIV-1 and HIV-2, and the current studies mainly focus on the
HIV type 1 virus. In the following contents in this dissertation, HIV refers to HIV type 1 virus.
HIV virus has a diameter about 100-120 nm and its structure is shown in Figure 1.1.2 HIV is in the
genus Lentivirus, which is part of the family Retroviridae. HIV virus is an enveloped virus,
meaning the virion is wrapped by lipid membrane. There are ~72 “knob” on the viral membrane,
which are the viral glycoprotein trimers.3 These proteins are cleaved inside the host cell from a
precursor called gp160 into two noncovalently bonded glycoproteins: external surface protein
gp120 and transmembrane protein gp41 which are responsible for virus binding to host cell
membrane and promoting viral and host cell membrane fusion, respectivley.4 The matrix proteins
forms a sphere underneath membrane to ensure the viral structure and integrity of the virion. There
are two identical positive single-stranded RNAs enclosed by a conical shaped capsid. The reverse
transcriptase and nucleocapsid proteins are also closely associated to the RNA.3, 5, 6

2

Figure 1.1 Cutaway schematic of the structure of an HIV virion.

3

Figure 1.2 (A)Stages of the HIV‑1 life cycle. The yellow boxes in the diagram show the possible
antiretroviral therapy mechanism.7 (B) Schematic infection model of HIV. The spikes on virion
are gp120/gp41 complex. (a) Binding of HIV on host cell, (b) Hemifusion of viral and host cell
membrane, (c) Pore formation, and (d) complete fusion and entry of viral genetic material into the
host cell. (C) Electron microscopy pictures of HIV infection process correspond to panel B. 8 (D)
Schematic infection model of an alternative HIV entry pathway via clathrin-mediated
endocytosis.8, 9

4

The infection cycle of HIV is shown in Figure 1.2 panel A. This process is initiated by the
binding of gp120 onto the host cell. HIV infects vital cells in the human immune system such as
CD4+ T cells and macrophages.3, 10, 11 The gp120 typically binds to the CD4 molecules on the host
cell where gp120 and CD4 forms a cavity-laden interface and this binding can cause structural
changes in gp120. This conformation change facilitates the coreceptor, such as CXCR4 and CCR5,
which are in the chemokine receptor family, to bind to a conserved binding site on gp120.12, 13
After the binding, gp120 is displaced and results in the exposure of gp41.The function of gp41 is
to catalysis the fusion between viral and host cell membrane and during the fusion process a large
conformational change happens.14, 15 At the end of the fusion, a fusion pore is formed and the
nucleoprotein complexes can get into the host cell. Figure 1.2 panel B and C are the schematic
diagram and electron microscopy pictures of the fusion process.8 On the other hand, endocytosis
is another alternative entry pathway for HIV. Electron microscopy results showed that HIV can
be endocytosed via clathrin-coated pits (Figure 1.2 D).8,

9

After fusion, the RNA is reverse

transcribed into DNA catalyzed by reverse transcriptase and then the DNA can get into the host
cell nucleus and be integrated into the host cell genome. Then the viral proteins can be expressed
and finally new viruses bud from the host cell membrane and start the new infection cycle.16 The
antiretroviral drugs are designed based on the virus replication process. To prevent the drug
resistance, antiretroviral drugs are usually administered in combination and help the patient to live
a near-normal life span.7

1.1.2 Structure of gp41
The HIV envelop glycoproteins forms a “knob” shape complex on the viral membrane in
a noncovalent manner where three molecules of gp120 form the “cap” and three molecules of gp41

5

form the “stem” (as shown in Figure 1.2 on the virus surface).7, 14 The function of gp120 includes
the target cell binding while gp41 fuses the viral and host cell membrane to release the RNA into
host cell. The gp120 and gp41 are the product from a precursor gp160 which is expressed from the
viral gene and be cleaved in the host cell .3 Since gp41 plays a key role in membrane fusion, it has
been a target protein in HIV vaccine and antiretroviral drug development. Gp41 is a
transmembrane protein with 345 amino acids. The schematic diagrams of full-length HIV gp41
and the corresponding amino acid sequence are shown in Figure 1.3. Numerous of previous work
illustrates that the gp41 undergoes a huge conformational change before and after fusion. X-ray
crystallography and cryo-electron microscopy are used to study the structure of gp41 and big
progresses has been made for reveal the structure and its possible fusion mechanism.14, 15, 17-19
Chan and his coworkers14 found out the high resolution X-ray crystallographic structure
using a protein-dissection approach by which they were able to obtain a water soluble, highly
stable complex consists of two peptide fragments denoted N36 and C34. The N36 peptide
corresponds to residue 546 – 581, which is in the NHR region, while C34 peptide are residue 628
– 661 and it is in the CHR region. The complex is a six-helical bundle (SHB) whose center consists
of a parallel trimeric coiled-coil of three N36 helices and three C34 helices wrap antiparallel
outside of the center trimer (Figure 1.4). This complex has a ~35 Å diameter and ~55 Å height.
The N36 coiled-coil adopts the “knobs-into-holes” packing, which allows -branched residues in
the complex center to pack into the cavities in the nearby helices. The three C34 helices pack into
the highly conserved and hydrophobic grooves on the N36 trimer surface. The analog single chain
protein N34(L6)C28 (residue 546 – 579 in NHR, six residue linker SGGRGG and residue 628 –
655 in CHR) also has confirmed the formation of -helix trimer and showed high termostability.20
This SHB structure is believed to be the structure of gp41 at post-fusion state.21, 22

6

Figure 1.3 (A) Schematic diagrams of full-length HIV gp41 and corresponding colors: FP  fusion
peptide, red; NHR  N-helix region, blue; Loop, grey; CHR  C-helix region, green; MPER 
membrane-proximal external-region, pink; TM  transmembrane domain, orange; and endo 
endodomain, white. (B) Amino acid sequences with colors matching segments in panel A except
for the endo domain in black. The sequence is from the HXB2 laboratory strain of HIV and have
the gp160 precursor residue numbering, 1-511 for gp120 and 512-856 for gp41.

7

Figure 1.4 (A)Sequence of the N36/C34 complex. (B) The end-on view of the N36/C34 complex
looking down the three-fold axis of the trimer. The N36 residues are in blue and the C34 residues
are in red. (C)The side view of the N36/C34 complex. The amino termini of the N36 helices point
toward the top of the page, while those of the C34 helices point toward the bottom.14

8

Other researchers15,

19, 22

studied the structure of soluble, cleaved HIV envelope

glycoprotein trimer by X-ray crystallography and cryo-electron microscopy with ~ 5 Å resolution.
In these studies, the gp120/gp41 complex were obtained from HIV virion and stabilized by
antibodies, and the structure was at a prefusion state of gp41. The schematic diagram and the
structure for the complex is shown in Figure 1.5. In this study15, gp41 NHR formed a ~50 Å helix
trimer bundle and an additional short helix extended but was kinked away from the trimer axis.
CHR had a ~60° angle relative to NHR and it wrapped around central trimer bottom (Figure 1.5 B
and C). This NHR/CHR interaction in pre-fusion state was completely different compare to the
post fusion state, where NHR and CHR formed closely packed SHB.14 Similar structures were
observed from cryo-electron microscopy results.19,

22

The difference of the NHR coiled-coil

structure in pre- and post- fusion stated is shown Figure 1.6.22 There was a 15° angle change in
each N-terminal helices from the pre-fusion state (Figure 1.6 A) to the post-fusion state (Figure
1.6 B). In addition, in the post-fusion state, the N-terminal helices became more compactly packed
and the NHR/CHR SHB was formed.
Although the X-ray crystallography and cryo-electron microscopy results illustrated a lot
details about the core domain of gp41 before and after fusion, the membrane-associate MPER and
TM domain were truncated in gp140 proteins, and the FP structure was missing.

9

Figure 1.5 (A) Schematic of the HIV gp140 construct studied15 in comparison to full-length gp160.
N-linked glycans are shown and numbered on their respective Asn residues. The FP, NHR, CHR,
MPER, transmembrane (TM), and endodomain elements in gp41 are indicated. The mutations are
shown in red, as well as the added N332 glycan site. The color coding is preserved in (B) and (C).
(B) Side view of the gp140 trimer. The main domains are labeled correspond to panel A and
glycans are shown as spheres. (C) gp41 in gp140 complex. Secondary structure determination was
ambiguous at the dashed line areas.

10

Figure 1.6 Compare of NHR coiled-coil trimer at pre- and post-fusion states. (A) Top view of the
locations of the three N-terminal helices at pre-fusion state derived from cryo-electron microscopy.
(B) Top view of the locations of the same helices in the post-fusion state derived by X-ray
crystallography. (C) Superposition of the arrangement of the three N-terminal helices in pre-(cyan)
and the post-fusion (magenta) conformation.

11

The fusion peptide (FP) is first 23 amino acid in the N-terminus of gp41 with sequence:
AVGIGALFLGFLGAAGSTMGARS. FP is a key domain for fusion, and the FP alone can
promote lipid vesicle fusion.23 Even the structure of FP was not clear in gp41 crystal or cryoelectron microscopy results, there were a lot of NMR research on the peptide alone in both
detergent micelles and lipid membrane. The FP formed Îą-helical structure in the SDS or DPC
detergent micelles24-26, however, other studies showed that FP formed antiparallel β-sheet structure
in physiologically relevant membrane environment.27-30 In membrane containing ∟30 mol %
cholesterol, the first 16 residues of FP had 0.30 fraction of antiparallel β sheet secondary structure
with residue 16→1/1→16 and 17→1/1→17 registries for adjacent strands (Figure 1.7 A) and
parallel β sheet with up to 0.15 fraction. The antiparallel conformation was also confirmed in a
larger gp41 construct.

27, 30

FP had two membrane locations: major deeply-inserted and minor

shallowly-inserted (Figure 1.7 B).

29

By studying the membrane location and fusogenicity of FP

wild type and V2E mutation, a positive correlation between the insertion depths and the fusion
activities was observed.28
Membrane proximal external region (MPER) is the ectodomain region that is closest to the
viral membrane. The MPER is 22 amino acid tryptophan-rich domain and the sequence is
ELDKWASLWNWFNITNWLWYIK. Liquid state NMR was used to study the MPER peptide in
DPC micelle and the result demonstrated a well-defined -helical peptide (Figure 1.8 A).

12

Figure 1.7 Models of FP in membrane. (A) Antiparallel β sheet FP with residue 16→1/1→16 or
17→1/1→17 registries for adjacent strands binding to membrane.27 The FP is shown as red lines
and the lipids are the blue sphere with gray lines.(B) Two membrane locations of FP studied by
13

C-2H REDOR: majorly deep and minorly shallow insertion. The FP is 13CO labeled on Gly_5,

Leu_12 and Gly_16 residues. The labels on both side of the diagram are the approximate
membrane locations of the 2H’s and

P’s where P  phosphorous in lipid; Chol_d6, Chol_d7,

31

PC_d4, PC_d8 and PC_d10 are deuterium in cholesterol or phosphocholine lipids. 29

13

Figure 1.8 Models of MPER and TM. (A) MPER665-683 in DPC micelle at pH 3.5 showing a
straight -helix;31 (B) MPER662-683 in DPC micelles at pH 6.6 showing a L-shape kink between
two -helices;32 (C) Crystallography result of NHR547–575-CHR630-662-MPER663-675 monomer33.
The CHR and MPER forms a continuous helix and the MPER in this structure ends at I675. (D)
Molecular dynamics (MD) simulation result of TM681-707 in membrane with Arg694 snorkeling
toward the membrane surface. TM peptide is shown in grey, Arg residues in blue, lipid molecules
in yellow and water molecules in red.34 (E) Liquid-state NMR result of MPER675-683-TM684-704
peptide. The MPER and TM form a continuous -helix with a turn at 690GGLV693 residues.35

14

Additionally, the MPER peptide showed an amphipathic structure where the aromatic
residues were positioned on one side of the helix while the polar residues were on the other side,
suggesting the MPER was attached to the viral membrane.31 A newer study using liquid-state NMR,
electron paramagnetic resonance (EPR), and surface plasmon resonance (SPR) techniques
revealed that MPER had two discrete helical segments with a central hinge, forming an L-shape
(Figure 1.8 B).32 A crystal structure of gp41 containing NHR, CHR and MPER region revealed a
trimeric, coiled-coil SHB and the MPER extended from CHR in continuation of a slightly bent
long helix (Figure 1.8 C).33
The transmembrane (TM) domain is a 22-residue domain that anchors the gp41 in the viral
membrane. Based on computer simulations data, it acquired a tilted Îą-helical conformation in
membrane and had tendency to form a trimer (Figure 1.8 D).34, 36 Liquid state NMR also revealed
that MPER and TM forms an uninterrupted long -helix (Figure 1.8 E).35 Even there were some
differences in the experimental conditions such as pH and the construct length, the MPER and TM
domains all showed high helicity. Although there is no high-resolution structure for the whole
gp41 protein, these isolated ectodomain or membrane segments structures provide a lot of
information and help further studies.

1.1.3 Possible membrane fusion mechanism
As concluded in the previous sections, after the gp120 binds to CD4 and coreceptor, gp41
is exposed and the fusion process starts. The gp41 undergoes a large conformational change: from
a NHR/60°-turn/CHR structure to become a NHR/180°-turn/CHR SHB structure. Like gp41,
distinct change also happens in membrane structures during fusion (Figure 1.9).37 The fusion starts
with the lipid contact between viral and host cell membranes outer leaflet followed by stalk

15

formation where the outer leaflets fuse while the inner leaflets stay unfused. Hemifusion
intermediate characterized by intermembrane lipid mixing and no contents mixing. This is
followed by breaking the hemifusion barrier and formation of a fusion pore.
There are several models being proposed to explain the mechanism of gp41 catalyzed
fusion.
Model 118: (1) gp120 binding and gp41 exposure; (2) formation of an “prehairpin
intermediate” (PHI) state followed by FP insertion into host cell membrane. The PHI is
hypothetically a NHR/0°-turn/CHR SHB structure and the NHR and CHR regions stay as trimer
bundle; (3) NHR and CHR in PHI state fold into SHB which consequently brings the two
membranes closer; (4) viral and host cell membrane hemifusion; (5) initial fusion pore formation;
and (6) fusion pore expansion (Figure 1.10). It was also proposed that the free energy released by
PHI folding to SHB is used to overcome the energy barrier during fusion process. However, there
are some evidence that is contradictory to this hypothesis. Either NHR, CHR, CHR+MPER or
NHR+loop+CHR peptides can inhibit fusion since these peptides were presumed to bind to the
PHI and stop the formation of SHB.20, 38, 39 These functional studies showed that these peptides
stopped the fusion pore expansion which demonstrated that the pore formation happened before
PHI → SHB folding. There were also evidence supporting that the hemifusion happens before
SHB forms.40 Thus Model 2 was proposed:

16

Figure 1.9 Stages of membrane fusion.

d
a

c

b

Figure 1.10 Schematic diagram of Model 1.18 (a) Trimer gp120 and gp41 at pre-fusion state; (b)
displacement of gp120 and PHI formation; (c) SHB formation; (d) gp41 at the SHB post-fusion
state. ‘A’ represents the transmembrane domain and ‘F’ represent the FP domain.

17

Model 230: (1) gp120 binding and gp41 exposure; (2) formation of PHI followed by FP
insertion into host cell membrane; (3) Viral and host cell membrane hemifusion; (4) initial fusion
pore formation; (5) PHI → SHB trimer folding; and (6) fusion pore expansion (shown in Figure
1.11). However, the oligomeric state studies at low pH or neutral pH with 6M GuHCl showed that
gp41 ectodomain formed stable hairpin monomers as well as stable hexamers, which is
contradictory with the trimeric PHI and SHB state in Model 2.41

Figure 1.11 Schematic diagram of Model 2.30 (a) Trimer gp120 and gp41 at pre-fusion state; (b)
PHI formation and IFP inserted in host cell membrane; (c) Viral and host cell membrane
hemifusion; (D)gp41 at the SHB post-fusion state. FP is shown in red, NHR in blue, CHR in green
and MPER in white.

18

Model 341: (1) gp120 binding and gp41 exposure; (2) dissociation of gp41 ectodomain into
monomers and formation of extended PHI followed by FP insertion into host cell membrane; (3)
PHI → hairpin monomer folding that brings the two membranes close together; (4) hemifusion;
(5) initial pore formation; (6) hairpin monomer → SHB trimer → hexamer ectodomain assembly;
and (7) fusion pore expansion (shown in Figure 1.12)

Figure 1.12 Membrane fusion Model 3 of gp41 ectodomain monomer and hexamer. The different
domains of gp41 are color coded the same as Figure 1.3 and the TM and endodomain are not
shown. One of the monomers is not displayed in steps 3−5. The initial gp41 structure of step 1 and
the final SHB structure of step 7 are based on high-resolution structures.14, 15, 41

19

As shown in Model 3, the PHI → SHB transition initializes the following hemifusion and
fusion pore formation. The PHI → SHB monomer folding was also supported by hyperthermostable ectodomain monomer. Besides, the stable hexamer is consistent with the requirement
of multiple gp160 trimers for membrane fusion and HIV infection.41, 42

20

1.2 Introduction to influenza hemagglutinin fusion peptide
1.2.1 Influenza virus structure and its infection pathway
Influenza virus causes contagious respiratory illness that infects the nose, throat, bronchi
and lungs. Not only for human, influenza viruses also infect other mammals and birds.43 According
to the World Health Organization’s report in 2009, even though influenza vaccine is the most
effective way to avoid infection and severe consequences, influenza virus still resulted in about 35 million cases of severe illness, and about 250,000 to 500,000 deaths worldwide. There were three
other major influenza pandemics the past 100 years: 1918, 1957 and 1968, and each pandemic
causes over one million of deaths. For example, in the 1918 Spanish flu, 500 million people were
infected and 50 million of deaths across the world. Flu vaccines are widely used in the United
States. However, people need to get vaccinated each year because of the influenza virus mutates
rapidly. This variability results from different mechanisms including point mutations (antigenic
drift) and gene reassortment (gene shift).44 Therefore, even with having a flu vaccine, influenza
viruses continue to pose a significant health risk to public worldwide.
Influenza virus has four types: A, B, C and D. Human influenza A virus is the most
widespread type and it is the cause influenza pandemics.45 According to Centers of Disease Control
and Prevention (CDC), Influenza A and B viruses cause seasonal epidemics, and influenza C only
causes a mild respiratory illness and are not thought to cause epidemics, while influenza D viruses
primarily affect cattle and are not known to infect or cause illness in people. The following
discussion in this dissertation will be focused on influenzas type A virus and will be abbreviated
as “influenza virus”.
Influenza virus is a member of orthomyxovirus family and it is an enveloped virus. The
viral membrane comes from the budding when the virion leaves the host cell. There are three

21

membrane proteins on the viral membrane: hemagglutinin (HA), neuraminidase (NA), and M2 ion
channel. HA is an integral membrane protein and there are ~ 400 copies of HA trimers per virion.
46, 47

HA is responsible for binding of virus to host cell and is involved in membrane fusion. HA is

first expressed in the host cell as a precursor called HA0. The HA0 is cleaved into two subunits,
HA1 and HA2 by a protease on a single Arginine residue. They are linked by a disulfide bond.
HA1 has ~328 residues, including a receptor-binding site which can bind to sialic acids on
carbohydrate side chains of cell-surface glycoproteins and glycolipids. HA2 has ~211 residues and
the first ~20 residues at the N-terminus of HA2 is the influenza fusion peptide (IFP) and its function
is essential for the membrane fusion.48-50 NA is also an integral membrane glycoprotein with ~ 100
copies of tetramer per virion. The function of NA is to cleave terminal sialic acid from
glycoproteins or glycolipids and free the virion form host cell receptors then the virus can spread
and infect other cells.44 HA and NA are highly mutable proteins and influenza A type virus is
divided into strains based on the subtype of these two proteins. There are 18 different HA subtypes
(H1 to H18) and 11 different NA (N1 to N11) subtypes and the strain of influenza is defined by
the combinations of HA and NA, for example: H1N1, H3N1 etc.48 M2 is a tetramer integral
membrane protein act as proton ion channel. It lowers the pH to uncoat the virus.51 The M1 protein
is the most abundant protein in virion and it forms a shell surrounding the virion nucleocapsids
inside of the envelope. The influenza viruses have eight unique segments of single-stranded RNA
and the RNAs are loosely encapsidated by NP, which is the second most abundant protein. Viral
polymerase proteins (PA, PB1 and PB2) are attached at the end of the genomic segments and
provides RNA-dependent RNA polymerase activity. The complex of viral RNA, NP and
polymerase proteins are called viral ribonucleoproteins (vRNP). The NS1(host antiviral response
antagonist) and NS2 (nuclear export protein) are not incorporated into progeny virions, but they

22

present in infected cell nucleus (NS1) and cytoplasm (NS2). NS2 is responsible to export viral
RNAs from the nucleus into the cytoplasm and pack viral RNAs into newly formed viruses.44, 52,
53

Figure 1.13 demonstrates the structure of a virion.
The expression and replication of influenza viruses is a multi-step process, shown in Figure

1.14. An influenza virion with cleavage-activated HA recognizes and binds to the host cell through
the binding of subunit on HA1 onto the terminal sialic acid of the host cell surface glycoproteins
and glycolipids. Then the virion is endocytosed by the host cell. Subsequent lowering of pH to 56 in the endosome causes conformational changes of HA and the fusion peptide is exposed to
trigger the fusion between viral membrane and endosomal membrane.54-56 The pH in virion is
lowered by M2 ion channel, removing the M1 proteins from vRNPs, which is also called uncoating.
The genome of the virus is released into the cytoplasm and then imported into the nucleus for viral
gene transcription and replication.57 RNAs are transcribed into mRNA and the translation happens
in the host cell cytosol. Newly synthesized viral proteins are assembled into RNPs in the nucleus
and transported to the host cell membrane, while synthesized membrane proteins are integrated
into the membrane. Finally, budding of new virus happens where host cell plasma membrane
becomes the viral envelope.44, 53

23

Figure 1.13 Schematic representation of the structure of influenza virus.53 Viral envelope is shown
as gray sphere and the membrane proteins (HA, NA and M2) are shown as spikes on the membrane.
Inside the membrane is M1 that surrounds the viral ribonucleoproteins (vRNP). There are eight
single-stranded, negative-sense RNA segments and each encoded one or two proteins.

24

Figure 1.14 Influenza viral life cycle.53

25

1.2.2 Structure of HA2
HA is synthesized as a single polypeptide chain called HA0 and cleaved to HA1 and HA2
two subunits, which are linked by a disulfide bond. There are six glycosylation sites on HA1 and
one on HA2 with ~13,000 MW.56 The HA1 has ~328 residues and HA2 has ~211 residues, and
the glycosylated site on HA2 is Asn154.58 HA1 has a sialic acid bind site while HA2 is a
transmembrane protein and it is responsible for catalyzing membrane fusion.57 The structure of
HA2 and its fusion mechanism is not fully understood. By studying HA2 may provide more
information for vaccine and drug development to combat the virus.
As discussed in section 1.2.1, the infection process of influenza virus undergoes an
endocytosis pathway. The pH of the endosome decreases from pH 7.4 to 5 or 6 and triggers the
conformation change of HA2. Extensive studies have been made to elucidate the structure of HA
at these two pHs, or in other words, the pre- and post-fusion state (Figure 1.15). At pre-fusion state,
the oligomeric state of HA is trimer. The crystal structure of HA2 ectodomain containing residue
38 – 170 is shown in Figure 1.15 panel A. The N-terminal -helix (29 Å) extended away from the
membrane and connected by an extended chain to a very long -helix (76 Å) that stretched back
towards the membrane. In the trimer, the long helix formed the coiled-coil core with two identical
helices from other subunits.47 At post-fusion state, HA2 also formed three-stranded -helix coiledcoil, approximately. The monomer HA2 from residue 38 - 175 was shown in Figure 1.15 panel B.
The center helix including residue 40 – 105 followed by a 180˚ turn and helix including residue
113 - 129 packed antiparallelly outside.49, 55 There was a big conformational change between these
two conformations (Figure 1.15 C). From pre- to post-fusion state, helix A was relocated on top
of the triple-stranded -helix coiled-coil. Loop B adopted an -helix conformation and region A
– C formed a ~100 Å long helix. A portion of helix CD (residue 106 - 112) refolded to form a loop

26

and region D was antiparallel to the center bundle. Thus, region E, F and G were packed in an
inverted orientation.55 The region E and F were no longer antiparallel -strands in the post fusion
state.49 Details about the fusion mechanism will be discussed in section 1.2.3.
The polypeptide chain of ~ 25 N-terminal residues of HA2 is called influenza fusion
peptide (IFP). The sequence of IFP is highly conserved and hydrophobic.59 IFP plays a vital role
in fusion activity. Even a single site mutation may eliminate the fusion activity.60 However, in
these crystal structures, the IFP was not shown. Isolated IFP has been widely studied to understand
its structure and function in fusion process. Liquid-state NMR, solid-state NMR as well as electron
spin resonance (EPR) have been used to illustrate IFP structure in detergent micelle or lipid
membrane at both physiology pH and low pH.61-65 These studies showed that IFP adopted a helixturn-helix structure in micelle or membranes without cholesterol. Two IFP subtypes have been
extensively studied and their sequence are shown as below:
H1:

GLFGAIAGFIEGGWTGMIDGWYG

H3:

GLFGAIAGFIENGWEGMIDGWYG

The difference in these two sequences are residue 12 and 15 (Labeled as bold). In H3 subtype, the
residue 12 is Asparagine instead of Glycine, and the residue 15 is Glutamic acid instead of
Threonine compared to H1 subtype.

27

Figure 1.15 (A) Pre-fusion structures of HA2 ectodomain (Protein Data bank entries 1RD8). This
structure includes residues 38 - 170 of HA2 at pH 7.5;47, 66 (B) Post-fusion structures of HA2
(Protein Data bank entries 1QU1). This structure includes residues 38 - 175 of HA2 at pH 5.49 (C)
The structure of pre-fusion (left) and post-fusion (right). The HA2 is divided into A – H regions
and the residue numbering is labeled. The -strand of HA1 is also shown as “1” and the disulfide
bond between 141 and 1372 is indicated.55

28

Tamm’s group studied the structure of the first 20 residues of H3 (denoted as H3_20 IFP)
in DPC micelle using liquid-state NMR.64 At pH 5 (Figure 1.16 A) IFP adopted a N-terminal helix
(residues Leu-2 to Ile-10) / turn (residue Glu-11 to Gly-13)/C-terminal helix structure (residues
Trp-14 to Ile-18) and this structure was referred as “open structure” with an interhelical angle ~
100°. At pH 7.4, it formed a N-terminal helix (residues Leu-2 to Ile-10) / turn (residue Glu-11 to
Gly-13)/C-terminal extended structure (residues Trp-14 to Gly-20) (Figure 1.16 B). The membrane
location of H3_20 IFP was also studied by EPR. The 20 residues were replaced by Cystine
individually and labeled with nitroxide groups. The result showed that H3_20 IFP was immerged
in an inverted V-shaped manner into lipid hydrophobic core, where the N- and C-terminus deeply
inserted while the turn is close to the aqueous surface. At pH 7.4, the N-terminal was ~ 5 Å from
the lipid phosphate group, and C-terminal was ~ 2 Å. When at pH 5, the N-terminal has ~ 11 Å
distance from the lipid phosphate group, while C-terminal ~ 10 Å. Additionally, H3_20 IFP is
inserted deeper at pH 5 compare to at pH 7.4. The result also suggests that a greater perturbation
of lipid bilayer at pH 5 and facilitate the fusion.
Another very different structure was proposed by Bax’s group when they studied the
H1_23 IFP in DPC micelle by liquid-state NMR.61 H1_23 IFP had a tightly packed anti-parallel
N-helix/turn/C-helix structure with a 158° interhelical angle at both pH 7.0 and pH 4 (Figure 1.16
C and D) and this structure was referred to “closed structure”. The closed structure also showed
an amphipathic property: the hydrophobic side chains were interacting with the hydrophobic core
of the micelle and the hydrophilic side was exposed to the solvent (Figure 1.16 D). The such
significant difference may result from the three additional C-terminal residues: Trp, Tyr and Gly.

29

Figure 1.16 NMR structures of IFP. (A) (B) H3_20 in DPC micelle at pH5 and pH 7.4 respectively.
Amino acids are labeled in the diagram. Backbone of the peptide is shown as ribbon
representations and side chains shown as stick models.64 (C) (D) Longitudinal view and lateral
view of H1_23 in DPC micelle at pH 7 and pH 4, with the hydrophobic side chains in yellow,
polar side chains in green, acidic side chains in red, and Gly residues shown in white van der Waals
representation;61 (E) (F) H1_23 and H3_20 in lipid membrane at pH 5 or pH 7. Carbon, nitrogen
and oxygen atoms in green, blue and red vertices. The dashed lines are between F9 N and G16 CO
with distances 3.9 Å and 5.5 Å respectively.62, 63

30

Due to the high curvature of detergent micelle, the IFP structures determined in DPC are
less physiologically relevant. Solid-state NMR techniques has been used to determine the IFP
structure in physiologically relevant lipid membrane. Sun from Weliky’s group studied the
structure of H3_20 IFP in lipid membrane and the result illustrated that at pH5.0, membraneassociate IFP adopted two distinct helix/turn/helix structure.65 Ghosh from Weliky’s group studied
the distance between H3_20 IFP Phe19N and Gly16CO at pH 5. There was ~ 0.6 mole fraction of
“semi-close” structure with 5.2 Å distance and ~ 0.4 mole fraction of “close” structure with 3.6 Å
distance.62 Further the interhelical geometry study of both H1_23 and H3_20 in lipid membrane
at pH 5 or pH 7 showed a mixture of “closed” and “semi-closed” structures (Figure 1.16 E and F).
Both structures showed N-helix/ turn/C-helix with different interhelical angel and they were also
amphipathic. It was proposed that IFP interacted with the membrane hydrophobic core with its
hydrophobic face while the hydrophilic face interacted with water. Semi-close structure had a
bigger hydrophobic area than close structure and larger population of semi-close structure presents
at pH 5 compared to pH 7. This is consistent with higher fusion activity at pH 5. H1_23 also
showed a higher fusion activity compared to H3_20. The reason was because that H1_23 had
additional WYG residues at C-terminal, and these hydrophobic residues increased the hydrophobic
surface area of H1_23 and lead to greater membrane perturbation and membrane fusion.63

31

1.2.3 Proposed membrane fusion mechanism
Figure 1.17 shows the sequence of events in membrane fusion promoted by HA2.67, 68 The
crystal structure of HA1 and HA2 in a prefusion conformation was determined by Wiley, Wilson
and Skehel in 1981(Figure 1.15 A).47, 69 This crystal structure of HA at pH 7.5 showed that HA is
a trimer. Three HA2 subunits anchored on the viral membrane and forms the coiled-coil trimer
while three HA1 subunits binds outside the HA2 core. After the HA1 sialic acid binding site binds
to the glycolipids or glycoprotein on host cell membrane, the virion is endocytosed and the pH
drops to ~5 in the endosome. The low pH triggers a large-scale conformational rearrangement,
resulting in HA1 separates from HA2 and HA2 is exposed to the target endosomal membrane. It
is proposed that during the conformation change, HA2 forms an extended intermediate and the IFP
on the N-terminus of HA2 interacts with endosomal.68, 70 Then the intermediate starts to collapse
and energy is released during the refolding process causing the membrane to contact each other.
The formation of the hemifusion stalk happens afterwards and finally the protein refolds to form
the stable post-fusion conformation. The post-fusion conformation was visualized in 1999 (Figure
1.15 B).49, 55 There is no crystal structures for stage B to D, neither experimental evidence for
energy releasing due to the protein refolding process. But some biochemical studies support many
of the proposed steps. Oligomeric state studies of HA2 ectodomain or full length HA2 showed
predominant trimer.71, 72 When destabilized with either heat or urea at neutral pH, HA2 had a
conformational change which showed strong fusion activity and was believed to be the evidence
of the extended intermediate state.73 HA2 without transmembrane domain caused lipid mixing but
no contents mixing, which meant the fusion stopped at hemifusion stalk and transmembrane
domain is crucial for the formation of fusion pore.74

32

Figure 1.17 Mechanism of membrane fusion promoted by HA2.67, 68 Host cell membrane is the
blue bilayer on top in each diagram and viral membrane on bottom. The IFP is in green and Nterminal part of ectodomain of HA2 is in red while the C-terminal domains are in navy. The HA1
is not shown in this figure. (A) In the prefusion state, the protein is anchored to the viral membrane
by a C-terminal transmembrane domain. (B) pH decreasing to pH 5 in endosome triggers a
conformational change resulting in an extended intermediate and exposing IFP to the target
membrane. (C) The intermediate collapses. (D) Hemifusion stalk. (E) Fusion pore formation. As
the hemifused bilayers open into a fusion pore, the final zipping up of the C-terminal ectodomain
segments snaps the refolded trimer into its post-fusion conformation, preventing the pore from
resealing.

33

1.3 Introduction solid-state NMR techniques for membrane proteins
Membrane proteins are the proteins that interact with biological membrane and they have
crucial functions such as membrane receptors and enzymes. However, due to the difficulty of
growing protein crystals, only a small fraction of protein structure is solved.
Solid-state nuclear magnetic resonance (NMR) is a powerful technique to determine high
resolution structure and function of large biomolecules. Comparing to widely used X-ray
crystallography and liquid-state NMR, solid-state NMR is specifically suitable for membrane
proteins, large proteins, protein aggregates and nucleic acids that cannot be crystallized or that are
too large for solution NMR spectroscopy.75, 76

1.3.1 Magic-Angle Spinning (MAS)
In solution-state NMR spectra, anisotropic effects are rarely observed because of the rapid
tumbling of the molecules in solution. Thus, the orientation of the molecules with respect to the
external magnetic field B0 is rapidly averaged out and results in sharp peaks.a For solid samples or
macro biomolecules, the tumbling is much slower which results in broad lines in NMR spectra
since all the orientations in the sample contribute to different spectral frequencies.77, 78 Magicangle spinning (MAS) is a routinely used techniques to achieve high resolution spectra. It can
remove the effects of chemical shift anisotropy and to assist in the removal of heteronuclear
dipolar-coupling effects. It is also used to narrow lines from quadrupolar nuclei and removing the
effects of homonuclear dipolar coupling.78

a

In this dissertation, letters or symbols representing vectors are displayed in bold, and letters or

symbols representing quantum mechanical operators have a "^" above them.

34

Figure 1.18 Geometry of the geometry of the 13C – 2H vector in solid state NMR sample under
MAS. The sample is spun rapidly in a cylindrical rotor about a spinning axis oriented at the magic
angle ( = 54.7) with respect to external magnetic field B0.

Figure 1.18 shows the geometry of the MAS. In solid-state NMR MAS experiments,
samples are packed in a cylindrical rotor and spun at high speed by an angle  with respect to B0.
This angle is also called the magic angle and it equals to 54.7. Angle  is the angle between the
C – 2H internuclear vector and B0, and  is the angle between the 13C – 2H internuclear vector

13

and the spinning axis. When the sample is spin at  = 54.7, then  varies with time as the molecule
rotates with the sample. The average of (3cos2 -1) over each rotor period is:
1

〈3𝑐𝑜𝑠 2 𝜃 − 1〉 = (3𝑐𝑜𝑠 2 𝛼 − 1)(3𝑐𝑜𝑠 2 𝛽 − 1)
2

1.1

In equation 1.1, β is fixed for a given nucleus in a rigid solid, and θ can be all possible values in a
powder sample. The α is fixed and set to 54.7, so over each rotor period:
〈3𝑐𝑜𝑠 2 𝜃 − 1〉 = 0

1.2

The secular Hamiltonian for 13C-2H dipolar coupling can be expressed as:
̂ℎ𝑒𝑡𝑒𝑟𝑜 = 𝜇0 ℎ𝛾2𝐶 𝛾3𝐷 (3𝑐𝑜𝑠 2 𝜃 − 1)(2𝐶̂𝑧 𝐷
̂𝑧 )
𝐻
16𝜋 𝑟

1.3

35

Where C represents 13C spin and D represents 2H spin,
Îź0 = permeability of the free space,
̂ and 𝑫
̂ z = the z component of spin operator 𝑪
̂ in a direction parallel to B0,
𝐶̂ z and 𝐷
 is the angle between the spin vector and B0,
r is the distance between spin C and D,
Îł is gyromagnetic ratio.
Therefore, the 13C-2H dipolar coupling is averaged out by MAS.

1.3.2 13C-2H Rotational-Echo Double-Resonance NMR (REDOR)
REDOR was developed by Gullion and Schafer and it is widely used MAS NMR
techniques for studying molecular structure in solid-state.29, 79-84 As discussed previously, MAS
can average out the heteronuclear dipolar interactions. However, in REDOR experiments, by using
simple rotor-synchronized  pulses, the heteronuclear dipolar interactions can be recovered. Since
the dipolar interaction is inversely proportional to the cube of the internuclear distance, the
distances information can be easily obtained. Another advantage of REDOR is the simplicity of
its pulse sequence and data analysis.79
13

C-2H REDOE pulse sequence is shown in Figure 1.19. 13C-2H REDOE is a three channel

experiments. At the beginning of the sequence, a /2 pulse is applied to rotate the 1H magnetization
from the B0 direction to the transverse plane. Then a 1H-13C cross polarization (CP) pulse sequence
is applied to transfers 1H transverse magnetization to 13C nucleus and to enhance the 13C signal.
Due to the various orientation of molecules, chemical bonds, internuclear vectors with respect to
B0, the nuclei in a powder sample have a distribution of Larmor frequencies and the resonance
offset field. Thus, a ramp CP is used to increase the efficiency of the magnetization transfer.

36

Figure 1.19 13C-2H REDOR NMR pulse sequence. The columns represent the /2 or  pulses. CP
= cross polarization that transfers 1H transverse magnetization to
signal. The CP is followed by a

13

13

C and can enhance the

13

C

C-2H dipolar evolution for a period of time which is called

dephasing time (). Adjacent 13C  pulses are separated by one rotor period as are adjacent 2H 
pulses. 13C is the detecting channel.79

After the 1H-13C CP, followed by a dephasing period () during which a series of 13C and
2

H  pulses and 1H decoupling field were applied. For each , two separate spectra are collected

and they are denoted as S0 and S1. In the S0 experiment, only 13C  pulses are applied at the end
of each rotor period except at the end of . In the S1 experiment, 13C  pulses are applied at the end
of each rotor period while 2H  pulses are applied in the middle of each rotor period. Then followed
by the 13C signal acquisition.
The 13C-2H dipolar coupling and chemical shift anisotropy are averaged out by MAS. In
REDOR experiments, the function of the

13

C and 2H  pulses are to flip the spin by 180. The

secular 13C-2H dipolar coupling is discussed in equation 1.3:

37

̂ℎ𝑒𝑡𝑒𝑟𝑜 =
𝐻

𝜇0 ℎ𝛾𝐶 𝛾𝐷
̂𝑧 )
(3𝑐𝑜𝑠 2 𝜃 − 1)(2𝐶̂𝑧 𝐷
16𝜋 2 𝑟 3

Thus, the

13

C and 2H  pulses change the sign of dipolar coupling. In S0 experiment, the rotor

synchronized 13C  pulses are applied at the end of each rotor period and it flips the sign of 𝐶̂𝑧 .
Their effect on the

13

C-2H dipolar coupling is shown in Figure 1.20. Additionally,

13

C pulses

refocus the 13C isotropic chemical shift. In S1 experiment, 2H  pulses are applied at the middle of
each rotor period and the 13C-2H dipolar coupling is recoupled. For an isolated 13C-2H spin pair,
the signal of S0 and S1 follow the relationship as:
𝑆1
𝑆0

= cos 𝜙

𝜙=

𝑁𝑐 𝑇𝑟 𝐷
2𝜋 2

1.4

√2 sin 2𝑏 𝑠𝑖𝑛𝑎

1.5

Where Nc is number of rotor period, Tr is sample rotation period, D is dipolar coupling, and a and
b are the azimuthal and polar angles of the internuclear vector with respect to the spinning axis.
For a powder sample, all internuclear orientations need to be considered, and all values of a and b
must be summed over. Thus, S1 is always smaller than S0.
The dephasing of REDOR is:
∆𝑆
𝑆0

=

𝑆0 −𝑆1

1.6

𝑆0

In this equation, S0 and S1 are the respective signal intensity of S0 and S1 experiments. The
dephasing is a function of dephasing time  and the dephasing buildup curve is plotting S/S0 vs


38

Figure 1.20 Evolution of dipolar coupling as a function of rotor period in REDOR experiments.
The dipolar coupling is averaged out over each rotor period by MAS. In S 0 experiment, rotor
synchronized 13C  pulses do not interfere with the MAS averaging of the heteronuclear dipolar
interaction. In S1 experiment, rotor-synchronized
dipolar interaction.

39

C and 2H  pulses re-introduce the

13

13

C-2H

Figure 1.21

13

C-2H REDOR spectra with a 40 ms dephasing time for the I4 peptide as well as

S/S0 vs τ. Black squares are the experimental dephasing of I4 peptide. Blue triangles are the bestfit calculated by the SIMPSON program (without 2H relaxation) with a 22 Hz
coupling. The red line is the best-fit exponential buildup.81

40

13

C-2H dipolar

The buildup can be fitted with SIMPSON simulation.85 Exponential fitting can also be used
for 13C-2H REDOR.29, 81 The exponential fitting function is A(1 – e-), in which A and  are fitting
parameters. The buildup rate  has a relationship with dipolar coupling D. The buildup extent A is
correlated to the fraction of

C nuclei with this coupling. The corresponding value of 1 − A is

13

correlated to the fraction of lab 13C nuclei with D ≈ 0.
Figure 1.21 shows the comparison between these two methods. I4 peptide is used as a
standard sample. It has a sequence AEAAAKEAAAKEAAAKAW and it has a has a regular 
helical structure. I4 peptide is

13

CO label at A9 and 2H label at A8. The isolated 13CO-2H spin

pairs all have the same internuclear separation r of 5.0 Å and with a corresponding dipolar coupling
D of 37 Hz.81 The exponential fitting result (red line) has a much smaller deviation from
experimental dephasing (black square) compared to sigmoidal shape SIMPSON simulation (blue
triangle). Additionally, the maximal values of ∟ (2/3) in SIMPSON differs from ~1 in exponential
fitting. The best-fit dipolar coupling D of 22 Hz is smaller than the expected d of 37 Hz.
The reason for all these differences are that the nonradiative transitions between the m =
Âą1 states and the m = 0 states of individual 2H nuclei during the dephasing period that are not
considered in the SIMPSON calculations. The 2H T1 relaxation times is ≈ 50 ms, which is
comparable to  in REDOR experiments. Thus, there are m = 0 ↔ m = ±1 2H transitions during
the dephasing time. For each 2H nucleus, it approximately spends two-thirds of the dephasing
period at m = Âą1 states and approximately one-third at m = 0 state. Since there is not buildup when
CO is at m = 0 state, so that the observed buildup rate  ≈ 2D/3.

13

The exponential fitting A × (1 – e-) for the experimental data results in r = 5 Å and = 24
Hz ≈ (2/3) (37 Hz) and 37 Hz is the dipolar coupling of I4 peptide. This shows that exponential

41

fitting is an excellent and simple fitting method for 13CO-2H REDOR. D, in units of Hz is given
by: 𝐷 =

𝜇0 ℎ𝛾𝐶 𝛾𝐷
16𝜋 3 𝑟 3

=

4642
𝑟3

𝐻𝑧. Thus, the internuclear distance r was calculated as 3√

4642 𝐻𝑧
3𝛾
2

Å.

1.3.3 Paramagnetic Relaxation Enhancement NMR (PRE)
Paramagnetic Relaxation Enhancement NMR (PRE) is first developed by Solomon in the
1955.86 Over the past decades, PRE has become a powerful method to provide long-range distance
information. In PRE experiments, paramagnetic species such as Mn2+, Gd2+, and nitroxide spin
labels have been used in solid-state NMR for studying protein membrane location.87-90 The PRE
arises from dipolar interactions between a nucleus of interest and the unpaired electrons of the
paramagnetic center which result in an increase in nuclear relaxation rates. Since the gyromagnetic
ratio of electron is three magnitudes greater than most nuclei, this method can be used to get
distance information in a range up to 35 Å, which is especially useful for biomacromolecules and
biosystems.87 The first applications of the PRE is to study spin-labeled lysozyme and bovine
pancreatic trypsin inhibitor in 1984, and the PRE effects were converted to approximate distance
restraints.91 In addition, PRE is also used to accelerate data acquisition for systems such as
biomolecules and human tissue, since PRE increases longitudinal relaxation T1 rates.92-95
Other techniques can also be used to study the protein topology in membrane, but they
have some limitations. The Nuclear Overhauser Effect (NOE) is widely used for protein structure
determination. However, NOE is based on short-range local interactions, which limits to ~6 Å
interproton distance.61, 96 1H spin diffusion NMR technique can be used to probe membrane protein
topology in membrane.97-99 This technique detects the rate of 1H magnetization transfer from
mobile lipids to the rigid proteins, but not suitable for mobile peptides. Beyond NMR, EPR
spectroscopy is also well established for investigating membrane location of peptides.64,

42

100

However, these approaches require the use of bulky spin probes, which may affect the insertion
depth or orientation of the peptide in membrane.
To study the membrane protein structure and topology in membrane, the paramagnetic
metal ions were added to protein-membrane complex and these ions bind at the membrane surface.
The unpaired electron in paramagnetic species enhance the relaxation of the nuclear spins. Let’s
take Mn2+ as an example. The paramagnetic Mn2+ contributes to faster dipolar transverse relaxation,
T2, is proposed by Solomon86 and Bloembergen101:
1
𝑇2

1

𝜇

2 𝛾2 𝜇2
𝐼

= 𝑊 15 (4𝜋0 )

𝑒𝑓𝑓 𝛽
𝑟6

2

3𝜏

13𝜏

(4𝜏𝑠 + 1+𝜔2𝑠 𝜏2 + 1+𝜔2𝑠𝜏2)
𝐼 𝑠

1.7

𝑒 𝑠

Where W is the local concentration of the Mn2+ ions,0 is the vacuum permeability, I is the
gyromagnetic ratio of nucleus spin I, eff is the effective magnetic moment of Mn2+ ions,  is the
Bohr magneton, r is the average electron-nucleus distance, I and e are the nucleus and election
Larmor frequencies. The correlation time s is the inverse sum of the electronic spin-lattice
relaxation time T1e; the rotational correlation time of the molecule r; and the residence time of the
Mn2+ near the nuclear spin m:
1
𝜏𝑠

1

1

1

1𝑒

𝑟

𝑚

=𝑇 +𝜏 +𝜏

1.8

Based on equation 1.5, T2 ∝ r6, thus the closer distance between the paramagnetic center and the
nucleus of interest, the shorter the T2, resulting in line broadening and peak suppression. By
measuring T2 of the nucleus, the distance information can be obtained.
The quadrupolar echo (quecho) solid-state NMR experiment is one of the most widely used
experiments for studying quadrupolar nuclei in solid samples.102, 103 This technique can be used to
measure the T2 relaxation time and to minimize effects from probe ring-down.104 The pulse
sequence is shown in Figure 1.22. At the beginning of the sequence, a /2 pulse with phase x is

43

applied. After a duration of time  the second /2 pulse with phase y or -y is applied to refocus
the magnetization. The echo maximum appears at τ2 after the second pulse. In actual experiments,
the τ2 is set shorter than the τ1 to obtain the maximum intensity. To determine the T2 relaxation
time, the decay of the acquired signal was measured as a function of synchronous incrementation
of t, and t = [1 + 2 + time being shifted]. By plotting the signal intensity I in FID vs t, the T2
relaxation time can be calculated by:
𝑡

𝐼(𝑡) = 𝐼(0)×𝑒𝑥𝑝 (𝑇 )

1.9

2

Figure 1.22 Quadrupolar echo (quecho) pulse sequence.

44

REFERENCES

45

REFERENCES

[1] Weiss, R. A. (1993) HOW DOES HIV CAUSE AIDS, Science 260, 1273-1279.
[2] Briggs, J. A. G., Simon, M. N., Gross, I., Krausslich, H. G., Fuller, S. D., Vogt, V. M., and
Johnson, M. C. (2004) The stoichiometry of Gag protein in HIV-1, Nat. Struct. Mol. Biol.
11, 672-675.
[3] Levy, J. A. (1993) PATHOGENESIS OF HUMAN-IMMUNODEFICIENCY-VIRUS
INFECTION, Microbiol. Rev. 57, 183-289.
[4] McCune, J. M., Rabin, L. B., Feinberg, M. B., Lieberman, M., Kosek, J. C., Reyes, G. R., and
Weissman, I. L. (1988) ENDOPROTEOLYTIC CLEAVAGE OF GP160 IS REQUIRED
FOR THE ACTIVATION OF HUMAN IMMUNODEFICIENCY VIRUS, Cell 53, 55-67.
[5] Gelderblom, H. R., Ozel, M., Hausmann, E. H. S., Winkel, T., Pauli, G., and Koch, M. A.
(1988) FINE-STRUCTURE OF HUMAN IMMUNODEFICIENCY VIRUS (HIV),
IMMUNOLOCALIZATION OF STRUCTURAL PROTEINS AND VIRUS-CELL
RELATION, Micron and Microscopica Acta 19, 41-60.
[6] Gelderblom, H. R., Ozel, M., and Pauli, G. (1989) MORPHOGENESIS AND
MORPHOLOGY OF HIV - STRUCTURE-FUNCTION RELATIONS, Arch. Virol. 106,
1-13.
[7] Laskey, S. B., and Siliciano, R. F. (2014) A mechanistic theory to explain the efficacy of
antiretroviral therapy, Nat. Rev. Microbiol. 12, 772-780.
[8] Grewe, C., Beck, A., and Gelderblom, H. R. (1990) HIV - EARLY VIRUS-CELL
INTERACTIONS, J. Acquir. Immune Defic. Syndr. Hum. Retrovirol. 3, 965-974.
[9] Miyauchi, K., Kim, Y., Latinovic, O., Morozov, V., and Melikyan, G. B. (2009) HIV Enters
Cells via Endocytosis and Dynamin-Dependent Fusion with Endosomes, Cell 137, 433444.
[10] Gottlieb, M. S., Schroff, R., Schanker, H. M., Weisman, J. D., Fan, P. T., Wolf, R. A., and
Saxon, A. (1981) PNEUMOCYSTIS-CARINII PNEUMONIA AND MUCOSAL
CANDIDIASIS IN PREVIOUSLY HEALTHY HOMOSEXUAL MEN - EVIDENCE OF
A NEW ACQUIRED CELLULAR IMMUNODEFICIENCY, N. Engl. J. Med. 305, 14251431.
[11] Dalgleish, A. G., Beverley, P. C. L., Clapham, P. R., Crawford, D. H., Greaves, M. F., and
Weiss, R. A. (1984) THE CD4 (T4) ANTIGEN IS AN ESSENTIAL COMPONENT OF
THE RECEPTOR FOR THE AIDS RETROVIRUS, Nature 312, 763-767.

46

[12] Kwong, P. D., Wyatt, R., Robinson, J., Sweet, R. W., Sodroski, J., and Hendrickson, W. A.
(1998) Structure of an HIV gp120 envelope glycoprotein in complex with the CD4 receptor
and a neutralizing human antibody, Nature 393, 648-659.
[13] Broder, C. C., and Dimitrov, D. S. (1996) HIV and the 7-transmembrane domain receptors,
Pathobiology 64, 171-179.
[14] Chan, D. C., Fass, D., Berger, J. M., and Kim, P. S. (1997) Core structure of gp41 from the
HIV envelope glycoprotein, Cell 89, 263-273.
[15] Julien, J. P., Cupo, A., Sok, D., Stanfield, R. L., Lyumkis, D., Deller, M. C., Klasse, P. J.,
Burton, D. R., Sanders, R. W., Moore, J. P., Ward, A. B., and Wilson, I. A. (2013) Crystal
Structure of a Soluble Cleaved HIV-1 Envelope Trimer, Science 342, 1477-1483.
[16] Frankel, A. D., and Young, J. A. T. (1998) HIV-1: Fifteen proteins and an RNA, Annu. Rev.
Biochem. 67, 1-25.
[17] Chan, D. C., and Kim, P. S. (1998) HIV entry and its inhibition, Cell 93, 681-684.
[18] Weissenhorn, W., Dessen, A., Harrison, S. C., Skehel, J. J., and Wiley, D. C. (1997) Atomic
structure of the ectodomain from HIV-1 gp41, Nature 387, 426-430.
[19] Lyumkis, D., Julien, J. P., de Val, N., Cupo, A., Potter, C. S., Klasse, P. J., Burton, D. R.,
Sanders, R. W., Moore, J. P., Carragher, B., Wilson, I. A., and Ward, A. B. (2013) CryoEM Structure of a Fully Glycosylated Soluble Cleaved HIV-1 Envelope Trimer, Science
342, 1484-1490.
[20] Lu, M., Ji, H., and Shen, S. (1999) Subdomain folding and biological activity of the core
structure from human immunodeficiency virus type 1 gp41: Implications for viral
membrane fusion, J. Virol. 73, 4433-4438.
[21] White, J. M., Delos, S. E., Brecher, M., and Schornberg, K. (2008) Structures and mechanisms
of viral membrane fusion proteins: Multiple variations on a common theme, Crit. Rev.
Biochem. Mol. Biol. 43, 189-219.
[22] Tran, E. E. H., Borgnia, M. J., Kuybeda, O., Schauder, D. M., Bartesaghi, A., Frank, G. A.,
Sapiro, G., Milne, J. L. S., and Subramaniam, S. (2012) Structural Mechanism of Trimeric
HIV-1 Envelope Glycoprotein Activation, PLoS Pathog. 8, 18.
[23] Gabrys, C. M., Qiang, W., Sun, Y., Xie, L., Schmick, S. D., and Weliky, D. P. (2013) SolidState Nuclear Magnetic Resonance Measurements of HIV Fusion Peptide (CO)-C-13 to
Lipid P-31 Proximities Support Similar Partially Inserted Membrane Locations of the alpha
Helical and beta Sheet Peptide Structures, J. Phys. Chem. A 117, 9848-9859.
[24] Gabrys, C. M., and Weliky, D. P. (2007) Chemical shift assignment and structural plasticity
of a HIV fusion peptide derivative in dodecylphosphocholine micelles, Biochim. Biophys.
Acta-Biomembr. 1768, 3225-3234.

47

[25] Jaroniec, C. P., Kaufman, J. D., Stahl, S. J., Viard, M., Blumenthal, R., Wingfield, P. T., and
Bax, A. (2005) Structure and dynamics of micelle-associated human immunodeficiency
virus gp41 fusion domain, Biochemistry 44, 16167-16180.
[26] Li, Y. L., and Tamm, L. K. (2007) Structure and plasticity of the human immunodeficiency
virus gp41 fusion domain in lipid micelles and bilayers, Biophys. J. 93, 876-885.
[27] Schmick, S. D., and Weliky, D. P. (2010) Major Antiparallel and Minor Parallel beta Sheet
Populations Detected in the Membrane-Associated Human Immunodeficiency Virus
Fusion Peptide, Biochemistry 49, 10623-10635.
[28] Qiang, W., Sun, Y., and Weliky, D. P. (2009) A strong correlation between fusogenicity and
membrane insertion depth of the HIV fusion peptide, Proc. Natl. Acad. Sci. U. S. A. 106,
15314-15319.
[29] Jia, L. H., Liang, S., Sackett, K., Xie, L., Ghosh, U., and Weliky, D. P. (2015) REDOR solidstate NMR as a probe of the membrane locations of membrane-associated peptides and
proteins, J. Magn. Reson. 253, 154-165.
[30] Sackett, K., Nethercott, M. J., Zheng, Z. X., and Weliky, D. P. (2014) Solid-State NMR
Spectroscopy of the HIV gp41 Membrane Fusion Protein Supports Intermolecular
Antiparallel 13 Sheet Fusion Peptide Structure in the Final Six-Helix Bundle State, J. Mol.
Biol. 426, 1077-1094.
[31] Schibli, D. J., Montelaro, R. C., and Vogel, H. J. (2001) The membrane-proximal tryptophanrich region of the HIV glycoprotein, gp41, forms a well-defined helix in
dodecylphosphocholine micelles, Biochemistry 40, 9570-9578.
[32] Sun, Z. Y. J., Oh, K. J., Kim, M. Y., Yu, J., Brusic, V., Song, L. K., Qiao, Z. S., Wang, J. H.,
Wagner, G., and Reinherz, E. L. (2008) HIV-1 broadly neutralizing antibody extracts its
epitope from a kinked gp41 ectodomain region on the viral membrane, Immunity 28, 5263.
[33] Shi, W. X., Bohon, J., Han, D. P., Habte, H., Qin, Y. L., Cho, M. W., and Chance, M. R.
(2010) Structural Characterization of HIV gp41 with the Membrane-proximal External
Region, J. Biol. Chem. 285, 24290-24298.
[34] Gangupomu, V. K., and Abrams, C. F. (2010) All-Atom Models of the Membrane-Spanning
Domain of HIV-1 gp41 from Metadynamics, Biophys. J. 99, 3438-3444.
[35] Apellaniz, B., Rujas, E., Serrano, S., Morante, K., Tsumoto, K., Caaveiro, J. M. M., Jimenez,
M. A., and Nieva, J. L. (2015) The Atomic Structure of the HIV-1 gp41 Transmembrane
Domain and Its Connection to the Immunogenic Membrane-proximal External Region, J.
Biol. Chem. 290, 12999-13015.
[36] Kim, J. H., Hartley, T. L., Curran, A. R., and Engelman, D. M. (2009) Molecular dynamics
studies of the transmembrane domain of gp41 from HIV-1, Biochim. Biophys. ActaBiomembr. 1788, 1804-1812.

48

[37] Chernomordik, L. V., and Kozlov, M. M. (2008) Mechanics of membrane fusion, Nat. Struct.
Mol. Biol. 15, 675-683.
[38] Markosyan, R. M., Cohen, F. S., and Melikyan, G. B. (2003) HIV-1 envelope proteins
complete their folding into six-helix bundles immediately after fusion pore formation, Mol.
Biol. Cell 14, 926-938.
[39] Wild, C. T., Shugars, D. C., Greenwell, T. K., McDanal, C. B., and Matthews, T. J. (1994)
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. 91, 9770-9774.
[40] Melikyan, G. B., Markosyan, R. M., Hemmati, H., Delmedico, M. K., Lambert, D. M., and
Cohen, F. S. (2000) Evidence that the transition of HIV-1 gp41 into a six-helix bundle, not
the bundle configuration, induces membrane fusion, J. Cell Biol. 151, 413-423.
[41] Banerjee, K., and Weliky, D. P. (2014) Folded Monomers and Hexamers of the Ectodomain
of the HIV gp41 Membrane Fusion Protein: Potential Roles in Fusion and Synergy
Between the Fusion Peptide, Hairpin, and Membrane-Proximal External Region,
Biochemistry 53, 7184-7198.
[42] Magnus, C., Rusert, P., Bonhoeffer, S., Trkola, A., and Regoes, R. R. (2009) Estimating the
Stoichiometry of Human Immunodeficiency Virus Entry, J. Virol. 83, 1523-1531.
[43] Yoon, S. W., Webby, R. J., and Webster, R. G. (2014) Evolution and Ecology of Influenza A
Viruses, In Influenza Pathogenesis and Control - Vol I (Compans, R. W., and Oldstone, M.
B. A., Eds.), pp 359-375, Springer Int Publishing Ag, Cham.
[44] Webster, R. G., Bean, W. J., Gorman, O. T., Chambers, T. M., and Kawaoka, Y. (1992)
EVOLUTION AND ECOLOGY OF INFLUENZA-A VIRUSES, Microbiol. Rev. 56, 152179.
[45] Claas, E. C. J., Sprenger, M. J. W., Kleter, G. E. M., Vanbeek, R., Quint, W. G. V., and
Masurel, N. (1992) TYPE-SPECIFIC IDENTIFICATION OF INFLUENZA VIRUS-A,
VIRUS-B AND VIRUS-C BY THE POLYMERASE CHAIN-REACTION, J. Virol.
Methods 39, 1-13.
[46] Harris, A., Cardone, G., Winkler, D. C., Heymann, J. B., Brecher, M., White, J. M., and
Steven, A. C. (2006) Influenza virus pleiomorphy characterized by cryoelectron
tomography, Proc. Natl. Acad. Sci. U. S. A. 103, 19123-19127.
[47] Wilson, I. A., Skehel, J. J., and Wiley, D. C. (1981) STRUCTURE OF THE
HEMAGGLUTININ MEMBRANE GLYCOPROTEIN OF INFLUENZA-VIRUS AT 3A RESOLUTION, Nature 289, 366-373.
[48] Gamblin, S. J., and Skehel, J. J. (2010) Influenza Hemagglutinin and Neuraminidase
Membrane Glycoproteins, J. Biol. Chem. 285, 28403-28409.

49

[49] Chen, J., Skehel, J. J., and Wiley, D. C. (1999) N- and C-terminal residues combine in the
fusion-pH influenza hemagglutinin HA(2) subunit to form an N cap that terminates the
triple-stranded coiled coil, Proc. Natl. Acad. Sci. U. S. A. 96, 8967-8972.
[50] Curtis-Fisk, J., Preston, C., Zheng, Z. X., Worden, R. M., and Weliky, D. P. (2007) Solidstate NMR structural measurements on the membrane-associated influenza fusion protein
ectodomain, J. Am. Chem. Soc. 129, 11320-+.
[51] Schnell, J. R., and Chou, J. J. (2008) Structure and mechanism of the M2 proton channel of
influenza A virus, Nature 451, 591-U512.
[52] O'Neill, R. E., Talon, J., and Palese, P. (1998) The influenza virus NEP (NS2 protein)
mediates the nuclear export of viral ribonucleoproteins, Embo J. 17, 288-296.
[53] Shi, Y., Wu, Y., Zhang, W., Qi, J. X., and Gao, G. F. (2014) Enabling the 'host jump':
structural determinants of receptor-binding specificity in influenza A viruses, Nat. Rev.
Microbiol. 12, 822-831.
[54] Lakadamyali, M., Rust, M. J., and Zhuang, X. W. (2004) Endocytosis of influenza viruses,
Microbes Infect. 6, 929-936.
[55] Bullough, P. A., Hughson, F. M., Skehel, J. J., and Wiley, D. C. (1994) STRUCTURE OF
INFLUENZA HEMAGGLUTININ AT THE PH OF MEMBRANE-FUSION, Nature 371,
37-43.
[56] Wiley, D. C., and Skehel, J. J. (1987) THE STRUCTURE AND FUNCTION OF THE
HEMAGGLUTININ MEMBRANE GLYCOPROTEIN OF INFLUENZA-VIRUS, Annu.
Rev. Biochem. 56, 365-394.
[57] Skehel, J. J., and Wiley, D. C. (2000) Receptor binding and membrane fusion in virus entry:
The influenza hemagglutinin, Annu. Rev. Biochem. 69, 531-569.
[58] Sriwilaijaroen, N., and Suzuki, Y. (2012) Molecular basis of the structure and function of H1
hemagglutinin of influenza virus, Proc. Jpn. Acad. Ser. B-Phys. Biol. Sci. 88, 226-249.
[59] Nobusawa, E., Aoyama, T., Kato, H., Suzuki, Y., Tateno, Y., and Nakajima, K. (1991)
COMPARISON OF COMPLETE AMINO-ACID-SEQUENCES AND RECEPTORBINDING PROPERTIES AMONG 13 SEROTYPES OF HEMAGGLUTININS OF
INFLUENZA A-VIRUSES, Virology 182, 475-485.
[60] Gething, M. J., Doms, R. W., York, D., and White, J. (1986) STUDIES ON THE
MECHANISM OF MEMBRANE-FUSION - SITE-SPECIFIC MUTAGENESIS OF THE
HEMAGGLUTININ OF INFLUENZA-VIRUS, J. Cell Biol. 102, 11-23.
[61] Lorieau, J. L., Louis, J. M., and Bax, A. (2010) The complete influenza hemagglutinin fusion
domain adopts a tight helical hairpin arrangement at the lipid:water interface, Proc. Natl.
Acad. Sci. U. S. A. 107, 11341-11346.

50

[62] Ghosh, U., Xie, L., and Weliky, D. P. (2013) Detection of closed influenza virus
hemagglutinin fusion peptide structures in membranes by backbone (CO)-C-13-N-15
rotational-echo double-resonance solid-state NMR, J. Biomol. NMR 55, 139-146.
[63] Ghosh, U., Xie, L., Jia, L. H., Liang, S., and Weliky, D. P. (2015) Closed and Semiclosed
Interhelical Structures in Membrane vs Closed and Open Structures in Detergent for the
Influenza Virus Hemagglutinin Fusion Peptide and Correlation of Hydrophobic Surface
Area with Fusion Catalysis, J. Am. Chem. Soc. 137, 7548-7551.
[64] Han, X., Bushweller, J. H., Cafiso, D. S., and Tamm, L. K. (2001) Membrane structure and
fusion-triggering conformational change of the fusion domain from influenza
hemagglutinin, Nat. Struct. Biol. 8, 715-720.
[65] Sun, Y., and Weliky, D. P. (2009) C-13-C-13 Correlation Spectroscopy of MembraneAssociated Influenza Virus Fusion Peptide Strongly Supports a Helix-Turn-Helix Motif
and Two Turn Conformations, J. Am. Chem. Soc. 131, 13228-13229.
[66] Chen, J., Lee, K. H., Steinhauer, D. A., Stevens, D. J., Skehel, J. J., and Wiley, D. C. (1998)
Structure of the hemagglutinin precursor cleavage site, a determinant of influenza
pathogenicity and the origin of the labile conformation, Cell 95, 409-417.
[67] Harrison, S. C. (2008) Viral membrane fusion, Nat. Struct. Mol. Biol. 15, 690-698.
[68] Floyd, D. L., Ragains, J. R., Skehel, J. J., Harrison, S. C., and van Oijen, A. M. (2008) Singleparticle kinetics of influenza virus membrane fusion, Proc. Natl. Acad. Sci. U. S. A. 105,
15382-15387.
[69] Wiley, D. C., Wilson, I. A., and Skehel, J. J. (1981) STRUCTURAL IDENTIFICATION OF
THE
ANTIBODY-BINDING
SITES
OF
HONG-KONG
INFLUENZA
HEMAGGLUTININ AND THEIR INVOLVEMENT IN ANTIGENIC VARIATION,
Nature 289, 373-378.
[70] Carr, C. M., and Kim, P. S. (1993) A SPRING-LOADED MECHANISM FOR THE
CONFORMATIONAL CHANGE OF INFLUENZA HEMAGGLUTININ, Cell 73, 823832.
[71] Chen, J., Skehel, J. J., and Wiley, D. C. (1998) A polar octapeptide fused to the N-terminal
fusion peptide solubilizes the influenza virus HA(2) subunit ectodomain, Biochemistry 37,
13643-13649.
[72] Ratnayake, P. U., Ekanayaka, E. A. P., Komanduru, S. S., and Weliky, D. P. (2016) Fulllength trimeric influenza virus hemagglutinin II membrane fusion protein and shorter
constructs lacking the fusion peptide or transmembrane domain: Hyperthermostability of
the full-length protein and the soluble ectodomain and fusion peptide make significant
contributions to fusion of membrane vesicles, Protein Expr. Purif. 117, 6-16.
[73] Carr, C. M., Chaudhry, C., and Kim, P. S. (1997) Influenza hemagglutinin is spring-loaded
by a metastable native conformation, Proc. Natl. Acad. Sci. U. S. A. 94, 14306-14313.

51

[74] Kemble, G. W., Danieli, T., and White, J. M. (1994) LIPID-ANCHORED INFLUENZA
HEMAGGLUTININ PROMOTES HEMIFUSION, NOT COMPLETE FUSION, Cell 76,
383-391.
[75] Griffin, R. G. (1998) Dipolar recoupling in MAS spectra of biological solids, Nat. Struct. Biol.
5, 508-512.
[76] Marassi, F. M., and Opella, S. J. (1998) NMR structural studies of membrane proteins, Curr.
Opin. Struct. Biol. 8, 640-648.
[77] Andrew, E. R., Bradbury, A., and Eades, R. G. (1958) NUCLEAR MAGNETIC
RESONANCE SPECTRA FROM A CRYSTAL ROTATED AT HIGH SPEED, Nature
182, 1659-1659.
[78] Duer, M. J. (2002) Solid-state NMR spectroscopy principles and applications, Blackwell
Science Ltd.
[79] Gullion, T. (1998) Introduction to rotational-echo, double-resonance NMR, Concepts Magn.
Resonance 10, 277-289.
[80] Gullion, T., and Schaefer, J. (1989) ROTATIONAL-ECHO DOUBLE-RESONANCE NMR,
J. Magn. Reson. 81, 196-200.
[81] Xie, L., Ghosh, U., Schmick, S. D., and Weliky, D. P. (2013) Residue-specific membrane
location of peptides and proteins using specifically and extensively deuterated lipids and
C-13-H-2 rotational-echo double-resonance solid-state NMR, J. Biomol. NMR 55, 11-17.
[82] Gullion, T., Kishore, R., and Asakura, T. (2003) Determining dihedral angles and local
structure in silk peptide by C-13-H-2 REDOR, J. Am. Chem. Soc. 125, 7510-7511.
[83] Schmidt, A., Kowalewski, T., and Schaefer, J. (1993) LOCAL PACKING IN GLASSY
POLYCARBONATE BY CARBON DEUTERIUM REDOR NMR, Macromolecules 26,
1729-1733.
[84] Mueller, K. T. (1995) ANALYTIC SOLUTIONS FOR THE TIME EVOLUTION OF
DIPOLAR-DEPHASING NMR SIGNALS, J. Magn. Reson. Ser. A 113, 81-93.
[85] Bak, M., Rasmussen, J. T., and Nielsen, N. C. (2000) SIMPSON: A general simulation
program for solid-state NMR spectroscopy, J. Magn. Reson. 147, 296-330.
[86] Solomon, I. (1955) RELAXATION PROCESSES IN A SYSTEM OF 2 SPINS, Physical
Review 99, 559-565.
[87] Clore, G. M., and Iwahara, J. (2009) Theory, Practice, and Applications of Paramagnetic
Relaxation Enhancement for the Characterization of Transient Low-Population States of
Biological Macromolecules and Their Complexes, Chem. Rev. 109, 4108-4139.

52

[88] Buffy, J. J., Hong, T., Yamaguchi, S., Waring, A. J., Lehrer, R. I., and Hong, M. (2003) Solidstate NMR investigation of the depth of insertion of protegrin-1 in lipid bilayers using
paramagnetic Mn2+, Biophys. J. 85, 2363-2373.
[89] Chu, S. D., Maltsev, S., Emwas, A. H., and Lorigan, G. A. (2010) Solid-state NMR
paramagnetic relaxation enhancement immersion depth studies in phospholipid bilayers, J.
Magn. Reson. 207, 89-94.
[90] Su, Y., Mani, R., and Hong, M. (2008) Asymmetric insertion of membrane proteins in lipid
bilayers by solid-state NMR paramagnetic relaxation enhancement: A cell-penetrating
peptide example, J. Am. Chem. Soc. 130, 8856-8864.
[91] Schmidt, P. G., and Kuntz, I. D. (1984) DISTANCE MEASUREMENTS IN SPINLABELED LYSOZYME, Biochemistry 23, 4261-4266.
[92] Cai, S., Seu, C., Kovacs, Z., Sherry, A. D., and Chen, Y. (2006) Sensitivity enhancement of
multidimensional NMR experiments by paramagnetic relaxation effects, J. Am. Chem. Soc.
128, 13474-13478.
[93] Linser, R., Chevelkov, V., Diehl, A., and Reif, B. (2007) Sensitivity enhancement using
paramagnetic relaxation in MAS solid-state NMR of perdeuterated proteins, J. Magn.
Reson. 189, 209-216.
[94] Caravan, P. (2006) Strategies for increasing the sensitivity of gadolinium based MRI contrast
agents, Chem. Soc. Rev. 35, 512-523.
[95] Flacke, S., Fischer, S., Scott, M. J., Fuhrhop, R. J., Allen, J. S., McLean, M., Winter, P., Sicard,
G. A., Gaffney, P. J., Wickline, S. A., and Lanza, G. M. (2001) Novel MRI contrast agent
for molecular imaging of fibrin implications for detecting vulnerable plaques, Circulation
104, 1280-1285.
[96] Lorieau, J. L., Louis, J. M., Schwieters, C. D., and Bax, A. (2012) pH-triggered, activatedstate conformations of the influenza hemagglutinin fusion peptide revealed by NMR, Proc.
Natl. Acad. Sci. U. S. A. 109, 19994-19999.
[97] Huster, D., Yao, X. L., and Hong, M. (2002) Membrane protein topology probed by H-1 spin
diffusion from lipids using solid-state NMR spectroscopy, J. Am. Chem. Soc. 124, 874-883.
[98] Kumashiro, K. K., Schmidt-Rohr, K., Murphy, O. J., Ouellette, K. L., Cramer, W. A., and
Thompson, L. K. (1998) A novel tool for probing membrane protein structure: Solid-state
NMR with proton spin diffusion and X-nucleus detection, J. Am. Chem. Soc. 120, 50435051.
[99] Gallagher, G. J., Hong, M., and Thompson, L. K. (2004) Solid-state NMR spin diffusion for
measurement of membrane-bound peptide structure: Gramicidin A, Biochemistry 43,
7899-7906.

53

[100] Macosko, J. C., Kim, C. H., and Shin, Y. K. (1997) The membrane topology of the fusion
peptide region of influenza hemagglutinin determined by spin-labeling EPR, J. Mol. Biol.
267, 1139-1148.
[101] Bloembergen, N. (1957) PROTON RELAXATION TIMES IN PARAMAGNETIC
SOLUTIONS, J. Chem. Phys. 27, 572-573.
[102] Davis, J. H., Jeffrey, K. R., Bloom, M., Valic, M. I., and Higgs, T. P. (1976)
QUADRUPOLAR ECHO DEUTERON MAGNETIC-RESONANCE SPECTROSCOPY
IN ORDERED HYDROCARBON CHAINS, Chem. Phys. Lett. 42, 390-394.
[103] Weisman, I. D., and Bennett, L. H. (1969) QUADRUPOLAR ECHOES IN SOLIDS,
Physical Review 181, 1341-&.
[104] Gabrys, C. M., Yang, R., Wasniewski, C. M., Yang, J., Canlas, C. G., Qiang, W., Sun, Y.,
and Weliky, D. P. (2010) Nuclear magnetic resonance evidence for retention of a lamellar
membrane phase with curvature in the presence of large quantities of the HIV fusion
peptide, Biochim. Biophys. Acta-Biomembr. 1798, 194-201.

54

Chapter 2
Materials and Methods

55

2.1 Materials
The DNA plasmids containing HM, HM_TM and FP_HM and FP_HM_TM genes were
ordered from GenScript (Piscataway, NJ). The protein expression cell Escherichia coli BL21(DE3)
strain was purchased from Novagen (Gibbstown, NJ). Luria-Bertani broth (LB) medium was
purchased from Dot Scientific (Burton, MI); isopropyl β-D-thiogalactopyranoside (IPTG) from
Goldbio (St. Louis, MO); Cobalt affinity resin from Thermo Scientific (Waltham, MA). Wang
resins, Fmoc protected amino acids and t-Boc protected amino acids were purchased from Peptides
International (Louisville, KY), Dupont (Wilmington, Delaware) and Sigma Aldrich (St. Louis,
MO). 1-13C Gly and 1-13C Ala were purchased from Cambridge Isotope Laboratories (Andover,
MA) and were N-Fmoc or N-t-Boc protected in our laboratory following literature procedures.1-3
1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC), 1-palmitoyl-2-oleoyl-sn-glycero-3[phospho-rac-(1-glycerol)] (sodium salt) (POPG), 1,2-Dipalmitoyl-sn-glycero-3-phosphocholine
(DPPG) and cholesterol were purchased from Avanti Polar Lipids (Alabaster, AL). Deuterium
labeled fatty acids Hexadecanoic-2,2-d2 Acid; Hexadecanoic-7,7,8,8-d4 Acid and Hexadecanoic15,15,16,16,16-d5 Acid were purchased from CDN Isotopes (Quebec, Canada) and shipped to
Avanti Polar Lipids (Alabaster, AL) to synthesize DPPC-D4, DPPC-D8 and DPPC-D10 lipids
(Structure shown in Figure 2.1). Other materials were obtained from Sigma-Aldrich (St. Louis,
MO).

56

Figure 2.1 (A) Structure of DPPC-D4, DPPC-D8 and DPPC-D10. (B) Approximate membrane
locations of the 2H’s in the membrane bilayer without protein. The lipid 2H are for the membrane
gel-phase.4

2.2 Gp41 constructs expression and purification
Figure 2.2 shows the schematic diagram and amino acid sequence of the four gp41
constructs purified in this study. Amino acid sequences of gp41 constructs are for the HXB2
laboratory strain of HIV and use gp160 numbering with color coding of different regions. HM
includes residues 535(M535C) – 581 + residue 628 – 683. FP_HM includes residue 512 – 581 +
residue 628 - 683. HM_TM includes residue 535(M535C) – 581+ residue 628 – 705. FP_HM_TM
includes residue 512 – 581 + residue 628 – 705. All constructs include a non-native sequence
SGGRGG in place of residue 582 – 627 and a non-native H6 affinity tag at their C-terminus that
is preceded by a G6LE or G8LE spacer. The spacer is needed for exposure of the affinity tag during
purification. The M535C mutation in HM and HM_TM is needed for native chemical ligation with
FP.
57

Figure 2.2 (A) Schematic diagrams of full-length HIV gp41 and the four gp41 constructs being
studied in this work with domains and corresponding colors: FP  fusion peptide, red; N-helix,
blue; Loop, grey; C-helix, green; MPER  membrane-proximal external-region, pink; TM 
transmembrane domain, orange; and endo  endodomain, white. The four constructs have nonnative SGGRGG replacing native residues 582-627. (B) Amino acid sequences with the same color
coding as panel A and the non-native C-terminal G6H6 or G8H6 in black. The H6 is for Co2+-affinity
chromatography and the G6/G8 are necessary spacers for exposure of the H6 tag.

58

DNA sequences of these four gp41 constructs are shown in Figure 2.3. All DNA inserts
were subcloned into pET-24a(+) vector that contains Lac operon and kanamycin resistance. The
plasmid was transformed into E. coli BL21(DE3) strain and then grew in 50 mL LB broth at 37 °C
overnight. The culture was mixed with 50% glycerol with 1:1 volume ratio and stored at -80 °C as
glycerol stock. The protein expression started with adding 50 L of the bacteria glycerol stock into
50 mL LB medium with Kanamycin 50 mg/L concentration. The 50 mg/L Kanamycin was used
to select for bacterial cells that contains the plasmid. After growth at 37 °C and 180 rpm shaking
overnight, the 50 mL culture was added to 1 L LB medium and kept growing for ~ 2 hr at the same
condition. When the OD600 of the culture increased to ~0.8, 2 mM Isopropyl β-D-1thiogalactopyranoside (IPTG) was added to start the induction. After 5 hr of induction, the cell
suspension was centrifuged at 9000 g, 4 °C for 10 min and the cell pellet was kept in -20 °C freezer.
5 g of wet cell was lysed by tip sonication in 40 mL phosphate-buffered saline (PBS) at pH
7.4. The insoluble fragments were pelleted down by centrifugation (48000 g, 4 °C for 20 min).
The pellet contains the desired recombinant protein (RP) since the RP is not soluble in PBS. The
PBS wash process was repeated two more times to remove the soluble proteins, molecules and
suspended membrane fragments which are only effectively pelleted by >100000 g. The RP pellet
was tip sonicated in 40 mL of PBS at pH 7.4 that also contained 8 M Urea, 0.5% sodium dodecyl
sulfate (SDS) and 0.8% Sodium lauroyl sarcosinate (Sarkosyl) until fully dissolved. 1 mL of Co2+
affinity resin was added to the solution and agitated for 2 hours at room temperature. After 2 hr
binding, the unbound proteins were removed by gravity filtration of the suspension. The following
steps were different to solubilize gp41 constructs in SDS or dodecylphosphocholine (DPC).

59

HM
TGTACGCTGACGGTCCAAGCACGTCAGCTGCTGAGCGGCATTGTGCAGCAACAGAACAATCTGCTGCGCGCGATC
GAAGCCCAACAGCATCTGCTGCAGCTGACCGTTTGGGGTATTAAACAACTGCAGGCTCGTATCCTGAGCGGCGGT
CGCGGCGGTTGGATGGAATGGGATCGTGAAATTAACAATTATACGAGCCTGATTCACTCTCTGATCGAAGAAAGT
CAAAACCAACAGGAGAAAAACGAACAGGAACTGCTGGAACTGGACAAATGGGCCTCCCTGTGGAACTGGTTTAAC
ATTACGAACTGGCTGTGGTACATCAAAGGCGGCGGTGGCGGTGGT

HM_TM
TGTACCCTGACGGTCCAAGCTCGCCAACTGCTGAGTGGTATCGTTCAACAACAAAACAATCTGCTGCGTGCTATC
GAAGCCCAACAACATCTGCTGCAGCTGACCGTGTGGGGCATTAAACAGCTGCAGGCCCGTATCCTGAGCGGCGGT
CGCGGCGGTTGGATGGAATGGGATCGTGAAATTAACAATTATACGAGCCTGATTCACTCTCTGATCGAAGAAAGT
CAGAATCAGCAAGAGAAAAACGAACAAGAACTGCTGGAACTGGACAAATGGGCCTCCCTGTGGAACTGGTTTAAT
ATTACCAACTGGCTGTGGTACATCAAACTGTTCATTATGATCGTTGGCGGTCTGGTCGGTCTGCGTATTGTGTTT
GCTGTCCTGAGTATTGTCGGTGGCGGTGGCGGCGGT

FP_HM
GCCGTGGGTATCGGTGCTCTGTTCCTGGGTTTCCTGGGTGCTGCTGGTTCGACGATGGGTGCCCGCTCAATGACG
CTGACGGTCCAAGCACGTCAGCTGCTGAGCGGCATTGTGCAGCAACAGAACAATCTGCTGCGCGCGATCGAAGCC
CAACAGCATCTGCTGCAGCTGACCGTTTGGGGTATTAAACAACTGCAGGCTCGTATCCTGAGCGGCGGTCGCGGC
GGTTGGATGGAATGGGATCGTGAAATTAACAATTATACGAGCCTGATTCACTCTCTGATCGAAGAAAGTCAAAAC
CAACAGGAGAAAAACGAACAGGAACTGCTGGAACTGGACAAATGGGCCTCCCTGTGGAACTGGTTTAACATTACG
AACTGGCTGTGGTACATCAAAGGCGGCGGTGGCGGTGGT

FP_HM_TM
GCGGTTGGCATTGGTGCGCTGTTCCTGGGCTTTCTGGGTGCGGCGGGTAGCACTATGGGTGCGCGTAGCATGACC
CTGACCGTTCAAGCGCGTCAACTGCTGAGCGGTATCGTGCAGCAACAGAACAACCTGCTGCGTGCGATTGAGGCG
CAACAGCACCTGCTGCAGCTGACCGTTTGGGGCATCAAGCAACTGCAGGCGCGTATTCTGAGCGGTGGCCGTGGT
GGCTGGATGGAGTGGGACCGTGAAATCAACAACTACACCAGCCTGATCCACAGCCTGATTGAGGAAAGCCAAAAC
CAACAGGAGAAGAACGAGCAGGAACTGCTGGAACTGGATAAATGGGCGAGCCTGTGGAACTGGTTCAACATCACC
AACTGGCTGTGGTACATCAAACTGTTCATCATGATCGTGGGTGGCCTGGTTGGTCTGCGTATCGTGTTTGCGGTT
CTGAGCATTGTTGGCGGTGGTGGTGGCGGTGGTGGT

Figure 2.3 DNA sequences of gp41 inserts. Each line is 75 nucleotides.

60

To solubilized proteins in SDS detergent, the Co2+ resin was washed with 31 mL aliquots
of the PBS/urea/SDS/Sarkosyl solution and subsequent gravity filtration to remove unbound
protein.5 Bound protein was eluted from the resin by 40.5 mL aliquots of the
PBS/urea/SDS/Sarkosyl solution with 250 mM imidazole, and gravity filtration. The eluent
fractions were pooled and then mixed with an equal volume of buffer that contained 10 mM TrisHCl at pH 8.0, 0.17% n-Decyl- β-D-Maltoside (DM), 2 mM EDTA, and 1 M L-arginine, with
subsequent agitation overnight at 4 °C. Arginine, urea, and detergents were removed by dialysis
against 10 mM Tris at pH 7.4 with 0.2% SDS. If the intermediate mixing step with the arginine
solution was skipped, RP precipitated during dialysis, which suggests that RP aggregates are
broken up by the arginine solution.
To solubilize proteins in DPC detergent, proteins need to be refolded on column to prevent
aggregation. After RP binds on Co2+ resin, column was washed by 31 mL of 20 mM sodium
phosphate at pH 7.4 with 0.25% DPC, with subsequent gravity filtration. RP was then eluted in
the pH 7.4 phosphate buffer with 0.25% DPC and 250 mM imidazole. The eluent fractions were
pooled and imidazole removed by dialysis against pH 7.4 phosphate buffer with 0.25% DPC, or
by dialysis against 20 mM sodium acetate buffer at pH 3.2 with 0.25% DPC. Purified RP
concentrations were determined using A280.

2.3 Circular Dichroism (CD)
Spectra were obtained with a Chirascan instrument (Applied Photophysics) at ambient
temperature. Proteins with 10-15 M concentration were added to a 1 mm path length quartz
cuvette. 190 nm – 260 nm spectral range was scanned with 0.5 nm steps and 1.5 s per step
averaging time. For each protein samples, spectra were averaged and the final spectra were the

61

(protein + buffer) – (buffer) difference. Thermal stability of folded protein was probed by CD in
5 °C increments over a 25-90 °C range using a J-810 instrument (Jasco) with a water circulation
bath. No visible precipitation was observed at any temperature or after cooling down.

2.4 Size-Exclusion Chromatography (SEC)
The chromatography was done using a DuoFlow Pathfinder 20 instrument (Bio-Rad) with
a Tricorn Superdex 200 Increase 10/300 GL column (GE Technologies). Flow rate was set to 0.3
mL/min and the detector was set to A280. Proteins were dialyzed against the running buffer,
concentrated to ~ 1 mg/mL and centrifuged to remove any precipitate before injected to the
instrument. There was no visible precipitate for samples in SDS and a very small precipitate for
samples in DPC. A ~100 L aliquot was injected for each run.

2.5 Protein induced vesicle fusion
The fusion activities of gp41 constructs were assessed by their ability to induce fusion
between unilamellar lipid vesicles.6 Two vesicle compositions were used in this assay: (1) 1palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC), 1-palmitoyl-2-oleoyl-sn-glycero-3phospho-(1'-rac-glycerol) (POPG) and Cholesterol (Chol) with 8:2:5 mole ratio; and (2)
POPC:Chol with 2:1mole ratio. The Chol mole fraction in both compositions is close to that of the
plasma membrane of host cells of HIV. The lipid vesicles with these two compositions are
negatively charged and neutral, respectively, so comparison between them allows assessment of
the impact of protein/vesicle electrostatics on fusion. To prepare unilamellar lipid vesicles, 1.6
mol of POPC, 0.4 mol of POPG and 1 mol of Chol or 2 mol of POPC and 1 mol of Chol
were dissolved in 1 mL chloroform and then the solvent was evaporated by dry nitrogen gas and

62

keep vacuum dry overnight. The dry lipids films were suspended in 2 mL of aqueous buffers (10
mM HEPES, 5 mM MES at pH 7.4 or 10 mM sodium formate at pH 3.2) and a homogeneous
suspension was created by 10 freeze/thaw cycles. Unilamellar vesicles were obtained by extrusion
of the suspension through a polycarbonate membrane with 100 nm diameter pores for 10 times.
For each composition, a set of companions “labeled” vesicles were also prepared that contained
additional 2 mol % of the fluorescent lipid NDB-PE and 2 mol % of the quenching lipid Rh-PE.
Labeled vesicles were mixed with unlabeled vesicles in 1:9 ratio. Considering a 30% lipid loss
during the extrusion, the final concentration of [POPC+POPG] lipids is ~150 M and chol is ~75
M. The solution was transferred to a quartz cuvette in a fluorimeter and subjected to constant
stirring at 37 °C. Fluorescence was monitored using 467 nm excitation, 530 nm detection, and 1 s
time increment. After measurement of the baseline fluorescence F0, a protein aliquot was added
and marked time t = 0. The stock solution contained [protein] ~ 40 M in 10 mM Tris buffer at
pH 7.4 with 150 mM NaCl and 0.2% SDS. The time-dependent fluorescence increase F(t) = F(t)
- F0 is resulted from protein-mediated fusion between a labeled and unlabeled vesicle and causing
a longer average distance between fluorophore-quencher lipids. The dead-time in the assay was
~10 s and final asymptotic fluorescence was usually reached by ~600 s. The maximum
fluorescence change (Fmax) was detected after addition of 12 L 10% Triton X-100 which
solubilized the vesicles. The percent fusion was calculated as [F(t)/Fmax] × 100. No fusion was
detected after the addition of the detergent solution without protein.

2.6 Solid phase peptide synthesis, purification and characterization
The IFP sequence used in this research is: GLFGAIAGFIENGWEGMIDGGGKKKKG.
The underlined residues are the first 20 residues of the N-terminal of HA2 subunit, and the

63

sequence is the H3 subtype of hemagglutinin.7 The IFP has a non-native C-terminal tag. The four
Lysine residues were designed to increase the peptide solubility in water, which makes the
purification and NMR sample preparation easier to perform. Four IFP samples were synthesized
with different isotopically labeling scheme. They are Leucine 2, Alanine 7 and Glycine 16 residues
carbonyl carbon 13C labeled and denoted as IFP_L2C, IFP_A7C and IFP_G16C, respectively.
IFP was manually synthesized by Fmoc solid phase peptide synthesis. The Fmoc peptide
synthesis started with Wang resin (polymer-bounded 4-Benzyloxybenzyl alcohol), which has one
Fmoc protected Glycine attached to it. The following steps added one amino acid to the resin, and
the peptide can be synthesized by repeating these steps:
1. Deprotection: 20% piperidine in DMF was added to the resin to remove the Fmoc group
from the existing peptide N-terminal.
2. Amino acid coupling: amino acid with HOBt, HBTU and DIEA in DMF can couple the
amino acid to unprotected peptide N-terminal.
3. Capping: 5% acetic anhydride, 2% DIEA and 0.2% HOBt in DMF was used to esterify
the N-terminus of the peptide and stop the following side reaction from the unreacted peptide.
Before doing the next step, the solution from the previous step was drained by filtration
and the resin was washed by DMF. After the desired sequence is synthesized, the peptide was
cleaved from the resin by using 5% water, 2% Tri-isopropyl silane, 2% thioanisole and 2%1, 2ethane-dithiol in TFA solution and precipitated in cold ether. The peptide was purified using
reversed phase HPLC equipped with a semi-preparative C18 column using water - acetonitrile
containing 0.1% TFA as mobile phase. TFA helps to maintain the acidic pH (water with 0.1% TFA
has a pH ~ 2) and neutralizes the carboxylate group present in the peptide. The purity of the peptide
was checked by mass spectrometry and resulted in ˃ 95% peptide purity.

64

2.7 solid state NMR sample preparation
The lipid composition was DPPC and DPPG with 4:1 ratio. This composition was chosen
because: (1) the PC lipids are the major fraction in influenza virus host cell, and (2) the host cell
membrane is negatively charged.8 50 mol lipids were dissolved in 2 mL chloroform and methanol
solution with a 9:1 volume ratio, and the solvent was removed by dry nitrogen gas flow and
vacuum pumping overnight. 3 mL of 10 mM HEPES and 5mM MES buffer at pH 5.0 was used
to hydrate the lipid film and followed by 10 times freeze-thaw cycles to make a homogeneous
suspension. The lipid-buffer suspension was extruded through a polycarbonate membrane with
100 nm pore size to get large unilamellar vesicles. 1 mol IFP was dissolved in 10 mL of the
HEPES/MES buffer and was added to lipid vesicles drop by drop, then agitate overnight. The
peptide-lipid complex was pelleted down by ultra-centrifugation at 100000g for 5 hours. The
proteo-liposome complex pellet was lyophilized overnight. Lyophilization helps to reduce sample
lost when pack the sample into NMR rotor. The sample was packed in NMR rotor and rehydrate
with 10 L of the HEPES/MES buffer overnight at room temperature.
In PRE experiments, MnCl2 was added in membrane samples as paramagnetic species. The
Mn2+ was added at the rehydration step followed by 5 freeze-thaw cycles to ensure the accurate
amount of Mn2+ was added and the Mn2+ ions were evenly distributed on both side of the
membrane.9

2.8 solid state NMR
Rotational Echo Double Resonance (REDOR) solid state NMR and Paramagnetic
Relaxation Enhancement (PRE) solid state NMR were used in this study.10,

11

Spectra were

obtained from a 9.4 T Agilent Infinity Plus spectrometer and triple - resonance MAS probe tuned

65

to 1H, 13C, and 2H frequencies. Figure 2.4A shows the pulse sequence of REDOR and quadrupolar
echo (quecho) pulse sequence in PRE experiment.
In REDOR experiment, the rotor was span at 10 kHz and -50 °C, cooling by nitrogen gas
flow and the expected sample temperature is ~ -30 °C. The REDOR experiments collects two sets
of data: S0 and S1. In both experiments: (1) 1H π/2 pulse; (2) 1H to 13C cross-polarization (CP); (3)
1

H decoupling; (4) 13C π pulses at the end of each rotor period; and (5) 13C detection. 2H π pulses

are applied in the middle of each rotor period in S1 experiment but absent in S0 experiment. The
spectra were processed with 100 Hz Gaussian line broadening and referenced to adamantane
methylene chemical shift, which is 40.5 ppm. The S0 and S1 intensities were calculated with a 3ppm width of the 13CO peak. The uncertainties were the RMSD’s of 6 spectral noise regions with
3-ppm widths. The REDOR dephasing is defined as:
∆𝑆/𝑆0 = (𝑆0 − 𝑆1 )/𝑆0

2.1

The dephasing is a function of dephasing time , whose duration is the after 1H to

13

C CP and

before 13C acquisition. The dephasing is plotted vs τ and the dipolar coupling (D) and internuclear
distance (r) between 13C and 2H can be calculated:
𝐷(𝐻𝑧) =

𝜇0 𝛾13𝐶 𝛾2𝐻 ℎ
16𝜋 2 𝑟 3

=

4642Å3 𝑠 −1

2.2

𝑟3

66

Figure 2.4 (A) 13C - 2H REDOR pulse sequence and (B) quecho pulse sequence.

The PRE experiments were designed for measuring the IFP-bounded membrane lipid acyl
chain location by using quadrupolar echo solid-state NMR experiments and the pulsed sequence
is shown in Figure 2.4 panel B. 2H spectra were acquired with different 1 and 2 and a fixed ( 1) value. Typically, the pulse length was set around 1.5 s; 1, 2 were set between 10 and 1000
s; the recycle delay was set to 1 s. Samples are detected at 25 °C (gel phase lipids) and 50 °C
(fluid phase lipids). In data processing, the FID needs to be shifted to make the maximum echo
signal at t = 0 and the total relaxation time is denoted as t and t = [1 + 2 + time being shifted].
Data processing was done by 1000 Hz exponential line broadening, baseline correction and data
shift. 2H T2 relaxation time is determined by the decay of acquired signal as a function of t. By
plotting the signal intensity (I) in FID vs t, the T2 relaxation time can be calculated by:

67

𝑡

2.3

𝐼(𝑡) = 𝐼(0)×𝑒𝑥𝑝 (𝑇 )
2

The T2 relaxation rate R2 can be calculated as:
1

2.4

𝑅2 = 𝑇

2

Thus, the internuclear distance information between 2H in deuterated lipid acyl chain and the
paramagnetic species Mn2+ on membrane surface can be obtained.12

68

REFERENCES

69

REFERENCES

[1] Chang, C. D., Waki, M., Ahmad, M., Meienhofer, J., Lundell, E. O., and Haug, J. D. (1980)
PREPARATION
AND
PROPERTIES
OF
N-ALPHA-9FLUORENYLMETHYLOXYCARBONYLAMINO ACIDS BEARING TERT-BUTYL
SIDE-CHAIN PROTECTION, Int. J. Pept. Protein Res. 15, 59-66.
[2] Lapatsanis, L., Milias, G., Froussios, K., and Kolovos, M. (1983) SYNTHESIS OF N-2,2,2(TRICHLOROETHOXYCARBONYL)-L-AMINO
ACIDS
AND
N-(9FLUORENYLMETHOXYCARBONYL)-L-AMINO
ACIDS
INVOLVING
SUCCINIMIDOXY ANION AS A LEAVING GROUP IN AMINO-ACID
PROTECTION, Synthesis, 671-673.
[3] Harris, R. B., and Wilson, I. B. (1984) TERT-BUTYL AMINOCARBONATE (TERTBUTYLOXYCARBONYLOXYAMINE) - A NEW ACYLATING REAGENT FOR
AMINES, Int. J. Pept. Protein Res. 23, 55-60.
[4] Jia, L. H., Liang, S., Sackett, K., Xie, L., Ghosh, U., and Weliky, D. P. (2015) REDOR solidstate NMR as a probe of the membrane locations of membrane-associated peptides and
proteins, J. Magn. Reson. 253, 154-165.
[5] Porath, J. (1992) IMMOBILIZED METAL-ION AFFINITY-CHROMATOGRAPHY, Protein
Expr. Purif. 3, 263-281.
[6] Struck, D. K., Hoekstra, D., and Pagano, R. E. (1981) USE OF RESONANCE ENERGYTRANSFER TO MONITOR MEMBRANE-FUSION, Biochemistry 20, 4093-4099.
[7] Nobusawa, E., Aoyama, T., Kato, H., Suzuki, Y., Tateno, Y., and Nakajima, K. (1991)
COMPARISON OF COMPLETE AMINO-ACID-SEQUENCES AND RECEPTORBINDING PROPERTIES AMONG 13 SEROTYPES OF HEMAGGLUTININS OF
INFLUENZA A-VIRUSES, Virology 182, 475-485.
[8] Ghosh, U., Xie, L., Jia, L. H., Liang, S., and Weliky, D. P. (2015) Closed and Semiclosed
Interhelical Structures in Membrane vs Closed and Open Structures in Detergent for the
Influenza Virus Hemagglutinin Fusion Peptide and Correlation of Hydrophobic Surface
Area with Fusion Catalysis, J. Am. Chem. Soc. 137, 7548-7551.
[9] Su, Y., Mani, R., and Hong, M. (2008) Asymmetric insertion of membrane proteins in lipid
bilayers by solid-state NMR paramagnetic relaxation enhancement: A cell-penetrating
peptide example, J. Am. Chem. Soc. 130, 8856-8864.
[10] Gullion, T. (1998) Introduction to rotational-echo, double-resonance NMR, Concepts Magn.
Resonance 10, 277-289.

70

[11] Clore, G. M., and Iwahara, J. (2009) Theory, Practice, and Applications of Paramagnetic
Relaxation Enhancement for the Characterization of Transient Low-Population States of
Biological Macromolecules and Their Complexes, Chem. Rev. 109, 4108-4139.
[12] Gabrys, C. M., Yang, R., Wasniewski, C. M., Yang, J., Canlas, C. G., Qiang, W., Sun, Y.,
and Weliky, D. P. (2010) Nuclear magnetic resonance evidence for retention of a lamellar
membrane phase with curvature in the presence of large quantities of the HIV fusion
peptide, Biochim. Biophys. Acta-Biomembr. 1798, 194-201.

71

Chapter 3
Expression, purification and functional characterization of
HIV gp41 ectodomain and transmembrane domain

72

3.1 Introduction
Human Immunodeficiency Virus (HIV) leads to the Acquired Immunodeficiency
Syndrome (AIDS), which is a disease that causes human immune system failure.1 HIV virus is
enveloped by lipid membrane obtained during viral budding from an infected host cell.2 On HIV
envelop, glycoprotein gp120 and gp41 form a “knob” shape complex in a noncovalent manner
where three molecules of gp120 forms the “cap” and three molecules of gp41 forms the “stem”.
The first step of HIV infection cycle is the binding of gp120 onto the host cell CD4 and
coreceptors, and this binding can cause conformational change and lead to gp120 displacement.35

gp41 is then exposed and it catalyzes the fusion between viral and host cell membrane. At the

end of the fusion, a fusion pore is formed and the viral nucleocapsid is released into the host cell
cytoplasm.6 The RNA is reverse transcribed into DNA catalyzed by reverse transcriptase and then
the viral proteins are expressed. Finally new viruses bud from the host cell membrane and start the
new infection cycle.7 This process is detailed discussed in chapter 1 and is also shown in Figure
1.2.8
gp41 is the only fusion protein in HIV and it plays a vital role in membrane fusion.
Numerous of previous studies illustrate that the gp41 undergoes a huge conformational change
before and after fusion. X-ray crystallography and cryo-electron microscopy are used to study the
structure of gp41 and big progresses has been made to reveal the structure and its possible fusion
mechanism.6, 9-11
Crystal and cryo-electron microscopy structure of gp41 at its pre-fusion state included a 50
Å long parallel bundle of three NHR helices in the center, and three CHR 40 Å long helix forming
a tripod (Figure 1.5). The monomer structure was NHR-helix/60°-turn/CHR-helix.9, 11, 12 X-ray
crystallography result of gp41 at post-fusion state was a six-helical bundle (SHB) state, in which

73

three NHR helices formed a parallel trimeric coiled-coil in the center and three CHR helices
wrapped antiparallel outside. The monomer structure changed to NHR-helix/180°-turn/CHR-helix
(Figure 1.4).6, 13
The FP, MPER and TM are the domains that interacting with either host cell or viral
membrane and they all play key roles during membrane fusion. However, the structures of these
domains were ether truncated or missing in crystallography and cryo-electron microscopy results.
The FP alone adopted antiparallel β-sheet structure in physiologically relevant membrane
environment.14 The functional study of FP also suggested its importance in membrane fusion.15
MPER is the ectodomain region that is closest to the viral membrane with two discrete  helical
segments with a central hinge structure.16, 17 TM is the domain that anchors the gp41 in the viral
membrane with a continuous long -helix with MPER. The structural study results of FP, MPER
and TM were shown in Figure 1.7 and 1.8.
The previous functional studies of gp41 also provide information regarding its structure.
FP induced vesicle fusion was ~10 times less when Val_2 was mutated to Glu_2, which was
consistent with much higher fusion and infection of HIV with wild-type compared to HIV with
the V2E mutation. This result supported the catalytic significance of the FP domain in gp41.18, 19
Solid-phase synthesized FP was monomeric in solution. To obtain FP-dimer or FP-trimer, a tag
containing cysteine and lysine at the peptide C-terminal was added. FP-dimer was obtained by
cysteine cross-linking between monomers and FP-trimer by synthesizing a peptide bond between
the peptide C-terminal carbonyl group in a dimer and a lysine -NH2 group in monomer. Lipid
mixing percentage of FP-trimer was 3 folds greater than FP-dimer and 10 folds greater than FPmonomer, which suggested that higher oligomeric constructs induce more vesicle fusion.20 Vesicle
fusion experiments for large gp41 constructs containing N47(L6)C39 (match the SHB crystal

74

structure in Figure 1.4) and N70(L6)C39 (with appending of FP domain) showed significant lipid
mixing for vesicle composition POPC:POPG:Chol (8:2:5) at pH 3.5, but a rapid loss in fusion
activity was observed as pH increase, and completely no fusion at pH 5.5. This suggested a strong
depend on electrostatic interactions between proteins and vesicles. At low pH, the positively
charged protein had electrostatic attraction with negatively charged lipids. As pH increased, the
protein charge changed from positive to negative and resulted in repulsion between protein and
vesicles, and lead to poor protein binding on vesicles and no fusion.21 By varying the fraction of
negatively charged PG in vesicles, the fusogenicity was also affected. With higher PG fraction, the
electrostatic repulsion between vesicles was greater and resulting in lower fusion.22
This study is aimed to have an insight of the gp41 structure and function correlation. In this
study, I have synthesized the gp41 protein that includes the gp41 construct of residue 512581+SGGRGG+628-705, and characterized its overall structure, oligomeric state as well as its
fusion function. The overall goal is to get a better understanding of the structure and fusion
mechanism of gp41 and may help the further vaccine and antiretroviral drug development.

3.2 Result and discussion
3.2.1 Protein solubilization and purification
The constructs and amino acid sequences of full-length HIV gp41 and the four gp41
constructs being studied in this work are shown in Figure 3.1. The constructs had a non-native 6residue sequence that replaced the SE loop and adjoining terminal regions of the SE N- and Chelices. This modified SE was chosen because of better solubility properties. It also showed that
is can retain hyperthermostable helical structure with Tm > 90 °C in HM.23 The protein expression
procedure is discussed in Chapter 2. After expression, the first step in recombinant protein (RP)
purification was cell lysis in PBS. The four gp41 constructs are insoluble in PBS, since SDS-PAGE
75

of the soluble lysate did not show any bands at corresponding molecular weight, as would be
expected for membrane proteins. Thus, after centrifugation at 48000 g, 4 °C for 20 min, the RP’s
were in the pellet. The pellet was resuspended in PBS and centrifugated (2×) to remove non-RP
material and this is consistent with visibly decrease in pellet size.
RP expression levels were assessed by static solid-state NMR. The cells were grown in LB
broth and then transferred to minimal medium, followed by IPTG induction and adding 1-13C Gly.
The procedure was developed from our group’s previous work.24 50 mL of LB with 0.5 mL of
glycerol stock of E. coli cells that contained a specific plasmid was incubated at 37 °C and 180
rpm shaking overnight. The cells were harvested by centrifugation at 10,000g at 4°C for 10 minutes
and then resuspended into a 250 mL baffled flask containing 50 mL of M9 minimal medium, 100
L of 1.0 M MgSO4, and 250 L of 50% glycerol at 180 rpm and 37 °C. 2.0 mM IPTG was added
after 1 hr. A dry mixture contained 10 mg of 13CO labeled glycine and 10 mg of each of the other
19 common amino acids were added at the same time. An equivalent mixture was added after one
hour of induction. After three hours of expression, the cells were harvested by centrifugation at
10,000 g and 4°C for 10 minutes. The wet cells were resuspended in PBS by tip sonication. The
pellet of the lysate after centrifugation was packed into NMR rotor and

13

C NMR spectra were

obtained (Figure 3.2). The chemical shifts of the peaks in spectra were ~ 175 ppm. By comparing
to 13C static NMR of proteins samples, these peaks were assigned to 13CO.25-27 The similar pellet
sizes and the comparable 13CO signal intensities evidenced similar quantities of expressed RP for
each construct.

76

Figure 3.1 (A) Full-length HIV gp41 and the four gp41 constructs. (B) Amino acid sequences with
the same color coding as panel A and the non-native C-terminal G6H6 or G8H6 in black.

77

Figure 3.2 Static 13C NMR spectra of lysate pellets labeled with 1-13C Gly. Each spectrum was
the sum of 1000 scans.

To find conditions that completely solubilized the RP-rich pellet, four different conditions
that contained PBS at pH 7.4 and either: (1) 8M Urea; (2) 6M GuHCl; (3) 8M Urea and 0.8%
Sarkosyl; or (4) 8M Urea, 0.5% SDS, and 0.8% Sarkosyl; were tested. The pellets were visually
soluble in these buffers after sonication, and the pellet size after centrifugation was similar. Each
solution was then subjected to Co2+-affinity chromatography, followed by SDS-PAGE (Figure 3.3
and Figure 3.4). The highest purity and yield were obtained for mixture (4) which was then used
for all subsequent protein purification.
All constructs have a non-native C-terminal H6 affinity tag preceded by either a G6LE or
G8LE spacer, where the spacer is required for RP binding to the Co2+ resin. According to previous
crystallography result, the soluble ectodomain (SE, which includes CHR, loop and NHR) by itself
likely still adopts helical hairpin structure,23, 28 so we expect that the spacer provides greater solvent
exposure of the H6 tag.
The longer G8LE spacer was required for FP_HM_TM binding to the resin. The affinity
purification eluent of FP_HM_TM with G6LE spacer showed no product in SDS-PAGE. The
reason might be the H6 tag buried in protein hydrophobic core, leading to poor affinity column

78

binding. After adding two other glycine residues to create a G8LE spacer, a band corresponds to
the MW of FP_HM_TM showed on SDS-PAGE, which evidences that a longer spacer increases
the exposure of H6 tag.
To find conditions without urea and Sarkosyl for which solubility was retained for all
purified RP’s, dialysis was done against a variety of buffers, and soluble RP’s were obtained for
10 mM Tris at pH 7.4 and 0.2% SDS. On the other hand, since PC lipids is a significant component
of HIV host cell membranes and the DPC has a similar phosphatidylcholine headgroup, the RP’s
were also dialyzed against buffers with DPC detergent. The RP’s precipitated quickly if dialyzed
directly against low- or neutral- pH buffers containing 0.25% DPC, which is not consistent with
other groups’ results.29, 30 One previous study suggested to refold RP on affinity column.31 In our
observation, all RP’s were soluble if the initial exchange into buffer with 0.25% DPC was done
with RP bound to Co2+-resin, followed by elution using buffer with DPC and 250 mM imidazole,
and then dialysis to remove the imidazole. The final solutions contained 0.25% DPC and either 20
mM sodium acetate buffer at pH 3.2 or 20 mM sodium phosphate buffer at pH 7.4. RP solubility
when initial exchange was done with resin-bound RP vs. RP precipitation when initial exchange
was done by dialysis of a RP solution suggests that RP aggregates can form quickly, but are not
the lowest free-energy state in 0.25% DPC.
Figure 3.4 displays SDS-PAGE of the four RP’s after dialysis into 10 mM Tris at pH 7.4
with 0.2% SDS. Bands corresponding to FP_HM, HM_TM, and FP_HM_TM were cut from a gel
and subjected to trypsin digestion and mass spectrometry. Peptides were identified from each of
the three bands that provided 80, 55, and 75% sequence coverage, respectively (Figure 3.5). In
Figure 3.5, the highlighted yellow residues were in a peptide that identified by mass spectrometry.

79

The highlighted green Met residues had masses consistent with oxidation and the highlighted green
Asn and Gln residues had masses consistent with deamination.
Typical purified yields of HM, FP_HM, HM_TM, and FP_HM_TM were respectively ~10,
0.5, 0.5, and 0.3 mg/L bacterial culture. Such significant difference in yields was due to the
differences in their binding to the Co2+-resin, since each construct had similar expressed RP
quantities. FP_HM_TM required a longer glycine spacer than the other constructs to bind to Co2+resin. This suggests occlusion of the H6 tag by FP and TM segments, which is sterically plausible
because of SE helical hairpin folding even in the mixtures with denaturant at high concentration.23

80

Figure 3.3 SDS-PAGE after Co2+-affinity chromatography of the solubilized pellet enriched in
inclusion body protein. The protein is FP_HM and different solubilization conditions for the pellet
are noted. Only the MW marker lane and relevant elution lane(s) are displayed.

Figure 3.4 SDS-PAGE of the purified HM (MW = 13.7 kDa), HM_TM (MW = 16.7 kDa), FP_HM
(MW = 16.5 kDa), and FP_HM_TM (MW = 18.9 kDa) using solubilization condition: PBS at pH
7.4 with 8M Urea, 0.5% SDS, and 0.8% Sarkosyl.
81

Figure 3.5 Sequence coverage of FP_HM, HM_TM, and FP_HM_TM after trypsin digestion.

3.2.2 Influence of FP and TM on hyperthermal Îą helical hairpin structure
Figure 3.6 shows the CD spectra at ambient temperature of the four gp41 constructs in: (A)
0.2% SDS at pH 7.4; (C) 0.25% DPC at pH 7.4; and (D) 0.25% DPC at pH 4.0. The protein
concentration and all spectra for a single buffer were acquired during the same day to obtain the
most quantitative comparison between constructs. Panel B displays 222 vs. temperature plots
derived from CD spectra in panel A (0.2% SDS at pH 7.4). Spectra at 25, 60, and 90 °C are
presented in Figure 3.7. The temperature-series of spectra for a single construct were all obtained
during the same day. The panel A vs. B differences between the ambient-temperature 222 values
for the same construct are in part due to use of different CD instruments. There is the same ordering
of 222 values with construct in both panels.
There are several trends among the ambient-temperature CD spectra in Figure 3.6 panel A,
C and D. All spectra have the characteristic shape of  helical secondary structure with minima
near 208 and 222 nm. This helicity is consistent with the helical hairpin structure of gp41 SE. The
82

| | values for a single construct in DPC are similar at low and neutral pH, which supports a pHindependent hairpin structure. The | DPC|/| SDS| ≈ 3/2, with some variation both among constructs
and with wavelength. This ratio suggests higher helicity in DPC vs. SDS.
The HM_TM construct has the largest | | values in all three buffer conditions, with |222|
≈ 28,000 (degrees-cm2/dmole-residue) in DPC that correlates with 85% average helicity. This is
equal to the average helicity calculated using a model of 100% helicity of 123/125 of the native
residues and 0% helicity for 2 native residues as well as 20 non-native residues, i.e. SGGRGG
loop and C-terminal G6LEH6 (Table 3.1). Nearly complete native helicity is consistent with a fullyfolded protein containing SE helical hairpin and TM helix structural elements. Similar analysis of
|222| of HM yields ~76% experimental helicity and ~10 non-helical native residues. These residues
are most likely at the N- and C-termini of the hairpin, based on reasoning that the TM domain
stabilizes hairpin helical structure.

83

Figure 3.6 Circular dichroism spectra of samples containing ~10 ÂľM protein concentration in
different buffer + detergent solutions: (A) 10 mM Tris at pH 7.4 and 0.2% SDS; (C) 20 mM
phosphate at pH 7.4 and 0.25% DPC; and (D) 20 mM acetate at pH 4.0 and 0.25% DPC. The
spectra in panels A, C, and D spectra were obtained at ambient temperature. (B) the 222 values
derived from spectra in 0.2% SDS at pH 7.4 and temperatures between 25 and 90 °C. The panel A
vs. B differences between the ambient-temperature 222 values of the same construct may reflect
measurement uncertainties in protein concentrations and use of different CD instruments.

84

Figure 3.7 Circular dichroism spectra of four gp41 constructs with ~10 ÂľM protein concentration
in 10 mM Tris at pH 7.4 and 0.2% SDS at 25, 60 and 90 °C. The panels are noted corresponds to
each construct.

85

Table 3.1 Analysis of CD spectra in 0.25% DPC at pH 4
Construct

Average fractional helicity a

Number non-helical residues b

HM

0.76

10

HM_TM

0.85

2

FP_HM

0.73

20

FP_HM_TM

0.73

24

a

Calculated using 100% helicity as222= 33000 degrees-cm2/dmole-residue

b

Calculated using Nnon-helical = Ntot – Nnon-native – fhelix/Ntot

The CD result showed a general trend that addition of the FP segment decreases average
helicity in both SDS and DPC, and this correlates with similar loss of helicity deduced from
comparative CD spectra of the shorter “HP” and “FP_HP” constructs.28 HP and HM have identical
N-terminal sequences including SGGRGG non-native loop, but HP lacks the 17 native C-terminal
residues of HM, as well as the non-native C-terminal G6LEH6. The average helicity of FP_HM
and FP_HM_TM in DPC at pH 4 is ~73%, which corresponds to non-helical structure for ~20
(FP_HM) and ~24 (FP_HM_TM) native residues. Most of these residues are likely within the FP,
based on ~15-25 non-helical residues when the 23-residue N-terminal FP segment is appended to
either HP, HM, or HM_TM. Even the  strand FP is consistent with previous CD result, it is not
consistent with earlier NMR chemical shifts.18, 28, 29 We don’t understand the discrepancy between

86

the CD and NMR data, but note that in membrane, the FP often adopts intermolecular antiparallel
 sheet structure.14
Figure 3.6 panel B shows only moderate decreases in |222| with temperature in the 25-90
°C range, which supports Tm ≥ 90 °C for the four constructs. HM and FP_HM exhibit similar
(|222|) ≈ 4000 degrees-cm2/dmole-residue over the 25-90 °C range, which is consistent with
assignment of the helical structure to the HM region. In some contrast, HM_TM exhibits (|222|)
≈ 5000, and FP_HM_TM exhibits (|222|) ≈ 3000 degrees-cm2/dmole-residue, so adding both the
FP and TM domains stabilizes the helical hairpin structure of HM. Close FP/TM contact has not
been detected by NMR, so increased thermostability may be due to location of the FP and TM
segments in the same micelle, so opening of the hairpin requires micelle deformation to maintain
detergent contact with both FP and TM.30, 32, 33 There isn’t this penalty if the construct contains
only the FP or the TM. Common features of most studies are the very high helicity of the SE and
Tm > 100 °C, particularly if the construct contains longer N- and C- helix segments with
complementarity between their hydrophobic surfaces. These features are exhibited with or without
the native loop, for solutions without detergent at low pH, and for solutions with detergent at low
and neutral pH.23, 34, 35 Similar result from our group showing moderate CD changes of the shorter
HP and FP_HP constructs, and the combined data with current result support hyperthermostable
hairpin SE structure for all constructs.23 By combining all data, the results support greater
thermostability of FP_HM_TM vs. more truncated constructs. The same thermostability
correlation was also observed for the influenza virus HA2 construct, whose sequence is nonhomologous with gp41, but performs a similar fusion function.36

87

3.2.3 Oligomeric states vary in different detergent
Figure 3.8 displays SEC of the gp41 constructs in: (A) SDS at pH 7.4; (B) DPC at pH 7.4;
and (C) DPC at pH 4.0. These detergents were chosen in part because all proteins are thermostable,
as assessed by their CD spectra. The running [protein] ≈ 10 M is comparable to the CD
experiments. All buffers also contained 150 mM NaCl to inhibit protein binding on column. The
0.2% SDS is ~5×CMC and the 0.25% DPC is ~8×CMC.37-39 There is <10% statistical probability
of protein occupation of a micelle, so oligomerization is likely due to protein/protein interaction
rather than crowding. The protein and detergent form complex and they migrate together on
column. The MW determined in SEC is for the complexes of protein + detergent (MWProt+Det).
Table 3.2 shows the MW of SEC peaks and their peak assignment. SEC in 0.2% SDS at
pH 7.4 (Figure 3.8 A) mainly has two oligomeric species with MWProt+Det ≈ 80 – 115 and 190
kDa. A dominant peak is observed for FP_HM at 82 kDa and for FP_HM_TM at 115 kDa. The 78
kDa peak is also observed for HM and the 115 kDa peak for HM_TM, and are accompanied by a
peak of approximately equal intensity at 190 kDa for both constructs. Assignments include: 80 and
115 kDa peaks – protein trimers, and 190 kDa peak – protein hexamers. These peak assignments
are made based on the known trimer hairpin structure of the SE domain at mM protein
concentration, and also considered earlier SEC that supported predominant hexamer species in 6M
GuHCl without detergent.23

88

Figure 3.8 SEC of gp41 constructs under the following conditions: (A) 10 mM Tris at pH 7.4, 150
mM NaCl, 0.2% SDS, and ambient temperature; (B) 20 mM phosphate at pH 7.4, 150 mM NaCl,
and 0.25% DPC at 4 °C; and (C) 20 mM acetate at pH 4.0, 150 mM NaCl, and 0.25% DPC at 4
°C. SEC was obtained with a Superdex 200-increase column, 1 mg/mL protein loading with ~10fold dilution in the column, and A280 detection. The arrows in the plots are at the elution volumes
of the MW standards, and some of the peaks are identified with dashed lines and with MW’s
calculated from interpolation between MW standards.

89

Table 3.2 Analysis of SEC traces a
Condition

Construct

MW (kDa)

Oligomer

Protein (kDa) Detergent (kDa)

HM

78

41

37

FP_HM b

82

49

33

trimer

SDS at pH 7.4

HM_TM

115

50

65

FP_HM_TM b

115

57

58

HM

186

82

104

100

97

164

136

hexamer
HM_TM

197

HM

300

HM

36

14

22

FP_HM

31

17

14

dodecamer

monomer

DPC at pH 7.4

DPC at pH 4.0

HM_TM

27

17

10

FP_HM_TM b

21

19

3

HM b

85

41

44

FP_HM

100

49

51

HM_TM

125

50

75

HM

36

14

22

trimer

monomer

90

Table 3.2 (cont’d)

a

FP_HM

33

17

16

HM_TM

34

17

17

FP_HM_TM

20

19

1

FP_HM_TM

36

19

17

FP_HM b

88

49

39

HM_TM

90

50

40

FP_HM_TM

90

57

33

HM_TM

180

100

80

trimer

hexamer

The MW of HM, FP_HM, HM_TM and FP_HM_TM are 13.7, 16.5, 16.7 and 18.9 kDa,

respectively
b

Dominant peak

The assignments are also based on reasonable calculated protein and SDS contributions to
the SEC peak masses. The 78 kDa peak is proposed to include HM trimer (41 kDa) and SDS (37
kDa), and the 82 kDa peak includes FP_HM trimer (49 kDa) and SDS (33 kDa). A SDS micelle
without protein has a MW of 23 kDa, and additional detergent mass is reasonably needed to solvate
protein.40 The 115 kDa peak includes either HM_TM (50 kDa) or FP_HM_TM (57 kDa) trimer,
and SDS (≈ 60 kDa). Relative to the 80 kDa peaks, the additional ~25 kDa SDS mass in the 115

91

kDa peaks is consistent with solvation of the TM segment by SDS. There is also correlation with
the SDS-PAGE in Figure 3.4. The gel shows ~5 kDa increase in monomer MW with inclusion of
TM vs. much smaller MW change for inclusion of FP. Hexamer assignment of the 190 kDa peak
correlates with HM (82 kDa) or HM_TM (100 kDa), and SDS (≈ 100 kDa), and dodecamer
assignment of the 300 kDa peak correlates with HM (164 kDa) and SDS (136 kDa).
SEC traces in DPC (Figure 3.8 B and C) exhibit some peaks with MW’s comparable to
those in SDS, and assignment is done by similar reasoning. For SEC in 0.25% DPC at pH 7.4,
HM, FP_HM and HM_TM peaks at 85 ~ 125 kDa assigned to trimers: HM (41 kDa) + DPC (44
kDa); FP_HM (49 kDa) + DPC (51 kDa); and HM_TM (50 kDa) + DPC (75 kDa). The DPC
micelle without protein is 25 kDa.41 The trimer assignment is supported by previous sedimentation
velocity analysis of the related gp41512-711 construct in DPC at pH 7.0. This construct includes the
full ectodomain and TM. The major species was a trimer (65 kDa) + DPC (62 kDa).31 For SEC in
0.25% DPC at pH 4.0, the major peaks of FP_HM, HM_TM, and FP_HM_TM are at 90 kDa and
assigned as trimer, with ≈ 50 kDa protein and ≈ 40 kDa DPC.
All DPC traces have peaks at lower MW’s that are assigned to monomer species. There is
often a defined peak at 35 kDa with calculated monomer protein (15 kDa) + DPC (20 kDa)
contributions, and sometimes a broader peak at lower MW’s. The monomer peaks are dominant
for FP_HM and FP_HM_TM at pH 7.4. The SEC monomer assignment is supported by earlier
sedimentation velocity data in DPC at pH 7 for several constructs like gp41529-656 and gp41546-683
that lack the FP and TM domains. Sedimentation coefficients evidenced monomer protein (10
kDa) + DPC (15 kDa).30
For DPC, there are high-MW peaks at pH 7.4 that are weak and broad, while there are
defined peaks at pH 4.0 including 55 kDa for HM, 130 kDa for HM and HM_TM, and 180 kDa

92

for HM_TM. The latter peak is assigned to HM_TM hexamers (100 kDa) and DPC (80 kDa), like
the corresponding 190 kDa peak of HM_TM in SDS at pH 7.4. I don’t have assignments for the
55 kDa and 130 kDa peaks.
Additional SEC traces are provided in Figure 3.9. Panel A displays traces that support
reproducibility for replicate samples in 0.2% SDS at pH 7.4. Panel B displays FP_HM and
HM_TM traces reproducibility in 0.25% DPC at pH 7.4. The dash lines in these two panels show
the alignment of the peaks and support high reproducibility of SEC experiments. Panel C is SEC
traces of HM in 0.25% DPC at pH 7.4 with the presents and absent of DTT reducing agent in the
injection sample and running buffer. These traces support the lack of cysteine cross-linking in
these proteins and little dependence on reducing agent.

93

A
0.2% SDS
pH 7.4

B
0.25% DPC
pH 7.4

C
0.25% DPC
pH 7.4

HM

Figure 3.9 SEC traces of replicate samples in: (A) 10 mM Tris buffer at pH 7.4 with 0.2% SDS
and 150 mM NaCl; and (B, C) 20 mM phosphate buffer at pH 7.4 with 0.25% DPC and 150 mM
NaCl. The (C) replicates differ in the presence vs. absence of 2 mM DTT reducing agent.
Constructs and their replicate traces are labeled and dashed lines were used to show peak
alignment.

94

3.2.4 Comparison of gp41 oligomeric states in SEC and AUC
SEC peaks of the present study support the gp41 SE with either monomer, trimer, hexamer,
or larger oligomeric states. The current result is compared to SEC data from earlier studies as well
as equilibrium and sedimentation velocity analytical ultracentrifugation (AUC) data.23, 29, 31, 42 In
these studies, the constructs all have Tm > 100 °C, which matches the stability of full-length SE.
The typical [total protein] ≈ 5 mM for all these SEC and AUC studies, which makes it more
reasonable to compare oligomeric interpretations.
Current SEC data supports predominant trimer fraction for all four constructs in SDS at
neutral pH, with some hexamer fraction for HM and HM_TM, and no monomer fraction for any
construct. There is a significant monomer fraction for all constructs in DPC at both low and neutral
pH, as well as trimer and larger oligomer fractions. These data suggest that monomer charges are
stabilized by the zwitterionic charges of DPC but not by anionic SDS. The shorter construct HP
showed predominant monomer faction at low pH without detergent, whereas large aggregates are
predominant at neutral pH.23 These different oligomeric states correlate with different magnitudes
of calculated protein charge, ~ +10e at low pH vs. –1e at neutral pH, with corresponding significant
difference in inter-monomer electrostatic repulsion. We don’t observe greater monomer fraction
in DPC at low vs. neutral pH which supports the significance of electrostatic interactions between
the protein and the DPC headgroups with pH-independent zwitterionic charges.
An earlier study used SEC and AUC techniques to study the gp41 512-705 construct in
DPC at low pH.29 The SEC exhibited monomer:trimer ratio ≈ 3:1, and the sedimentation
equilibrium AUC data were interpreted to support monomer ↔ trimer Ka ≈ 1012 M–2 which would
correspond to a monomer:trimer ratio ≈ 1:4. This contradictory was not explained in this literation.
Our SEC of the closely-related FP_HM_TM in DPC at low pH exhibits monomer:trimer ratio ≈

95

1:1 which is intermediate between the two previous results. For the 538-705 SE+MPER+TM
construct, and the 538-665 SE-only construct, sedimentation equilibrium data in DPC at low pH
are interpreted to support Ka ≈ 1010 M–2.33 Combined consideration of all Ka’s suggests that FP
and TM segments stabilize the trimer vs. monomer, even though they don’t directly interact. There
is some similarity with our observation that FP_HM_TM is the most thermostable of the four fullytrimeric constructs in SDS at neutral pH. For 538-665 at low pH without detergent, sedimentation
equilibrium Ka’s of 5×1011 M–2 and 7×1012 M–2 have been independently reported, and likely
reflect the magnitude of uncertainty for this technique.33,

43

The general consensus from the

literature is that monomer is stabilized by low pH and also stabilized by DPC detergent. Monomer
units may also be stabilized by membrane containing phosphatidylcholine lipids.
Sedimentation velocity AUC has been applied to a construct very similar to FP_HM_TM
in DPC.31 Although Ka’s are not derived from sedimentation velocity, the data exhibited a
sedimentation coefficient that increased abruptly at pH 5, and was interpreted to support monomer
below pH 5 and trimer above pH 5. We did not observe this behavior in SEC and do not understand
the discrepancy between methods. The AUC result correlates with the reduction in calculated
protein charge from +9e to +3e between pH 4 and 5.
The constructs of the present study were designed based on SE structures to have N- and
C-helix segment lengths that yielded Tm > 100 °C, like the full-length protein. There are earlier
studies on constructs with similar or shorter helix lengths, with some surprising and sometimes
conflicting results about monomer vs. trimer fractions in DPC. For example, the longer 528579/SGGRGG/628-683 construct at 400 M concentration and low pH exhibits a tumbling time
derived from NMR that is consistent with predominant monomer.30 This corresponds to Ka ≤ 106
M–2 which is >100× smaller than the Ka determined by sedimentation equilibrium for the shorter

96

538-579/SGGRGG/628-665 construct.33, 43 At neutral pH, a dramatically larger Ka ≥ 4×1010 M–2
has been reported for the longer construct based on sedimentation velocity data, whereas much
smaller Ka ≤ 2×108 M–2 were reported for either 546-579/SGGRGG/628-683 or 528579/SGGRGG/628-655, which are constructs with respectively a shorter N- or C-helix.31, 42

3.2.5 Hairpin protein-induced vesicle fusion at both physiologic and low pH
Figure 3.10 shows the time-courses of vesicle fusion induced by the four different gp41
proteins for vesicles composed of either POPC:POPG:Chol (8:2:5) or POPC:Chol (2:1), at either
pH 3.2 or pH 7.4. The data were all acquired on the same day with the same protein stocks. There
is Âą2% typical variation in long-time fusion extent for replicate assays. POPC and Chol are
included to represent some of the physicochemical characteristics of the host cell membrane. For
this membrane, PC is a common lipid headgroup, and Chol is ~0.3 mole fraction of total lipid.
POPG has a calculated charge of –1 at both pH’s, and is included to represent the ~0.15 mole
fraction of anionic lipids that are primarily located in the inner leaflet of the host membrane.44, 45
Neutral POPC:Chol (2:1) may be more representative of the outer leaflet initially encountered by
gp41 during HIV/cell fusion.
Vesicle fusion was probed by inter-vesicle lipid mixing detected after addition of 40 M
protein solubilized in 10 mM Tris buffer at pH 7.4, with 0.2% SDS and 150 mM NaCl. This buffer
was also used for SEC in Figure 3.8 A, and these SEC data support initial protein trimers for all
constructs, and additional hexamer populations for HM and HM_TM. No fusion is observed for
buffer without protein, so fusion is due to protein/vesicle rather than SDS/vesicle interaction.

97

A
40
30

% Fusion

20

HM

15

HM_TM

10

35
30

FP_HM

25

PC + PG + Chol pH 3.2

40

FP_HM_TM

35

% Fusion

B

PC + Chol pH 3.2

5

25

FP_HM_TM

20
15

FP_HM
HM
HM_TM

10
5

0

0
0

100

200

300

400

500

600

0

100

Time (s)

300

400

500

600

Time (s)

C

D

40

40

PC + Chol pH 7.4

35
30

30

25

25

20
15

FP_HM_TM

10

FP_HM
HM_TM
HM

5
0
0

100

200

300

400

500

PC + PG + Chol pH 7.4

35

% Fusion

% Fusion

200

20
15

FP_HM_TM
FP_HM
HM_TM

10
5
0

600

HM
0

100

200

Time

300

400

500

600

Time

Figure 3.10 Vesicle fusion assays of gp41 proteins. Fusion was initiated by addition of an aliquot
of protein stock solution at 0 s, and subsequent fusion was monitored by increased fluorescence
associated with inter-vesicle lipid mixing. The stock contained 40 M protein in buffer at pH 7.4
with 0.2% SDS, and the protein + vesicle mixture contained [protein] = 0.5 M, [POPC+POPG]
= 150 M, and vesicle molar compositions and pH’s: (A) POPC:Chol = 2:1 at pH 3.2; (B)
POPC:POPG:Chol = 8:2:5 at pH 3.2; (C) POPC:Chol = 2:1 at pH 7.4; and (D) POPC:POPG:Chol
= 8:2:5 at pH 7.4. Fusion wasn’t observed for any vesicle composition after addition of an aliquot
of buffer + 0.2% SDS without protein. The assay dead time is ~5 s.

98

Figure 3.10 panel C and D showed the gp41-induced vesicle fusion at pH 7.4, and the
fusion of neutral vesicles at neutral pH is the condition similar HIV/cell fusion. There was a study
from Chang’s group showing the gp41512-665, which is similar to FP-HM, had a fusion activity that
was comparable to our data with neutral vesicles and at neutral pH.46 However, there were two
unusual result in that work: (1) the fusion activity of gp41512-665 did not increase as the protein
concentration increased; (2) the solubility was very high as ~1 mM in water.47 From our groups
previous experiments, the fusion activity always increases as the concentration of protein
increases, and the solubility of gp41 constructs in water is ˂ 1M. For most of the previous studies,
fusion required negatively-charged vesicles.21, 22 These previous results were interpreted to support
attractive electrostatics as a requirement for protein/vesicle binding. For the present study, HM
didn’t induce fusion at neutral pH, which is consistent with these previous results, but larger
constructs induced appreciable fusion.23 FP_HM_TM exhibited the greatest fusion extent, while
FP_HM and HM_TM had smaller extents that were similar to one another. These data support a
positive correlation between protein hydrophobicity and vesicle fusion. Fusion was moderatelyhigher for anionic vs. neutral vesicles, which evidences that electrostatic repulsion between anionic
vesicles does not affect fusion, and is also consistent with little bulk electrostatic interaction
between a vesicle and the nearly neutral protein, whose calculated charge is ~ –1 at pH 7.4.
The pH 3.2 data provides a comparison with earlier studies on shorter gp41 constructs at
low pH. One common feature of all four conditions of Figure 3.10 is highest fusion extent for
FP_HM_TM, which supports the hydrophobicity/fusion correlation. Fusion extent at low pH was
generally comparable or greater than at neutral pH. This correlates with different magnitudes of
protein charge at low vs. neutral pH, +12 vs. –1. Interestingly, the greater extent at low pH is likely

99

not due to direct protein/vesicle attraction, because extent was also greater for neutral vs. anionic
vesicles. This result contrasts with the reverse trend at neutral pH.

3.2.6 FP_HM_TM structural model
Figure 3.11 displays a medium-resolution structural model for FP_HM_TM which
incorporates the Table 3.1 analysis of the circular dichroism spectra as well as previous results.6,
13, 48

This model likely reflects the final gp41 state during fusion, based on Tm > 90 °C (Figure 3.6

B). A monomer is displayed for clarity but the model should also be valid for a trimer bundle or a
hexamer (dimer of trimer bundles). The HM region is primarily hairpin structure that contains
helices over residues 536-581 and 628-675. This structure is supported by very high helicity of
HM and HM_TM, and by a previous crystal structure.6, 13 The C-terminal MPER and TM regions
are also highly helical, based on the HM_TM CD spectrum, and on structures in detergent of
peptides corresponding to MPER and/or TM sequences.48 These structures are typically
continuous helices, but there are likely breaks in helical structure for the larger protein because of
the topological constraints of membrane interface and traversal locations for the MPER and TM
domains, respectively.
The FP region is represented as extended and  strand structure, based both on reduced
average helicity for constructs that include the FP, and on earlier NMR and infrared data that
evidence FP antiparallel  sheet structure in membrane.14, 49-51 Such structure is reasonable for a
hexamer for which the strands from the two trimers are interleaved. NMR data also support a
distribution of N-terminal antiparallel  sheet registries, with significant populations of registries
like 512→527/527→512 for adjacent strands.14 FP insertion in a single leaflet is evidenced by

100

NMR contacts between multiple FP residues and lipid tails.49 There isn’t close contact between
the FP and the MPER or between the FP and the TM, as evidenced by previous NMR studies.30, 32

683

MPER

N-helix

581

C-helix

628

536

675

512

TM

Membrane

FP

705

Figure 3.11 Structural model of FP_HM_TM based on circular dichroism spectra of the four
constructs as well as other data.6, 14, 30, 48 A monomer is shown for clarity but the model should be
valid for trimers and hexamers. Approximate residue numbers are displayed.

101

3.2.7 Relationship between vesicle fusion and HIV/cell fusion
The vesicle fusion data of the present study provides information about the relative
membrane perturbations by different constructs in defined oligomeric states. This may describe
potential contributions of these states to viral gp41-induced fusion. In the pre-fusion complex of
gp41 with gp120, electron densities are interpreted to support an interior bundle of trimeric Nhelices, and C-helices separated from the bundle.9 The trimer of hairpins structure has been
observed for gp41 without gp120, and large ectodomain constructs exhibit a hyperthermostable
monomer at ~5 M concentration with helices indistinguishable from the monomer units of the
trimer. This underlies the hairpin monomer and trimer structures in Figure 3.11 and 3.12.
For the present study, stock protein was principally hyperthermostable trimers and the
protein:(PC+PG+Chol) ratio was ~1:450. Each vesicle has a diameter d of 100 nm, and the
membrane thickness b is ~ 5 nm. Thus, the surface area including the vesicle outer and inner leaflet
can be calculated as:
𝑑 2

2

𝑑

𝐴 = 4𝜋 [(2 ) + ( 2 − 𝑏) ] = 56834 𝑛𝑚2

3.1

The surface area of PC or PG head group is ~ 0.7 nm2 and Chol is ~ 0.4 nm2.52 The PC+PG:Chol
ratio is 2:1. The average surface area a of each lipid molecule is:
2

1

𝑎 = 3 ×0.7 𝑛𝑚2 + 3 ×0.4 𝑛𝑚2 = 0.6 𝑛𝑚2

3.2

The number of lipid molecules in each vesicle is:
𝐴

𝑁𝑡𝑜𝑡 = 𝑎 ≈ 95,000

3.3

This implies ~210 protein molecules or ~70 protein trimers per vesicle when there is quantitative
protein binding, and smaller copy number with reduced binding. This is comparable to the ~15

102

trimers/virion, with significant microscopy and functional evidence that trimers are spatially
clustered during fusion.8, 53
Earlier studies of vesicle fusion induced by constructs like HM, and FP_HM often showed
strong dependences on pH and membrane charge that were interpreted as supporting a large
contribution to fusion from protein/vesicle electrostatic attraction. There was similarly much
greater leakage of anionic vesicles at low vs. neutral pH.21-23 For the present study, electrostatic
effects were much less pronounced, as evidenced by much smaller dependence of fusion extents
on low vs. neutral pH and anionic vs. neutral vesicles (Figure 3.10). Larger constructs induced
significant fusion at neutral pH of both neutral and anionic vesicles, which reflects expected
physiologic conditions of HIV/host cell fusion. Fusion under physiologic conditions for the present
but not earlier studies may be due to inclusion of more hydrophobic segments in the present study,
and also stock solutions with predominant trimer in the present study vs. monomer in previous
studies.54 Hydrophobic perturbation of the membrane is increased by both effects, and is magnified
in the fusion rate via the Arrhenius Law, assuming that perturbations reduce activation energy by
making membranes more like the fusion transition state. The contribution of the hydrophobic
effect is evidenced by highest fusion for FP_HM_TM. In addition, the HIV TM sequence is fairly
conserved,55 including the central R696 snorkeling towards membrane surface, which likely
contributes membrane perturbation.56
Apposition of HIV and host cell membranes is likely aided by conversion from orthogonal
N-helix/C-helix geometry in the pre-fusion state to antiparallel geometry in the hairpin state. It is
also known that peptides corresponding to N- or C-helix regions inhibit fusion up to the final pore
expansion step, which is after inter-membrane lipid mixing and pore formation (Figure 3.12 B).13,
53, 57

The present study shows that the trimer of hairpins is effective at inducing vesicle fusion

103

under physiologic conditions, particularly when both the TM and FP segments are included in the
construct. The fusion efficiency of the trimer is based on increased local concentration of TM and
FP, and correlates with much higher vesicle fusion induced by FP’s that are cross-linked at their
C-termini, with topology similar to that in the hairpin trimer.20 The inhibitory peptides would likely
not bind to the trimer of hairpins and we propose that they instead bind to monomer hairpins
formed after dissociation (Figure 3.12). Such dissociation has been known for twenty years, with
additional recent data from our and other groups, and could plausibly occur during the ~1-3 minute
HIV/cell fusion time.8, 23, 30, 32 The bound peptides prevent re-association of the fusion-efficient
trimer. For constructs like FP_HM_TM in DPC detergent at low pH, the 3 monomer ↔ trimer Ka
≈ 1012 M–2, so that massmonomer ≈ masstrimer when [total protein] ≈ 1 M. The membrane Ka hasn’t
been measured, but there are comparable statistical-average inter-protein molecular separations of
~100 nm for [bulk protein] ≈ 1 M, and ~50 nm for ~15 protein trimers in a 100 nm diameter
virion. Previous work from our and other groups supports the monomer as the hyperthermostable
helical folding unit and it is therefore reasonable that the hairpin is the lowest-free-energy
structure.23, 30 Asynchronous folding of monomer protein into hairpin brings the two membranes
into apposition and is likely topologically easier than concerted folding of a trimer into a six-helix
bundle. This may therefore be evolutionary advantage of retaining some hairpin monomer stability
relative to trimer.

104

Figure 3.12 Schematic illustrating (A) trimer and (B) monomer respectively favored in the
absence and presence of peptide inhibitor. Panel B displays “C34” inhibitor which contains Chelix residues 628-661. The sequence color coding matches Figure 3.1 and 3.11, and loops
between structured regions are not displayed for clarity. The FP’s from different trimers or
monomers adopt antiparallel  sheet structure. Fusion is enhanced in panel A vs. B because of
greater clustering of membrane-perturbing protein regions in the trimer vs. monomer. This
enhancement exists for the displayed hemifusion state as well as membrane states that precede
hemifusion.

105

3.3 Conclusion
The construct of HIV gp41 membrane fusion protein including the whole ectodomain and
transmembrane domain, and shorter constructs have been expressed, purified and stabilized in
physiology buffers. The constructs adopt  helical SE and TM, and non-helical FP in SDS and
DPC detergents, and they are all hyperthermostable with Tm > 90 °C. The oligomeric states of
these proteins vary in different detergent buffer: predominant trimer for all constructs and some
hexamer fraction for HM and HM_TM protein in SDS at pH 7.4; and mixtures of monomer, trimer,
and higher-order oligomer protein in DPC at pH 4.0 and 7.4. Substantial protein-induced vesicle
fusion was observed, including fusion of neutral vesicles at neutral pH, which are the conditions
similar HIV/cell fusion. Vesicle fusion by a gp41 ectodomain construct has rarely been observed
under these conditions, and is aided by inclusion of both the FP and TM, and by protein which is
predominantly trimer rather than monomer. Current data was integrated with existing data, and a
structural model and some new interpretations of these data were proposed: (1) gp41 has a
monomer ↔ trimer thermodynamic equilibrium; (2) monomer hairpins are formed from trimer
dissociation. Asynchronous folding of monomer gp41 into hairpin is likely topologically easier
than concerted folding of a trimer into a six-helix bundle; (3) trimer has higher fusion efficiency
since higher local concentration of TM and FP and greater membrane perturbation.

106

REFERENCES

107

REFERENCES

[1] Weiss, R. A. (1993) HOW DOES HIV CAUSE AIDS, Science 260, 1273-1279.
[2] Levy, J. A. (1993) PATHOGENESIS OF HUMAN-IMMUNODEFICIENCY-VIRUS
INFECTION, Microbiol. Rev. 57, 183-289.
[3] Dalgleish, A. G., Beverley, P. C. L., Clapham, P. R., Crawford, D. H., Greaves, M. F., and
Weiss, R. A. (1984) THE CD4 (T4) ANTIGEN IS AN ESSENTIAL COMPONENT OF
THE RECEPTOR FOR THE AIDS RETROVIRUS, Nature 312, 763-767.
[4] Kwong, P. D., Wyatt, R., Robinson, J., Sweet, R. W., Sodroski, J., and Hendrickson, W. A.
(1998) Structure of an HIV gp120 envelope glycoprotein in complex with the CD4 receptor
and a neutralizing human antibody, Nature 393, 648-659.
[5] Broder, C. C., and Dimitrov, D. S. (1996) HIV and the 7-transmembrane domain receptors,
Pathobiology 64, 171-179.
[6] Chan, D. C., Fass, D., Berger, J. M., and Kim, P. S. (1997) Core structure of gp41 from the
HIV envelope glycoprotein, Cell 89, 263-273.
[7] Frankel, A. D., and Young, J. A. T. (1998) HIV-1: Fifteen proteins and an RNA, Annu. Rev.
Biochem. 67, 1-25.
[8] Grewe, C., Beck, A., and Gelderblom, H. R. (1990) HIV - EARLY VIRUS-CELL
INTERACTIONS, J. Acquir. Immune Defic. Syndr. Hum. Retrovirol. 3, 965-974.
[9] Julien, J. P., Cupo, A., Sok, D., Stanfield, R. L., Lyumkis, D., Deller, M. C., Klasse, P. J.,
Burton, D. R., Sanders, R. W., Moore, J. P., Ward, A. B., and Wilson, I. A. (2013) Crystal
Structure of a Soluble Cleaved HIV-1 Envelope Trimer, Science 342, 1477-1483.
[10] Weissenhorn, W., Dessen, A., Harrison, S. C., Skehel, J. J., and Wiley, D. C. (1997) Atomic
structure of the ectodomain from HIV-1 gp41, Nature 387, 426-430.
[11] Lyumkis, D., Julien, J. P., de Val, N., Cupo, A., Potter, C. S., Klasse, P. J., Burton, D. R.,
Sanders, R. W., Moore, J. P., Carragher, B., Wilson, I. A., and Ward, A. B. (2013) CryoEM Structure of a Fully Glycosylated Soluble Cleaved HIV-1 Envelope Trimer, Science
342, 1484-1490.
[12] Tran, E. E. H., Borgnia, M. J., Kuybeda, O., Schauder, D. M., Bartesaghi, A., Frank, G. A.,
Sapiro, G., Milne, J. L. S., and Subramaniam, S. (2012) Structural Mechanism of Trimeric
HIV-1 Envelope Glycoprotein Activation, PLoS Pathog. 8, 18.
[13] Lu, M., Ji, H., and Shen, S. (1999) Subdomain folding and biological activity of the core
structure from human immunodeficiency virus type 1 gp41: Implications for viral
membrane fusion, J. Virol. 73, 4433-4438.

108

[14] Schmick, S. D., and Weliky, D. P. (2010) Major Antiparallel and Minor Parallel beta Sheet
Populations Detected in the Membrane-Associated Human Immunodeficiency Virus
Fusion Peptide, Biochemistry 49, 10623-10635.
[15] Qiang, W., Sun, Y., and Weliky, D. P. (2009) A strong correlation between fusogenicity and
membrane insertion depth of the HIV fusion peptide, Proc. Natl. Acad. Sci. U. S. A. 106,
15314-15319.
[16] Shi, W. X., Bohon, J., Han, D. P., Habte, H., Qin, Y. L., Cho, M. W., and Chance, M. R.
(2010) Structural Characterization of HIV gp41 with the Membrane-proximal External
Region, J. Biol. Chem. 285, 24290-24298.
[17] Sun, Z. Y. J., Oh, K. J., Kim, M. Y., Yu, J., Brusic, V., Song, L. K., Qiao, Z. S., Wang, J. H.,
Wagner, G., and Reinherz, E. L. (2008) HIV-1 broadly neutralizing antibody extracts its
epitope from a kinked gp41 ectodomain region on the viral membrane, Immunity 28, 5263.
[18] Gabrys, C. M., Qiang, W., Sun, Y., Xie, L., Schmick, S. D., and Weliky, D. P. (2013) SolidState Nuclear Magnetic Resonance Measurements of HIV Fusion Peptide (CO)-C-13 to
Lipid P-31 Proximities Support Similar Partially Inserted Membrane Locations of the alpha
Helical and beta Sheet Peptide Structures, J. Phys. Chem. A 117, 9848-9859.
[19] Freed, E. O., Delwart, E. L., Buchschacher, G. L., and Panganiban, A. T. (1992) 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. 89, 70-74.
[20] Qiang, W., and Weliky, D. P. (2009) HIV Fusion Peptide and Its Cross-Linked Oligomers:
Efficient Syntheses, Significance of the Trimer in Fusion Activity, Correlation of beta
Strand Conformation with Membrane Cholesterol, and Proximity to Lipid Headgroups,
Biochemistry 48, 289-301.
[21] Sackett, K., TerBush, A., and Weliky, D. P. (2011) HIV gp41 six-helix bundle constructs
induce rapid vesicle fusion at pH 3.5 and little fusion at pH 7.0: understanding pH
dependence of protein aggregation, membrane binding, and electrostatics, and implications
for HIV-host cell fusion, Eur. Biophys. J. Biophys. Lett. 40, 489-502.
[22] Ratnayake, P. U., Sackett, K., Nethercott, M. J., and Weliky, D. P. (2015) pH-dependent
vesicle fusion induced by the ectodomain of the human immunodeficiency virus membrane
fusion protein gp41: Two kinetically distinct processes and fully-membrane-associated
gp41 with predominant beta sheet fusion peptide conformation, Biochim. Biophys. ActaBiomembr. 1848, 289-298.
[23] Banerjee, K., and Weliky, D. P. (2014) Folded Monomers and Hexamers of the Ectodomain
of the HIV gp41 Membrane Fusion Protein: Potential Roles in Fusion and Synergy
Between the Fusion Peptide, Hairpin, and Membrane-Proximal External Region,
Biochemistry 53, 7184-7198.

109

[24] Vogel, E. P., and Weliky, D. P. (2013) Quantitation of Recombinant Protein in Whole Cells
and Cell Extracts via Solid-State NMR Spectroscopy, Biochemistry 52, 4285-4287.
[25] Zhang, H. Y., Neal, S., and Wishart, D. S. (2003) RefDB: A database of uniformly referenced
protein chemical shifts, J. Biomol. NMR 25, 173-195.
[26] Naito, A., Nagao, T., Norisada, K., Mizuno, T., Tuzi, S., and Saito, H. (2000) Conformation
and dynamics of melittin bound to magnetically oriented lipid bilayers by solid-state P-31
and C-13 NMR spectroscopy, Biophys. J. 78, 2405-2417.
[27] Costa, P. R., Kocisko, D. A., Sun, B. Q., Lansbury, P. T., and Griffin, R. G. (1997)
Determination of peptide amide configuration in a model amyloid fibril by solid-state
NMR, J. Am. Chem. Soc. 119, 10487-10493.
[28] Sackett, K., Nethercott, M. J., Shai, Y., and Weliky, D. P. (2009) Hairpin Folding of HIV
gp41 Abrogates Lipid Mixing Function at Physiologic pH and Inhibits Lipid Mixing by
Exposed gp41 Constructs, Biochemistry 48, 2714-2722.
[29] Lakomek, N. A., Kaufman, J. D., Stahl, S. J., Louis, J. M., Grishaev, A., Wingfield, P. T., and
Bax, A. (2013) Internal Dynamics of the Homotrimeric HIV-1 Viral Coat Protein gp41 on
Multiple Time Scales, Angew. Chem.-Int. Edit. 52, 3911-3915.
[30] Roche, J., Louis, J. M., Aniana, A., Ghirlando, R., and Bax, A. (2015) Complete dissociation
of the HIV-1 gp41 ectodomain and membrane proximal regions upon phospholipid binding,
J. Biomol. NMR 61, 235-248.
[31] Dai, Z., Tao, Y. S., Liu, N. N., Brenowitz, M. D., Girvin, M. E., and Lai, J. R. (2015)
Conditional Trimerization and Lytic Activity of HIV-1 gp41 Variants Containing the
Membrane-Associated Segments, Biochemistry 54, 1589-1599.
[32] Roche, J., Louis, J. M., Grishaev, A., Ying, J. F., and Bax, A. (2014) Dissociation of the
trimeric gp41 ectodomain at the lipid-water interface suggests an active role in HIV-1 Envmediated membrane fusion, Proc. Natl. Acad. Sci. U. S. A. 111, 3425-3430.
[33] Lakomek, N. A., Kaufman, J. D., Stah, S. J., and Wingfield, P. T. (2014) HIV-1 Envelope
Protein gp41: An NMR Study of Dodecyl Phosphocholine Embedded gp41 Reveals a
Dynamic Prefusion Intermediate Conformation, Structure 22, 1311-1321.
[34] Weissenhorn, W., Wharton, S. A., Calder, L. J., Earl, P. L., Moss, B., Aliprandis, E., Skehel,
J. J., and Wiley, D. C. (1996) The ectodomain of HIV-1 env subunit gp41 forms a soluble,
alpha-helical, rod-like oligomer in the absence of gp120 and the N-terminal fusion peptide,
Embo J. 15, 1507-1514.
[35] Buzon, V., Natrajan, G., Schibli, D., Campelo, F., Kozlov, M. M., and Weissenhorn, W. (2010)
Crystal Structure of HIV-1 gp41 Including Both Fusion Peptide and Membrane Proximal
External Regions, PLoS Pathog. 6, 7.

110

[36] Ratnayake, P. U., Ekanayaka, E. A. P., Komanduru, S. S., and Weliky, D. P. (2016) Fulllength trimeric influenza virus hemagglutinin II membrane fusion protein and shorter
constructs lacking the fusion peptide or transmembrane domain: Hyperthermostability of
the full-length protein and the soluble ectodomain and fusion peptide make significant
contributions to fusion of membrane vesicles, Protein Expr. Purif. 117, 6-16.
[37] Corrin, M. L., and Harkins, W. D. (1947) THE EFFECT OF SALTS ON THE CRITICAL
CONCENTRATION FOR THE FORMATION OF MICELLES IN COLLOIDAL
ELECTROLYTES, J. Am. Chem. Soc. 69, 683-688.
[38] Thongngam, M., and McClements, D. J. (2005) Influence of pH, ionic strength, and
temperature on self-association and interactions of sodium dodecyl sulfate in the absence
and presence of chitosan, Langmuir 21, 79-86.
[39] Palladino, P., Rossi, F., and Ragone, R. (2010) Effective Critical Micellar Concentration of a
Zwitterionic Detergent: A Fluorimetric Study on n-Dodecyl Phosphocholine, J. Fluoresc.
20, 191-196.
[40] Hayashi, S., and Ikeda, S. (1980) MICELLE SIZE AND SHAPE OF SODIUM DODECYLSULFATE IN CONCENTRATED NACL SOLUTIONS, J. Phys. Chem. 84, 744-751.
[41] le Maire, M., Champeil, P., and Moller, J. V. (2000) Interaction of membrane proteins and
lipids with solubilizing detergents, Biochim. Biophys. Acta-Biomembr. 1508, 86-111.
[42] Louis, J. M., Baber, J. L., Ghirlando, R., Aniana, A., Bax, A., and Roche, J. (2016) Insights
into the Conformation of the Membrane Proximal Regions Critical to the Trimerization of
the HIV-1 gp41 Ectodomain Bound to Dodecyl Phosphocholine Micelles, PLoS One 11,
21.
[43] Wingfield, P. T., Stahl, S. J., Kaufman, J., Zlotnick, A., Hyde, C. C., Gronenborn, A. M., and
Clore, G. M. (1997) The extracellular domain of immunodeficiency virus gp41 protein:
Expression in Escherichia coli, purification, and crystallization, Protein Sci. 6, 1653-1660.
[44] Brugger, B., Glass, B., Haberkant, P., Leibrecht, I., Wieland, F. T., and Krausslich, H. G.
(2006) The HIV lipidome: A raft with an unusual composition, Proc. Natl. Acad. Sci. U. S.
A. 103, 2641-2646.
[45] Lorizate, M., Sachsenheimer, T., Glass, B., Habermann, A., Gerl, M. J., Krausslich, H. G.,
and Brugger, B. (2013) Comparative lipidomics analysis of HIV-1 particles and their
producer cell membrane in different cell lines, Cell Microbiol. 15, 292-304.
[46] Cheng, S. F., Chien, M. P., Lin, C. H., Chang, C. C., Lin, C. H., Liu, Y. T., and Chang, D. K.
(2010) The fusion peptide domain is the primary membrane-inserted region and enhances
membrane interaction of the ectodomain of HIV-1 gp41, Mol. Membr. Biol. 27, 31-44.
[47] Lin, C. H., Chang, C. C., Cheng, S. F., and Chang, D. K. (2008) The application of
perfluorooctanoate to investigate trimerization of the human immunodeficiency virus-1
gp41 ectodomain by electrophoresis, Electrophoresis 29, 3175-3182.

111

[48] Apellaniz, B., Rujas, E., Serrano, S., Morante, K., Tsumoto, K., Caaveiro, J. M. M., Jimenez,
M. A., and Nieva, J. L. (2015) The Atomic Structure of the HIV-1 gp41 Transmembrane
Domain and Its Connection to the Immunogenic Membrane-proximal External Region, J.
Biol. Chem. 290, 12999-13015.
[49] Jia, L. H., Liang, S., Sackett, K., Xie, L., Ghosh, U., and Weliky, D. P. (2015) REDOR solidstate NMR as a probe of the membrane locations of membrane-associated peptides and
proteins, J. Magn. Reson. 253, 154-165.
[50] Sackett, K., Nethercott, M. J., Zheng, Z. X., and Weliky, D. P. (2014) Solid-State NMR
Spectroscopy of the HIV gp41 Membrane Fusion Protein Supports Intermolecular
Antiparallel 13 Sheet Fusion Peptide Structure in the Final Six-Helix Bundle State, J. Mol.
Biol. 426, 1077-1094.
[51] Martin, I., Schaal, H., Scheid, A., and Ruysschaert, J. M. (1996) 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. 70, 298-304.
[52] Sardan, M., Kilinc, M., Genc, R., Tekinay, A. B., and Guler, M. O. (2013) Cell penetrating
peptide amphiphile integrated liposomal systems for enhanced delivery of anticancer drugs
to tumor cells, Faraday Discuss. 166, 269-283.
[53] Markosyan, R. M., Cohen, F. S., and Melikyan, G. B. (2003) HIV-1 envelope proteins
complete their folding into six-helix bundles immediately after fusion pore formation, Mol.
Biol. Cell 14, 926-938.
[54] Lev, N., Fridmann-Sirkis, Y., Blank, L., Bitler, A., Epand, R. F., Epand, R. M., and Shai, Y.
(2009) Conformational Stability and Membrane Interaction of the Full-Length Ectodomain
of HIV-1 gp41: Implication for Mode of Action, Biochemistry 48, 3166-3175.
[55] Ramakrishnan, R., Mehta, R., Sundaravaradan, V., Davis, T., and Ahmad, N. (2006)
Characterization of HIV-I envelope gp4I genetic diversity and functional domains
following perinatal transmission, Retrovirology 3, 18.
[56] Gangupomu, V. K., and Abrams, C. F. (2010) All-Atom Models of the Membrane-Spanning
Domain of HIV-1 gp41 from Metadynamics, Biophys. J. 99, 3438-3444.
[57] Wild, C. T., Shugars, D. C., Greenwell, T. K., McDanal, C. B., and Matthews, T. J. (1994)
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. 91, 9770-9774.

112

Chapter 4
IFP membrane location studied by 13C-2H REDOR NMR

113

4.1 Introduction
Influenza virus infects respiratory epithelial cells and causes contagious respiratory illness.
Influenza virus is an enveloped virus which means the virion is covered by lipid bilayer. HA is a
viral integral membrane protein and has two disulfide bounded subunits: HA1 and HA2.1-3
Influenza virus infection is a multi-step process and starts with HA1 binds to the sialic acid on host
cell membrane. Virion is then endocytosed. The low pH in the endosome triggers the
conformational change of HA and HA2 catalyzes the fusion between viral membrane and
endosomal membrane, allowing the viral genetic material to get into the host cell.4, 5 The ~20 Nterminus residues of HA2 is called fusion peptide (IFP) and it is highly conserved.6 Functional
studies of IFP shows that it plays a vital role in membrane fusion, even a single mutation can
eliminates its fusion activity.7-9 IFP is also believed to be the only segment in HA that binds to the
endosomal membrane.10 However, there is no crystal structure of IFP domain in HA, neither
evidence of its fusion mechanism. Polypeptide IFP has been extensively studied as fusion model,
and the previous research revealed the structure of IFP in detergent micelles and lipid vesicles.1116

Lipid membrane is more physiology relevant environment compared to detergent micelles, thus

this study is focusing on the IFP-membrane interaction. Ghosh from Weliky’s group found out
that IFP always adopts two different N-helix/turn/C-helix structures in membrane, denoted as
“semi-closed” and “closed” structures.15 The details have been discussed in chapter 1.
Several models have been proposed either by experiment results or molecular dynamics
(MD) simulations to settle the question of IFP membrane location. EPR experiments were done
by Tamm’s group. The result showed that H3_20 IFP is immerged in membrane as an inverted Vshaped, where the N- and C-terminus deeply inserted into hydrophobic core while the turn was
close to the aqueous surface. H3_20 IFP was inserted deeper at pH 5 compare to at pH 7.4 (Figure

114

4.1 A). The result also suggests that a greater perturbation of lipid bilayer at pH 5 and facilitate the
fusion.11 H1 and H3 membrane associated IFP structure was studied by solid-state NMR,
exhibiting amphipathic property: one side of IFP is hydrophobic and the other side is hydrophilic
(Figure 4.2 B). It was proposed that IFP hydrophobic sidechains inserted into lipid hydrophobic
core and the hydrophilic side facing aqueous layer.14, 15
Molecular dynamic simulations are used to understand the details in fusion for both peptide
and lipid membrane. In 2013, Kasson’s group performed MD simulations of IFP in lipid bilayers.
They observed a strong relationship between lipid tail protrusion and the ability of fusion activity,
and the tail protrusion is important to the transition state for fusion. It was also predicted the
kinked helix structure is more fusogenic than straight or hairpin-like structures (Figure 4.1 C).17
This is consistent with the solid-state NMR results from Weliky’s group, which show a higher
fusion activity at low pH when the semi-closed structure population is higher compared to closed
structure (Figure 4.2).15 People also suggested that the lipid head group intrusion contributes to the
perturbation of membrane and lipid mixing.18
Lazaridis’s gourp performed MD simulations of the “closed” H1_23 IFP in lipid membrane
and they found out that the IFP deeply inserted in one membrane leaflet (Figure 4.1 D). Even Nterminal helix is buried more deeply into the hydrophobic membrane interior than the C-terminal
helix, the N-terminus is solvent accessible. Several water molecules penetrate the membrane
interface and interact with it. The pH also influences the orientation of the IFP. On average, the
IFP plane is more tilted in the membrane at neutral pH than lower pH, because of the protonation
of acidic residues E11 and D19 at low pH allows the C-terminal helix to insert deeper, lowering
the rotation angle.19

115

Figure 4.1 (A) The 20 lowest energy conformers of the H3_20 IFP in lipid bilayers by the EPR
data at pH 5 (red) and 7.4 (yellow). A phospholipid is shown for reference and nitrogen, oxygen
and phosphorus atoms are colored in yellow, red and green, respectively. The white line represents
the level of the lipid phosphate groups. The hydrophobic hydrocarbon is in dark gray and interface
is in light gray.11 (B) Membrane locations of closed structure H1_23 IFP. Dashed lines are the
hydrocarbon core.15 (C) Lipid tail protrusion induced by IFP. IFP, lipid tails and phosphates are in
green, gray and orange, respectively. A thin gray plane shows the average phosphorus position in
the upper leaflet. One lipid is shown in sticks, with an acyl tail protruding into the polar layer.17
(D) IFP deeply inserted in one membrane leaflet at low pH. Hydrophobic residues are highlighted
in yellow and hydrophilic residues in red. Phosphorus and nitrogen atoms are in tan and blue,
respectively.19 (E) Membrane-spanning conformation model of IFP in membrane at low pH. IFP,
phosphate group, amine group and acyl chain are in purple, orange, blue and green, respectively.20
(F) IFP binds on membrane surface at neutral pH. N-terminal helix, C-terminal helix and
hydrophobic side chains are in red, blue and white, respectively.21

116

Victor et al. discovered a different model in their MD simulation of IFP in membrane.
These simulations revealed that the peptide became deeply inserted into the membrane and adopted
a perpendicular or tilted orientation relative to the membrane plane, extending from one leaflet to
the other and contacting the lipid headgroups on both sides of the membrane. This model is also
called membrane-spanning conformation, which had not been predicted by previous MD
simulation studies. The peptide insertion had a strong effect on the membrane, lowering the bilayer
thickness, disordering nearby lipids, and promoting lipid tail protrusion (Figure 4.1 E).20
Another model of membrane surface was proposed recently by Tajkhorshid’s group. The
MD simulation result showed that hydrophobic residues Leu-2, Phe-3, Ile-6, Phe-9, and Ile-10 in
N-terminal helix are buried in membrane hydrophobic core, while in C-terminal helix, amphipathic
side chain Trp-14 and Met-17 are buried, whereas Trp-21 is mostly located at the headgroup region.
N-terminal helix inserts much deeper into the membrane than C-terminal helix, suggesting that the
N-terminal helix was responsible for hydrophobic anchoring of the peptide into the membrane. Cterminal helix, in contrast, is found to establish major amphipathic interactions at the interfacial
region thereby contributing to binding strength of IFP (Figure 4.1 F).21
This chapter presents an investigation of the IFP membrane location studied by

13

C-2H

REDOR solid-state NMR. It is vital to detailed characterize the IFP topology in membrane since
IFP is the only segment in HA that interact with target membrane and it is high relevance to the
mechanism of viral entry into the host cell. The overall goal is to get a better understanding of how
IFP promotes membrane fusion and provide more information for understanding fusion
mechanism.
To measure the IFP membrane location, distance was measured between carbonyl carbons
in peptide backbone and lipid acyl chain using

13

117

C-2H REDOR. H3_20 IFP was synthesized

manually by Fmoc solid-phase peptide synthesis as described in chapter 2. The IFP constructs
were 13C site-specifically labeled at either Leu_2, Ala_7 or Gly_16 residues. The lipid membrane
used in this study had a composition of DPPC:DPPG 4:1 ratio as a mimic of host cell membrane.
The DPPC lipid was 2H labeled at different acyl chain positions and denoted as DPPC_D4,
DPPC_D8 and DPPC_D10 (structure in Figure 2.1). This labeling scheme was chosen based on
previous peptide membrane location studies.22

4.2 Result and discussion
4.2.1 13CO-2H REDOR spectra, buildups, and fittings
The H3_20 IFP amino acid sequence is GLFGAIAGFIENGWEGMIDGGGKKKKG
where the underlined part is the original sequence and the C-terminal tag is added intended to
increase the solubility of IFP. The labeled amino acids are shown in Figure 4.2 and the deuterium
atoms in lipid membrane are shown in Figure 2.1.

Figure 4.2 H2_20 IFP sequence, closed structure model and semi-closed model. Labeled amino
acids are in red in sequence and orange in the cartoon model.

118

Figure 4.3

13

CO–2H REDOR data of samples that contain either IFP_L2c in membrane with

DPPC:DPPG 4:1 ratio. Samples were prepared with DPPC_D4, DPPC_D8 and DPPC_D10, and
the corresponding data are displayed in purple, green and red, respectively. (A) S 0 (black) and S1
(colored) REDOR spectra for  = 40 ms. (B) and (C) Plots of (S/S0) vs  are displayed for the 
helical and  strand peaks. The solid lines are best-fits to A(1 – e-) and the fitting parameters are
in Table 4.1.

119

Figure 4.4

13

CO–2H REDOR data of samples that contain either IFP_A7c in membrane with

DPPC:DPPG 4:1 ratio. Samples were prepared with DPPC_D4, DPPC_D8 and DPPC_D10, and
the corresponding data are displayed in purple, green and red, respectively. (A) S 0 (black) and S1
(colored) REDOR spectra for  = 40 ms. (B) and (C) Plots of (S/S0) vs  are displayed for the 
helical and  strand peaks. The solid lines are best-fits to A(1 – e-) and the fitting parameters are
in Table 4.1.

120

Figure 4.5

13

CO–2H REDOR data of samples that contain either IFP_G16c in membrane with

DPPC:DPPG 4:1 ratio. Samples were prepared with DPPC_D4, DPPC_D8 and DPPC_D10, and
the corresponding data are displayed in purple, green and red, respectively. (A) S 0 (black) and S1
(colored) REDOR spectra for  = 40 ms. (B) Plot of (S/S0) vs  are displayed for the  helical
peak. The solid lines are best-fits to A(1 – e-) and the fitting parameters are in Table 4.1.

121

Table 4.1 Best-fit exponential buildup parameters for 13CO–2H REDOR of IFP in DPPC and
DPPG a b

a

Peptide

Peak

Membrane

A

(Hz)

r (Å)

L2c



DPPC_D4

0.36 (15)

34 (24)

4 (2)

L2c c



DPPC_D10

1.00 (15)

38 (10)

4.3 (5)

L2c



DPPC_D8

0.72 (10)

14 (3)

6.0 (5)

L2c c



DPPC_D10

0.55 (23)

27 (18)

5 (2)

A7c



DPPC_D8

0.34 (10)

21 (9)

5 (1)

A7c c



DPPC_D10

0.99 (10)

30 (5)

4.7 (3)

A7c



DPPC_D8

0.47 (6)

25 (5)

5.0 (4)

A7c



DPPC_D10

0.90 (19)

23 (7)

5.1 (7)

G16c



DPPC_D8

1.00 (27)

12 (4)

6.4 (9)

G16c c



DPPC_D10

1.00 (28)

22 (8)

5.2 (8)

Samples are prepared by aqueous binding method with 1 mol IFP, 40 mol deuterated DPPC

and 10 mol DPPG
b

Data that with big error and unreasonable fitting is not shown

c

Samples with substantial dephasing

The spectra of the IFP L2c, A7c and G16c and their fitting are shown in Table 4.1, Figure
4.3, Figure 4.4 and Figure 4.5 respectively. The spectra were obtained at ~-30 °C and the S/S0
dephasing buildup was fitted to A(1 – e-), in which A and  are fitting parameters. A represents
the fraction of protein with dipolar coupling D ≈ 3/2 which is based on equal time spent in the

122

three 2H m states during  because of T1 relaxation. (1 - A) is the fraction protein with D ≈ 0. D,
in units of Hz is given by: 𝐷 =

𝜇0 ℎ𝛾𝐶 𝛾𝐷
16𝜋 3 𝑟 3

. Thus, the 13C – 2H dipolar coupling in Hz is 𝐷 =

The internuclear distance r was calculated as 3√

4642 𝐻𝑧
3𝛾
2

4642
𝑟3

𝐻𝑧.

Å. The 13CO chemical shifts are correlated

to the peptide backbone secondary structures and thus the secondary structure can be determined
by comparing the chemical shifts in spectra to literature.23, 24
A9_13CO, A8_2H labeled I4 peptide was used as a standard compound, the IFP-membrane
interaction has some differences. (1) The membrane environment is locally non-crystalline so that
the distance between the 13CO in IFP and a particular 2H in membrane may vary among peptide
molecules even if all the IFP have the same membrane location. (2) Each IFP is surrounded by
several lipid molecules, which means the S/S0 is a sum of multi-spin geometry. Since D ∝ r-3 for
each spin pair, the S/S0 buildup is dominated by the D associated with the closest 2H. Overall,
these considerations for peptide 13CO–lipid 2H REDOR imply that fitting parameters will be semiquantitatively rather than quantitatively related to membrane location.22, 25
The L2c and A7c have two peaks in DPPC-D4, DPPC-D8 and DPPC-D10 membrane,
however, G16c only exhibits one peak. The L2c, A7c and G16c spectra have peaks with respective
177, 178 and 175 ppm chemical shifts and are assigned to  helical structure of L2, A7 and G16
residues. The peaks at 174 and 175 ppm in L2c and A7c spectra are assigned to  sheet structure.24
Although all samples were prepared using the same protocol, there was some variation in the :
population ratio. L2  helical peak population in three lipids are D10 > D8 > D4. A7 helical
peak population are D8 > D10 > D4.
As for G16, there is only one broader peak at 175 ppm and this peak is assigned to 
structure based on two reasons: (1) The chemical shift of  sheet Gly should have a chemical shift

123

of ~170 ppm25 and there is no peak or shoulder correspond to this region, and (2) the chemical
shifts of lipid carbonyl carbon natural abundance is also at ~ 175 ppm, and the broader peak may
resulted from the overlapping of 13CO in IFP and natural abundance in lipids. 1,2-di-O-tetradecylsn-glycero-3-phosphocholine (DTPC) and 1,2-di-O-tetradecyl-sn-glycero-3-phospho-(1'-racglycerol) (DTPG) are lipids that do not have carbonyl group. By comparing the full width at half
maximum (FWHM) of G16c S0 in DPPC:DPPG membrane to DTPC:DTPG membrane, the natural
abundance contribution can be obtained (shown in Table 4.2). IFP in DTPC:DTPG has 4.6 ppm
FWHM at 2 ms and 4.3 ppm FWHM at 40 ms. Even though IFP in DPPC:DPPG has an average
4.9 ppm FWHM at both dephasing time, which indicates that lipid carbonyl carbon natural
abundance indeed contributes to broader lines, the L2 and A7 still showed a much narrower
FWHM of ~ 2 ppm. Thus, lipid natural abundance is not the main reason of broad G16 peak. The
reason might be because the Gly has less steric restriction than other residues.
Previous studies of IFP in detergent micelle and lipid membrane both showed that the IFP
adopts a N-helix (residues 1–11)/turn/C-helix (residues 13–19) structure. The two helices are in
close contact and this structure is denoted as “closed” structure. In lipid membrane, IFP also adopts
a “semi-closed” structure, which is also a helix/turn/helix structure but with a wider interhelical
angle.12, 15 There is not much information about the  structure. It might be an  sheet structure at
residue 1 – 11 region and  helix at residue 13 – 19.

124

Table 4.2 FWHM of membrane associated IFP a

a

Labeling

2 ms

40 ms

G16c – D4



5.2

G16c – D8



5.2

G16c – D10



4.2

G16c – F9n b



4.3

IFP with DPPC:DPPG 4:1 at pH 5. S0 spectra of 2 ms or 40 ms dephasing time were processed

with 20 Hz line broadening, baseline DC offset correction and baseline polynomial correction of
the order 5. The FWHM values are in unit ppm.
b

IFP with DTPC:DTPG 4:1 at pH 5. The data is obtained from Ujjayini Ghosh’s dissertation.

The  peak dephasing buildups for L2c, A7c and G16c in membrane containing
DPPC_D10 is comparable large and buildups in DPPC_D8 or DPPC_D4 are smaller. As shown
in Table 4.1, the best-fit A ≈ 1 for all three residues with the DPPC_D10 lipid supports a single
membrane location of IFP. Considering the IFP structure, L2 13CO is on the hydrophobic face and
A7, G16 13CO nuclei are on the hydrophilic faces of the IFP. However, they all showed a similar
4 – 5 Å internuclear distance with the D10 deuterium, which is consistent with van der Waals
contact.
For membrane that contains DPPC_D10, there are smaller  than  buildups for L2c and
A7c. This suggests there are smaller  than  buildups, which supports an overall shallower
location for  IFP. Additionally, the fitting results of A7c in D10 and D8 show rD10 ≈ rD8 and AD10
≈ 2AD8. This supports two different membrane locations for IFP with a major population with deep

125

insertion in contact with D10 2H’s and a minor population with shallower insertion in contact with
D8 2H’s.25

4.2.2 Effects of sample preparation method and lipid charge
The result discussed in 4.2.1 was IFP in negatively charge DPPC and DPPG lipid
membrane prepared by aqueous binding (AB) and the protocol is described in Chapter 2. This
method was designed as incorporation during viral fusion. Organic co-solubilization (OC) sample
preparation method was also used for IFP in neutral DPPC membrane. Table 4.3 and Figure 4.5
displays the  = 40 ms 13C-2H REDOR spectra and fittings of IFP L2c in 100 % DPPC-D10 and
DPPC-D8 lipids. This method is to achieve thermodynamic equilibrium integration of the two
components. The details process is as below.
50 mole of DPPC-D10 or DPPC-D8 lipids was dissolved in chloroform and then solvent
was removed by dry nitrogen gas flow and vacuum pumping overnight. 1 mol dry IFP_L2c was
added to lipid film and then dissolved in a solvent mixture containing 2,2,2-trifluoroethanol,
1,1,1,3,3,3-hexafluoroisopropanol, and chloroform with 2:2:3 volume ratio followed by
subsequent solvent removal via nitrogen gas and vacuum pumping. 3 mL of 10 mM HEPES and
5mM MES buffer at pH 5.0 was used to hydrate the film and followed by 10 times freeze-thaw
cycles to make a homogeneous suspension. Another 20 mL buffer was added followed by ultracentrifugation at 100000g for 5 hours at 4 °C to pellet the peptide-lipid complex.
As shown in Figure 4.6 panel A, the chemical shifts of L2c in DPPC-D8 and DPPC-D10
are 174 ppm which is correspond to the  strand structure only. This result is different from the
situation where L2c binds to negatively charged membrane (DPPC:DPPG 4:1 ratio) in aqueous
phase and the peptide adopts both  helix and  strands structure.

126

Figure 4.6 13CO–2H REDOR data of samples that contain either IFP_L2c in membrane with DPPC
prepared by organic co-solubilization method. Samples were prepared with DPPC_D8 and
DPPC_D10, and the corresponding data are displayed in green and red, respectively. (A) S0 (black)
and S1 (colored) REDOR spectra for  = 40 ms. (B) Plot of (S/S0) vs  are displayed for the 
strand peak. The solid lines are best-fits to A(1 – e-) and the fitting parameters are in Table 4.3.

127

Table 4.3 Best-fit exponential buildup parameters for 13CO–2H REDOR of IFP in DPPC a

a

Peptide

Peak

Membrane

A

(Hz)

r (Å)

L2cb



DPPC_D8

1.00 (29)

26 (11)

5 (1)

L2c



DPPC_D10

0.22 (7)

22 (10)

5 (1)

Samples are prepared by organic co-solubilization method with 1 mol IFP and 50 mol

deuterated DPPC.
b

Samples with substantial dephasing

The difference in signal to noise ratio of D8 and D10 spectra (Figure 4.6 A right panel) is
due to the amount of IFP that bound to the membrane. The D8 sample contained 0.65 mol peptide,
while the D10 sample contained 1.08 mol peptide. The signal to noise of D8 and D10 samples at
2 s dephasing time are the same, while the acquisition number of D8 is three times greater than
D10. Thus the signal of D10 is 1.7 times greater than D8, which is consistent with the 1.08:0.65
mole ratio.
L2c has a much higher buildup rate in DPPC_D8 lipids than in DPPC_C10 lipids (Figure
4.6 B). The

13

CO-2H distances of L2c with D10 and D8 are both ≈ 5 Å, and AD10 ≈ 1 is much

greater than AD8 ≈ 0.22 (Table 4.3). This suggest that when IFP and neutral lipid were prepared by
organic co-solubilization method, the L2

13

CO is predominantly contacting D8 2H’s. However,

when the sample is prepared by aqueous binding method and with negatively charged lipids, rD10
≈ 5 Å is closer than rD8 ≈ 6 Å, and AD10 > AD8. Besides, comparing the dephasing of D8, (S/S0)AB
= 0.3 is smaller than (S/S0)OC = 0.7 at large . All the information showed that the IFP L2c has
close contact with D8 2H’s in neutral membrane, while close contact with D10 2H’s in negatively
charged membrane.

128

The sample preparation methods and the charge of the lipids can be the reasons that cause
such significant difference. Aqueous binding process is a mimic of actual infection, while organic
co-solubilization is to achieve thermodynamic equilibrium integration of the two components.
Previous study of HIV gp41 fusion peptide (FP) in lipid membrane containing
DPPC:DPPG:Cholesterol with 8:2:5 ratio composition using aqueous binding and organic cosolubilization sample preparation methods obtained both  sheet structure spectra and similar
extent of buildups.25 As for the charge effect, in that paper, organic co-solubilization prepared FP
with neutral PC only membrane and negative PC:PG membrane were also compared. The two
spectra both showed  sheet structure spectra and qualitatively similar buildups. Thus, from the
results, the sample preparation and lipid composition did not affect the peptide secondary structure
and membrane location of HIV fusion peptide.
However, when IFP was prepared by different sample preparation methods and with
different lipid composition, both secondary structure and membrane location were change. When
prepared by aqueous binding with negatively charged membrane, IFP adopts both  strand and 
sheet structures. The IFP showed fast dephasing buildup with D10, on the other hand, IFP has
small buildups with all D4, D8 and D10. When prepared by organic co-solubilization with neutral
membrane, IFP only adopts  sheet structure and showed high buildup with D8. The reason for
different secondary structures and different  structure membrane locations maybe due to the
sample preparation methods. The organic co-solubilization method achieves the thermodynamic
favorable integration of the IFP and membrane, while the aqueous binding method leads to kinetic
favorable conformation.
The electrostatic interaction between positively charged IFP and negatively charged DPPG
lipid at pH 5.0 also affects the binding percentage. As mentioned in Chapter 2, the NMR sample

129

was the centrifugation pellet containing membrane + bound peptide and the unbound IFP exists in
the supernatant. After ultra-centrifugation, the supernatant A280 absorbance was measured to
quantify the amount of unbounded IFP. For samples prepared by aqueous binding method, no IFP
exist in supernatant (A280 = 0), which means 100% binding. But for organic co- solubilization,
only 60% to 80% binding percentage was observed.
The DPPC:DPPG 4:1 composition and aqueous binding was chosen for most of the
samples based on two reasons: (1) PC:PG 4:1 combination is a better mimic of host cell membrane
since the human cell membrane has similar fraction of negatively charged lipids; 26 (2) aqueous
binding is similar to the viral infection process, in which the IFP is exposed during conformational
change of HA and then binds to the host cell membrane.

4.2.3 Close contact of  helical IFP and 2H in DPPC_D10
IFP with N-helix/turn/C-helix structure appears to have a single membrane location as
evidenced by rapid L2c, A7c and G16c buildups in membrane with DPPC_D10 and correlate bestfit A ≈ 1 (Figure 4.3 and Table 4.1). Much smaller buildups were observed in membrane with
DPPC_D4 and DPPC_D8. There are close contacts between the lipid acyl chain tail and  helical
IFP in both N-helix and C-helix region. This may be relevant for fusion catalysis because of local
perturbation of the membrane bilayer with consequent reduced activation energy to the highly
perturbed fusion transition state.
By comparing the REDOR result to the current MD simulations (Figure 4.1 C - F), close
contacts of all the L2, A7 and G16 residues with lipid acyl chain tail is inconsistent with N-helix
deeply inserted while C-helix at membrane surface (Figure 4.1 D) and inconsistent with interfacial
location of a IFP (Figure 4.1 F).19, 21 However, based on the current REDOR results, it is hard to

130

distinguish whether IFP adopts a membrane-spanning conformation or IFP promotes lipid acyl
chain protrusion. The possible IFP membrane location models are shown in Figure 4.7.
Both models may facilitate membrane fusion by perturbing local membrane structure. In
membrane spanning model, the IFP inserts deeply in membrane hydrophobic center and break the
continuous membrane structure (Figure 4.7 A), and in lipid protrusion model, the lipid acyl chain
is induced to protrude and contact with water (Figure 4.7 B). When IFP bonds to host cell
membrane, this perturbation may facilitate the lipid acyl chain to interact with viral outer leaflet
and thus induce hemifusion.

Figure 4.7 Possible models for IFP topology in membrane. The lipid molecules are shown in blue,
D10 2H atoms in lipid acyl chain shown as red dots, IFP in green and labeled residues in orange.
(A) Membrane spanning model and (B) lipid tail protrusion model.

131

4.3 Conclusion
13

C-2H REDOR solid-state NMR is used to study the IFP membrane location. In this work,

H3_20 IFP was synthesized manually by Fmoc solid-phase peptide synthesis, and 13CO labeled at
either Leu_2, Ala_7 or Gly_16 residues. IFP was bonded to acyl chain deuterated lipid vesicle
with a composition of DPPC:DPPG 4:1 ratio as a mimic of host cell membrane. The

13

CO-2H

dipolar coupling was fitted based on REDOR dephasing and internuclear distances were calculated.
The IFP adopts major  helical, minor  strand secondary structure in PC/PG membrane.
The  helical IFP’s with respectively 13CO labeled Leu-2, Ala-7 and Gly-16 all show close contacts
with the lipid acyl chain tail, suggesting IFP has strong interaction with the membrane: it either
has a membrane-spanning confirmation, or it promotes lipid trail protrusion. IFP bounded lipid
membrane structure studied by paramagnetic relaxation enhancement (PRE) solid-state NMR will
be discussed in Chapter 5 and provides more information about the detailed IFP membrane
location model.

132

REFERENCES

133

REFERENCES

[1] Webster, R. G., Bean, W. J., Gorman, O. T., Chambers, T. M., and Kawaoka, Y. (1992)
EVOLUTION AND ECOLOGY OF INFLUENZA-A VIRUSES, Microbiol. Rev. 56, 152179.
[2] Wilson, I. A., Skehel, J. J., and Wiley, D. C. (1981) STRUCTURE OF THE
HEMAGGLUTININ MEMBRANE GLYCOPROTEIN OF INFLUENZA-VIRUS AT 3A RESOLUTION, Nature 289, 366-373.
[3] Bullough, P. A., Hughson, F. M., Skehel, J. J., and Wiley, D. C. (1994) STRUCTURE OF
INFLUENZA HEMAGGLUTININ AT THE PH OF MEMBRANE-FUSION, Nature 371,
37-43.
[4] Kemble, G. W., Danieli, T., and White, J. M. (1994) LIPID-ANCHORED INFLUENZA
HEMAGGLUTININ PROMOTES HEMIFUSION, NOT COMPLETE FUSION, Cell 76,
383-391.
[5] Skehel, J. J., and Wiley, D. C. (2000) Receptor binding and membrane fusion in virus entry:
The influenza hemagglutinin, Annu. Rev. Biochem. 69, 531-569.
[6] Nobusawa, E., Aoyama, T., Kato, H., Suzuki, Y., Tateno, Y., and Nakajima, K. (1991)
COMPARISON OF COMPLETE AMINO-ACID-SEQUENCES AND RECEPTORBINDING PROPERTIES AMONG 13 SEROTYPES OF HEMAGGLUTININS OF
INFLUENZA A-VIRUSES, Virology 182, 475-485.
[7] Gething, M. J., Doms, R. W., York, D., and White, J. (1986) STUDIES ON THE
MECHANISM OF MEMBRANE-FUSION - SITE-SPECIFIC MUTAGENESIS OF THE
HEMAGGLUTININ OF INFLUENZA-VIRUS, J. Cell Biol. 102, 11-23.
[8] Qiao, H., Armstrong, R. T., Melikyan, G. B., Cohen, F. S., and White, J. M. (1999) A specific
point mutant at position 1 of the influenza hemagglutinin fusion peptide displays a
hemifusion phenotype, Mol. Biol. Cell 10, 2759-2769.
[9] Steinhauer, D. A., Wharton, S. A., Skehel, J. J., and Wiley, D. C. (1995) STUDIES OF THE
MEMBRANE-FUSION ACTIVITIES OF FUSION PEPTIDE MUTANTS OF
INFLUENZA-VIRUS HEMAGGLUTININ, J. Virol. 69, 6643-6651.
[10] Durrer, P., Galli, C., Hoenke, S., Corti, C., Gluck, R., Vorherr, T., and Brunner, J. (1996) H+induced membrane insertion of influenza virus hemagglutinin involves the HA2 aminoterminal fusion peptide but not the coiled coil region, J. Biol. Chem. 271, 13417-13421.
[11] Han, X., Bushweller, J. H., Cafiso, D. S., and Tamm, L. K. (2001) Membrane structure and
fusion-triggering conformational change of the fusion domain from influenza
hemagglutinin, Nat. Struct. Biol. 8, 715-720.

134

[12] Lorieau, J. L., Louis, J. M., and Bax, A. (2010) The complete influenza hemagglutinin fusion
domain adopts a tight helical hairpin arrangement at the lipid:water interface, Proc. Natl.
Acad. Sci. U. S. A. 107, 11341-11346.
[13] Sun, Y., and Weliky, D. P. (2009) C-13-C-13 Correlation Spectroscopy of MembraneAssociated Influenza Virus Fusion Peptide Strongly Supports a Helix-Turn-Helix Motif
and Two Turn Conformations, J. Am. Chem. Soc. 131, 13228-13229.
[14] Ghosh, U., Xie, L., and Weliky, D. P. (2013) Detection of closed influenza virus
hemagglutinin fusion peptide structures in membranes by backbone (CO)-C-13-N-15
rotational-echo double-resonance solid-state NMR, J. Biomol. NMR 55, 139-146.
[15] Ghosh, U., Xie, L., Jia, L. H., Liang, S., and Weliky, D. P. (2015) Closed and Semiclosed
Interhelical Structures in Membrane vs Closed and Open Structures in Detergent for the
Influenza Virus Hemagglutinin Fusion Peptide and Correlation of Hydrophobic Surface
Area with Fusion Catalysis, J. Am. Chem. Soc. 137, 7548-7551.
[16] Epand, R. M. (2003) Fusion peptides and the mechanism of viral fusion, Biochim. Biophys.
Acta-Biomembr. 1614, 116-121.
[17] Larsson, P., and Kasson, P. M. (2013) Lipid Tail Protrusion in Simulations Predicts Fusogenic
Activity of Influenza Fusion Peptide Mutants and Conformational Models, PLoS Comput.
Biol. 9, 9.
[18] Legare, S., and Laggue, P. (2014) The influenza fusion peptide promotes lipid polar head
intrusion through hydrogen bonding withphosphates and N-terminal membrane insertion
depth, Proteins 82, 2118-2127.
[19] Brice, A. R., and Lazaridis, T. (2014) Structure and Dynamics of a Fusion Peptide Helical
Hairpin on the Membrane Surface: Comparison of Molecular Simulations and NMR, J.
Phys. Chem. B 118, 4461-4470.
[20] Victor, B. L., Lousa, D., Antunes, J. M., and Soares, C. M. (2015) Self-Assembly Molecular
Dynamics Simulations Shed Light into the Interaction of the Influenza Fusion Peptide with
a Membrane Bilayer, J. Chem Inf. Model. 55, 795-805.
[21] Baylon, J. L., and Tajkhorshid, E. (2015) Capturing Spontaneous Membrane Insertion of the
Influenza Virus Hemagglutinin Fusion Peptide, J. Phys. Chem. B 119, 7882-7893.
[22] Xie, L., Ghosh, U., Schmick, S. D., and Weliky, D. P. (2013) Residue-specific membrane
location of peptides and proteins using specifically and extensively deuterated lipids and
C-13-H-2 rotational-echo double-resonance solid-state NMR, J. Biomol. NMR 55, 11-17.
[23] Morcombe, C. R., and Zilm, K. W. (2003) Chemical shift referencing in MAS solid state
NMR, J. Magn. Reson. 162, 479-486.
[24] Zhang, H. Y., Neal, S., and Wishart, D. S. (2003) RefDB: A database of uniformly referenced
protein chemical shifts, J. Biomol. NMR 25, 173-195.

135

[25] Jia, L. H., Liang, S., Sackett, K., Xie, L., Ghosh, U., and Weliky, D. P. (2015) REDOR solidstate NMR as a probe of the membrane locations of membrane-associated peptides and
proteins, J. Magn. Reson. 253, 154-165.
[26] van Meer, G., Voelker, D. R., and Feigenson, G. W. (2008) Membrane lipids: where they are
and how they behave, Nat. Rev. Mol. Cell Biol. 9, 112-124.

136

Chapter 5
IFP effects on membrane studied by 2H paramagnetic
relaxation enhancement (PRE) solid-state NMR

137

5.1 Introduction
In Chapter 4, REDOR NMR results showed that the Leu2, Ala7 and Gly16 residues in IFP
have close contact with lipid acyl chain tail. According to the current models, the REDOR data is
consistent with two of them: membrane-spanning model and lipid acyl chain protrusion model,
and the possible IFP membrane location models are shown in Figure 4.7. To further study the
membrane structure perturbed by IFP, Paramagnetic Relaxation Enhancement (PRE) solid-state
NMR is used. The information about IFP-membrane interaction may help us to understand the
fusion process and mechanism.
Solomon developed the PRE in 1955.1 PRE experiments utilizes the high gyromagnetic
ratio of unpaired electron in paramagnetic species to enhance dipolar interactions between the
nuclear of interest and the electron and can be used to get distance information in a range up to 35
Å. Thus, it is a very powerful tool for long-range distance determination, and for data acquisition
acceleration, especially suitable for biomolecules and human tissue.2, 3
Previous studies used paramagnetic species such as Mn2+, Gd2+, and nitroxide spin in solid
state NMR for studying protein membrane location. For example, to study the Protegrin-1 peptide
membrane location,

13

C site-specifically labeled the peptide was bonded to membrane and the

signal attenuation caused by Mn2+ was measured. Paramagnetic Mn2+ induces distance-dependent
line broadening and signal attenuation of the nuclear spin signals by enhancing the spin-spin
relaxation rate (1/T2). The Mn2+ location in a lipid vesicle system is well defined: it only attaches
to membrane surface and does not penetrate the membrane. By comparing the signal decrease in
peptide to the signal decrease in lipid carbons, the relative depth of the peptide with respect to the
lipid can be extracted.4 This method can also be used to study the asymmetric insertion of
membrane proteins in membrane by only applying Mn2+ ions on the outer but not the inner leaflet

138

of lipid bilayers.5 Another example is the location of cholesterol in membrane determined by
comparing the 13C spin-lattice relaxation rates of phospholipid and cholesterol in the presence of
paramagnetic Gd3+.6
In this study, PRE was used to illustrate the effect of IFP on membrane structure and also
to solve the model of IFP insertion in membrane.
PRE was chosen based on its long distance detection range. Besides, compare to other
techniques such as NOE, 1H spin diffusion or EPR, PRE has less limitations. The NOE detection
range is ~ 5 Å, which is too short compared to the 34 Å membrane hydrophobic core thickness.
To study the peptide insertion in membrane, 1H spin diffusion NMR is another approach. The
magnetization transfer rate from lipid methyl protons in the center of the bilayer to the target
peptide is measured in this experiment, and this transfer is facilitated by 1H-1H dipolar coupling.
There is a significant difference in the magnetization transfer rates between a rigid peptide and the
mobile lipids at ambient temperature. The spin diffusion from the lipid methyl protons to a rigid
peptide close to the center of the bilayer is rapid. In contrast, for a peptide that bind on membrane
surface, the methyl 1H magnetization must first diffuse through the lipid acyl chains before
reaching the protein. Since spin diffusion in the lipids is extremely slow because of motionaveraged 1H-1H dipolar couplings, the surface-bound protein receives little 1H magnetization from
the methyl protons and showed low transfer rate. If the peptide does not have enough rigidity, the
spin diffusion from methyl proton to peptide is also slow, then it is hard to distinguish the diffusion
to peptide or lipid since they have closer rate. Thus, to distinguish the peptide location on
membrane surface or membrane hydrophobic core, a rigid the peptide is required.7 EPR
experiments required to introduce of bulky spin probes, which may perturb the membrane packing
and complicate the data interpretation.8

139

In the experiment, lipid membrane with composition of DPPC:DPPG 4:1 ratio was used,
and this fraction of negatively charged lipid was chosen as a mimic the HIV host cell membrane.9
To study the membrane structure, DPPC-D8 and DPPC-D10 lipids were chosen, as they are
deuterated on lipid acyl chain as shown in Figure 2.1. Deuterium atoms in DPPC-D8 and
DPPC_D10 are at center of membrane leaflet and center of membrane hydrophobic core, thus their
positions in membrane can provide information of the lipid acyl chain conformation. Paramagnetic
species, Mn2+, was added to vesicles. Labeling scheme shown in Figure 5.1.
The PRE arises from dipolar interactions between the 2H in lipid acyl chain and the
unpaired electrons in Mn2+ and contributes to faster T2 relaxation which can be described as:
1
𝑇2

1

𝜇

2 𝛾2 𝜇2

= 𝑅2 = 𝑊 15 (4𝜋0 )

𝐷 𝑒𝑓𝑓 𝛽
𝑟6

2

3𝜏

13𝜏

(4𝜏𝑠 + 1+𝜔2𝑠 𝜏2 + 1+𝜔2𝑠𝜏2 )
𝐷 𝑠

5.1

𝑒 𝑠

Where R2 is the T2 relaxation rate; W is the local concentration of the Mn2+ ions;0 is the vacuum
permeability; D is the gyromagnetic ratio of 2H; eff is the effective magnetic moment of Mn2+
ions;  is the Bohr magneton; r is the average electron-nucleus distance; D and e are the 2H and
election Larmor frequencies. The correlation time s is the inverse sum of the electronic spin-lattice
relaxation time T1e; the rotational correlation time of the molecule r; and the residence time of the
Mn2+ near 2H m:
1
𝜏𝑠

1

1

1

1𝑒

𝑟

𝑚

=𝑇 +𝜏 +𝜏

5.2

Since Mn2+ only binds on membrane surface, it can represent the aqueous layer. By measuring 2H
Mn2+ induced R2 increase, the distance of 2H to membrane surface can be obtained.
Deuterium T2 relaxation rate was determined by quadrupolar echo (quecho) NMR
experiment (pulse sequence shown in Figure 5.2). The details have been discussed in chapter 1

140

and chapter 2. The signal intensity (I) in FID was plotted vs t, and t = [1 + 2 + time being shifted].
The T2 relaxation time can be fitted by:
𝑡

5.3

𝐼(𝑡) = 𝐼(0)×𝑒𝑥𝑝 (𝑇 )
2

Where 𝐼(0) and T2 are fitting parameters. 𝐼(𝑡) is the FID intensity at t = [1 + 2 + time being
shifted]; 𝐼(0) is the FID intensity at t=0; and T2 is the transverse relaxation time or spin-spin
relaxation time. The R2 is calculated by:
1

5.4

𝑅2 = 𝑇

2

R2 of (1) pure lipid; (2) pure lipid with Mn2+; (3) peptide-bounded lipid; and (4) peptidebounded lipid with Mn2+ were measured. If smaller R2 difference is observed between (1) and (2)
than (3) and (4), the 2H atoms in IFP-bounded membrane have a closer distance to membrane
surface. If the R2 difference between (1) (2) and (3) (4) is similar, no lipid protrusion is promoted
by IFP. Thus, the distance information between 2H and membrane surface can be obtained.

141

Figure 5.1 Labeling scheme of PRE experiments. Paramagnetic species are Mn2+ ions, and they
bind to the surface of lipid membrane. Lipid composition used was DPPC:DPPG 4:1 mole ratio.
DPPC-D8 and DPPC-D10 were used. The hydrophobic core is 34 Å, which is a comparable length
with the PRE detection range. The Mn2+ ions are labeled in red, lipid molecules in blue and
deuterium atoms shown as red dots. (A) Lipid acyl chain without protrusion. (B) Lipid acyl chain
protrudes towards aqueous surface.

Figure 5.2 Quecho pulse sequence. 2H spectra were acquired with different 1 and 2 and a fixed
(1 -2) value. Typically, the pulse length was set ≈1.5 s; 1, 2 were set between 10 and 1000
s; the recycle delay was set to 1 s. Samples are detected at 25 °C (gel phase lipids) and 50 °C
(fluid phase lipids).

142

5.2 Result and discussion
5.2.1 Sample preparation methods
The lipid vesicle preparation and peptide aqueous binding methods are discussed in chapter
2. The relaxation rate in PRE experiments depends on the concentration of Mn2+ ions. Thus, the
quantity of Mn2+ that binds on membrane surface is crucial. Two sample preparation methods were
tested:
Method 1. According to literature, Mn2+ ions all binds to the lipid membrane surface if the
mole percentage of Mn2+ is smaller than 50%.4 To prepare lipid vesicle with 20 mole % Mn2+,
lipid film containing 20 mol of DPPC (D8 or D10 deuterated) and 5 mol DPPG was hydrated
with 2 mL 10 mM HEPES and 5mM MES buffer at pH 5.0 containing 5 mole of Mn2+, followed
by 10 freeze-thaw cycle and extrusion. 1 mol IFP in the same buffer was added to vesicles, then
ultracentrifugation at 100000g for 5 hours. Presumably all Mn2+ ions should bind to the membrane
surface and no Mn2+ left in supernatant. The pellet was lyophilized, packed into NMR rotor and
rehydrate with 5 L pH 5 buffer.
Method 2. Lipid film containing 20 mol of DPPC (D8 or D10 deuterated) and 5 mol
DPPG lipid was hydrated with 2 mL 10 mM HEPES and 5mM MES buffer at pH 5.0, followed
by 10 freeze-thaw cycle and extrusion. 1 mol IFP in the same buffer was added to vesicles, then
ultracentrifugation at 100000g for 5 hours. The pellet was lyophilized, packed into NMR rotor and
rehydrate with 5 L 10 mM HEPES and 5mM MES buffer that contains 5 mole of Mn2+. Five
freeze-thaw cycles were done to evenly distribute the Mn2+ on both side of the membrane.
The key difference in these two methods are when to introduce Mn2+ ions. In Method 1,
Mn2+ was added to rehydrate the lipid film, thus, Mn2+ binds to membrane surface during

143

preparation of unilamellar vesicles, while in Method 2 Mn2+ was added at when pack the sample
into NMR rotor.
The binding result in Method 1 was tested. Mn2+ in solution can be oxidized by NaIO4,
following this equation:
2𝑀𝑛2+ + 5𝐼𝑂4− + 3𝐻2 𝑂 → 2𝑀𝑛𝑂4− + 5𝐼𝑂3− + 6𝐻 +
The product permanganate ion has an intense purple color and can be determined by A525. The
supernatant after ultracentrifugation was mixed with access amount of NaIO4 and boiled, then the
absorbance at 525 nm was measured. By comparing the Mn2+ quantity in supernatant to the control
experiment (1.25 mole Mn2+ dissolved in the same volume as the supernatant), the mole of Mn2+
that binds to membrane can be calculated. Five conditions for DPPC-D8 or DPPC-D10, and with
or without IFP were tested. The results showed Mn2+ mole percentage to lipid molecules are
respectively 16%, 15%, 9%, 12% and 3%, not as expected 20%. The uncertainty of Mn2+ binding
causes unpredictable change in T2 relaxation time measurements.
Mn2+ was added at the last step in Method 2, thus the Mn2+ quantity is consistent from
sample to sample. In Method 2, IFP solution was added to unilamellar vesicles with 100 nm
diameter. IFP bond to vesicles and due to the fusion activity of IFP, the vesicles may have various
diameter. Five cycles of freeze-thaw were applied to reach Mn2+ distributions equilibrium on both
side of the membrane.10,

11

It was also reported that repeating freeze-thaw cycles can form

unilamellar vesicles with diameter ˂ 200 nm.12 Thus, the freeze-thaw effect on membrane structure
should be neglectable.

144

5.2.2 2H FID, spectra and T2 relaxation time of pure lipid membrane
To obtain molecular level view of how IFP modulates lipid organization, the pure lipid
structure and T2 relaxation rate was studied. Solid-state 2H NMR FID and spectra of DPPCD10:DPPG or DPPC-D8:DPPG (D10 and D8 as abbreviation respectively) with 4:1 ratio at 25 or
50 °C were obtained and analyzed. Figure 5.3 – Figure 5.6 show the FID, stacked FID and stacked
spectra with synchronize increasing 1 and 2 values. As shown in these figures, both the FID
intensity and spectra integration decrease as 1 and 2 increase, and the decrease in signal follows
exponential decay.
The 2H NMR spectra D10 or D8 at 25 or 50 °C are shown in Figure 5.7. The peak splitting
analysis is in Table 5.1. The 2H spectra at 50 °C in Figure 5.7 showed the sharp cut off signal
indicate that the sample is in the liquid phase, which is consistent with the 41 °C theoretical phase
transition temperature of pure DPPC and DPPG lipids. The big differences in spectra line shape at
25 and 50 °C also illustrate a membrane phase change.
The spectrum of D10 at 50 °C had the smallest peak splitting of 2.6 kHz, which is from the
terminal methyl group of the lipid acyl chain. The broader 9.9 kHz splitting came from the
methylene groups. The 25.9 kHz splitting in D8 was greater than the 9.9 kHz in D10, since when
the methylene groups that are further from acyl chain terminal, a larger splitting occurs. At lower
temperature, both D10 and D8 are in gel phase and had larger peak splitting compare to liquid
phase. The methylene peak -CD2 is not resolved at 25 °C (Table 5.1).

145

Figure 5.3 D10 at 25 °C. Top: FID. Middle: FID stacked plots with increasing 1 and 2. Bottom:
spectra stacked plots. All spectra were acquired with 5000 scans, processed with 300 Hz
exponential line broadening, data shift = -11, and baseline correction.

146

Figure 5.4 D10 at 50 °C. Top: FID. Middle: FID stacked plots with increasing 1 and 2. Bottom:
spectra stacked plots. All spectra were acquired with 5000 scans, processed with 300 Hz
exponential line broadening, data shift = -11, and baseline correction.

147

Figure 5.5 D8 at 25 °C. Top: FID. Middle: FID stacked plots with increasing 1 and 2. Bottom:
spectra stacked plots. All spectra were acquired with 5000 scans, processed with 1000 Hz
exponential line broadening, data shift = -11, and baseline correction.

148

Figure 5.6 D8 at 50 °C. Top: FID. Middle: FID stacked plots with increasing 1 and 2. Bottom:
spectra stacked plots. All spectra were acquired with 5000 scans, processed with 1000 Hz
exponential line broadening, data shift = -11, and baseline correction.

149

Figure 5.7 2H NMR spectra D10 or D8 at 25 or 50 °C. Spectra were acquired with 5000 scans,
data shift = -11, and baseline correction. D10 spectra were processed with 300 Hz exponential line
broadening, and D8 with 1000 Hz exponential line broadening.

150

Table 5.1 2H NMR spectra peak splitting (kHz)

a

25 °C

50 °C

D10 a

11.8

2.6 / 9.9

D8

50.6

25.9

Methylene peak -CD2 of D10 at 25 °C was not determined because of the lack of the well resolved

-CD2 peaks.

Figure 5.8 Quecho experimental (black squares) and best fit (red lines) plots of D10 and D8 at 25
and 50 °C.

151

Table 5.2 Best-fit 2H T2 (s) of D10 and D8 lipids measured by quecho experiment a

a

25 °C

50 °C

D10

516 (14)

1937 (36)

D8

310 (2)

688 (21)

The T2 relaxation time values are in unit s. The uncertainties are shown in parenthesis

The T2 relaxation rate was fitted by equation 5.3.
𝑡

5.3

𝐼(𝑡) = 𝐼(0)×𝑒𝑥𝑝 (𝑇 )
2

Where 𝐼(0) and T2 are fitting parameters. 𝐼(𝑡) is the FID intensity at t = [1 + 2 + time being
shifted]; time being shifted ≈ 22 s in these experiments; 𝐼(0) is the FID intensity at t=0; and T2
is the transverse relaxation time or spin-spin relaxation time.13
Figure 5.8 is shown as the FID intensity decay and their best fitting. The T 2 values from
exponential fitting are shown in Table 5.2. From 25 to 50 °C, the T2 relaxation time of D10
increased from 516 to 1937 s, which is almost four folds longer. As for D8, the T2 relaxation time
increased by two folds, from 310 to 688 s. Even the increase differs in these two samples, but it
can be concluded that the T2 has a positive as correlation with temperature. Additionally, D8 had
a T2 of 310 s at 25 °C, which is shorter compare to the 516 s T2 of D10 at the same temperature.
The trend is also true at 50 °C, T2 of D8 was 688 s, shorter than D10’s 1937 s.

152

5.2.3 The effect of Mn2+ concentration on lipid T2
As discussed in equation 5.1:
1
𝑇2

1

𝜇

2 𝛾2 𝜇2

= 𝑅2 = 𝑊 15 (4𝜋0 )

𝐷 𝑒𝑓𝑓 𝛽
𝑟6

2

3𝜏

13𝜏

(4𝜏𝑠 + 1+𝜔2𝑠 𝜏2 + 1+𝜔2𝑠𝜏2 )
𝐷 𝑠

5.1

𝑒 𝑠

The concentration of Mn2+ (W) affects the T2 and R2. High concentration of Mn2+ may also affect
the membrane structure. Therefore, in this section, the concentration effect is discussed. The
stacked D10 2H spectra with different concentration of Mn2+ were shown in Figure 5.9. The left
panel shows the spectra at 25 °C and the right panel shows the spectra at 50 °C. From top to bottom
are D10 with 20%, 5%, 1%, 0.2% and pure lipids. The peak splitting in each spectrum was shown
in Table 5.3.
At 25 °C, the peak splitting of the methyl groups fell in 11.8 – 12 kHz range, not effected
by the increasing concentration of Mn2+. By looking at the spectra, the line shapes of various
concentration of Mn2+ were overall the same. However, at 50 °C, when lipid in fluid phase, the
line shape was greatly influenced by Mn2+. The linewidths were narrower with smaller Mn2+
concentration and spectrum was better resolved. This is also evidenced by a decreasing peak
splitting of both methyl and methylene groups. When Mn2+ concentration decrease from 5% to
0%, the peak splitting of methyl group slightly decreased from 2.9 to 2.5 kHz, and 11 to 10 kHz
for methylene group. However, with 20% Mn2+, the shape as well as peak width did not follow the
trend. The peak was much broader but the peak splitting was smaller. The reason for contradictory
result may due to the high concentration of Mn2+ on membrane surface changed the membrane
structure.

153

25 °C

50 °C

20%

5%

1%

0.2%

pure lipids

Figure 5.9 2H NMR spectra of D10 with different mole percentage of Mn2+.

154

Table 5.3 D10 2H NMR spectra peak splitting affected by Mn2+ concentration a
25 °C

a

b

50 °C

Methyl -CD3

Methylene -CD2 b

Methyl -CD3

Methylene -CD2

D10_20% Mn2+

11.9

-

2.1

9.2

D10_5% Mn2+

12.0

-

2.8

11.0

D10_1% Mn2+

11.8

-

2.9

11.3

D10_0.2% Mn2+

11.9

39.4

2.5

9.5

D10

11.8

-

2.6

9.8

The numbers represents the peak splitting in unit kHz.
Some methylene peak splitting at 25 °C was not determined because of the lack of the well

resolved -CD2 peaks.

Table 5.4 Best-fit 2H T2 (s) of D10 with various Mn2+ concentration a

a

25 °C

50 °C

D10_20%

477 (10)

245 (5)

D10_5%

471 (2)

593 (18)

D10_1%

458 (9)

1035 (13)

D10_0.2%

520 (10)

1018 (9)

D10

516 (14)

1937 (36)

The T2 relaxation time values are in unit s. The uncertainties are shown in parenthesis.

155

Figure 5.10 Quecho experimental (black squares) and best fit (red lines) plots of D10 with various
concentration of Mn2+ at 25 °C.

156

Figure 5.11 Quecho experimental (black squares) and best fit (red lines) plots of D10 with various
concentration of Mn2+ at 50 °C.

157

The fitting of T2 relaxation time of D10 lipid with various Mn2+ concentration is shown in
Figure 5.10, 5.11 and Table 5.4. With 0 – 20% Mn2+, the T2 relaxation time at gel phase was in
470 – 520 s range, which is consistent with the line shape and peak splitting similarities at 25 °C.
At fluid phase, D10 with 20% Mn2+ had the shortest T2 = 245 s. With decrease in Mn2+
concentration from 5%, 1%, 0.2% to 0%, the T2 values increased to 593, 1035, 1018 and 1937 s
respectively. This is due to with higher concentration Mn2+, the average distance between unpaired
electron in Mn2+ and 2H is shorter.
IFP bounded membrane samples were prepared with 20% and 5% Mn2+ and will be shown
in the following sections. However, from the previous data, the big differences in line shape and
peak splitting of D10 with 20% Mn2+ illustrate that high concentration of Mn2+ may affect the
overall membrane structure. Therefore, the R2 difference between IFP bounded membrane with
and without 20% Mn2+ may not only due to the addition of IFP, since Mn2+ also changes the
membrane. The 1% and 0.2% Mn2+ D10 samples were prepared most recently and the IFP bounded
membrane samples correspond to these two Mn2+ concentration were not prepared.

5.2.4 FID, spectra and T2 of IFP-bounded membrane with 5% Mn2+
Figure 5.12 and 5.13 are shown as examples of the FID, stacked FID and stacked spectra
of IFP-bounded membrane at gel of liquid phase. The spectra of pure D10, D10 with 5% Mn2+,
IFP-bounded D10, and IFP-bounded D10 Mn2+ (denoted as D10, D10_Mn, IFP_D10, and
FIP_D10_Mn) and pure D8, D8 with 5% Mn2+, IFP-bounded D8, and IFP-bounded D8 Mn2+
(denoted as D8, D8_Mn, IFP_D8, and FIP_D8_Mn) at 25 or 50 °C are shown in Figure 5.14 and
5.15. The lipid composition were DPPC:DPPG with 4:1 mol ratio. All the samples were prepared
using buffer containing 10 mM HEPES and 5mM MES at pH 5.0. The IFP-bounded lipid samples

158

contained 20 mol DPPC-D10 or DPPC-D8, 5 mol DPPG and 1 mol IFP and were prepared by
aqueous binding method (As discussed in Chapter 2). The Mn2+ containing samples are prepared
separately from Mn2+ free samples. The sample preparation method is the same for samples with
or without Mn2+. The difference is the rehydration step. For example, D10 is rehydrated with pH
5 buffer and D10_Mn is rehydrated with pH 5 buffer and additional 5% mol Mn2+.
Figure 5.12 and 5.13 show the IFP_D10 sample at 25 and 50 °C. The top panels in these
two figures are the FID at gel and fluid phase, which showed high similarity to the FID of pure
D10 at these temperatures (Figure 5.3 and 5.4). The middle and bottom panels show the stacked
FID intensity as well as stack spectra, and they both decrease as 1 and 2 increase, and the decrease
in signal follows exponential decay. The FID, stacked FID and stacked spectra of D10_Mn,
IFP_D10_Mn, D8_Mn, IFP_D8, and IFP_D8_Mn all showed similar decay in signal (not shown
in this dissertation).
The spectra of all samples at both 25 and 50 °C are shown in Figure 5.14 and 5.15. The
peak splitting and fitted T2 are shown in Table 5.5 – 5.8. For D10 at gel phase and for D8 at liquid
phase, adding IFP or Mn2+ or both does not change the line shape in the spectra. However, for D10
at liquid phase, adding both IFP and Mn2+ lead to a less resolved spectrum and the methylene peak
splitting cannot be obtained. As for D8 at 25 °C, addition of IFP or Mn2+ or both result in narrower
lines. Besides, with same quantity of lipid, same number of scans and same date processing, IFP
and Mn2+ cause the D8 spectra signal to noise decrease. Even there were slightly differences in
these 2H NMR spectra, we still can conclude that addition of IFP or Mn2+ to DPPC+DPPG lipid
does not change the lamellar membrane phase.

159

Figure 5.12 IFP_D10 at 25 °C. Top: FID. Middle: FID stacked plots with increasing 1 and 2.
Bottom: spectra stacked plots. All spectra were acquired with 5000 scans, processed with 300 Hz
exponential line broadening, data shift = -11, and baseline correction.

160

Figure 5.13 IFP_D10 at 50 °C. Top: FID. Middle: FID stacked plots with increasing 1 and 2.
Bottom: spectra stacked plots. All spectra were acquired with 5000 scans, processed with 300 Hz
exponential line broadening, data shift = -11, and baseline correction.

161

25 °C

50 °C

D10

D10 Mn

IFP D10

IFP D10 Mn

Figure 5.14 2H NMR spectra of D10 samples at 25 and 50 °C with 5% Mn2+.

162

25 °C

50 °C

D8

D8 Mn

IFP D8

IFP D8 Mn

Figure 5.15 2H NMR spectra of D8 samples at 25 and 50 °C with 5% Mn2+.

163

Table 5.5 IFP-bounded D10 membrane 2H NMR spectra peak splitting with 5% Mn2+ a
25 °C

50 °C

Methyl -CD3

Methyl -CD3

Methylene -CD2

D10

11.8

2.6

9.8

D10_Mn

12.0

2.8

11.0

IFP_D10

11.1

2.7

10.4

IFP_D10_Mn b

10.8

2.6

-

a

The numbers represents the peak splitting in unit kHz.

b

The methylene peak splitting was not determined because of the lack of the well resolved -CD2

peaks.

Table 5.6 IFP-bounded D8 membrane 2H NMR spectra peak splitting with 5% Mn2+ a

a

25 °C

50 °C

Methylene -CD2

Methylene -CD2

D8

48.1

26.1

D8_Mn

36.3

24.6

IFP_D8

40.4

24.2

IFP_D8_Mn

35.3

25.5

The numbers represents the peak splitting in unit kHz.

164

Figure 5.16 Quecho experimental (black squares) and best fit (red lines) plots of D10, D10_Mn,
IFP_D10 and IFP_D10_Mn with 5% Mn2+ at 25 °C.

165

Figure 5.17 Quecho experimental (black squares) and best fit (red lines) plots of D10, D10_Mn,
IFP_D10 and IFP_D10_Mn with 5% Mn2+ at 50 °C.

166

Figure 5.18 Quecho experimental (black squares) and best fit (red lines) plots of D8, D8_Mn,
IFP_D8 and IFP_D8_Mn with 5% Mn2+ at 25 °C.

167

Figure 5.19 Quecho experimental (black squares) and best fit (red lines) plots of D8, D8_Mn,
IFP_D8 and IFP_D8_Mn with 5% Mn2+ at 50 °C.

168

Table 5.7 Best-fit 2H T2 (s) of IFP-bounded membrane with 5% Mn2+
25 °C

50 °C

D10

516 (14)

1937 (36)

D10_Mn

471 (2)

593 (18)

IFP_D10

385 (7)

1116 (42)

IFP_D10_Mn

408 (6)

578 (17)

D8

310 (2)

688 (21)

D8_Mn

243 (4)

259 (9)

IFP_D8

169 (3)

278 (6)

IFP_D8_Mn

163 (3)

244 (5)

Table 5.8 Best-fit 2H T2 relaxation rate (R2, kHz) of IFP-bounded membrane with 5% Mn2+
25 °C

50 °C

D10

1.94 (5)

0.52 (1)

D10_Mn

2.12 (1)

1.69 (5)

IFP_D10

2.60 (5)

0.90 (4)

IFP_D10_Mn

2.45 (4)

1.73 (5)

D8

3.23 (2)

1.45 (5)

D8_Mn

4.12 (7)

3.86 (14)

IFP_D8

5.92 (11)

3.60 (8)

IFP_D8_Mn

6.13 (12)

4.10 (9)

169

The D10 peak splitting were presented in Table 5.5. At 25 °C, the peak splitting of pure
D10 was 11.8 kHz, which is close to the 12.0 kHz peak splitting of D10_Mn. As discussed in
section 5.2.3, concentration of Mn2+ did not affect the D10 peak splitting at gel phase. However,
IFP_D10 had a 11.1 kHz peak splitting, which is ≈ 1 kHz narrower than D10. The IFP_D10_Mn
also showed this narrowing. As for 50 °C, there were slightly changings in methyl peak splitting,
but Mn2+ and IFP increased the methylene peak splitting of D10_Mn and IFP_D10 by 1.2 and 0.6
kHz, respectively. The trend of IFP_D10_Mn was not clear due to the methylene peak was not
well resolved.
Mn2+ ions and IFP decreased the peak splitting of D8 both 25 and 50 °C (Table 5.6). The
addition of Mn2+ to D8 membrane decrease the peak splitting from 48.1 to 36.3 kHz at 25 °C, and
from 26.1 to 24.6 at 50 °C. The IFP also reduced the splitting by 7.7 and 1.9 kHz at 25 and 50 °C
respectively. It can be concluded that IFP changes the order of the lipid acyl chain.
Figure 5.16 – 5.19 and Table 5.7 – 5.8 showed the best-fit plots, T2 and R2 values for the
eight samples at different temperature. For pure D10 lipid at 25 °C, it had a T2 relaxation rate (R2)
of 1.94 kHz, which is close to the R2 of D10_Mn. This suggests that the addition of Mn2+ does not
affect the D10 2H’s T2 relaxation rate. As for IFP bounded membrane, the IFP_D10 and
IFP_D10_Mn also showed close R2 values. Therefore, both IFP and Mn2+ does not change the D10
2

H’s membrane location. The D8 had a R2 of ~ 3.2 kHz and D8_Mn had a R2 of ~ 4.1 kHz, which

shows that the addition of Mn2+ leads to a 0.7 kHz R2 increase. This increase is due to the spinnuclear coupling between paramagnetic election in Mn2+ and the D8 2H’s. IFP_D8 and
IFP_D8_Mn both had a R2 of ~ 6 kHz. By comparing the D8 data, IFP induced less R2 increase,
which indicates the average distance between IFP bounded D8 2H’s and Mn2+ is longer compare

170

to IFP free D8. Thus, the results at gel phase showed IFP does not decrease the average distance
between D10 and D8 2H’s and aqueous layer and are inconsistent with lipid protrusion model.
At fluid phase, the effect of Mn2+ on D8 and D10 also has similar trends. D10 has a R2 of
0.52 kHz and the addition of Mn2+ increase the R2 by 1.2 kHz. Since at liquid phase, the lipid acyl
chain is not packed as ordered at gel phase and has more motion, there is a greater fraction of D10
2

H’s locate closer the membrane surface compare to gel phase. Thus, the R2 increases when Mn2+

is introduced, while D10 and D10_Mn have similar R2 value at gel phase. IFP bounded membrane
with Mn2+ showed a 0.8 kHz increase in R2 compare to IFP_D10, and this increase is smaller than
the 1.2 kHz of D10 and D10_Mn. On the other hand, Mn2+ increase the R2 of pure D8 lipid by 2.4
kHz, from 2.1 to 5.6 kHz, and 0.5 kHz for IFP-bounded D8. The D8 also showed a smaller increase
of IFP_D8 than D8. Since the IFP-bounded D10 and D8 did not have a greater R2 increase with
and without Mn2+ compared to pure membrane, the IFP does not decrease the average distance
between D10 or D8 deuterium atoms and the aqueous layer. Thus, it can be concluded that the IFP
does not induce lipid tail protrusion at both gel phase and liquid phase.
As for the reproducibility, the pure D8 membrane at 25 °C T2 was measured with different
1 and 2 values, and the fitting was shown in Figure 5.20. The result R2 were 3.23 and 3.53 kHz,
respectively, with a ~ 0.3 kHz difference.

171

Figure 5.20 Reproducibility of D8 at 25°C. The quecho experimental (black squares) and best fit
(red lines) plots as well as the fitted T2 and R2 values are shown.

5.2.5 T2 of IFP-bounded membrane with 20% Mn2+
The spectra of pure D10, D10 with 20% Mn2+, IFP-bounded D10, and IFP-bounded D10
Mn2+ (denoted as D10, D10_Mn’, IFP_D10, and FIP_D10_Mn’) at 25 or 50 °C are shown in
Figure 5.21, and the fitting of D10_Mn’ and FIP_D10_Mn’ at both temperatures are shown in
Figure 5.22. The peak splitting, fitted T2 and R2 are shown in Table 5.9 – 5.11.
At gel phase, the addition 20% Mn2+ did not change the line shape or peak splitting of the
spectra, which is consistent with the 5% Mn2+ results. However, at liquid phase, the 20% Mn2+
greatly affected the spectra with a smaller peak splitting of both methyl and methylene groups.
With both 20% Mn2+ and IFP, the doublet was even not well resolved.
The D10 at 25 °C showed a R2 of D10 ≈ R2 of D10_Mn’ and R2 of IFP_D10 ≈ R2 of
IFP_D10_Mn’. At 50 °C the addition of Mn2+ increase the R2 by ~3.6 kHz in D10 and ~2.6 kHz
in IFP bounded D10. This trend is consistent with the 5% Mn2+ data and both show no IFP induced
lipid tail protrusion at gel phase or liquid phase.

172

D10

D10 Mn’

IFP D10

IFP D10 Mn’

Figure 5.21 2H NMR spectra of D10 samples at 25 and 50 °C with 20% Mn2+.

173

Figure 5.22 Quecho experimental (black squares) and best fit (red lines) plots of D10_Mn’ and
IFP_D10_Mn’ with 20% Mn2+ at 25 °C and 50 °C.

174

Table 5.9 IFP-bounded D10 membrane 2H NMR spectra peak splitting (kHz) with 20% Mn2+
25 °C

50 °C

Methyl -CD3

Methyl -CD3

Methylene -CD2

D10

11.8

2.6

9.8

D10_Mn’

11.9

2.1

9.2

IFP_D10

11.1

2.7

10.4

IFP_D10_Mn’

11.0

-

-

Table 5.10 Best-fit 2H T2 (s) of IFP-bounded membrane with 20% Mn2+
25 °C

50 °C

D10

516 (14)

1937 (36)

D10_Mn’

477 (10)

245 (5)

IFP_D10

385 (7)

1116 (42)

IFP_D10_Mn’

355 (11)

283 (7)

Table 5.11 Best-fit 2H T2 relaxation rate (R2, in unit kHz) of IFP-bounded membrane with 20%
Mn2+
25 °C

50 °C

D10

1.94 (5)

0.52 (1)

D10_Mn’

2.10 (4)

4.08 (9)

IFP_D10

2.60 (5)

0.90 (4)

IFP_D10_Mn’

2.82 (9)

3.53 (9)

175

5.2.6 IFP bounded membrane structure
The T2 relaxation rate of the D10 and D8 samples at both gel phase and liquid phase showed
that IFP does not promote lipid protrusion. This is concluded from the evidence that the Mn2+
induced R2 increase is the same for IFP free D10 and IFP bounded D10, and the Mn2+ induced R2
increase is the smaller for IFP bounded D8 than IFP free D8 at gel phase. At liquid phase, both
IFP bounded D8 or D10 showed smaller Mn2+ induced R2 increase than membrane only.
In Chapter 4, the IFP topology in membrane at gel phase was studied by REDOR solidstate NMR and the result showed close contacts between the lipid acyl chain terminus with Gly_2,
Ala_7 and Gly_16 residues at gel phase, which is consistent with either the membrane spanning
model or the lipid protrusion model (Figure 4.1). The membrane structure studied by PRE solidstate NMR experiments showed the IFP does not promote lipid protrusion at both gel phase and
liquid phase. Thus, the only possible model of IFP is the membrane spanning conformation. A
model of IFP topology in membrane is proposed by integrating the REDOR and PRE results, and
is shown in Figure 5.23. The red dots on the lipid acyl chain represents the positions of D10 and
D8 2H’s.
To my knowledge, the IFP membrane spanning conformation is only proposed by two
other groups and evidenced by simulation results done by Victor et al. and Worch et al. 14, 15 In
Victor and Worch’s simulation, the membrane and IFP were randomly distributed in a box of water
and allowed to spontaneously assemble. Even though this simulation process was not a mimic of
the actual infection, it was an efficient and unbiased way of determining the insertion mode of
membrane-interacting IFP. The results showed that the dominant IPF conformation is a deeply
buried helical hairpin, and it is the most probable and maybe the lowest free energy configuration.
There is one discrepancy in these two studies of the orientation of IFP. Victor et al. showed a 47°

176

angle between peptide N-terminal helix axis and the membrane plane normal, while Worch et al.
observed a parallel orientation.14, 15
There have been several studies for the IFP orientation in membrane from both our group
as well as other groups. Wasniewski from our group used solid-state NMR to study the 15N labeled
Ala7 and Phe3 IFP in oriented membrane. The NMR

15

N peak was sharp when the membrane

normal is parallel to the external magnetic field, and poor signal was obtained when the membrane
normal is perpendicular to the external magnetic field. His results suggested that the IFP adopts an
orientation approximately perpendicular to membrane normal.16 In Yan Sun’s dissertation, she
also presented the study of the IFP N-terminal helix orientation in bicelle with
DTPC:DMPC:DHPC 48:1:15 ratio by solid-state NMR. She observed that the IFP N-helix has a
~45° relative to membrane bicelle normal and this angle is independent of sample pH. This result
is consistent with the ATR-FTIR experimental data showing this angle to be 45° and 41° from
Luneberg et al. and Wu et al., respectively.17, 18 Thus, in Figure 5.23, the IFP is shown with a ~45°
tilt angle with respect to membrane normal, and this conformation is based on the MD simulation,
solid-state NMR and ATR-FTIR results discussed above.14, 17, 18

177

Figure 5.23 Model for IFP-bounded membrane. The lipid molecules are shown as blue sphere
with two lines, 2H atoms in lipid acyl chain shown as red dots, IFP in green, labeled amino acids
in orange and Mn2+ binds on membrane surface.

178

5.3 Conclusion
Secondary structure and conformation of IFP have been intensively studied in both DPC
micelle and lipid membrane.8, 19-22 It has come to an agreement that IFP adopts a helix-turn-helix
structure in membrane. In the recent study from our group, the IFP structure is independent of pH,
HA subtype and sequence length, and there are always two distinct conformation: closed and semiclosed.
However, there has been arguments about the IFP membrane location. In some of the
previous studies, the IFP is reported to be parallel to the membrane plane, and located near the
lipid-water interface.22-25 In this study, the REDOR and PRE solid-state NMR experiments were
designed to testify this hypothesis. The REDOR experiments discussed in Chapter 4 showed close
contacts between the lipid acyl chain tail with Gly_2, Ala_7 and Gly_16 residues. In this chapter,
PRE solid-state NMR was used to study the IFP bounded membrane structure.
Sample preparation methods, effect of Mn2+ concentration on T2 relaxation time and Mn2+
induced R2 of IFP bounded membrane were studied and discussed. Based on both 5% and 20%
Mn2+ PRE results, it can be concluded that IFP does not promote lipid protrusion at both gel phase
and liquid phase, which is evidenced by that the R2 difference with and without Mn2+ is smaller
for IFP free membrane than IFP bounded membrane, meaning IFP does not induced smaller the
average distance between lipid acyl chain and aqueous layer. By integrating these results, a IFPmembrane interaction model is proposed, in which the IFP deeply inserted in membrane
hydrophobic core and the N-terminal helix has a ~45° with respect to membrane normal.

179

REFERENCES

180

REFERENCES

[1] Solomon, I. (1955) RELAXATION PROCESSES IN A SYSTEM OF 2 SPINS, Physical
Review 99, 559-565.
[2] Clore, G. M., and Iwahara, J. (2009) Theory, Practice, and Applications of Paramagnetic
Relaxation Enhancement for the Characterization of Transient Low-Population States of
Biological Macromolecules and Their Complexes, Chem. Rev. 109, 4108-4139.
[3] Cai, S., Seu, C., Kovacs, Z., Sherry, A. D., and Chen, Y. (2006) Sensitivity enhancement of
multidimensional NMR experiments by paramagnetic relaxation effects, J. Am. Chem.
Soc. 128, 13474-13478.
[4] Buffy, J. J., Hong, T., Yamaguchi, S., Waring, A. J., Lehrer, R. I., and Hong, M. (2003)
Solid-state NMR investigation of the depth of insertion of protegrin-1 in lipid bilayers
using paramagnetic Mn2+, Biophys. J. 85, 2363-2373.
[5] Su, Y., Mani, R., and Hong, M. (2008) Asymmetric insertion of membrane proteins in lipid
bilayers by solid-state NMR paramagnetic relaxation enhancement: A cell-penetrating
peptide example, J. Am. Chem. Soc. 130, 8856-8864.
[6] Villalain, J. (1996) Location of cholesterol in model membranes by magic-angle-samplespinning NMR, Eur. J. Biochem. 241, 586-593.
[7] Huster, D., Yao, X. L., and Hong, M. (2002) Membrane protein topology probed by H-1 spin
diffusion from lipids using solid-state NMR spectroscopy, J. Am. Chem. Soc. 124, 874883.
[8] Han, X., Bushweller, J. H., Cafiso, D. S., and Tamm, L. K. (2001) Membrane structure and
fusion-triggering conformational change of the fusion domain from influenza
hemagglutinin, Nat. Struct. Biol. 8, 715-720.
[9] van Meer, G., Voelker, D. R., and Feigenson, G. W. (2008) Membrane lipids: where they are
and how they behave, Nat. Rev. Mol. Cell Biol. 9, 112-124.
[10] Mayer, L. D., Hope, M. J., Cullis, P. R., and Janoff, A. S. (1985) SOLUTE
DISTRIBUTIONS AND TRAPPING EFFICIENCIES OBSERVED IN FREEZETHAWED MULTILAMELLAR VESICLES, Biochimica Et Biophysica Acta 817, 193196.
[11] Hope, M. J., Bally, M. B., Mayer, L. D., Janoff, A. S., and Cullis, P. R. (1986)
GENERATION OF MULTILAMELLAR AND UNILAMELLAR PHOSPHOLIPIDVESICLES, Chem. Phys. Lipids 40, 89-107.
[12] Traikia, M., Warschawski, D. E., Recouvreur, M., Cartaud, J., and Devaux, P. F. (2000)
Formation of unilamellar vesicles by repetitive freeze-thaw cycles: characterization by

181

electron microscopy and P-31-nuclear magnetic resonance, Eur. Biophys. J. Biophys.
Lett. 29, 184-195.
[13] Gabrys, C. M., Yang, R., Wasniewski, C. M., Yang, J., Canlas, C. G., Qiang, W., Sun, Y.,
and Weliky, D. P. (2010) Nuclear magnetic resonance evidence for retention of a lamellar
membrane phase with curvature in the presence of large quantities of the HIV fusion
peptide, Biochim. Biophys. Acta-Biomembr. 1798, 194-201.
[14] Victor, B. L., Lousa, D., Antunes, J. M., and Soares, C. M. (2015) Self-Assembly Molecular
Dynamics Simulations Shed Light into the Interaction of the Influenza Fusion Peptide
with a Membrane Bilayer, J. Chem Inf. Model. 55, 795-805.
[15] Worch, R., Krupa, J., Filipek, A., Szymaniec, A., and Setny, P. (2017) Three conserved Cterminal residues of influenza fusion peptide alter its behavior at the membrane interface,
Biochim. Biophys. Acta-Gen. Subj. 1861, 97-105.
[16] Wasniewski, C. M., Parkanzky, P. D., Bodner, M. L., and Weliky, D. P. (2004) Solid-state
nuclear magnetic resonance studies of HIV and influenza fusion peptide orientations in
membrane bilayers using stacked glass plate samples, Chem. Phys. Lipids 132, 89-100.
[17] Wu, C. W., Cheng, S. F., Huang, W. N., Trivedi, V. D., Veeramuthu, B., Kantchev, A. B.,
Wu, W. G., and Chang, D. K. (2003) Effects of alterations of the amino-terminal glycine
of influenza hemagglutinin fusion peptide on its structure, organization and membrane
interactions, Biochim. Biophys. Acta-Biomembr. 1612, 41-51.
[18] Luneberg, J., Martin, I., Nussler, F., Ruysschaert, J. M., and Herrmann, A. (1995)
STRUCTURE AND TOPOLOGY OF THE INFLUENZA-VIRUS FUSION PEPTIDE
IN LIPID BILAYERS, J. Biol. Chem. 270, 27606-27614.
[19] Lorieau, J. L., Louis, J. M., and Bax, A. (2010) The complete influenza hemagglutinin
fusion domain adopts a tight helical hairpin arrangement at the lipid:water interface,
Proc. Natl. Acad. Sci. U. S. A. 107, 11341-11346.
[20] Lorieau, J. L., Louis, J. M., Schwieters, C. D., and Bax, A. (2012) pH-triggered, activatedstate conformations of the influenza hemagglutinin fusion peptide revealed by NMR,
Proc. Natl. Acad. Sci. U. S. A. 109, 19994-19999.
[21] Ghosh, U., Xie, L., and Weliky, D. P. (2013) Detection of closed influenza virus
hemagglutinin fusion peptide structures in membranes by backbone (CO)-C-13-N-15
rotational-echo double-resonance solid-state NMR, J. Biomol. NMR 55, 139-146.
[22] Ghosh, U., Xie, L., Jia, L. H., Liang, S., and Weliky, D. P. (2015) Closed and Semiclosed
Interhelical Structures in Membrane vs Closed and Open Structures in Detergent for the
Influenza Virus Hemagglutinin Fusion Peptide and Correlation of Hydrophobic Surface
Area with Fusion Catalysis, J. Am. Chem. Soc. 137, 7548-7551.

182

[23] Larsson, P., and Kasson, P. M. (2013) Lipid Tail Protrusion in Simulations Predicts
Fusogenic Activity of Influenza Fusion Peptide Mutants and Conformational Models,
PLoS Comput. Biol. 9, 9.
[24] Brice, A. R., and Lazaridis, T. (2014) Structure and Dynamics of a Fusion Peptide Helical
Hairpin on the Membrane Surface: Comparison of Molecular Simulations and NMR, J.
Phys. Chem. B 118, 4461-4470.
[25] Baylon, J. L., and Tajkhorshid, E. (2015) Capturing Spontaneous Membrane Insertion of the
Influenza Virus Hemagglutinin Fusion Peptide, J. Phys. Chem. B 119, 7882-7893.

183

Chapter 6
Summary and future directions

184

My research during the past five years has been focused on HIV gp41 and influenza
hemagglutinin (HA). These two proteins are responsible for the initial step of virus infection by
catalyzing the fusion between viral and host cell membrane. HIV infection happens on the surface
of immune cells. After the HIV glycoprotein gp120 binds to the host cell, gp41 is exposed and
fusion between host cell membrane and viral membrane starts. The fusion peptide (FP) and
transmembrane (TM) domains are the segments that binds to membrane and are critical for fusion.1
As for influenza virus, the virion is endocytosed by host cell and the pH drop inside endosome
triggers a conformational change in HA. The ~25 N-terminal residues of HA subunit 2 is fusion
peptide domain (IFP). IFP binds to endosomal membrane and is necessary for fusion.2 However,
the structure of gp41 and HA as well as the infection mechanism are still not fully understood. The
overall goal is to understand the gp41 and IFP structure and their interaction with membrane
Chapter 3 mainly discussed the structure and function of HIV gp41 fusion protein. A gp41
construct including the whole ectodomain and transmembrane domain, and three other shorter
constructs were expressed, purified and stabilized in physiology buffers containing SDS or DPC
detergents at neutral or low pH’s. The constructs adopt  helical SE and TM, and non-helical FP
in both detergents, and they are all hyperthermostable with Tm > 90 °C. The oligomeric states of
these proteins vary in different detergent buffer: predominant trimer for all constructs and some
hexamer fraction for HM and HM_TM protein in SDS at pH 7.4; and mixtures of monomer, trimer,
and higher-order oligomer protein in DPC at pH 4.0 and 7.4. Substantial protein-induced vesicle
fusion was observed, including fusion at the condition similar to HIV/cell fusion. The fusion
activity is aided by the membrane associate FP and TM domains, and by protein which is
predominantly trimer rather than monomer.

185

We proposed that there is no contact between FP and TM domains in detergent solutions
according to current CD data. This can be testified by further hydrogen-deuterium exchange
experiments.3 To perform the experiment, the gp41 proteins is mixed with D2O and incubate for a
certain time, then the H-D exchange is quenched by lowering the pH to ~2 and decrease
temperature to 0 °C. The deuterated protein is digested by pepsin and the digested peptides are
injected into mass spectrometry. The molecular weight of the peptides can be analyzed. If the FP
and TM formed a complex in the final SHB structure, the exchange rate of deuterons present in
the FP region of FP_HM_TM can be different from the exchange rate of deuterons present in the
FP region of FP_HM.
The FP and TM interaction in membrane also helps to understand the conformation of gp41
at the post-fusion state. Future studied can be done to determine the proximity of FP_HM_TM
incorporated in membrane using

13

C-15N REDOR solid-state NMR. The FP_HM_TM can be

obtained by native chemical ligation of 13C labeled FP peptide and 15N labeled HM_TM. The FP
peptide can be synthesized by solid-phase peptide synthesis with site-specific labeling, while the
HM_TM can be labeled with

15

N Phe residue by bacterial expression.4 There are three Phe in

HM_TM: one of them in MPER and two in TM. The Phe in MPER could be mutated to other
residues to avoid the FP and MPER interaction being observed. The result of a big dephasing
buildup of the 13C-15N REDOR indicates close contact between FP and TM, and small dephasing
buildup indicates no contact. The possible problem lies in the expression of

15

N Phe labeled

HM_TM and other residues might also get labeled due to E.coli scrambling. It is possible that
except for Phe, other residues are also labeled. If the yield of ligation experiment is high, it is worth
trying to ligate FP and TM to HM.

186

Previous studies of IFP were done in both detergent and membrane, and it has come to an
agreement that IFP adopts a helix-turn-helix structure in membrane.5, 6 The studies shown in
Chapter 4 and 5 are trying to illustrate the IFP membrane location. The 13C-2H REDOR solid-state
NMR results showed that the IFP adopts major  helical, minor  strand secondary structure in
PC/PG membrane. The Leu-2, Ala-7 and Gly-16 in  helical IFP all show close contacts with the
lipid acyl chain tail, suggesting IFP has strong interaction with the membrane. The PRE solid-state
NMR is used to study the IFP bounded lipid membrane structure. The T2 relaxation time and rate
were measured for membrane with or without IFP and with or without Mn2+. The results showed
not IFP induced lipid acyl chain protrusion at both gel phase and liquid phase, which is evidenced
by that the Mn2+ induced R2 increase is smaller or equal for IFP bounded membrane than IFP free
membrane. This suggest that the IFP does not induce a smaller the average distance between lipid
acyl chain and aqueous layer. By comparing the REDOR and PRE results to the existing IFP
topology models, a IFP membrane spanning model was proposed, in which IFP N-terminal helix
adopts a 45° angle with respect to membrane normal.
It will be interesting to study the IFP mutant membrane location. The IFP is highly
conserved and even a single site mutation might affect the IFP fusion activity. For example,
mutation of Gly_1 to Val, Glu, Gln or Lys completely eliminates the fusion.7 The secondary
structure, membrane location and peptide-membrane interaction might be different from the wild
type and can provides more information about the possible fusion process. The experiment can be
done by the 13C-2H REDOR experiment of site specifically labeled IFP mutant and deuterated lipid
membrane. The membrane structure perturbed by the IFP mutant can be studied by PRE with
deuterated lipids and Mn2+ ions.

187

It will also be meaningful to study the IFP in a larger construct and how the IFP interacts
with membrane. The large construct such as FHA2, which includes the soluble ectodomain SHA2
and fusion peptide, can be obtained by ligate 13C labeled FP peptide and unlabeled SHA2. The IFP
interaction with the soluble ectodomain is also very interesting. The 13C labeled FP peptide can be
ligated with deuterated SHA2. Compared to isolated IFP, these experiments are better mimic of
the actual infection.

188

APPENDICES

189

APPENDIX A

Location of NMR data
Figure 4.3
(a) /home/khare0/mb4b/data/Shuang/13C2H/IFP/L2C/031214 (L2c-D4)
(b) /home/khare0/mb4b/data/Shuang/13C2H/IFP/L2C/031814 (L2c-D8)
(c) /home/khare0/mb4b/data/Shuang/13C2H/IFP/L2C/032714 (L2c-D10)
Figure 4.4
(a) /home/khare0/mb4b/data/Shuang/13C2H/IFP/A7C/041614 (A7c-D4)
(b) /home/khare0/mb4b/data/Shuang/13C2H/IFP/A7C/050614 (A7c-D8)
(c) /home/khare0/mb4b/data/Shuang/13C2H/IFP/A7C/052114 (A7c-D10)
Figure 4.5
(a) /home/khare0/mb4b/data/Shuang/13C2H/IFP/G16C/111314 (G16c-D10)
(b) /home/khare0/mb4b/data/Shuang/13C2H/IFP/G16C/112414 (G16c-D8)
(c) /home/khare0/mb4b/data/Shuang/13C2H/IFP/G16C/120214 (G16c-D4)
Figure 4.6
(a) /home/hapi0/mb4b/data/Shuang/13C2H/IFP/IFP_040213 (L2c-D10)
(b) /home/hapi0/mb4b/data/Shuang/13C2H/IFP/IFP_091113 (L2c-D8)
Figure 5.3
/home/khare0/mb4b/data/Shuang/quecho/repeat_100316/D10/100316_25_T2
Figure 5.4
/home/khare0/mb4b/data/Shuang/quecho/repeat_100316/D10/101116_50_T2
Figure 5.5
/home/khare0/mb4b/data/Shuang/quecho/repeat_100316/D8/121216_25_T2

190

Figure 5.6
/home/khare0/mb4b/data/Shuang/quecho/repeat_100316/D8/120316_50_T2
Figure 5.8
(a) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/D10/100316_25_T2 (D10 25°C)
(b) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/D10/101116_50_T2 (D10 50°C)
(c) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/D8/121216_25_T2 (D8 25°C)
(d) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/D8/120316_50_T2 (D8 50°C)
Figure 5.10
(a) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/D10_Mn/101216_25_T2 (20%
Mn2+ 25°C)
(b) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/D10_Mn/103116_25_T2_5%
(5% Mn2+ 25°C)
(c) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/D10_Mn/020117_25_T2_1%
(1% Mn2+ 25°C)
(d) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/D10_Mn/032417_25_T2_0.2%
(0.2% Mn2+ 25°C)

191

Figure 5.11
(a) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/D10_Mn/102216_50_T2 (20%
Mn2+ 50°C)
(b) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/D10_Mn/110116_50_T2_5%
(5% Mn2+ 50°C)
(c) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/D10_Mn/020217_50_T2_1%
(1% Mn2+ 50°C)
(d) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/D10_Mn/031417_50_T2_0.2%
(0.2% Mn2+ 50°C)
Figure 5.12
/home/khare0/mb4b/data/Shuang/quecho/repeat_100316/IFP_D10/100616_25_T2
Figure 5.13
/home/khare0/mb4b/data/Shuang/quecho/repeat_101016/IFP_D10/100616_50_T2
Figure 5.16
(a) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/D10/100316_25_T2 (D10 25°C)
(b) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/D10_Mn/103116_25_T2
(D10_Mn 25°C)
(c) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/IFP_D10_Mn/100616_25_T2
(IFP_D10 25°C)
(d) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/IFP_D10_Mn/112916_25_T2
(IFP_D10_Mn 25°C)

192

Figure 5.17
(a) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/D10/101116_50_T2 (D10 50°C)
(b) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/D10_Mn/110116_50_T2
(D10_Mn 50°C)
(c) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/IFP_D10_Mn/101016_50_T2
(IFP_D10 50°C)
(d) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/IFP_D10_Mn/120116_50_T2
(IFP_D10_Mn 50°C)
Figure 5.18
(a) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/D8/121216_25_T2 (D8 25°C)
(b) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/D8_Mn/020717_25_T2
(D8_Mn 25°C)
(c) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/IFP_D8_Mn/112416_25_T2
(IFP_D8 25°C)
(d) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/IFP_D8_Mn/121616_25_T2
(IFP_D8_Mn 25°C)
Figure 5.19
(a) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/D8/120316_50_T2 (D8 50°C)
(b) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/D8_Mn/020816_50_T2
(D8_Mn 50°C)
(c) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/IFP_D8_Mn/112616_50_T2
(IFP_D8 50°C)
(d) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/IFP_D8_Mn/120416_50_T2
(IFP_D8_Mn 50°C)

193

Figure 5.20
(a) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/D8/121216_25_T2 (D8 25°C)
(b) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/D8/111716_25_T2 (D8 25°C
replica)
Figure 5.22
(a) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/D10_Mn/101216_25_T2
(D10_Mn’ 25°C)
(b) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/D10_Mn/102216_50_T2
(D10_Mn’ 50°C)
(c) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/IFP_D10_Mn/101716_25_T2
(IFP_D10_Mn’ 25°C)
(d) /home/khare0/mb4b/data/Shuang/quecho/repeat_100316/IFP_D10_Mn/102016_50_T2
(IFP_D10_Mn’ 50°C)

194

APPENDIX B

Solid-state NMR raw data
Table B1 IFP L7c in DPPC-D4 studied by REDRO raw data (Figure 4.3)
Dephasing

Error of

Error of
S0

time (s)

S1

error of S1

S/ S0

S0

S/ S0


2

9.1931

0.344

8.5472

0.211

0.07026

0.04168

8

15.7928

0.337

15.0415

0.331

0.04757

0.02919

16

17.1303

0.441

14.6836

0.438

0.14283

0.03377

24

6.4586

0.419

4.9174

0.429

0.23863

0.08278

32

15.4763

0.666

11.2919

0.582

0.27037

0.04899

40

7.0760

0.521

5.7963

0.407

0.18085

0.08334

48

12.7943

0.757

9.7390

0.680

0.23880

0.06966


2

16.07

0.344

14.6996

0.211

0.085276914

0.02358

8

39.8608

0.337

32.8811

0.331

0.175101854

0.01084

16

45.6214

0.441

37.2641

0.438

0.183188153

0.01243

24

15.8156

0.419

12.4953

0.429

0.209938289

0.03426

32

37.3028

0.666

28.8126

0.582

0.227602218

0.02082

40

19.622

0.521

14.6103

0.407

0.255412292

0.02865

48

35.483

0.757

27.5546

0.68

0.223442212

0.02533

195

Table B2 IFP L7c in DPPC-D8 studied by REDRO raw data (Figure 4.3)
Dephasing

Error of
S0

Error of S0

S1

error of S1

S/ S0

time (s)

S/ S0


2

16.1121

0.374

16.0004

0.288

0.0069

0.0292

8

12.4556

0.396

12.1574

0.273

0.0239

0.0380

16

15.8868

0.426

14.5082

0.304

0.0868

0.0311

24

14.1491

0.552

12.2086

0.445

0.1371

0.0461

32

22.4405

0.848

16.6942

0.722

0.2561

0.0427

40

14.8187

0.794

10.4636

0.448

0.2939

0.0484

48

16.2268

0.811

10.4543

0.444

0.3557

0.0423


2

21.078

0.374

20.7092

0.288

0.0175

0.0221

8

17.9311

0.396

16.7442

0.273

0.0662

0.0256

16

25.1182

0.426

21.3658

0.304

0.1494

0.0188

24

24.7114

0.552

19.5980

0.445

0.2069

0.0253

32

34.3695

0.848

24.8544

0.722

0.2768

0.0276

40

26.5748

0.794

18.3733

0.448

0.3086

0.0267

48

27.5736

0.811

17.7389

0.444

0.3567

0.0248

196

Table B3 IFP L7c in DPPC-D10 studied by REDRO raw data (Figure 4.3)
Dephasing

Error of
S0

Error of S0

S1

error of S1

S/ S0

time (s)

S/ S0


2

65.2858

0.695

63.6866

0.775

0.0245

0.0158

8

35.2148

0.439

26.6633

0.630

0.2428

0.0202

16

21.8837

0.507

11.6344

0.598

0.4684

0.0300

24

33.8663

0.849

14.1174

0.704

0.5831

0.0233

32

30.6851

0.934

8.0556

0.616

0.7375

0.0216

40

28.1228

0.836

4.9677

0.975

0.8234

0.0351

48

14.1873

0.679

2.9976

0.907

0.7887

0.0647


2

41.0415

0.695

40.4178

0.775

0.0152

0.0252

8

24.2896

0.439

21.1622

0.630

0.1288

0.0303

16

13.6729

0.507

11.0643

0.598

0.1908

0.0530

24

21.7212

0.849

17.0868

0.704

0.2134

0.0447

32

19.8017

0.934

12.6815

0.616

0.3596

0.0434

40

16.9414

0.836

12.0296

0.975

0.2899

0.0674

48

10.7682

0.679

5.8910

0.907

0.4529

0.0910

197

Table B4 IFP A7c in DPPC-D4 studied by REDRO raw data (Figure 4.4)
Dephasing

Error of

Error of
S0

time (s)

S1

error of S1

S/ S0

S0

S/ S0


2

32.4448

1.220

31.9501

1.224

0.01525

0.05286

8

20.7829

0.381

18.8372

0.500

0.09362

0.02924

16

30.6827

0.891

28.0396

1.076

0.08614

0.04398

24

29.6244

0.754

26.2857

0.635

0.11270

0.03114

32

33.4121

0.911

30.8848

1.256

0.07564

0.04526

40

20.4066

0.566

18.7614

0.566

0.08062

0.03768

48

21.8580

0.739

18.5341

0.622

0.15207

0.04039


2

106.8569

1.220

97.4845

1.224

0.087709825

0.01548

8

73.9308

0.381

62.6787

0.500

0.152197731

0.00805

16

98.2642

0.891

84.3455

1.076

0.141645686

0.01343

24

110.911

0.754

94.9251

0.635

0.144132683

0.00816

32

108.6387

0.911

91.2165

1.256

0.160368267

0.01354

40

79.2688

0.566

67.6721

0.566

0.146295894

0.00939

48

96.2936

0.739

84.2791

0.622

0.124769455

0.00932

198

Table B5 IFP A7c in DPPC-D8 studied by REDRO raw data (Figure 4.4)
Dephasing

Error of

Error of
S0

time (s)

S1

error of S1

S/ S0

S0

S/ S0


2

49.7243

0.676

48.7030

0.449

0.02054

0.01609

8

83.8645

0.762

79.6153

0.593

0.05067

0.01115

16

87.1904

0.894

79.4545

0.580

0.08872

0.01147

24

36.4263

0.573

31.0494

0.436

0.14761

0.01797

32

41.2316

0.323

34.0730

0.432

0.17362

0.01232

40

33.4359

0.521

26.2444

0.450

0.21508

0.01819

48

37.6289

0.855

30.4702

0.430

0.19024

0.02166


2

55.9898

0.676

54.2555

0.449

0.030975285

0.01418

8

83.9161

0.762

75.6585

0.593

0.098403048

0.01081

16

82.008

0.894

69.3302

0.580

0.154592235

0.01162

24

30.8507

0.573

24.5143

0.436

0.205389181

0.02043

32

35.8539

0.323

26.9150

0.432

0.249314579

0.01382

40

28.0622

0.521

19.5341

0.450

0.303899908

0.02060

48

31.482

0.855

20.5213

0.430

0.348157677

0.02236

199

Table B6 IFP A7c in DPPC-D10 studied by REDRO raw data (Figure 4.4)
Dephasing

Error of

Error of
S0

time (s)

S1

error of S1

S/ S0

S0

S/ S0


2

26.6149

0.494

26.9447

0.394

-0.01239

0.02392

8

35.6549

0.405

28.3532

0.515

0.20479

0.01704

16

36.4014

0.444

22.7760

0.447

0.37431

0.01446

24

38.2096

0.591

18.1990

0.440

0.52371

0.01367

32

31.6442

0.700

11.8022

0.786

0.62703

0.02617

40

29.5316

0.930

9.2844

0.536

0.68561

0.02067

48

28.7216

0.585

7.1254

0.597

0.75191

0.02139


2

38.4733

0.494

38.3586

0.394

0.002981288

0.01639

8

46.8079

0.405

38.6348

0.515

0.174609414

0.01312

16

44.6926

0.444

31.9632

0.447

0.284821201

0.01227

24

45.2512

0.591

27.5809

0.440

0.390493512

0.01257

32

37.1126

0.700

20.7705

0.786

0.440338322

0.02366

40

33.4251

0.930

16.8915

0.536

0.494646239

0.02133

48

32.1653

0.585

11.2943

0.597

0.648866947

0.01963

200

Table B7 IFP G16c in DPPC-D4, DPPC-D8 and DPPC-D10 studied by REDRO raw data
(Figure 4.5)
Dephasing

Error of

Error of
S0

time (s)

S1

error of S1

S/ S0

S0

S/ S0

D4
2

56.0648

0.745

49.3945

0.509

0.1190

0.015

8

98.5737

1.214

77.5882

0.843

0.2129

0.013

16

122.5577

1.676

92.4948

1.358

0.2453

0.015

24

99.5157

1.134

70.2728

1.115

0.2939

0.014

32

71.7066

0.751

48.0781

0.847

0.3295

0.014

40

81.6746

1.721

52.4972

1.411

0.3572

0.022

48

72.1971

2.262

44.8710

1.380

0.3785

0.027

D8
2

112.9807

1.277

111.6880

1.554

0.0114

1.554

8

78.7638

1.295

72.4339

1.110

0.0804

1.110

16

134.6262

1.054

110.6980

1.673

0.1777

1.673

24

82.9444

1.573

62.2264

0.835

0.2498

0.835

32

73.8451

1.162

49.5225

0.932

0.3294

0.932

40

78.8261

1.347

46.5444

1.210

0.4095

1.210

48

54.4701

1.357

29.4906

2.056

0.4586

2.056

0.768

0.0046

0.0139

D10
2

67.3786

0.544

67.0672

201

Table B7 (cont’d)
8

200.2962

1.524

172.6975

1.769

0.1378

0.0110

16

154.4760

1.964

111.3407

1.682

0.2792

0.0142

24

73.3691

1.227

44.2168

1.721

0.3973

0.0255

32

79.8731

1.820

38.4816

1.493

0.5182

0.0217

40

81.1137

1.483

29.2690

2.181

0.6392

0.0277

48

45.4283

1.254

14.3887

1.796

0.6833

0.0405

202

Table B8 IFP L2c in pure DPPC-D8 and DPPC-D10 studied by REDRO raw data (Figure 4.6)
Dephasing

Error of

Error of
S0

time (s)

S1

error of S1

S/ S0

S0

S/ S0

D8
2

55.3897

0.483

53.8299

0.546

0.02816

0.01300

8

57.2080

0.647

46.9816

0.565

0.17876

0.01356

16

34.3314

0.719

21.0580

0.510

0.38663

0.01964

24

30.6357

0.757

14.8135

0.809

0.51646

0.02898

32

20.1005

1.000

6.6848

0.958

0.66743

0.05045

40

18.1662

0.840

5.3939

0.574

0.70308

0.03445

48

12.0504

1.040

2.5912

0.866

0.78497

0.07422

D10
2

37.1279

0.360

37.0288

0.264

0.00267

0.01200

8

57.5783

0.598

55.1622

0.438

0.04196

0.01252

16

39.0030

0.364

36.3823

0.331

0.06719

0.01216

24

39.2912

0.522

35.8535

0.442

0.08749

0.01654

32

43.3655

0.779

38.0072

0.653

0.12356

0.02179

40

56.7507

0.812

50.0904

0.756

0.11736

0.01836

48

47.8901

0.821

39.9486

1.178

0.16583

0.02845

203

Table B9 D10 with various concentration of Mn2+ at 25 °C studied by PRE (Figure 5.10)
1

FID

2

1

t

FID

2

t

intensity

intensity

20%

5%

30

11

73

16741177

30

11

73

25351580

80

61

173

13945282

80

61

173

22215764

130

111

273

10734524

130

111

273

18618496

180

161

373

8184182

180

161

373

15270932

230

211

473

6424096

230

211

473

12326184

280

261

573

5254574

280

261

573

9971276

330

311

673

4295137

330

311

673

7971368

380

361

773

3641079

380

361

773

6485960

430

411

873

2973930

430

411

873

5196364

480

461

973

2472762

480

461

973

4233112

530

511

1073

1995562

530

511

1073

3442028

580

561

1173

1728693

580

561

1173

2793020

630

611

1273

1317678

630

611

1273

2176876

680

661

1373

1163524

680

661

1373

1782552

1%

0.2%

30

11

73

15349736

30

11

73

2698428

80

61

173

13316156

80

61

173

2380908

130

111

273

11215504

130

111

273

2131156

204

Table B9 (cont’d)
180

161

373

9158632

180

161

373

1798660

230

211

473

7401352

230

211

473

1507128

280

261

573

5819964

280

261

573

1268288

330

311

673

4591952

330

311

673

1032948

380

361

773

3657716

380

361

773

834864

430

411

873

2909052

430

411

873

632496

480

461

973

2238292

480

461

973

542252

530

511

1073

1720156

530

511

1073

458456

580

561

1173

1337784

580

561

1173

369484

630

611

1273

1045780

630

611

1273

328884

680

661

1373

926568

680

661

1373

241948

205

Table B10 D10 with various concentration of Mn2+ at 50 °C studied by PRE (Figure 5.11)
1

FID

2

1

t

FID

2

t

intensity

intensity

20%

5%

30

11

73

14691436

30

11

71

20660388

60

41

133

13212680

80

61

171

17831868

90

71

193

10819840

130

111

271

14636092

120

101

253

8627060

180

161

371

11931888

150

131

313

6565180

230

211

471

9883984

180

161

373

4996032

280

261

571

8210824

210

191

433

3810804

330

311

671

6839976

240

221

493

2813868

380

361

771

5855812

270

251

553

2169360

430

411

871

4972756

300

281

613

1846720

480

461

971

4448160

330

311

673

1494368

530

511

1071

3905876

360

341

733

1236584

580

561

1171

3421528

390

371

793

1138232

630

611

1271

2957996

420

401

853

891008

680

661

1371

2635640

1%

0.2%

30

11

73

16205600

30

11

73

2698428

80

61

173

14598764

80

61

173

2380908

130

111

273

13389080

130

111

273

2131156

206

Table B10 (cont’d)
180

161

373

11880292

180

161

373

1798660

230

211

473

10761116

230

211

473

1507128

280

261

573

9770600

280

261

573

1268288

330

311

673

9035032

330

311

673

1032948

380

361

773

8015876

380

361

773

834864

430

411

873

7253204

430

411

873

632496

480

461

973

6605856

480

461

973

542252

530

511

1073

6191140

530

511

1073

458456

580

561

1173

5625868

580

561

1173

369484

630

611

1273

5300064

630

611

1273

328884

680

661

1373

4779708

680

661

1373

241948

207

Table B11 D10 at 25 °C studied by PRE (Figure 5.16)
1

FID

2

1

t

FID

2

t

intensity

intensity

D10

D10_Mn

30

11

73

22939148

30

11

73

25351580

80

61

173

21274456

80

61

173

22215764

130

111

273

18905892

130

111

273

18618496

180

161

373

16142057

180

161

373

15270932

230

211

473

13634104

230

211

473

12326184

280

261

573

11328846

280

261

573

9971276

330

311

673

9301617

330

311

673

7971368

380

361

773

7763542

380

361

773

6485960

430

411

873

6041013

430

411

873

5196364

480

461

973

4843347

480

461

973

4233112

530

511

1073

3912750

530

511

1073

3442028

580

561

1173

2934186

580

561

1173

2793020

630

611

1273

2366845

630

611

1273

2176876

680

661

1373

1896893

680

661

1373

1782552

IFP_D10

IFP_D10_Mn

30

11

73

23445574

30

11

73

23175468

80

61

173

19328814

80

61

173

19028900

130

111

273

15634267

130

111

273

14593728

208

Table B11 (cont’d)
180

161

373

12078985

180

161

373

11093240

230

211

473

9278228

230

211

473

8564448

280

261

573

7090318

280

261

573

6690932

330

311

673

5377061

330

311

673

5133752

380

361

773

3980167

380

361

773

4107736

430

411

873

3159236

430

411

873

3417952

480

461

973

2316542

480

461

973

2700456

530

511

1073

1747176

530

511

1073

2219808

580

561

1173

1311151

580

561

1173

1763332

630

611

1273

878248

630

611

1273

1515724

680

661

1373

716006

680

661

1373

1178916

209

Table B12 D10 at 50 °C studied by PRE (Figure 5.17)
1

FID

2

1

t

FID

2

t

intensity

intensity

D10

D10_Mn

30

11

73

20429904

30

11

71

20660388

180

161

373

13719244

80

61

171

17831868

330

311

673

10529842

130

111

271

14636092

480

461

973

8826575

180

161

371

11931888

630

611

1273

7400605

230

211

471

9883984

780

761

1573

6286838

280

261

571

8210824

930

911

1873

5476900

330

311

671

6839976

1080

1061

2173

4598277

380

361

771

5855812

1230

1211

2473

4091422

430

411

871

4972756

1380

1361

2773

3490750

480

461

971

4448160

1530

1511

3073

3099347

530

511

1071

3905876

1680

1661

3373

2597324

580

561

1171

3421528

1830

1811

3673

2302469

630

611

1271

2957996

1980

1961

3973

2087342

680

661

1371

2635640

IFP_D10

IFP_D10_Mn

30

11

71

23213666

30

11

73

23224660

180

161

371

17338260

80

61

173

19154584

330

311

671

12472822

130

111

273

15390656

210

Table B12 (cont’d)
480

461

971

9379849

180

161

373

12821624

630

611

1271

7100031

230

211

473

10820300

780

761

1571

5680466

280

261

573

9088668

930

911

1871

4467065

330

311

673

7790152

1080

1061

2171

3699964

380

361

773

6816512

1230

1211

2471

3080141

430

411

873

6083032

1380

1361

2771

2630065

480

461

973

5225736

1530

1511

3071

2197503

530

511

1073

4604304

1680

1661

3371

1979658

580

561

1173

3897424

1830

1811

3671

1723330

630

611

1273

3619868

1980

1961

3971

1508588

680

661

1373

3472196

211

Table B13 D8 at 25 °C studied by PRE (Figure 5.18)
1

FID

2

1

t

FID

2

t

intensity

intensity

D8

D8_Mn

30

11

73

18979972

30

11

73

11574004

70

51

153

16777240

50

31

113

10196976

110

91

233

13911604

70

51

153

8835052

150

131

313

10937824

90

71

193

7362160

190

171

393

8483648

110

91

233

6240424

230

211

473

6575180

130

111

273

5226200

270

251

553

4970600

150

131

313

4401616

310

291

633

3794728

170

151

353

3733872

350

331

713

2939836

190

171

393

3079276

390

371

793

2267060

210

191

433

2653260

430

411

873

1710868

230

211

473

2231720

470

451

953

1355856

250

231

513

1849952

510

491

1033

1055080

270

251

553

1571860

550

531

1113

763484

290

271

593

1325920

IFP_D8

IFP_D8_Mn

30

11

73

12914244

30

11

73

6839532

50

31

113

10577272

40

21

93

5753080

70

51

153

8350124

50

31

113

5323424

212

Table B13 (cont’d)
90

71

193

6890136

60

41

133

4575368

110

91

233

5205524

70

51

153

4000256

130

111

273

4006824

80

61

173

3575728

150

131

313

3083364

90

71

193

3069020

170

151

353

2477144

100

81

213

2829832

190

171

393

1932796

110

91

233

2362168

210

191

433

1444700

120

101

253

2328236

230

211

473

1081860

130

111

273

1952584

250

231

513

828176

140

121

293

1810428

270

251

553

646252

150

131

313

1692552

290

271

593

547804

160

141

333

1401204

213

Table B14 D8 at 50 °C studied by PRE (Figure 5.19)
1

FID

2

1

t

FID

2

t

intensity

intensity

D8

D8_Mn

30

11

73

18297716

30

11

73

11284620

80

61

173

15195036

50

31

113

9828736

130

111

273

11656272

70

51

153

8352444

180

161

373

9759664

90

71

193

6843708

230

211

473

8197936

110

91

233

5652852

280

261

573

6931136

130

111

273

4777816

330

311

673

5993344

150

131

313

4096996

380

361

773

5302892

170

151

353

3857288

430

411

873

4594732

190

171

393

3393380

480

461

973

4150888

210

191

433

3050220

530

511

1073

3699508

230

211

473

2728400

580

561

1173

3100776

250

231

513

2376980

630

611

1273

2909236

270

251

553

2285012

680

661

1373

2614820

290

271

593

2203524

IFP_D8

IFP_D8_Mn

30

11

73

11008496

30

11

73

14011488

60

41

133

8294992

50

31

113

11421860

90

71

193

6992826

70

51

153

10259204

214

Table B14 (cont’d)
120

101

253

5824356

90

71

193

8121512

150

131

313

4443080

110

91

233

7090256

180

161

373

3288032

130

111

273

5868476

210

191

433

2868336

150

131

313

5026988

240

221

493

2434644

170

151

353

4249668

270

251

553

1960872

190

171

393

3619308

300

281

613

1614376

210

191

433

3141236

330

311

673

1341708

230

211

473

2861672

360

341

733

1080152

250

231

513

2471484

390

371

793

978816

270

251

553

2148172

420

401

853

831404

290

271

593

1961748

215

Table B15 D8 reproducibility at 25 °C studied by PRE (Figure 5.20)
1

FID

2

1

t

2

FID
t

intensity

intensity

D8

D8_replicate

30

11

73

18297716

30

11

73

14017444

80

61

173

15195036

80

61

173

11658036

130

111

273

11656272

130

111

273

8644612

180

161

373

9759664

180

161

373

6239496

230

211

473

8197936

230

211

473

4303348

280

261

573

6931136

280

261

573

3077584

330

311

673

5993344

330

311

673

2068132

380

361

773

5302892

380

361

773

1385992

430

411

873

4594732

430

411

873

1012780

480

461

973

4150888

480

461

973

709404

530

511

1073

3699508

530

511

1073

493972

580

561

1173

3100776

580

561

1173

446048

630

611

1273

2909236

630

611

1273

315104

680

661

1373

2614820

680

661

1373

215124

216

Table B16 D10 with 20% Mn2+ studied by PRE (Figure 5.22)
1

2

FID

1

t

2

FID
t

intensity

intensity

D10_Mn’ at 25 °C

IFP_D10_Mn’ at 25 °C

30

11

73

16741177

30

11

73

20422396

80

61

173

13945282

80

61

173

15695484

130

111

273

10734524

130

111

273

11204164

180

161

373

8184182

180

161

373

8035720

230

211

473

6424096

230

211

473

5927740

280

261

573

5254574

280

261

573

4633164

330

311

673

4295137

330

311

673

3916324

380

361

773

3641079

380

361

773

3123200

430

411

873

2973930

430

411

873

2584748

480

461

973

2472762

480

461

973

2130140

530

511

1073

1995562

530

511

1073

1679304

580

561

1173

1728693

580

561

1173

1335496

630

611

1273

1317678

630

611

1273

1149484

680

661

1373

1163524

680

661

1373

896388

D10_Mn’ at 50 °C

IFP_D10_Mn’ at 50 °C

30

11

73

14691436

30

11

73

18792088

60

41

133

13212680

60

41

133

16079740

90

71

193

10819840

90

71

193

13179672

217

Table B16 (cont’d)
120

101

253

8627060

120

101

253

10427644

150

131

313

6565180

150

131

313

8130728

180

161

373

4996032

180

161

373

6341088

210

191

433

3810804

210

191

433

5036468

240

221

493

2813868

240

221

493

4125032

270

251

553

2169360

270

251

553

3381408

300

281

613

1846720

300

281

613

2802932

330

311

673

1494368

330

311

673

2594356

360

341

733

1236584

360

341

733

2240276

390

371

793

1138232

390

371

793

2142636

420

401

853

891008

420

401

853

1885260

218

APPENDIX C

Additional PRE data
In this section, the PRE data of samples prepared by Method 1 as described in Chapter 5
and with 20% Mn2+ are shown. Since the amount of Mn2+ that binds to membrane surface is
unknown, the data is not reliable. Besides, the 2H chanel is not on resonance, the data also showed
fluctuation in echo intensity. All the samples contains 20 mole DPPC-D10 or DPPC-D8, 5 mole
DPPG. The IFP bounded membrane samples contains 1 mole IFP. The 20% Mn2+ are introduced
before extrusion.

Table C1 D10 samples at 25 °C
1

FID

2

1

t

FID

2

t

intensity

intensity

D10

D10_Mn

100

50

212

2427552

100

50

212

1194804

120

70

252

2012156

120

70

252

979348

140

90

292

1854984

140

90

292

915672

160

110

332

1593588

160

110

332

962132

180

130

372

1370424

180

130

372

901556

200

150

412

1259636

200

150

412

714420

220

170

452

1112872

220

170

452

691116

240

190

492

951628

240

190

492

588460

219

Table C1 (cont’d)
260

210

532

905464

260

210

532

585264

280

230

572

797932

280

230

572

483384

IFP_D10

IFP_D10_Mn

100

50

212

580796

100

50

212

1588944

120

70

252

333272

120

70

252

1158172

140

90

292

208024

140

90

292

844656

160

110

332

119904

160

110

332

846028

180

130

372

76492

180

130

372

947076

200

150

412

13092

200

150

412

921324

220

170

452

16912

220

170

452

1008936

240

190

492

1580

240

190

492

1061484

260

210

532

-8500

260

210

532

1054540

280

230

572

-14516

280

230

572

1016268

File locations of data shown in this table are:
(d) /home/khare0/mb4b/data/Shuang/quecho/D10/121514_array_25 (D10)
(e) /home/khare0/mb4b/data/Shuang/quecho/D10_Mn/122214_array_25 (D10_Mn)
(f) /home/khare0/mb4b/data/Shuang/quecho/IFP_D10/123014_array_25 (IFP_D10)
(g) /home/khare0/mb4b/data/Shuang/quecho/IFP_D10_Mn/010615_array_25
(IFP_D10_Mn)

220

Table C2 D10 samples at 50 °C
1

FID

2

1

t

FID

2

t

intensity

intensity

D10

D10_Mn

100

50

212

1638944

100

50

212

842488

120

70

252

1265572

120

70

252

1043264

140

90

292

977224

140

90

292

970648

160

110

332

787708

160

110

332

879260

180

130

372

661260

180

130

372

892140

200

150

412

527400

200

150

412

747548

220

170

452

415532

220

170

452

639164

240

190

492

369428

240

190

492

590300

260

210

532

338308

260

210

532

445184

280

230

572

235244

280

230

572

424200

IFP_D10

IFP_D10_Mn

100

50

212

451708

100

50

204

1845600

120

70

252

876512

120

70

244

1590460

140

90

292

898724

140

90

284

1433212

160

110

332

777956

160

110

324

1016560

180

130

372

651280

180

130

364

835572

200

150

412

628744

200

150

404

840264

220

170

452

605104

220

170

444

862324

221

Table C2 (cont’d)
240

190

492

398048

240

190

484

885156

260

210

532

240316

260

210

524

941796

280

230

572

117136

280

230

564

955076

File locations of data shown in this table are:
(a) /home/khare0/mb4b/data/Shuang/quecho/D10/121614_array_50 (D10)
(b) /home/khare0/mb4b/data/Shuang/quecho/D10_Mn/122714_array_50 (D10_Mn)
(c) /home/khare0/mb4b/data/Shuang/quecho/IFP_D10/123114_array_50 (IFP_D10)
(d) /home/khare0/mb4b/data/Shuang/quecho/IFP_D10_Mn/010715_array_50
(IFP_D10_Mn)

222

Table C3 D8 samples at 25 °C
1

FID

2

1

t

FID

2

t

intensity

intensity

D8

D8_Mn

100

50

212

3327919

100

50

212

2228917

120

70

252

2961711

120

70

252

1806647

140

90

292

2621768

140

90

292

1458881

160

110

332

2257266

160

110

332

1258533

180

130

372

1924261

180

130

372

992903

200

150

412

1731391

200

150

412

871658

220

170

452

1573543

220

170

452

788846

240

190

492

1312340

240

190

492

699777

260

210

532

1100150

260

210

532

528551

280

230

572

968456

280

230

572

488626

IFP_D8

IFP_D8_Mn

100

50

212

5243646

100

50

212

3119518

120

70

252

4598274

120

70

252

2432473

140

90

292

4009754

140

90

292

1999605

160

110

332

3486237

160

110

332

1608763

180

130

372

2977790

180

130

372

1350645

200

150

412

2560767

200

150

412

1172473

220

170

452

2196623

220

170

452

1001531

223

Table C3 (cont’d)
240

190

492

1879294

240

190

492

871345

260

210

532

1582613

260

210

532

683008

280

230

572

1370021

280

230

572

579342

300

250

612

519509

320

270

652

448488

340

290

692

390976

File locations of data shown in this table are:
(a) /home/hapi0/mb4c/data/Shuang/quecho/D8/012015_array_25 (D8)
(b) /home/hapi0/mb4c/data/Shuang/quecho/D8_Mn/012215_array_25 (D8_Mn)
(c) /home/hapi0/mb4c/data/Shuang/quecho/IFP_D8/012515_array_25 (IFP_D8)
(d) /home/hapi0/mb4c/data/Shuang/quecho/IFP_D8_Mn/013015_array_25 (IFP_D8_Mn)

224

Table C4 D8 samples at 50 °C
1

FID

2

1

t

FID

2

t

intensity

intensity

D8

D8_Mn

100

50

212

1891933

100

50

212

2591942

120

70

252

1305385

120

70

252

1684197

140

90

292

889396

140

90

292

1121416

160

110

332

644175

160

110

332

729318

180

130

372

502305

180

130

372

577378

200

150

412

349133

200

150

412

421690

220

170

452

274068

220

170

452

268906

240

190

492

257280

240

190

492

285795

260

210

532

147527

260

210

532

265650

280

230

572

175812

280

230

572

201262

IFP_D8

IFP_D8_Mn

100

50

212

2496467

100

50

212

2600675

120

70

252

1992008

120

70

252

1581667

140

90

292

1563270

140

90

292

1063378

160

110

332

1295604

160

110

332

681049

180

130

372

1080933

180

130

372

491643

200

150

412

927278

200

150

412

347409

220

170

452

792277

220

170

452

250122

225

Table C4 (cont’d)
240

190

492

652293

240

190

492

243333

260

210

532

573384

260

210

532

238036

280

230

572

504266

280

230

572

153824

File locations of data shown in this table are:
(a) /home/hapi0/mb4c/data/Shuang/quecho/D8/012115_array_50 (D8)
(b) /home/hapi0/mb4c/data/Shuang/quecho/D8_Mn/012315_array_50 (D8_Mn)
(c) /home/hapi0/mb4c/data/Shuang/quecho/IFP_D8/012615_array_50 (IFP_D8)
(d) /home/hapi0/mb4c/data/Shuang/quecho/IFP_D8_Mn/012815_array_50 (IFP_D8_Mn)

226

REFERENCES

227

REFERENCES

[1] Levy, J. A. (1993) PATHOGENESIS OF HUMAN-IMMUNODEFICIENCY-VIRUS
INFECTION, Microbiol. Rev. 57, 183-289.
[2] Webster, R. G., Bean, W. J., Gorman, O. T., Chambers, T. M., and Kawaoka, Y. (1992)
EVOLUTION AND ECOLOGY OF INFLUENZA-A VIRUSES, Microbiol. Rev. 56,
152-179.
[3] Engen, J. R. (2009) Analysis of Protein Conformation and Dynamics by
Hydrogen/Deuterium Exchange MS, Anal. Chem. 81, 7870-7875.
[4] Curtis-Fisk, J., Spencer, R. M., and Weliky, D. P. (2008) Isotopically labeled expression in
E-coli, purification, and refolding of the full ectodomain of the influenza virus membrane
fusion protein, Protein Expr. Purif. 61, 212-219.
[5] Ghosh, U., Xie, L., Jia, L. H., Liang, S., and Weliky, D. P. (2015) Closed and Semiclosed
Interhelical Structures in Membrane vs Closed and Open Structures in Detergent for the
Influenza Virus Hemagglutinin Fusion Peptide and Correlation of Hydrophobic Surface
Area with Fusion Catalysis, J. Am. Chem. Soc. 137, 7548-7551.
[6] Lorieau, J. L., Louis, J. M., and Bax, A. (2010) The complete influenza hemagglutinin fusion
domain adopts a tight helical hairpin arrangement at the lipid:water interface, Proc. Natl.
Acad. Sci. U. S. A. 107, 11341-11346.
[7] Qiao, H., Armstrong, R. T., Melikyan, G. B., Cohen, F. S., and White, J. M. (1999) A
specific point mutant at position 1 of the influenza hemagglutinin fusion peptide displays
a hemifusion phenotype, Mol. Biol. Cell 10, 2759-2769.

228