University of Southern California Dissertations and Theses

Integrin β3TM structural investigations using phospholipid nanodiscs

(USC Thesis Other)

Integrin β3TM structural investigations using phospholipid nanodiscs

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Integrin β3TM Structural Investigations
Using Phospholipid Nanodiscs
By
Benjamin Frey
A Thesis Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(BIOCHEMISTRY AND MOLECULAR MEDICINE)
August 2017
Copyright 2017 Benjamin Frey
Table of Contents
Dedication i
Acknowledgements ii
Listed Figures iii
Nomenclature iv
Background v
Chapter 1: Introduction 1
1.1 Specific Aims 1
1.2 Overview 4
1.2.1 Integrin Overview 5
1.2.2 αIIbβ3 Heterodimeric interface 10
Chapter 2: Materials and Methods 16
2.1 DNA Expression, Mutagenesis & Sequence Analytics 16
2.2 Protein Expression & Purification 21
2.2.1
15
N MBP-β3TM preparation for nanodisc sample 25
2.3 MBP-β3TM Phospholipid Nanodiscs using MSP1D1ΔH5 30
Chapter 3: Results 35
3.1 Size Exclusion Chromatography 35
3.1.2 Aggregation Products 38
3.2 SDS PAGE Analytics 40
45
46
50
50
55
3.2.1 Size Exclusion including SLAS
3.3 Nanodisc Embedded NMR Spectra
Chapter 4: Discussion
4.1 Size Exclusion Chromatography
Chapter 5: Conclusion
References
5 7
i
Dedication
To my wonderful loving mother Linnea Frey
To my wonderful loving father and his partner Jeffery Frey and Garrel Renick.
To my beloved Fiancé Paige Louise Harrison.
To my awesome brothers Jeffery and Charles Frey
To my fabulous former mentors who inspire, guide and encourage me:
Dr. Michael Henzl
Dr. John Tanner
Dr. David Emerich
Mr. Cason Jones
Dr. Gary Stacey
Dr. Bing Stacey
Dr. Henry Nguyen
To the members of the Missouri Department of Elementary and Secondary Education:
Mr. Louis Gatewood
Mr. Duane Schumate
Ms. Lisa Miller
Ms. Diana Loesch
Ms. Roberta Rynning
Mr. Stephen Carter
To my incredible brilliant hardworking newfound daughter Audrey Rose.
ii
Acknowledgements
This work is made possible through the guidance, support and encouragement
from Dr. Tobias S. Ulmer and Mr. Alan J. Situ at Keck School of Medicine of USC.
I appreciate the time and commitment my thesis committee dedicated to this
project, to Dr. Ansgar Siemer and Dr. Ralf Langen of the Protein Structure Center at
Keck School of Medicine of USC.
I would further like to thank Dr. Thomas Schmidt at the National Institutes of
Health for his previous literary contributions and commitment toward this research.
I would like to thank the University of Southern California and the Keck School
of Medicine for the use of research facilities.
I would like to thank Dr. Lucas Sušac at the University of California, Los Angeles
for his time, expert advice and support in this project.
iii
Figures, Charts and Tables
7
9
11
12
14
22
22
24
28
35
39
41, 42
45
Figure I1: Integrin structure
Figure I2: Modalities of Integrin Structure Alteration
Table I1: Genes and Respective Protein Names of the Integrin Family
Figure I3: Components of Integrin αIIbβ3 & activation partners
Figure I4: Bicelle Embedded NMR Structure of the Integrin β3
Chart MM1: Growth media recipes
Chart MM2: Antibiotic & IPTG dosages
Chart MM3: IMAC 3L MBP-β3TM 500mM imidazole elution
Chart MM4: IMAC 3L MSP1D1ΔH5 300mM imidazole elution
Figure R1: MBP-B3TM Nanodisc Size exclusion Chromatogram
Figure R2 SDS PAGE adding SLAS
Figure R3 Signal Intensities (1) and (2)
Figure R4 - Nanodisc Elution SLAS S-200
Figure R5 – HN TROSY 47
iv
Nomenclature
GP IIa/IIIb Integrin αIIbβ3
CSAT complex Integrin αIIbβ3
ECM Extra cellular matrix
NMR Nuclear Magnetic Resonance
ApoA1 Apolipoprotein A1 / Membrane Scaffold Protein
MSP1D1ΔH5 Membrane Scaffold protein D1, truncated H5 variant
MBPβ3TM
or MBP-β3TM Maltose Binding Protein Integrin Beta 3 Transmembrane
fusion construct with A711P and G690C mutations
β3TM Integrin Beta 3 Transmembrane with A711P and G690C
mutations
DMPC 1,2-Dimyristoyl-sn-glycero-3-phosphocholine
LPXTG Staphylococcus aureus Sortase
TXA2 Thromboxane A2
NBP N-terminal Integrin linker region proline at position 685
ITGA Integrin Alpha Subunit Gene Family
ITGB Integrin Beta Subunit Gene Family
CD61 Peptide name for Integrin Beta 3
kDa Kilodaltons
Theoretical PI Theoretical Isoelectric Point
ORF Open Reading Frame
E.coli Escherichia coli
TEV Tobacco Etch Virus / Tobacco Etch Virus Protease
mM / uM MilliMolar / MicroMolar
v
Background
Integrins are a family of cell surface receptors that provide a means for interaction
between the cell and the extracellular matrix. As noted by Hynes, Integrins were
conceptualized in the 1970s, as a cellular component that links the extracellular matrix to
actin within the cytoskeleton was demonstrated through several binding studies
thereafter. In 1986 Integrin, then known as glycoprotein IIa/IIIb or the CSAT complex,
was first structurally described as a heterodimer (15).
Clinical significance drove early investigations of integrin αIIbβ3 (GPIIa/IIIb) as it was
found to bind fibrinogen in the context of clotting. Advances in cloning and antibody
development allowed researchers to better study a peptide as large as integrin and in latter
studies Integrin complexes with a differing range of substrate specificities were
identified. In 1992 Weisel demonstrated αIIbβ3 bound to fibrinogen induces a
conformational change within the extracellular domain. By 2000 all 24 currently known
variants were identified (Show figure) as well as a major binding partner, Talin, was
identified binding the cytosolic faces of β tails. Structural studies using subdomains of
Talin revealed binding regions in the cytosolic β3 domain using liquid state NMR in
2003 by Ulmer (1).
Takagi in 2001 demonstrated the significance of transmembrane helical proximity in
terms of αVβ1 binding and activation. Helices joined together by disulfide linkage were
unable to bind fibronectin wherein removal of the C-terminal allowed transmembrane
helices to migrate within the bilayer and resume binding.
vi
As revealed above, signal specific cytoskeletal to ECM affixation occurs through
conformational changes in the integrin heterodimer. This non-obligate heterodimer
consisting of variable alpha and beta subunits bind a wide array of ECM peptides:
fibronectin, laminins and collagens. From the outside inward the structure consists of a
large extracellular binding domain, a pliable linker region, a pair of single pass
transmembrane helices and a small intracellular cytosolic domain noted for binding Talin.
Transmembrane helical interactions play a vital role in understanding the differential in
structure between activated and resting state integrin. In the resting state, transmembrane
helices of alpha and beta integrin will interact driven by the closely associated C-termini
of α and β subunit ectodomains (10, 11). Disturbing these interactions moves the integrin
heterodimer into the active conformation (10). Previous transmembrane structures of
monomeric αIIb and β3 in bicelle samples (8) have been determined however it is
unknown if these structures are physiologically relevant and representative of active or
resting form (10).
1
Chapter 1: Introduction
1.1 Specific Aims
1) To propose and demonstrate a model more closely representing the physiological
environment for the transmembrane region of Integrin Beta 3.
2) To provide insight toward side chain and lipid interactions contained within a
larger plane of helical-membrane dispersion.
3) To introduce the usage of phospholipid nanodiscs in studying the integrin
transmembrane helices, providing more distinguished peaks and a stronger liquid
state NMR spectra.
The usage of phospholipid nanodiscs was previously used by Ginsberg in 2013 to
describe β3 transmembrane insertion as it applies to binding for Talin (13). As a similar
helical tilt is observed in Ginsberg’s study which also follows earlier studies in bicelle
samples; it is important to structurally confirm the helical deposition of the β3
transmembrane region in a phospholipid nanodisc environment.
The larger plane of interactions within a phospholipid nanodisc offers a more accurate
representation of helix-lipid interaction as well as imparting structural order within the
sample as held together by ApoA1 peptides. Micelle and bicelle bound peptides do not
offer the same degree of fluidity and range of motion for lipids aside from the annular
lipid-helical interactions. Native bilayer lipids are constantly in motion and the larger
plane of over 5nm in diameter for a nanodisc provides space for motion in simulating a
more native lipid bilayer.
2
This investigation uses self-assembling phospholipid nanodiscs held together by a
truncated version of ApoA1, or MSP1D1, named MSP1D1Δ5. Additionally, this point
mutation study investigates the affectation along the heterodimeric interface as correlated
to impact between transmembrane helix-helix interactions. Specifically at points
previously studied by Lau in 2009 (5), detailing positions A711P and G690C; bicelle
embedded structures are found in figure I4, 29.
Developments of Phospholipid Nanodiscs
Structurally a phospholipid nanodisc is a discoidal surface of a synthetic phospholipid
bilayer held together by surrounding amphipathic α-helical peptides, known as ApoA1 or
MSP (16). These self-assembling structures were demonstrated by Bayburt in 2002.
Earliest investigation between the ApoA1/MSP and the lysosomal phospholipid DMPC
used in making self-assembling nanodiscs was in 1976 by Atkinson through small-angle
X-ray scattering techniques. Atkinson’s findings indicated a perpendicular orientation
between DMPC and ApoA (17).
In 1979 a model of a unilamellar bilayer disks containing DMPC and encircled by
lysosomal ApoA peptides (aka MSP) was put forth by Wlodawer using electron
microscopy following differential scanning calorimetric studies (18). Following Bayburt,
in 2004 Denisov sought out to more clearly characterize a discoidal structure belted by
lysosomal proteins through their alteration. Through truncating up to 22 N-terminal
residues and by adding additional 22 residue repeats (19), smaller and larger nanodiscs
3
could be formed respectively. A truncated variant of the membrane scaffolding protein
(MSP) named helix 5, or MSP1D1ΔH5 is used as it is noted to produce a more stable
discoidal structure.
A well-documented structural study of the integral membrane protein titled human anion
voltage dependent anion channel (VDAC-1) completed in 2009 (21) details a close to
native structure as provided within a nanodisc bilayer using liquid state NMR. In the
same year Ginsberg adapted nanodisc assembly procedures for αIIbβ3 and confirmed
(22) visually through cryoelectron microscopy. A follow up study in 2012 further
characterized Talin binding and helical dispersion nature of Integrin αIIbβ3 (13), the β3
transmembrane helix was recently confirmed within a nanodisc by Ulmer in liquid state
NMR.
Efforts to stabilize bacteriorhodopsin for dispersion into nanodiscs have yielded a variant
held together by an 18 residue long amphipathic alpha helix dubbed 18A (23). The 18A
peptide is described to behave similarly in vivo to ApoA1, and can form nanodiscs using
DMPC with a membrane surface as small as 61.4 Å in diameter.
Advances in creating more stable nanodisc samples capable of increasing NMR signal
and peak distinction were explained by Wagner in late 2016. The newer method explains
the use of covalently circularized ApoA1 peptides affixed to a cupric chip by Glycine and
LPXTG linker (24) and could be a viable candidate for increasing resolution for αIIbβ3
nanodisc embedded NMR signal.
4
1.2 Overview
The Integrin Heterodimer - αIIbβ3 Complex formation & activation
While inactive the β3 (GPIIIb) and αIIb (GPIIa) transmembrane helices are found in
close proximity interacting non-covalentley (6). Platelet activation can stem from both
autocrine and paracrine signaling events. Each platelet contains 80,000 copies of integrin
αIIbβ3 while an inside-out signaling cascade activates the complex allowing the receptor
ectodomain to become adhesive (1, 5). ADP and TXA2 bind to receptors on the platelet
surface which, in turn, causes intracellular Ca2+ levels to rise stemming from
endoplasmic reticulum stores being released. The free Ca2+ binds to the ectodomain of
both alpha and beta chains creating an activated complex, known as GPIIa/IIIb.
Clinical implications
Historically, αIIbβ3 (or GPIIa/IIIb) was a clinical target of interest since it was
characterized in binding platelets in cases of ischemic cardiovascular disease. In the late
90s studies revealed the integrin αIIbβ3 complex as a receptor for fibrinogen (20). The
αIIbβ3 complex is formed in a calcium dependent interaction activated by ADP which, in
turn, drives binding to fibrinogen (Calvete, 95 and Shatti, 99). Formation of the active
GPIIa/IIIb complex was found to be the critical step in endothelial adherence and platelet
aggregation; thus making the GPIIa/IIIb receptor a target for many later clotting
medications.
5
Medications against the GPIIa/IIIb receptor are a clinical target for clotting medications
found to reduce heart attack and stroke; commercial names for these agonists are as
follows: ReoPro, Integrilin, and Aggrastat.
αIIbβ3 is also a target in a number illnesses including: Glanzmann Thrombasthenia (25) a
bleeding disorder stemming from low expression of αIIbβ3, and Thrombocytopenia, a
disorder characterized by low blood platelet counts.
1.2.1 Integrin Overview
Integrins are a family of heterodimeric membrane proteins that play a critical role in a
multitude of signaling, adhesion and cell development events. The two chains are held
together by non-covalent interactions found in the transmembrane helical region for each
α and β strand respectively (see figure I1). Recent studies indicate signaling across the
cell membrane which regulates cellular migration (7). There are two modalities of
integrin function: biochemically in response to adhesion and mechanical affixation of
cytoskeletal components to the extracellular domain. In each of these capacities a
structural change indicates a specific binding or adhesion event.
Integrins have a distinct characteristics in bidirectional signaling capability as well as
modular variability allowing cells to display a specific integrin profile as necessary. This
is accomplished through changing membrane dispersion of the type of integrin alpha and
beta chains. There are 18 alpha and 8 different beta subunits sorted according to substrate
specificity. The integrin complex is also modular, a total of 24 different known receptor
6
combinations have been described (5); this study focuses on the αIIbβ3 heterodimer,
specifically the transmembrane region of the β3 strand.
Structure
The variable heterodimer integrin can range in molar mass between 90-160 kDa
consisting of an endodomain between 40-70 residues while the ectodomain can reach
upwards of 700+ residues. The heterodimer consists of two chains, Alpha and Beta, both
having a single pass transmembrane helical domain separating the endo and ectodomains
across the lipid bilayer (1). Alone the β3 chain’s transmembrane α-helical region is 29
residues spanning positions 693 and 721, terminating at K716 described as the β-tail
domain (1, 30). The αIIb transmembrane segment spans positions 966-993 containing a
24 residue helix ending at position K989 named the Calf domain (1, 5, 8, 30).
Thermodynamically, four states of the heterodimer have been described (1) to detail
Integrin bidirectional signaling behavior. Figure I1 below quantifiably correlates the
Gibbs free energy of contribution of the transmembrane complex and the ectodomain
conformation while introducing coupling factor accounting for mutual stabilization or
minimal affinity depending on the directionality of a binding/signaling event.
7
Figure I1 (1) –Integrin structure
Figure I1 (1): A – Structural depiction of integrin αIIbβ3. Inactive on the left, sourced from PDB 3fcs and 2k9j.
Active on the right, sourced from PDB 2vdl, 3fcs, 2k1a, and 2rmz. (30) B - Primary sequence alignments of
Integrin alpha subunit transmembrane domains reveal highly conserved residues at αIIb positions: P975, G976,
K989, and G991-R995. A distinguishing position G972 is only found in αIIb, is considered one of the 3 key gly
positions whose packing interactions drive the outer membrane clasp, the other two being αIIbG976 and β3G708.
The intermembrane clasp is stabilized by an assortment of key interactions. Interhelical packing between G991,
F992 and F993 stabilizes α–β TM including electrostatic interactions between αIIbR995-β3D723. The domain
labels are derived from known ecto- and transmembrane domain structures (1, 5). C – Integrin bidirectional
signaling thermodynamics are described mathematically through illustration occupying at least four conformational
states. | Inside>Out (IO) signaling an intracellular agonist with an affinity no lesser than ΔG°IO = ΔG°TM + f·ΔG°E
disrupts the ectodomain-stabilized transmembrane complex. Structural fallout from this disruption carries forward
toward the resting ectodomain allowing destabilization and binding of an extracellular ligand equated to an affinity
of at least (1 − f)·ΔG°E | Outside>In (OI) signaling an extracellular agonist must deliver ΔG°OI = f·ΔG°TM + ΔG°E
in order to disrupt the transmembrane complex and elicit a signaling event (1). D – Structural alignment of the αIIb
Calf2 domain with the GB3 domain.
8
Spatial differences within the bilayer are present as helix for αIIb has been shown to
disperse vertically in the membrane while the β3 helix is disbursed horizontally at an
angle ranging from 22-28 degrees.
The most common variants Beta-1, Beta-2 and Beta-3 all share an isoleucine in the first
position on the extracellular face as well as aliphatic residues on the intracellular face.
Inactive Integrin presents a drastically different structure than the bound and active form.
In total, three discrete conformations have been described: inactive, primed (active),
bound (Figure I2, 28, 30). This is especially apparent in the large ectodomain and within
lipid bilayer (28). In the inactive form the alpha and beta strands are closely associated
both within the transmembrane region and the cytosolic endodomain. The ectodomain
bends across the thigh and calf1 regions in the alpha chain, and between the EGF1,
Hybrid and PSI regions within the beta chain. While the N-terminal domains of each
strand are bent inward toward the Calf2 (alpha) and β-tail (beta) domains. Talin binding
impacts flexible knee regions, which have a range of motion described as ‘breathing’ (4,
28).
Binding Talin to the cytosolic region (Figure I2, 28, 30) drives loosening of the leg
restraints and allows further motion outward, the natural outward movement of the hybrid
domain accompanies this loosening motion and heterodimer achieves primed position.
While primed, integrin is fully extended and ready to accept ligand. It is noted that
epitopes of antibodies are also exposed in this Talin-induced conformational change (28).
9
Figure I2: Modalities of Integrin Structure Alteration
Figure I2 (28). A five paneled depiction of the conformational changes in the context of Talin binding in the
integrin heterodimer: A- constricted inactive integrin. B- Talin binding loosens the constricted form allowing
for a swing into primed positon. C- Active bound integrin displaying space between the transmembrane
helices. D- Tensioning across the heterodimer in actin polymerization between cytoskeletal and ECM
polymers. E- Focal adhesion between ECM and Actin associated endodomain triggering intercellular
signaling in an inside-out direction.
10
1.2.2 αIIbβ3 Heterodimeric interface
Heterodimeric interactions between the two right-handed transmembrane helices are vital
to the integrin family’s structural integrity and signaling activity. The αIIbβ3 variant is
noted to have a heterodimeric interface between positions 968 and 986 (6) for the alpha
strand and positions 693-704 for the beta. As mentioned in figure I1, glycine packing at
six key residues drives the formation of the inner and outer membrane clasp (5) Key
positions indicated in the heterodimeric interface as shown in figure I1. Later studies of
αIIbβ3 using bicelle and liquid state NMR describes two interaction interfaces for the
membrane bound helical regions.
Earlier studies revealed a pliable and finely tuned ecto-TM coupling found to modulate a
signaling threshold for integrin receptors though integrin activation assays (1).
Specifically an N-terminal proline named NBP (P685) plays a crucial role in this linker
flexibility.
Alpha Chain
There are 20 known mammalian genes for the integrin alpha chain (12). This family,
prefixed as ITGA(*) contains variable suffixes (*) –listed: 1-11, D, E, L, M, V, 2B, X.
Table 1 below references each known integrin alpha chain gene family with its known
protein. There are also alternative names by which several integrin peptides are known
which are excluded.
11
Table I1: Genes and Respective Protein Names of the Integrin Family
GENE PROTEIN GENE PROTEIN
ITGA1 CD49a ITGB1 CD29
ITGA2 CD49b ITGB2 CD18
ITGA3 CD49c ITGB3 CD61
ITGA4 CD49d ITGB4 CD104
ITGA5 CD49e ITGB5 ITGB5
ITGA6 CD49f ITGB6 ITGB6
ITGA7 ITGA7 ITGB7 ITGB7
ITGA8 ITGA8 ITGB8 ITGB8
ITGA9 ITGA9
ITGA10 ITGA10
ITGA11 ITGA11
ITGAD CD11D
ITGAE CD103
ITGAL CD11a
ITGAM CD11b
ITGAV CD51
ITGA2B CD41
ITGAX CD11c
Table I1 (9, 12) – Provides known and most current gene
names for Integrin. Protein names with respect to the gene
encoded are provided in the adjacent column. Common, former
and historical names for genes are excluded. The highlighted
genes and peptides are the focus in this presentation.
12
The transmembrane helical partner of β3 is αIIb, encoded within mammalian gene
ITGA2B and transcribed into peptide CD41 (Table I1), is found within platelets (9). For
the sake of uniform clarity as this family is also known as glycoprotein IIb/IIIa the usage
of this name, its code GPIIa/IIIa will no longer be used in place for Integrin αIIb, or
αIIbβ3 when describing the heterodimer.
Figure I3: Components of Integrin αIIbβ3 & activation partners
Figure I3 (27) Integrin αIIbβ3 bound to its ligand in extended active conformation allows for distinction of the
heterodimeric subunits. Endodomain associations with Talin and Vinculin provide a linkage to F-actin while
polymerizing cell response complexes in the context of outside-in signaling.
13
Traveling inward from the N-terminus there are four described ectodomains of the
activated alpha chain: the beta propeller, thigh+Genu, Calf1 and Calf2. The single pass
transmembrane helical domain follows the Calf2 region, is 27 residues in length and is
dispersed perpendicularly into the lipid bilayer. Following the transmembrane helical
domain is the shorter of the two chains C-terminal endodomain. A structural
distinguishing feature of the αIIbβ3 dimer is the lack of an αA domain, and the βA
domain is noted as the key binding site, associating with the α beta propeller region (7,
13, 30).
Beta Chain
There are eight known genes for the integrin beta chain labeled as such ITGB1-8, for the
transcribed peptides spanning ITGB5 through ITGB8 the namesake is that of the
encoding gene. For genes ITGB1-4 a prefix of CD is used to name the transcribed
peptides and are as follows in respective order: CD29, CD18, CD61, and CD104. As
indicated above the focus of this structural study, β3, is encoded within ITGB3, is
transcribed into peptide CD61. Other prior common names for β3 are GP3A or GPIIIa;
these names will no longer be used for the purpose of uniform clarity in this study (9, 12).
14
Figure I4: Bicelle Embedded NMR Structure of the Integrin β3
Moving inward from the N-terminal outer reaches of integrin’s beta chain five domains
are commonly described and depicted in figure I3 (27, 30) (7, 13, 27, 28, 30): βA, Hybrid
domain, PSI domain, four EFG domains, β-tail domain followed by the pliable linker
region (1) adjoining the single-pass transmembrane helix. Figure I3 (27, 30) illustrates
the endodomain of the β-chain, the larger of the two in the heterodimer, as it is bound to
both Talin and Vinculin mobilizing F-actin in the context of outside-in signaling. Figure
Figure I4 (29) the integrin 𝛽 3(Ala711Pro) mutation can compensate for Lys716X mutations
by introducing a helix tilt between the inner and outer membrane clasp of the integrin αIIbβ3
transmembrane domains. The mutant form (green) is compared to wild type (red)
15
I4 (29) provides a wild type versus mutant comparison of the Integrin β3 transmembrane
helix.
To continue, methods for development of nanodisc embedded integrin αIIb, β3 and
αIIbβ3 and the inherent considerations in their preparation are described in detail. A
NMR generated structure of a bicelle embedded β3 A711P G690C transmembrane helix
will be used as a comparison for the derived spectra of a nanodisc embedded helix from
the same peptide. Followed by an overview of potential alternative nanodisc sample
preparation describing benefits and potential drawbacks of such implementation.
16
Chapter 2: Materials and Methods
2.1 DNA Expression, Mutagenesis & Sequence Analytics
Integrin β3TM G690C A711P
Following design and expression protocols as detailed by Ulmer (1), the primary
sequence for the Integrin β3 G690C A711P TM construct is as follows:
685 - GESPKCPDILVVLLSVMGAILLIGLAPLLIWKLLITIHDRKEF - 727
The α-helical region is highlighted in cyan. This primary peptide sequence is 43 residues
in length, has a molecular weight of 4739 Daltons, a theoretical PI of 6.75, and has an
extinction coefficient of 5500 (Expasy). This primary sequence is resultant from the
following primers:
5’-CAT GAT CGT AAA GAA TTT TAG AAA TTT GAA GAA GAA CGT- 3’
5’- ACG TTC TTC TTC AAA TTT CTA AAA TTC TTT ACG ATC ATG - 3’
The seventh codon, highlighted in red, codes for a stop at position A728 sequence
resulting in termination at position F727, six residues into the endodomain. The sequence
for β3 listed above was copied out of the plasmid encoding the full length β3 sequence
with both point mutations completed: pet44a-MBP-β3-G690C-A711P-FL.
Nomenclature note:
For brevity this sequence will be described as MBP-β3TM for the transmembrane
construct and MBP-β3FL for the full length peptide in further discussion in this section
although the two key point mutations A711P and G690C remain (see section on
nomenclature).
17
The initial amino acid sequence for pet44a-MBP-β3FL the red highlighted region
represents the mutation point, and stop codon at F727, introduced with the above primers:
CAT ATG AAA ATC CAT CAC CAT CAC CAT CAC GAA GAA GGT AAA CTG GTA ATC TGG
ATTAACGGCG ATAAAGGCTA TAACGGTCTC
GCTGAAGTCG GTAAGAAATT CGAGAAAGAT ACCGGAATTA AAGTCACCGT TGAGCATCCG GATAAACTGG
AAGAGAAATT CCCACAGGTT GCGGCAACTG GCGATGGCCC TGACATTATC TTCTGGGCAC ACGACCGCTT
TGGTGGCTAC GCTCAATCTG GCCTGTTGGC TGAAATCACC CCGGACAAAG CGTTCCAGGA CAAGCTGTAT
CCGTTTACCT GGGATGCCGT ACGTTACAAC GGCAAGCTGA TTGCTTACCC GATCGCTGTT GAAGCGTTAT
CGCTGATTTA TAACAAAGAT CTGCTGCCGA ACCCGCCAAA AACCTGGGAA GAGATCCCGG CGCTGGATAA
AGAACTGAAA GCGAAAGGTA AGAGCGCGCT GATGTTCAAC CTGCAAGAAC CGTACTTCAC CTGGCCGCTG
ATTGCTGCTG ACGGGGGTTA TGCGTTCAAG TATGAAAACG GCAAGTACGA CATTAAAGAC GTGGGCGTGG
ATAACGCTGG CGCGAAAGCG GGTCTGACCT TCCTGGTTGA CCTGATTAAA AACAAACACA TGAATGCAGA
CACCGATTAC TCCATCGCAG AAGCTGCCTT TAATAAAGGC GAAACAGCGA TGACCATCAA CGGCCCGTGG
GCATGGTCCA ACATCGACAC CAGCAAAGTG AATTATGGTG TAACGGTACT GCCGACCTTC AAGGGTCAAC
CATCCAAACC GTTCGTTGGC GTGCTGAGCG CAGGTATTAA CGCCGCCAGT CCGAACAAAG AGCTGGCAAA
AGAGTTCCTC GAAAACTATC TGCTGACTGA TGAAGGTCTG GAAGCGGTTA ATAAAGACAA ACCGCTGGGT
GCCGTAGCGC TGAAGTCTTA CGAGGAAGAG TTGGCGAAAG ATCCACGTAT TGCCGCCACC ATGGAAAACG
CCCAGAAAGG TGAAATCATG CCGAACATCC CGCAGATGTC CGCTTTCTGG TATGCCGTGC GTACTGCGGT
GATCAACGCC GCCAGCGGTC GTCAGACTGT CGATGAAGCC CTGAAAGACG CGCAGACTAA
T TCG AGC TCG AAC AAC AAC AAC AAT GGT TCT TCT CAT CAT CAC CAT CAC CAT TCG AAT TCT TCT GGT
GGA TCC GAA AAC CTG TAT TTC CAG GGC GAA AGC CCG AAA TGC CCG GAT ATT CTG GTG GTG CTG CTG
AGC GTG ATG GGC GCG ATT CTG CTG ATT GGC CTG GCG CCG CTG CTG ATT TGG AAA CTG CTG ATT ACC
ATT CAT GAT CGC AAA GAA TTT gcg
AAATTTGAAG AAGAACGTGC GCGTGCGAAA TGGGATACCG CGAACAACCC GCTGTATAAA GAAGCGACCA GCACCTTTAC
CAACATTACC
TATCGTGGCA CCTGACTCGA G
The bolded and underlined segment of the sequence above represents the 43 codons
encoding the β3 transmembrane sequence spanning G684 through F727. The yellow
highlighted codons are a BamHI restriction site followed by a tobacco etch virus protease
recognition sequence highlighted in green.
18
MBP-β3TM Site-Directed Mutagenesis
The resultant sequence from single nucleotide polymorphism:
CAT ATG AAA ATC CAT CAC CAT CAC CAT CAC GAA GAA GGT AAA CTG GTA ATC TGG
ATTAACGGCG ATAAAGGCTA TAACGGTCTC
GCTGAAGTCG GTAAGAAATT CGAGAAAGAT ACCGGAATTA AAGTCACCGT TGAGCATCCG GATAAACTGG
AAGAGAAATT CCCACAGGTT GCGGCAACTG GCGATGGCCC TGACATTATC TTCTGGGCAC ACGACCGCTT
TGGTGGCTAC GCTCAATCTG GCCTGTTGGC TGAAATCACC CCGGACAAAG CGTTCCAGGA CAAGCTGTAT
CCGTTTACCT GGGATGCCGT ACGTTACAAC GGCAAGCTGA TTGCTTACCC GATCGCTGTT GAAGCGTTAT
CGCTGATTTA TAACAAAGAT CTGCTGCCGA ACCCGCCAAA AACCTGGGAA GAGATCCCGG CGCTGGATAA
AGAACTGAAA GCGAAAGGTA AGAGCGCGCT GATGTTCAAC CTGCAAGAAC CGTACTTCAC CTGGCCGCTG
ATTGCTGCTG ACGGGGGTTA TGCGTTCAAG TATGAAAACG GCAAGTACGA CATTAAAGAC GTGGGCGTGG
ATAACGCTGG CGCGAAAGCG GGTCTGACCT TCCTGGTTGA CCTGATTAAA AACAAACACA TGAATGCAGA
CACCGATTAC TCCATCGCAG AAGCTGCCTT TAATAAAGGC GAAACAGCGA TGACCATCAA CGGCCCGTGG
GCATGGTCCA ACATCGACAC CAGCAAAGTG AATTATGGTG TAACGGTACT GCCGACCTTC AAGGGTCAAC
CATCCAAACC GTTCGTTGGC GTGCTGAGCG CAGGTATTAA CGCCGCCAGT CCGAACAAAG AGCTGGCAAA
AGAGTTCCTC GAAAACTATC TGCTGACTGA TGAAGGTCTG GAAGCGGTTA ATAAAGACAA ACCGCTGGGT
GCCGTAGCGC TGAAGTCTTA CGAGGAAGAG TTGGCGAAAG ATCCACGTAT TGCCGCCACC ATGGAAAACG
CCCAGAAAGG TGAAATCATG CCGAACATCC CGCAGATGTC CGCTTTCTGG TATGCCGTGC GTACTGCGGT
GATCAACGCC GCCAGCGGTC GTCAGACTGT CGATGAAGCC CTGAAAGACG CGCAGACTAA
T TCG AGC TCG AAC AAC AAC AAC AAT GGT TCT TCT CAT CAT CAC CAT CAC CAT TCG AAT TCT TCT GGT
GGA TCC GAA AAC CTG TAT TTC CAG GGC GAA AGC CCG AAA TGC CCG GAT ATT CTG GTG GTG CTG CTG
AGC GTG ATG GGC GCG ATT CTG CTG ATT GGC CTG GCG CCG CTG CTG ATT TGG AAA CTG CTG ATT ACC
ATT CAT GAT CGC AAA GAA TTT gcg
Preparation:
Primers were prepared to 20mM in stock solution and subsequently 2mM in the 50ul
PCR quick change procedure.
The 50 ul PCR Quick Change mixture contains: 2mM FWD primer, 2mM BWD primer,
50ng pet44a-MBP-β3FL, 25ul PFU.
Thermo-cycle parameters:
98C / 2 min  62C / 30 sec  72C / 100 sec  (35x repeat)
Post-Amplification:
Using a Machery-Nagel PCR cleanup kit (Ref # 740609.250) the resultant amplicon was
eluted into 30ul buffer NE
tm
.
19
Transformation & Sequence analysis
Using the eluate pet44a-GB3R2-β3TM from amplification, transformation was
completed into e.coli species Xl10 DE3 with bacterial resistance for ampicillin in LB-
agar. After 16hrs growth at 37C two colonies were cultured in 2ml LB with ampicillin
and subsequently grown for 6-8hrs. Plasmids of each growth were extracted and purified
using a Machery-Nagel NucleoSpin
tm
plasmid kit (Ref # 740588.250) and prepared for
sequencing. Sequences were confirmed using a T7 primer (Addgene).
To prevent aggregation and to aid in digestion with TEV protease this sequence was
fused to an N-terminal maltose binding protein.
MSP1D1ΔH5
The following pet28a vector map for Human Apolipoprotein A1 was ordered from
Addgene as a bacterial media stab (Addgene Cat # 71714):
20
The pet28a plasmid conferred kanamycin resistance, is indicated as a low copy construct
within E.coli DH5alpha. Including the target sequence the vector is 5816nt while
MSP1D1ΔH5 spans 543nt between positions 5122-5625. The transcribed following insert
sequence, 1033nt in size outlines the obtained MSP1D1ΔH5 sequence along with a 6X
histidine tag (yellow), a TEV recognition sequence (red) as well as T7 primer and
terminator sequences indicated in green. The ORF frame 3 spans positions 48-602 each
terminal position is highlighted in magenta. The latter image is that of the linearized
insert map (Addgene).
1 TTCCCCTCTA GAAATAATTT TGTTTAACTT TAAGAAGGAG ATATACCATG 50
51 GGCAGCAGCC ATCATCATCA TCATCATGAA AACCTGTATT TTCAGGGCAG 100
101 CACCTTTAGC AAACTGCGTG AACAGCTGGG CCCGGTGACC CAGGAATTTT 150
151 GGGATAACCT GGAAAAAGAA ACCGAAGGCC TGCGTCAGGA AATGAGCAAA 200
201 GATCTGGAAG AGGTGAAAGC GAAAGTGCAG CCGTATCTGG ATGACTTTCA 250
251 GAAAAAATGG CAGGAAGAGA TGGAACTGTA TCGTCAGAAA GTGGAACCGC 300
301 TGGGCGAAGA GATGCGTGAT CGTGCGCGTG CGCATGTGGA TGCGCTGCGT 350
351 ACCCATCTGG CGCCGTATAG CGATGAACTG CGTCAGCGTC TGGCGGCCCG 400
401 TCTGGAAGCG CTGAAAGAAA ACGGCGGTGC GCGTCTGGCG GAATATCATG 450
451 CGAAAGCGAC CGAACATCTG AGCACCCTGA GCGAAAAAGC GAAACCGGCG 500
501 CTGGAAGATC TGCGTCAGGG CCTGCTGCCG GTGCTGGAAA GCTTTAAAGT 550
551 GAGCTTTCTG AGCGCGCTGG AAGAGTATAC CAAAAAACTG AACACCCAGT 600
601 AAGAGCTCCG TCGACAAGCT TGCGGCCGCA CTCGAGCACC ACCACCACCA 650
651 CCACTGAGAT CCGGCTGCTA ACAAAGCCCG AAAGGAAGCT GAGTTGGCTG 700
701 CTGCCACCGC TGAGCAATAA CTAGCATAAC CCCTTGGGGC CTCTAAACGG 750
751 GTCTTGAGGG GTTTTTTGCT GAAAGGAGGA ACTATATCCG GATTGGCGAA 800
801 TGGGACGCGC CCTGTAGCGG CGCATTAAGC GCGGCGGGTG TGGTGGTTAC 850
851 GCGCAGCGTG ACCGCTACAC TTGCCAGCGC CCTAGCGCCC GCTCCTTTCG 900
901 CTTTCTTCCC TTCCTTTCTC GCCACGTTCG CCGGCTTTCC CCGTCAAGCT 950
951 CTAAATCGGG GGCTCCCTTT AGGGTTCCGA TTTAGTGCTT TACGGCACCT 1000
1001 CGACCCCAAA AAACTTGATT AGGGTGATGG TTC – 1033
21
Cloning Considerations:
In order to increase cloning efficiency for subsequent usage in higher quantity expression
DH5alpha cells from the provided bacterial stab were plated on LB media containing
Kanamycin. Two colonies were selected, for growth at 37C for 6-8hrs and subsequent
purification followed by the Machery-Nagel kit described above to extract plasmids. The
obtained plasmids sequences were then confirmed through using a T7 sequencing method
and were subsequently transformed into E.coli strain Xl10 as this line has been noted to
produce higher cloning efficiency.
2.2 Protein Expression & Purification
15
N Integrin β3TM
The plasmid pet44a-MBP-β3TM described above was used in transformation into E.coli
strain Bl21 DE3 with chloramphenicol resistance for peptide expression in M9 minimal
media. Colonies grown on an LB plate with ampicillin (100ug/ml) and chloramphenicol
(25ug/ml) were chosen on between the 1
st
and 3
rd
day after growth for further expression
in a 2ml LB primer culture with the aforementioned bacterial stringencies.
M9 Minimal Media Labeled Culture
The primer culture was grown at 37C for 6-8hrs at 200 RPM. Afterwards 750ul was
pelleted by centrifugation, resuspended in 750ul of M9 minimal media, and transferred
into 75ml (25ml per L growth culture) of M9 minimal media with
15
N ammonium
chloride. See charts MM1 and MM2 for media growth conditions.
22
Chart MM1: Growth media recipes
Luria Broth Growth Media Amount per liter
LB Stock 20g
DI Water 1000ml
Antibiotic resistance Varies per construct see chart MM2
M9 Minimal Media – 15N Amount per liter
DI Water 930ml
15
NH4Cl2 9.35mM
Na2HPO4 – 7H2O
KH2PO4
NaCl
MgSO4
CaCl2
Thiamine
22mM
22mM
8.6mM
1mM
0.1mM
1ug
Glucose 1%
Ampicillin
Chloramphenicol
*from stock solution (chart MM2)
0.5ml
1ml
*see chart MM2 for dosages
Media Sterilization (all types) Autoclave liquid setting 30min @ 121C
Chart MM2: Antibiotic & IPTG dosages
Antibiotic name Stock solution dosage Working solution dosage
Ampicillin 100mg/ml 100ug/ml
Chloramphenicol 25mg/ml 25ug/ml
Kanamycin 50mg/ml 50ug/ml
IPTG 1M 500uM
After 16-18hrs of growth at 37C, 200rpm starting cultures were pelleted once more at
2500g for 5min, and resuspended in 5ml/L 15N M9 minimal media growth culture, hence
5ml/L was used to inoculate each growth culture. Growth cultures were also grown at
37C, 200rpm for 5-8hrs depending on OD520. At an OD520 between 0.9-1.0 the lac
operon was induced in the peak of the log phase of the cultures using IPTG, see chart
MM2 for dosage. After induction growth cultures were grown in the same conditions for
4 hours and then stored at 4C overnight.
23
Immobilized metal ion affinity chromatography (Ni(II)SO4 / Imidazole)
Cultures stored overnight were pelleted at 3500 x g for 20min at 4C, re-suspended in lysis
buffer (300mM NaCl, 100mM SDS, 50mM Sodium Phosphate, 20mM imidazole, 2mM
β-mercaptoethanol – pH readjusted to 7.40). The suspended cells were lysed using 75W
pulse sonication until no longer viscous, filtered lysate was then loaded on a 5ml charged
GE IMAC HP column at 1.0ml/min at a pressure not exceeding 14mPa. The ionic
detergent sodium dodecyl sulfate is used for optimal efficacy in degradation of the cell
membrane and solubilization membranous proteins; in this case Integrin’s β3
transmembrane helix is the target peptide.
Column Prep:
The 5ml GE HiTrap
tm
IMAC HP columns are stored in 20% Ethanol at 4C, are prepared
and utilized on an AKTA prime plus FPLC (GE product code 11001313) as follows:
6CV mqH2O wash, 1CV charge with 100mM Ni(II)SO4, 6CV mqH2O rinse, 6CV Wash
buffer I equilibration. Rinse with wash buffer I (300mM NaCl, 50mM Sodium Phosphate
pH 7.40, 25mM SDS) followed loading of lysate to remove excess SDS from the resin
bed. Remaining SDS was washed with 10 CV of 8M Urea while nonspecific histidine
rich impurities were eluted in a stepwise gradient spanning 20-60mM Imidizole. MBP-
β3TM was eluted with 500mM imidazole elution buffer (8M Urea, 500mM imidazole,
300mM NaCl, 50mM Sodium Phosphate pH 7.40). The following chromatogram, MM3,
is what is derived from this assay with a growth culture of 3L:
24
Chart MM3: IMAC 3L MBP-β3TM 500mM imidazole elution
15N MBP-β3TM peptide analysis
Including the maltose binding fusion peptide, 6X polyhistidine tag and TEV recognition
sequence the primary structure of this β3TM construct is as follows (Expasy):
10 20 30 40 50 60
HMKIHHHHHH EEGKLVIWIN GDKGYNGLAE VGKKFEKDTG IKVTVEHPDK LEEKFPQVAA
70 80 90 100 110 120
TGDGPDIIFW AHDRFGGYAQ SGLLAEITPD KAFQDKLYPF TWDAVRYNGK LIAYPIAVEA
130 140 150 160 170 180
LSLIYNKDLL PNPPKTWEEI PALDKELKAK GKSALMFNLQ EPYFTWPLIA ADGGYAFKYE
190 200 210 220 230 240
NGKYDIKDVG VDNAGAKAGL TFLVDLIKNK HMNADTDYSI AEAAFNKGET AMTINGPWAW
250 260 270 280 290 300
SNIDTSKVNY GVTVLPTFKG QPSKPFVGVL SAGINAASPN KELAKEFLEN YLLTDEGLEA
310 320 330 340 350 360
VNKDKPLGAV ALKSYEEELA KDPRIAATME NAQKGEIMPN IPQMSAFWYA VRTAVINAAS
370 380 390 400 410 420
GRQTVDEALK DAQTNSSSNN NNNGSSHHHH HHSNSSGGSE NLYFQGESPK CPDILVVLLS
430 440
VMGAILLIGL APLLIWKLLI TIHDRKEFA
Chart MM3: depicts IMAC chromatogram as displayed in Unicorn
tm
PrimeVeiw V5.0. After 220min / 240ml
the sharp peak at approximately 2500 mAu 280nm depicts elution of MBPβ3TM. This target construct eluted
to a yield of 12-16mg/L in 500mM imidazole. A description of each eluate peak not including the target
construct is as follows: 1. 60min excess 100mM Ni(II)SO4 flow through 2. 75-150min unbound loaded lysate
flow through 3. 175min SDS flow through as plateau drops 4. 175-220min non-specific imidazole binding
elution impurities.
25
The full MBP-β3TM contains the β3 transmembrane region at the c-terminal spanning
positions 685-727:
Number of amino acids = 449
Molecular mass = 49 kDa
Extinction coefficient = 73340
Theoretical PI = 5.87
2.2.1
15
N MBP-β3TM preparation for nanodisc sample
The elution product MBP-β3TM was dialyzed at 4C for 24-48hrs in the following buffer
for optimal preparation in nanodisc sample testing: 100mM NaCl, 20mM Tris/HCl pH
7.50, 5mM EDTA. A subsequent buffer of a similar composition adding 20mM SLAS
was also prepared prior to preparation of a nanodisc embedded peptide; this was in order
to avoid formation of aggregates in solution.
For testing purposes and single labeled sample preparation the uncleaved fusion protein
containing a 50 kDa maltose binding protein was utilized. Keeping this domain also aided
in distinction between differential mass and spatial displacement species in later size
exclusion separation assays. See the section covering descriptions on purifying a single
population of nanodisc samples for further details.
After 24-48hrs of 4C dialysis samples were evaluated for purity by SDS Page (gel not
pictured) a lane containing a sole band at 50 kDa with impurities proved most suitable for
usage in continued trials. It is to be noted at molarities nearing the solubility limit in the
given dialysis buffer aggregates form producing additional impurity bands on the SDS
26
gel. Hence this sample peptide was stored at 4C for up to 3 weeks and concentrated to the
target molarity prior to usage in a nanodisc sample, while further trials also included
20mM SLAS to avoid aggregation.
MSP1D1ΔH5
Following purifications as set forth by Ginsberg (13, 22) in preparation of phospholipid
nanodiscs for cryoelectron microscopy MSP1D1ΔH5 was produced to a mass of 19-
21mg/L in the following procedure.
Cell culture
E.coli strain Bl21 DE3 gold described having the following genotype by Aligent
technologies:
E. coli B F– ompT hsdS(rB – mB – ) dcm+ Tetr gal λ(DE3) endA The
This high level expression strain (Aligent) was used as they lack the Lon and OmpT
proteases that could prematurely target and degrade the construct of interest.
A single colony from a transformation plate was used after 24hrs growth at 37C to
inoculate a Luria Broth starter culture, see tables MM1 and MM2 for details on culture
parameters and antibiotic resistances. 25ml of starter culture per liter of growth culture
was used. After 16hrs growth the starting culture was pelleted and resuspended in 5ml/L
growth culture media and 5ml/L of the resuspension was used to inoculate growth
cultures of Luria Broth and grown at 37C 180rpm. At an OD 520 of 0.9-1.0 the cultures
27
were induced using IPTG, see chart MM2 for details on dosage. After induction cultures
were grown for 3hrs at 37C, 180rpm and then stored overnight at 4C.
Immobilized metal ion affinity chromatography (Ni(II)SO4 / Imidazole)
Cell lysis
Similarly to the integrin construct protocol cultures were pelleted and resuspended in the
following ice cold lysis buffer on ice: 50 mM Na-Phosphate pH 7.50, 300 mM NaCl, 1%
Triton X-100, 1 mM PMSF. The resuspended cells were allowed to thaw on ice at room
temperature for no longer than 10min before adding 2mM β-mercaptoethanol. Sonication
at 48W with cooling interruptions followed the addition of the reducing agent, the
suspension was not allowed to warm to room temperature. After sonic lysis the
suspension was centrifuged for 20 min at 20,000 rpm and filtered before being applied to
the column.
GE HiTrap
tm
IMAC HP
A 5ml GE HiTrap
tm
IMAC HP column stored at 4C in 20% ethanol was prepared
similarly as described for the integrin construct above; followed by 6 CV equilibration
buffer: 50mM Na-Phosphate pH 7.5, 300mM NaCl, 1% Triton X-100. The lysate was
kept on ice as it was applied to the column and subsequently washed out with 10CV of
equilibration buffer. The nonionic detergent Triton X-100 was washed with 5 CV of the
same buffer substituting 50mM Na-Cholate in place of 1% Triton X-100, this was
followed by a 5CV wash of equilibration buffer containing 20mM imidazole to elute non-
specific Ni(II)SO4 binding impurities. Approximately 10-12ml of the target peptide was
eluted after 2CV of 300mM elution buffer shown below in chart MM4.
28
Chart MM4: IMAC 3L MSP1D1ΔH5 300mM imidazole elution
TEV digestion & Anion Exchange chromatography
The eluted MSP1D1ΔH5 was dialyzed at 4C for 24-48hrs in 4L of the following buffer:
20 mM Tris pH 8.5, 25 mM NaCl, 0.5 mM EDTA. The peptide concentration was
measured before adding a 1/150 molar ratio of TEV protease for 4hr at RT.
GE HiTrap
tm
Q sepharose
As with the IMAC procedures a 5ml sepharose column was used with a GE AKTA
PrimePlus FPLC system. The 5ml GE HiTrap
tm
Q sepharose column was used in anionic
Manual Run 0:10_UV Manual Run 0:10_Logbook
500
1000
1500
2000
2500
mAu
0 50 100 150 200 250 300 350 400 ml
Method Run 9/30/2016, 7:56:50 AM Pacific Standard Time, Method : , Result Flow 0.0 ml/min Flow 2.0 ml/min Flow 4.0 ml/min Flow 0.0 ml/min
Chart MM4 depicts a Unicorn PrimeView
tm
A280 chromatogram from a GE AKTA PrimePlus FPLC
system used in purification of MSP1D1ΔH5. This chromatogram depicts a plateau signal representing
the loading of the target lysate between 110-220ml. The drop in signal at 225ml is due to the Na-
Cholate wash. The peaks just before 300ml are those of non-specific poly histidine containing
impurities. The sharp peak after 400ml is target peptide.
29
exchange to further purify the digested MSP1D1ΔH5 from the TEV protease and the
cleaved strand. This column was equilibrated in 20mM Tris/HCl pH 8.5, 25mM NaCl
before loading the cleaved product. A stepwise gradient of 20%, 40%, and 60% was
applied using 20mM Tris/HCl pH 8.5, 500mM NaCl. Fraction peaks of each percentage
elution were collected, analyzed by SDS PAGE, and those with uniform purity were
pooled. The pooled eluate peptides were dialyzed overnight at 4C against 100mM NaCl,
20mM Tris/HCl pH 7.50, 5mM EDTA. The resulting purification scheme yields 19-
21mg/l MSP1D1ΔH5, this peptide was stored for future use at -20C.
MSP1D1ΔH5 Peptide Analysis
The transcribed ORF for MSP1D1ΔH5 primary sequence is as follows:
10 20 30 40 50 60
MGSSHHHHHH ENLYFQGSTF SKLREQLGPV TQEFWDNLEK ETEGLRQEMS KDLEEVKAKV
70 80 90 100 110 120
QPYLDDFQKK WQEEMELYRQ KVEPLGEEMR DRARAHVDAL RTHLAPYSDE LRQRLAARLE
130 140 150 160 170 180
ALKENGGARL AEYHAKATEH LSTLSEKAKP ALEDLRQGLL PVLESFKVSF LSALEEYTKK
LNTQ
Before digestion with TEV protease MSP1D1ΔH5 is described as follows (Expasy):
Number of amino acids = 184
Molecular mass = 24168.1 Da
Extinction coefficient = 19940
Theoretical PI = 5.94
30
The resultant sequence from TEV protease digestion:
10 20 30 40 50 60
GSTFSKLREQ LGPVTQEFWD NLEKETEGLR QEMSKDLEEV KAKVQPYLDD FQKKWQEEME
70 80 90 100 110 120
LYRQKVEPLG EEMRDRARAH VDALRTHLAP YSDELRQRLA ARLEALKENG GARLAEYHAK
130 140 150 160
ATEHLSTLSE KAKPALEDLR QGLLPVLESF KVSFLSALEE YTKKLNTQ
Number of amino acids = 168
Molecular mass = 19488 Da
Extinction coefficient = 18450
Theoretical PI = 5.54
2.3 MBP-β3TM Phospholipid Nanodiscs using MSP1D1ΔH5
Sample preparation
A 5ml sample using the following recipe was used to prepare phospholipid nanodiscs as
described (5, 13) previously:
- 200uM MSP1D1ΔH5
- 32uM MBP-β3TM
- 20mM Na-Cholate
- 10mM DMPC
The buffer used to solubilize the mixture: 100mM NaCl, 20mM Tris/HCl pH 7.50, 5mM
EDTA, an identical repeat mixture was prepared containing 20mM SLAS. The addition
31
of SLAS was critical to avoid aggregation of the reconstituted peptide while entering the
lipid bilayer and is depicted in the following results segment.
The solution was incubated for 1hr at room temperature before being added to Biobeads
SM-2 detergent. Biobeads SM-2 were prepared in solubilization buffer and degassed for
30min prior to usage. The nanodisc sample was later washed in these Biobeads at room
temperature overnight on an end to end shaker.
Nanodisc Size Exclusion Chromatography
The nanodisc mixture was taken up from the Biobeads solution and the Biobeads washed
with and equal volume of solubilization buffer. Prior to loading on the column the
mixture was concentrated to near 0.5ml.
GE HiPrep
tm
16/60 Sephacryl S-200 HR
A standard sephacryl 16/60 S-200 size exclusion column was fixed to a GE AKTA
PrimePlus FPLC system and equilibrated with 2CV of the following buffer: 20mM
HEPES pH 7.40, 50mM NaCl, 2mM CaCl2, 0.2% Sodium Azide. Nanodisc solutions
were injected through a 2ml superloop at a pressure less than 0.2mPa and loaded onto this
column. Fractions of 1ml were collected once an A280 signal breached 10mAu, fractions
were analyzed by SDS PAGE as well as respective band intensities using ImageJ peak
integration. See figures R1 and R2 for details.
2
H-
15
N NMR Sample
32
Double labeled MBP-β3TM prepared in methods depicted above with two crucial
modifications:
1) Per chart MM1 every additive constituent of M9 minimal media was combined,
lyophilized and solubilized in D2O prior to usage.
2) The H2O component, 930ml, was replaced with D2O.
Sample Preparation
NMR samples were prepared to 1mM in 320ul in 6% D2O including 0.025 NaN3. The
D2O is used as a field frequency lock while the Azide component acts as an antimicrobial
preservative. The solution was encased in a variable restricted cylindrical NMR sample
tube (Shigemi).
Doubly labeled peptides and resultant NMR samples were prepared by Mr. Alan Situ.
2
H-
15
N TROSY Measurements
A Brunker-Advance
tm
700 MHz NMR spectrometer was utilized in collecting TROSY
measurements. This spectrometer is equipped with a cryo-probe, shielded z-gradient and
operates 5 channels. NMR measures signal arising from nuclear spins within a large
magnetic field, and is best accomplished when that field is homogeneous. This signal is
derived from an applied sequence of radio frequency pulses which are varied in sequence
in specific time intervals. An obtained signal measurement is the sum total of these radio
frequencies emitted by the magnetically active sample nuclei.
33
The transverse relaxation time explains the time it takes for exponential decay of a
respective sample nuclei signal. Through Fourier Transform analysis is possible of
spectra containing resonance lines signifying each emitted radio frequency. Transverse
relaxation time is inversely proportional to the linewidth of the emission, which in turn is
proportional to the size of the objective. Samples with a larger mass and radius of
gyration have shorter transverse relaxation times producing wider bands in a resultant
resonance spectra.
An optimized measurement for such explain sample types, TROSY, also known as
Transverse Relaxation-Optomized Spectroscopy. At 700 MHz an optimal TROSY effect
is achieved. Thus alleviating excessive rapid carbonyl relaxation, improving
measurements of residual dipolar couplings and detecting scalar couplings across H-
bonds. These three factors taken together generates enhanced spectral quality (31).
Structural spectra were collected and recorded by Dr. Tobias Ulmer.
SDS Gel preparation, staining and image peak integration
ExpressPlus
TM
PAGE gels (Cat: M42012) using MOPS were obtained from Gene Script.
Gels contained 16% (w/v) acrylamide resolving phase. Sample wells of 60ul were
available, while only 30ul per sample well was loaded to prevent overflow and cross
contamination. PageRuler
TM
unstained protein ladder with a range of 200-10kDa in 8-
16% Tris-Glycine was used at a volume of 8ul per gel. Samples of 25ul were prepared
with 5ul of 6X SDS DTT marker, loaded in each respective and catalogued lane and ran
for 1hr and 12 min at 120V. Completed gels were stained in coomassie blue staining
34
solution for 16-18hrs overnight followed by a destain of 10% methanol, 10% glacial
acetic acid, 80% mqH2O for one hour and stored in deionized water afterward.
SDS gel images were collected on an Epson
TM
Perfection Photo V300 flatbed scanner
and band intensity peaks were analyzed on ImageJ viewer using an original image.
35
Chapter 3: Results
3.1 Size Exclusion Chromatography
The following figure R1 depicts the resultant chromatogram resolving differential species
in a nanodisc sample sans SLAS (sodium lauryl sarcosinate). A fused maltose binding
protein allowed one to discern sizes on the resolving column. The apparent 50 kDa mass
of the maltose binding protein fusion peptide contrasts well against an empty 80 kDa
nanodisc (42kDa scaffold and 39 kDa DMPC) as well as a singly loaded 130 kDa and
potentially doubly loaded 180 kDa nanodisc. It is to be noted the resolution limit of a
sephacryl S-200 column is 30kDa.
Figure R1: A chromatogram of the resultant nanodisc digestion and subsequent detergent removal. The loaded
sample did not include SLAS to prevent aggregation. The overlaid SDS resolving gel depicts the noted peak
numbers with approximate molecular weight/radius of gyration per the calibrated Sephacryl S-200 column. Peaks
1, 2 and 3 are labeled in the chromatogram as they correspond to a specific SDS gel lane. The lane in the left pane
outlined in the rectangle was selected for signal peak intensity integration, shown in figure R3.
Figure R1 MBP-B3TM Nanodisc Size Exclusion Chromatogram
36
Nanodisc size exclusion and SDS PAGE analytics sans SLAS
When nanodisc samples are resolved by SDS PAGE, such as those in figure R2, two
bands are visible. The upper band at 50 kDa depicts the fused target peptide, the
predominant signal stemming from the maltose binding protein. The lower band at 22
kDa is that of the scaffolding ApoA1 peptides. Since the nanodiscs in this study and in
previous cryoelectron microscopy studies (22) are described to have two scaffolding
proteins encircling the lipid bilayer which contains a single target peptide; one should
expect a 2:1 ratio of integrated peak intensity when comparing the two bands.
The resultant chromatogram displayed in figure R1 depicts a nanodisc trial where
excessive aggregation of either the maltose binding proteins, the scaffolding proteins or
both together have produced a number of peaks. This is a sub-optimal outcome as one
hopes to derive a single distinct peak, more closely achieved in figure R2, depicting
fractions with species listed below:
- Empty nanodiscs
- Singly loaded nanodiscs
- Multiple loaded nanodiscs.
Figure R1 - Peak 3
These three species are what have been described previously through size exclusion
chromatograms with loaded nanodisc samples (22). The selected sample per figure R1
labeled “3” was one that most closely resembles the nanodisc fraction of closest
molecular weight. This sample elutes at 44ml, correlates to a range within 120-150kDa,
37
and presents distinct bands on an SDS gel for each peptide component in the nanodisc
sample.
As shown in figure R2 the gel band intensity peak integration a signal of 17328 exists for
the scaffolding peptide MSP1D1ΔH5, and 1352 for the target peptide. A rough estimate
for this most pure fraction would provide a 13:1 ratio with respect to the following
species ΔH5:β3TM. Arithmetically if every single target peptide in sample 3 were to
have correctly incorporated into the phospholipid bilayer of the nanodisc there would still
be eleven parts unoccupied nanodiscs present in solution. This broad estimate however
does not account for the third criteria listed above regarding multiply loaded nanodiscs,
which could arise given an aggregation or oligomerization event.
Figure R1 – Peaks 1, 2
Peak 1 eluting as early as 25 ml from a Sephacryl S-200 column is very near the upper
range of the column’s resolution capabilities, 250kDa-5kDa in range. An estimate in the
molecular weight of the eluted species per a prior calibration would signify a radius of
gyration and/or a molar mass of 190kDa at the very least.
The corresponding SDS PAGE bands clearly depicted in figure R2 and overlaid near
their respective peak in figure R1 present a fraction with containing the majority of the
loaded sample for each peptide. The lane adjacent to the left of the lane for peak 1 is that
of the concentrated nanodisc fraction prior to loading onto the S-200 column. From the
gel sample, a heavy band near 50kDa is that of the maltose binding protein fused to the
integrin peptide; this band closely resembles that of the loaded fraction.
38
A latter band near 20kDa is that of the MSP1D1ΔH5 scaffold peptide. The first lane to
the right of the ladder is that of the un-concentrated initial nanodisc mixture prior to
addition of Biobeads-SM2. This lane shows a faint band at 50kDa, the target peptide at
32uM, while the bottom band is far more distinct containing 6.25 times the molarity of
scaffold protein. Upon washing and subsequent concentration prior to loading onto the
column the predominant band becomes the target peptide. A similar predominating band
appears in analysis of fraction 1, 25ml off the S-200 column. Also present in the gel
resolution of fraction 1 as with the initial loaded sample are faint bands near 100kDa,
which are not present in the initial nanodisc mixture lane. Another impurity, degradation
or aggregate product band is present near the 20kDa MSP1D1ΔH5 band for fraction 1.
Peak 2 contains a single faint band near 50kDa resembling a species containing an
untraceable amount of scaffold peptide. This peak comes out at 35ml, between 100-200
kDa, per the calibration of the column. The sole product is that containing a 50kDa
species on the gel.
3.1.2 Aggregation products
The lane resembling loaded concentrate of the nanodisc solution presents evidence of
aggregation in solution prior to being loaded on the size exclusion column as present in
the higher 100kDa band. This is further confirmed in the SDS gel lane analysis of the
eluate peak 1 containing a heavy target peptide peak eluting far earlier than expected for
a group containing independently non-associate species at 50kDa. The addition of the
ionic detergent sodium lauryl sarcosinate (SLAS) mitigated the aggregation effect as
shown in figures R2 and R3. Additionally 20mM SLAS in solution allowed eluate
39
species to arise closer to the optimal 2:1 MSP1D1ΔH5 to MBPβ3TM range upon analysis
of SDS gel band peak analysis as indicated in Figure R3.
As in figure R2 added ionic detergent SLAS significantly alters the initial nanodisc
solution. This is evident in a comparison of the two lanes labeled “concentrate ND pre S-
200,” and with the omission of an eluting sample around 190 kDa. Figure R4 further
corroborates this absence in a chromatogram. Another key component lies within the
Figure R2: Resultant side-by-side SDS PAGE gel comparison of nanodisc fractions eluting at similar time points
from a sephacryl S-200 size exclusion column. Per figures R1 and R3, a major peak arose at 44ml. The left gel
depicts aggregation products forming before loading onto the S-200 column and a lane/fraction resembling much
of the loaded concentrate eluting at a higher than anticipated molecular weight. The right gel is an identical trial
with the addition of 20mM SLAS, the resulting gel lanes display a gradient of eluate products
Figure R2 SDS PAGE adding SLAS
40
nature of the eluate species; appearing in a smooth gradient (right gel Figure R2) as
opposed to a single event with resulting fallout (left gel R2).
Heavier bands at 100 kDa in the left gel of figure R2 indicate a higher proportion of
aggregate products which appear to elute with most of the target peptide in the earliest
fraction. The latter gel including SLAS, which mitigates the appearance of aggregates as
little to no S-200 eluate samples portray a 100 kDa band. Although the initial concentrate
loaded does have a 100 kDa band, comparatively the intensity is visibly lighter. Ionic
detergent also affected the amount and distribution of scaffold peptide available with
relation to target peptide. Nanodisc samples eluting without SLAS first displayed a
smaller signal profile within the loaded and first eluting fractions. This is in contrast to
SLAS samples. When concentrated, the samples including SLAS portrayed a heavier
staining band at the molecular weight equal to that of the scaffold peptide. Furthermore
subsequent eluate nanodisc samples portray a different elution pattern. Eluting in a
gradient starting lighter and becoming heavier with later fractions.
3.2 SDS PAGE Analytics
Following up from the previous visible SDS lane overview, both the target and scaffold
peptide portray distinct intensity patterns with the addition of the ionic detergent. Figure
R3 details this nature through the rolling circle integration image analytics in ImageJ
software. One can also target signal from potential pre-loaded aggregate species to
partially gauge the efficacy of the ionic detergent in creating solution homogeneity. This
is accomplished through identifying and measuring the integrative peak for the bands
41
near 100 kDa on the SDS gels as described above. The following figure portrays signal
intensity measurements for both nanodisc samples with and without 20mM SLAS present
in solution.
Figure R3 Signal Intensities (1)
42
Figure R3 (1, 2): Screenshots of ImageJ rolling circle image analysis. From top to bottom: (1) Two analyses of
100 kDa bands indicative of aggregate species in nanodisc pre-loaded sample, (2) Two analyses of signal present
in an eluate fraction present after 44ml of loading onto sephacryl S-200 column, corresponding to a molecular
weight > 120 kDa.
(Top) The selected gel, portrayed as the leftmost in Figure R2. A resultant plot and integration analysis spectra per
ImageJ software is the rightmost spectral plot with the lane selected for analysis as outlined in yellow with the
number 1. The center window displays a raw uncalibrated signal from rolling circle integration of the commassie
blue stained SDS gel. Resultant signal = 5644.
(2
nd
) the selected gel and outlined lane of interest in yellow, portrayed as the rightmost in Figure R2 containing
nanodisc samples including SLAS. To its right a spectral plot resultant from rolling circle integration per ImageJ
software. The center window displays a raw uncalibrated signal from said integration. Resultant signal = 1666.
(3
rd
) similarly to the top image, SDS gel leftmost from figure R2. A ladder eluting lane is now selected for
analysis as outlined in the yellow rectangle. A plot of signal and integration analysis per ImageJ software. Two
peaks are selected 50 kDa, MBPβ3TM and ~20 kDa, MSP1D1ΔH5. Degenerate signal intensity is omitted from
integrative analysis as outlined in with the perpendicular lines. The resultant signal intensities respectively = 4123
and 23163.
(4
th
/bottom) rightmost gel in Figure R2 with the 44ml / >120 kDa lane selected within the yellow rectangle. The
resultant plot of the entire lane using ImageJ rolling circle integration. Perpendicular lines restrict measurement to
the peak of signal excluding degenerate signal. Two peaks are selected 50 kDa, MBPβ3TM and ~20 kDa,
MSP1D1ΔH5. The resultant signal intensities respectively = 11460 and 28012.
Figure R3 Signal Intensities (2)
43
Loaded aggregates
The approximate signals are from a concentrate sample from 5ml to less than 600ul in
both trials. The integrated signal intensity for the initial nanodisc sample at 100 kDa is
5644 as compared to 1666 for the sample including 20mM SLAS. The difference in
intensity between non-SLAS and SLAS samples is 3.38 to 1. As previously explained a
predominant peak from an intense signal above 20 kDa also details a strong presence of
MSP1D1ΔH5 in solution. This is not apparent in samples lacking SLAS. These figures
are also noted to be a conservative estimate as they also include various aggregate
degradative signal above 100 kDa included in the signal integration for the top image
analysis.
Intensities at 44ml / > 120 kDa
Elution fractions in both cases do display a gradient of species present that fall within the
three criteria explained above. However less pronounced in the absence of SLAS.
Selected lane intensities in both cases have similar degradation product signal degeneracy
which must be excluded as nanodisc sample impurity in measuring signal intensity. A
degree of heterogeneity within self-assembled nanodisc samples in this case exists and is
partially diminished with the addition of SLAS. This is most clearly depicted in figures
R3 and R4. One must note in Figure R3 (2) both peaks display a shoulder, resulting from
degradation of aggregates in solution. This is most pronounced in samples sans SLAS.
The third analysis down has a signal spike for the intensity measurement at 50 kDa,
MBPβ3TM. Accompanying this spike is a more pronounced shoulder in the spectra of
44
the 20 kDa, MSP1D1ΔH5 band. The addition of SLAS both diminishes the 50 kDa signal
spike and the pronounced shoulder evident at 20 kDa.
The ratio of target peptide signal MBPβ3TM to that of the nanodisc assembly peptide
MSP1D1ΔH5 is also significantly altered including SLAS. This is detailed in Figure R3
(2). Nanodisc samples eluting at 44ml from a sephacryl S-200 column without SLAS had
respective signals of target to scaffold peptides as follows:
4123: 23163
This translates as 1: 5.62 of MBPβ3TM: MSP1D1ΔH5. This further reduces to 1: 2.80 of
MBPβ3TM per phospholipid nanodisc. The optimal target for a most pure sample
incorporating a predominant species of 1: 2 target to scaffold peptide and naturally 1: 1
target to nanodisc.
Contrasted with samples including 20mM SLAS, subtracting degeneracy and shoulders
in staining signals the values of target: scaffold peptides are as follows:
11460: 28012
Translating to 1: 2.44 of MBPβ3TM: MSP1D1ΔH5. Accounting for target peptide per
nanodisc leaves the ratio toward 1: 1.22. This draws the solution homogeneity closer
toward the intended goal as singly loaded nanodiscs predominate in the 44 and 45ml
fractions. The 45ml fraction values contain the second highest population of the singly
loaded nanodisc species while the values are not expressly stated. This fraction elutes at
the apex of the 3
rd
peak in the non SLAS trial as well as the apex of the 2
nd
peak of the
detergent laden sample.
45
3.2.1 Size Exclusion including SLAS
The following figure further depicts the effect of SLAS on the A280 chromatogram of a
nanodisc sample containing 32uM MBPβ3TM. One obvious difference exists in the
absence of an early peak at 25ml corresponding to a higher mass or lack of linear
structure. Confirmation of two, and not three separate species in solution is shown on the
SDS gel per Figure R2.
Figure R4 (next page): An A280 Chromatogram resulting from an injection of a phospholipid nanodisc
concentrate including SLAS to reduce aggregate species. Two peaks are numbered along with respective post-load
elution volume. The scale below provides a rough estimate for molecular weight or apparent radius of gyration as
per a molecular weight calibration of the column. Sharp peaks separation is noted by the manufacturer
(GEHealthCare) to be accurate to as little as 30 kDa. A distinct initial peak at 35ml is present followed by an
equally distinguished 45ml peak. As compared to Figure R1 there is less of a tail region eluting after 45ml.
Figure R4 Nanodisc Elution SLAS S-200
46
Stark differences in the elution behavior through size exclusion are present in nanodisc
samples including SLAS. Both chromatograms do share peaks at 35 and 45ml however
the signal intensity and clarity is much improved in the latter trial. There is also a
diminished post-peak hump around 50ml. Analysis of 1ml fractions on an SDS gel gives
rise to a smoother gradient of elution species depicted in Figures R1 and R2 as well.
3.3 Nanodisc Embedded NMR Spectra
Moving forward, HN TROSY spectra comprised of nanodisc samples pooled from 44
and 45ml fractions post size exclusion details the impacts of SLAS on structural data
collection. However when compared to samples of the same peptide within a bicelle
sample signal occlusion is apparent. In double labeled
2
H
15
N TROSY experiments a loss
in peak distinction primarily within the peptide backbone region is visible. This results in
a nanodisc sample and resultant spectra not usable in structural generation proceedings.
Figure R5 provides two β3TM HT TROSY spectra derived from samples within
phospholipid nanodiscs and later within a bicelle sample. Lipid contents of the following
bicelle samples in the HN TROSY spectra are as follows: 350mM DMPC, 175mM
POPC.
47
In examination between the derived HN TROSY spectra bicelle samples still provide a
superior signal to phospholipid nanodiscs. Nanodisc samples also include SLAS to
prevent excessive aggregation of the target peptide. Starting from the peak labeled Indole
H1 and moving clockwise the following observations can be made.
- A single clean peak arises at 1H - 10.35, 105 - 15N PPM, this is indicative of H-N
resonance within indole within W715 in the bicelle spectra. The same peak produces a
doublet of peaks differing in size, the larger being non circular.
Figure R5: Side-by-side comparison of HN or H1 TROSY Spectra. To the left are bicelle embedded β3TM
samples, to the right are nanodisc embedded β3TM samples. Peaks with respect to their HN chemical shift are
noted in read across each readout. Also a comment on signal clarity is pointed out for the nanodisc spectra. Both
spectra were collected by a Brunker Advance
tm
700 megahertz liquid state NMR.
Figure R5 - HN TROSY
48
- Glycine peak at 1H – 8.67, 108 15N PPM is small clear and sharp within a bicelle
sample. Although distinct in nanodiscs there is less of a sharp peak as well as some
neighboring signal noise just to the left of the peak in question.
- R724 Tail Hε at 1H – 7.25, 110 15N PPM is smoothly and clearly articulated within the
bicelle sample. Much of this signal is different within a nanodisc counterpart. There is a
broadened signal, potentially arising from two overlaid or directly adjacent peaks at this
position.
- 711P Adjacent HN at 1H – 7.0, 155.5 15N PPM is clear and present within the left
spectra representing the bicelle sample’s ability to clearly portray the impact of such a
key point mutation. This peak is not distinguishable for the nanodisc sample and is easily
lost within the background at the similar coordinates.
- HN peptide backbone peaks within ranges 1H: 7.6-8.7, 15N: 114-124 PPM. There are
several distinct peaks that offer little if any signal overlay within the bicelle derived
spectra. Very clear spires arise distinguishing each individual HN signal within the
peptide backbone. Those which are close also have a discernable border. The
corresponding ranges are starkly different for the nanodisc trial. There are fainter signals
for each HN peptide peak; and those that do arise are occluded producing few, if any,
distinguishable signals. It is also noted that the scattering of the signal within this more
populated region if the spectra is not clearly consistent between the two samples.
Whereas the three aforementioned peak signals are tightly associated with their
respective coordinates. This nature for this denser region seemingly does not hold true.
49
- A differential in signal clarity is also clearly observable between bicelle and nanodisc
samples. The latter spectra contains a label per Figure R5. There appears a noticeable
amount of background noise, potentially due to the reduction in global sample signal.
- The final residue in the target peptide F727: 1H – 7.7, 123 15N PPM has a somewhat
shifted signal in the nanodisc sample as compared to the bicelle. Additionally, as this
peak is notably the most distinct in TROSY (31), there is evidence of excess noise and
occlusion in the nanodisc.
Taken together, size exclusion and the resultant SDS gel analytics portray alleviation of
aggregation using SLAS. However the amount of aggregation is not mitigated to the
extent to create a suitable sample for usage in a multi-spectral structural study. This is
clearly outlined in Figure R5.
50
Chapter 4: Discussion
4.1 Size Exclusion Chromatography
Portrayed in Figures R1 and R4 A280 chromatogram behavior varied drastically with the
addition of the ionic detergent SLAS to the nanodisc sample. Specifically the reduction of
the peak labeled “1” eluting at 25ml in Figure R1. Although three peaks are visible in the
first chromatogram, they are not as distinct and steep as in the SLAS elution profile.
Steep elution peaks are indicative of a more single species within a solution; and such
broad peaks as seen in the initial chromatogram detail a solution with a higher degree of
heterogeneity. As explained along with the presentation of Figure R1 three distinct
species arise in such a heterogeneous solution more specifically these species are as
follows:
1) Aggregates of target and scaffold peptide, multi-loaded nanodiscs, malformed
aggregate nanodiscs
2) Singly loaded, well formed nanodiscs
3) Well formed nanodiscs lacking a loaded peptide, single peptides not incorporated
in nanodiscs, smaller peptide (scaffold protein) aggregates.
The presence of three peaks post size exclusion suggests three distinct populations which
more likely fulfill the three criteria listed above. Resolving the components of each peak
by SDS PAGE shed light on such differential species eluting in these three peaks. It is
noted that aggregates of target-target, target-scaffold and scaffold-scaffold peptides could
very well exist in a solution lacking an ionic detergent. This is further corroborated in
SDS PAGE and band intensity analysis.
51
Species 1) likely elutes first as peak1 as aggregates of multiple 50kDa could reach the
200kDa range with an increasingly large radius of gyration. With the exception of multi-
loaded nanodiscs, which likely are the major contributor to peak 2.
Species 2) potentially eluted at the strongest peak 3 at 45ml resembling 120kDa off of the
calibrated S-200 column.
Species 3) species elute in a gradient and not in a single event in size exclusion. The gel
resembles a stronger fraction of empty, mal-formed and smaller scaffold peptide
aggregates that persist in peak 3. This is especially evident in the plateau tail region of the
same peak following the fall of the 45ml signal.
With the addition of SLAS to the nanodisc self-assembling mixture the size exclusion
A280 eluate profile is greatly altered. No peak at 25ml exists, indicating a low
undetectable amount of aggregates and malformed nanodiscs found in larger species 1.
The peak at 35ml suggests there are multi-loaded nanodiscs by molecular weight and
SDS gel analyses alone. However the occlusion in the H1 TROSY spectra could suggest
aggregates of singly loaded nanodiscs where a single scaffold peptide interferes with the
disc or between singly loaded discs. A sharp well defined peak at 45ml / 120kDa presents
a strong indication of singly loaded nanodiscs. Gel band intensity analyses do support this
assertion.
As stated above describing species 3, there could persist a significant population of empty
nanodiscs co-eluting with the singly loaded population. A consequence of this lies within
the H1 TROSY spectra. The diminished signal and increased noise ratio could suggest
the presence of this co-eluate species.
52
SDS PAGE
Figure R2 presents starkly different gel profiles for each trial. Without SLAS a sizeable
portion of scaffold peptide is lost before using Biobeads. This explains the great
difference in concentrate sample loaded onto the S-200 column. As SLAS prevents
maltose binding and presumably scaffold peptide aggregation, nanodiscs have a higher
potential of correct assembly when including ionic detergents.
From the outset the assembly mixture was lacking a key component in the ionic
detergent. At such high solution molarities the scaffold peptide is more apt to aggregate
rather than dimerize in a head-to-toe fashion in a cylindrical shape in forming a nanodisc.
Additionally the target peptide was just below its solubility limit for the buffers used.
Hence an inciting aggregation event could have occurred with the target peptide
impacting the other peptide in solution in a runaway aggregation event.
What potentially occurred in solution, as per the left gel in Figure R2, was:
1) A runaway aggregation event. A loss of most material in washes with Biobeads.
2) A disproportionate fraction of high MW species persists in the concentrate-loaded
solution. Visible in the very heavy 50kDa and noticeable 100kDa bands.
The fallout of these two events is clearly depicted in the early, ultra-high molecular
weight / radius of gyration species eluting from the size exclusion column.
Alleviation of the inciting aggregation event is accomplished in adding 20mM SLAS to
the near-solubility limited MBPβ3TM target peptide prior to addition into the nanodisc
creation sample. The right gel in Figure R2 shows a much more homogeneous solution
and elution profile. Instead of an imbalance of species loaded on the column, a solution
53
that follows suit for the initial molarities added exists. This is complemented with a
smooth gradient representing multi and singly loaded nanodiscs.
In support of this proposition gel band intensities reflect the same nature of solution. The
optimal ratio of scaffold peptide (MSP1D1ΔH5) to target (MBPβ3TM) is as stated
previously, 2:1. Without SLAS this ratio is greater than 5:1. Such a figure supports the
persistence of a large portion of aggregates of target and scaffold peptides or multiply
loaded nanodiscs. Adding SLAS reduces the ratio to roughly half that at 2.5:1.
Another note in analyzing the gel bands lies within degenerate signal from peptide
degradation. As the figures mentioned above are derived from signal plotting and
integration, the data was excluded to a single distinct peak while leaving out multi-band
or shoulders in the signal plot. SLAS also reduced the amount of signal per band that
appeared as a degradation product. This is most clear in figure R3 in examining the peak
intensities. In this capacity the ionic detergent allowed the highly concentrated species to
persist long enough in the self-assembly phase to coalesce into a loaded nanodisc.
HN TROSY
Although the addition of an anti-aggregation agent in solution may have allowed the
nanodisc sample achieve a molarity suitable for an NMR sample the quality of the
sample at this time too low.
As per Figure R5 only very few peaks are distinct enough in an HN TROSY spectra.
There is also a significant amount of occlusion both within the peptide backbone
grouped peaks and the individual species signals.
54
A few causes could create such a consequence. This TROSY spectra could be that of two
groupings of samples overlaid. This suggests the target peptide did not efficiently and
uniformly enter the bilayer of the nanodisc upon creation. Another potential pitfall for
the nanodisc sample creation could lie within the same root cause as proposed in the
sample lacking SLAS. An unintentional interaction between scaffold peptide and target
peptide could exist subtly in solution. In this case even a minimal amount of this species
co-eluting with desired species could create noise and occlusion.
As for the diminished signal apparent in the nanodisc sample as compared to the bicelle
counterpart; the persistence of empty nanodiscs and residual aggregates as explained just
above could create larger noise. Greater signal to noise effectively could occlude and
diminish the target spectra.
55
Chapter 5: Conclusion
Understanding the multitude of structural, reactionary and spatial adjustments inherent in
the Integrin complex is primary to elucidating causes of several prevalent afflictions. The
CDC states one in every four deaths is due to cardiovascular disease. A major factor in
platelet activation inherent in cardiovascular disease and clotting disorders; Integrin
αIIbβ3 and its structural nature holds the key to further understanding this affliction.
From the internal binding of Talin driving a dissociation of the internal clamp > the
loosening of the electrostatic tight inter-helical interactions > external linker clamp
opening in the scope of inside out signaling (13) to a similar extracellular binding event
driving internal complex association recruiting vinculin and actin in the scope of outside
in signaling; one significant structural event must occur. The transmembrane helices
must create a pliable interface across the heterodimeric interface (5). Moreover these
helices must have mobility within the phospholipid bilayer on short notice. Such
pliability has been demonstrated and characterized (1) leading toward upcoming studies
of point mutations affectations on helical dispersion in the cell membrane. The truly
unique nature of the two transmembrane helices present in integrin tell a much larger
story detailing cellular responsivity and mobility within its environment.
Structural characterization of these transmembrane helices is best achieved in a bicelle
environment (29). Context and physiological relevance of such structures may come into
question in such a sample (10); and this study seeks to address this concern. Through
recreating a phospholipid nanodisc embedded sample of Integrin β3TM as previously
investigated using cryoelectron-microscopy (22) this study investigated the challenges in
56
deriving a high resolution structure. In the scope of these proceedings one is able to
confirm similar findings in both SDS PAGE and size exclusion chromatography while
moving forward with preparation of a potentially improved sample.
To reiterate phospholipid nanodiscs provide key benefits as compared to bicelle
embedded environments (13, 16, 19, 21). In providing a stabilized bilayer nanodiscs
allow for a more native dispersion for intermembrane peptides. This is due to the fluid
mosaic that can persist within the nanodisc. Furthermore lipid-helical interactions
surrounding Integrin β3TM beyond the annular nested environment in a bilayer allow for
a more physiologically relevant sample. While being able to vary the size of the nanodisc
(19, 21) could allow for future embedding of both αIIb and β3 transmembrane segments.
Greater sample temperature and tumbling stability could allow for warmer or longer
NMR spectral collections. Thus giving rise to an improved dataset for structural
derivation (32, 33).
This study investigated a potential pitfall in creating a nanodisc embedded sample with a
target peptide so close to a solubility limit being prone to aggregation. A method to
overcome such limitations has been explored and could be implemented as part of a fix
in future nanodisc trials. Moving forward, two alternative solutions are briefly proposed
that can be added as a potential curative measures for encountered limitations in this
study.
Although more costly, covalently circularized nanodiscs could potentially stabilize and
create a homogeneous sample solution (24). Using a linearized nanodisc width peptide
linked via a 6X His tag to a Cu
2+
chip, adding Sortase frees the linearized peptide in a
fashion allowing a uniform discoidal structure to persist. Other proposed developments
57
utilizing a small peptide 18 residues in length allow for a nanodisc to persist on a
significantly smaller scale (23). Embedding target peptides within these conditions could
potentially alleviate aggregation products while achieving sample homogeneity at a
lesser cost.
One can speculate that significant membrane protein studies could truly benefit from the
usage of phospholipid nanodisc environments given the detailed benefits above. Without
adopting an entirely new platform of sample preparation two subsequent studies could
improve the sample derivation in a phospholipid nanodisc. The first would include
investigating differing molarities and differing ionic detergents than SLAS in mitigating
aggregates in solution. The second involves assembling the sample mixture and utilizing
a second pass of IMAC to remove potential occluding species and empty nanodiscs.
References
1. Schmidt, T., Ye, F., Situ, A. J., An, W., Ginsberg, M. H., & Ulmer, T. S. (2016). A Conserved
Ectodomain-Transmembrane Domain Linker Motif Tunes the Allosteric Regulation of Cell
Surface Receptors. Journal of Biological Chemistry, 291(34), 17536-17546
2. Zhu, J., Luo, B. H., Xiao, T., Zhang, C., Nishida, N., & Springer, T. A. (2008). Structure of a
complete integrin ectodomain in a physiologic resting state and activation and deactivation by
applied forces. Molecular cell, 32(6), 849-861.
3. Iwamoto, D. V., & Calderwood, D. A. (2015). Regulation of integrin-mediated adhesions. Current
opinion in cell biology, 36, 41-47.
4. Campbell, I. D. (2009). Structural studies of integrin activation (Doctoral dissertation, University
of Oxford).
5. Lau TL, Kim C, Ginsberg MH, Ulmer TS 2009. The structure of the integrin αIIbβ3
transmembrane complex explains integrin transmembrane signalling. EMBO J 28: 1351–1361
6. Campbell, I. D., & Humphries, M. J. (2011). Integrin structure, activation, and interactions. Cold
Spring Harbor perspectives in biology, 3(3), a004994
7. Ginsberg, M. H. (2014). Integrin activation. BMB reports, 47(12), 655.
8. Lau, T. L., Dua, V., & Ulmer, T. S. (2008). Structure of the integrin αIIb transmembrane segment.
Journal of Biological Chemistry, 283(23), 16162-16168.
9. Shattil, S. J. (1999). Signaling through platelet integrin aIIbb3: Inside-out, outside-in, and
sideways. Thrombosis and haemostasis, 82(2), 318-325.
58
10. Zhu J, et al. (2009) The structure of a receptor with two associating transmembrane domains on
the cell surface: Integrin αIIbβ3. Mol Cell 34(2):234–249.
11. Berger, B. W., Kulp, D. W., Span, L. M., DeGrado, J. L., Billings, P. C., Senes, A., ... & DeGrado,
W. F. (2010). Consensus motif for integrin transmembrane helix association. Proceedings of the
National Academy of Sciences, 107(2), 703-708.
12. Hynes RO (2002). "Integrins: bidirectional, allosteric signaling machines". Cell. 110 (6): 673–87.
doi:10.1016/S0092-8674(02)00971-6. PMID 12297042.
13. Kim, C., Ye, F., Hu, X., & Ginsberg, M. H. (2012). Talin activates integrins by altering the
topology of the β transmembrane domain. J Cell Biol, 197(5), 605-611.
14. Hynes, R. O. (2004). The emergence of integrins: a personal and historical perspective. Matrix
biology: journal of the International Society for Matrix Biology, 23(6), 333.
15. Cosgrove, L.J., Sandrin, M.S., Rajasekariah, P., McKenzie, I.F.C., 1986. A genomic clone
encoding the chain of the OKM1, LFA-1, and platelet glycoprotein IIb–IIIa molecules. Proc. Natl.
Acad. Sci. U. S. A. 83, 752 – 756.
16. Bayburt, T. H., Grinkova, Y. V., & Sligar, S. G. (2002). Self-assembly of discoidal phospholipid
bilayer nanoparticles with membrane scaffold proteins. Nano Letters, 2(8), 853-856.
17. ATKINSON, D., SMITH, H. M., DICKSON, J., & AUSTIN, J. P. (1976). Interaction of
Apoprotein from Porcine High ‐Density Lipoprotein with Dimyristoyl Lecithin. European journal
of biochemistry, 64(2), 541-547.
18. Wlodawer, A., Segrest, J. P., Chung, B. H., Chiovetti, R., & Weinstein, J. N. (1979). High-density
lipoprotein recombinants: evidence for a bicycle tire micelle structure obtained by neutron
scattering and electron microscopy. FEBS letters, 104(2), 231-235.
19. Denisov, I. G., Grinkova, Y. V., Lazarides, A. A., & Sligar, S. G. (2004). Directed self-assembly
of monodisperse phospholipid bilayer Nanodiscs with controlled size. Journal of the American
Chemical Society, 126(11), 3477-3487.
20. Vickers, J. D. (1999). Binding of polymerizing fibrin to integrin alpha IIb beta3 on chymotrypsin-
treated rabbit platelets decreases phosphatidylinositol 4, 5-bisphosphate and increases cytoskeletal
actin. Platelets, 10(4), 228-237.
21. Raschle, T., Hiller, S., Yu, T. Y., Rice, A. J., Walz, T., & Wagner, G. (2009). Structural and
functional characterization of the integral membrane protein VDAC-1 in lipid bilayer nanodiscs.
Journal of the American Chemical Society, 131(49), 17777.
22. Ye, F., Hu, G., Taylor, D., Ratnikov, B., Bobkov, A. A., McLean, M. A., ... & Ginsberg, M. H.
(2010). Recreation of the terminal events in physiological integrin activation. The Journal of cell
biology, 188(1), 157-173.
23. Midtgaard, S. R., Pedersen, M. C., Kirkensgaard, J. J. K., Sørensen, K. K., Mortensen, K., Jensen,
K. J., & Arleth, L. (2014). Self-assembling peptides form nanodiscs that stabilize membrane
proteins. Soft Matter, 10(5), 738-752.
24. Nasr, M. L., Baptista, D., Strauss, M., Sun, Z. Y. J., Grigoriu, S., Huser, S., ... & Wagner, G.
(2016). Covalently circularized nanodiscs for studying membrane proteins and viral entry. Nature
Methods.
25. Buitrago, L., Rendon, A., Liang, Y., Simeoni, I., Negri, A., Filizola, M., ... & Bariana, T. (2015).
αIIbβ3 variants defined by next-generation sequencing: Predicting variants likely to cause
Glanzmann thrombasthenia. Proceedings of the National Academy of Sciences, 112(15), E1898-
E1907.
26. Giudice, T. L., Lombardi, F., Santorelli, F. M., Kawarai, T., & Orlacchio, A. (2014). Hereditary
spastic paraplegia: clinical-genetic characteristics and evolving molecular mechanisms.
Experimental neurology, 261, 518-53.
27. Smith, H. W., & Marshall, C. J. (2010). Regulation of cell signalling by uPAR. Nature reviews
Molecular cell biology, 11(1), 23-36.
28. Askari, J. A., Buckley, P. A., Mould, A. P., & Humphries, M. J. (2009). Linking integrin
conformation to function. Journal of cell science, 122(2), 165-170.
59
29. Kim, C., Schmidt, T., Cho, E. G., Ye, F., Ulmer, T. S., & Ginsberg, M. H. (2012). Basic amino-
acid side chains regulate transmembrane integrin signalling. Nature, 481(7380), 209-213.
30. Dai, A., Ye, F., Taylor, D. W., Hu, G., Ginsberg, M. H., & Taylor, K. A. (2015). The structure of a
full-length membrane-embedded integrin bound to a physiological ligand. Journal of Biological
Chemistry, 290(45), 27168-27175.
31. Fernández, C., & Wider, G. (2003). TROSY in NMR studies of the structure and function of large
biological macromolecules. Current opinion in structural biology, 13(5), 570-580.
32. Sušac, L., Horst, R., & Wüthrich, K. (2014). Solution ‐NMR Characterization of Outer ‐Membrane
Protein A from E. coli in Lipid Bilayer Nanodiscs and Detergent Micelles. Chembiochem, 15(7),
995-1000.
33. Kucharska, I., Edrington, T. C., Liang, B., & Tamm, L. K. (2015). Optimizing nanodiscs and
bicelles for solution NMR studies of two β-barrel membrane proteins. Journal of biomolecular
NMR, 61(3-4), 261-274.

Abstract (if available)

Abstract Integrins are a vital cellular membrane variable heterodimer that modulate inside-out/outside-in signaling, cellular adhesion and motility. The transmembrane helix of Integrin's beta chain is distinct as it is dispersed with a 28 degree tilt compared to the perpendicularly dispersed alpha chain helix. As the Integrin heterodimer is noted in a number of variable three dimensional spacial arrangements depending on activity and signaling context it is crucial to understand lipid helical anchoring in order to better understand structural implications of Integrin relative to signaling context. Previous studies conducted by Ulmer and Ginsberg determined a structure of the both Integrin transmembrane helices within a bicellar environment using liquid state NMR. Although these structures are high resolution there still remains uncertainty surrounding the signaling context and the physiological relevance of such structures. These proceedings introduce an alternative platform for studying lipid-anchored transmembrane helices of Integrin in using phospholipid nanodiscs. In producing an optimized phospholipid bilayer sample offering a more close representation of the physiological state this study puts forth methods whereby Integrin transmembrane helices can be structurally characterized in a more native environment.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Molecular basis of CD33 receptor function, and effects of a destabilizing transmembrane motif on αXβ2 integrin

PDF

Dual effects of transmembrane proline residues on integrin function

PDF

Study of integrin αIIbβ3 transmembrane and cytosolic domain interaction and the expression of extracellular domain by Pichia Pastoris

PDF

Expression and purification of extracellular domain of human CD33 in Escherichia coli and Pichia pastoris

PDF

Mechanistic studies of the disintegrin contortrostatin and characterization of the recombinant protein vicrostatin

PDF

The structure and function of membrane curving proteins on different membrane shapes and their regulation by post-translational modifications

Asset Metadata

Creator Frey, Benjamin Brandreth (author)

Core Title Integrin β3TM structural investigations using phospholipid nanodiscs

School Keck School of Medicine

Degree Master of Science

Degree Program Biochemistry and Molecular Medicine

Publication Date 07/19/2017

Defense Date 05/24/2017

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag integrin,Investigations,lipid helical anchoring,liquid state NMR,NMR,OAI-PMH Harvest,phospholipid nanodiscs,structural,using,β3TM

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Ulmer, Tobias (committee chair), Langen, Ralf (committee member), Siemer, Angsar (committee member)

Creator Email bbfxb2@gmail.com,Benjambf@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-403977

Unique identifier UC11264229

Identifier etd-FreyBenjam-5547.pdf (filename),usctheses-c40-403977 (legacy record id)

Legacy Identifier etd-FreyBenjam-5547.pdf

Dmrecord 403977

Document Type Thesis

Rights Frey, Benjamin Brandreth

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA