University of Southern California Dissertations and Theses

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

(USC Thesis Other)

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

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Molecular basis of CD33 receptor function, and effects of a
destabilizing transmembrane motif on αXβ2 integrin
By Han Vu
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(MEDICAL BIOPHYSICS)
August 2024

ii
Dedication
To my family:
My parents, for all their love, and the sacrifices they’ve made to raise me in a new world
My older siblings, for the help they’ve given, and examples they’ve set throughout my life
My husband, for being my pillar of support as I navigate my present and future

iii
Acknowledgements
First of all, I want to thank my PI, Dr. Tobias Ulmer. It is my wholehearted belief that any
PhD or Master’s student could only be so lucky to have him as a mentor. Not only is he a brilliant
scientist with incredible insight and knowledge, he is a patient, caring, and gentle advisor,
especially at times when I needed it most. He is always available for questions or help, and makes
grueling moments bearable. My PhD would have been impossible to complete without him.
I would also like to thank the other members of my thesis committee, Dr. Ansgar Siemer
and Dr. Ralf Langen. They are a ready font of knowledge, and they ask sharp questions and have
provided resources to nudge me to be a better researcher. Interacting with them during the
Medical Biophysics seminar courses has been a pleasure. I would like to acknowledge Dr. Siemer
in particular. We have worked together very closely, being next door to each other both in the
lab space and in the NMR room, where toiling over the helium regeneration system has brought
us closer.
I am also thankful for everyone in the Ulmer lab that I have been able to work with. Our
Master’s students, Mingke (Sabrina) Wu, Dai (Daisy) Zhai, Xuhang Dai, Jiaqi (Alex) Xiao, and
Meirui Qian have all been amazing to work with. Mingke welcomed me in to the lab during my
first year, and Dai and I worked closely together to troubleshoot code, and swapped homemade
treats. Jiaqi and Xuhang were fun to joke around with, and Meirui has been a diligent student as
she trains to take my place. Of course, I have to give special thanks to Alan Situ, our lab manager.
Alan is the backbone of the lab, and has helped me so much with training on equipment and
running experiments. It was delightful talking with him about life as we waited on experiments.
I want to thank everyone on the first floor of the Zilkha Neurogenetic Institute for being
such wonderful people. From Beni Ramos-Cruz and Ramiro Montano down the hall to Sayuri
Pacheco, Dr. Silvia Cervantes, Dr. Dhanya, Connor Hurd, Dr. Alexander Falk, Dr. Nitin Pandy, Dr.
Jobin Varkey, Dr. Anoop Rawat, Dr. Mario Isas, Elissa Fultz, Joshua Lugo, and everyone else in the
labs next door, the collaborative and open community here has made ZNI a great workplace that
cannot be matched.
Thank you also to the PIBBS community. I have made life-long friends among those in my
cohort. I have leaned on and commiserated with Ariel Vonk, Ashley Del Dosso, Muniza Junaid,
Rosanna Calderon, Youngjoo Lee, and Heather Ogana. We have laughed and cried together at
the beach, during the holidays, at each other’s weddings, over the course of childbirth, and more
– without them, my PhD would have been unbearably lonely. I would also be remiss to leave out
the Medical Biophysics program. Angela Albanese, Benjamin Fixman, Hannah Hartman, and
Janielle Cuala make science enjoyable to think about more deeply.
I want to thank National Institutes of Health for providing the funding for the research
taking place in the Ulmer lab. I also want to thank USC and the director of the PIBBS program, Dr.
Ite Offringa and the PIBBS administrative staff, past and present: Joyce Perez, Bami Andrada,
Leslie Ann Licazo, Domonique Walker, Karina Recinos, and Marisela Zuniga, for being available to
talk, and running such a wonderful program that somehow each year, brings together cohorts
full of amazing people.
Lastly, I would like to thank my family for all their support. My parents have sacrificed so
much and worked so hard to allow me to succeed in the United States. I am especially
appreciative of my father’s continued encouragement, particularly as he progresses along the

iv
disease that I address in this thesis. My siblings have all given me advice and helped me navigate
life as a first-generation Vietnamese American. My in-laws have welcomed me into their family
and been a second support system. And finally, my husband, David Ghermezi, cheers me up when
I am discouraged, and has provided me perfect love.

v
Table of Contents

DEDICATION………………………………………………………………………………………………………………………………...ii
ACKNOWLEDGEMENTS………………………………………………………………………………………………………………..iii
LIST OF TABLES……………………………….…………………………………………………………………………………….........vi
LIST OF FIGURES……………………………………………….…………………………………………………………………........vii
ABSTRACT………………………………………………………………………………………………………………………………….viii
INTRODUCTION………………………………………………………………………………….……………………………..…………1
CHAPTER 1 – The CD33 receptor dimerizes and utilizes lipid raft clustering as a signaling
mechanism, plus implications for the wider Siglec receptor family…………………………………............10
1.1 INTRODUCTION………………………………………………………………………………………………………………10
1.2 MATERIALS AND METHODS…………………………………………………………………………………………….17
1.3 RESULTS………………………………………………………………………………………………………………………….22
1.4 DISCUSSION……………………………………………………………………………………………………………………37
1.5 CONCLUSIONS………………………………………………………………………………………………………………..46
CHAPTER 2 – Prolines and charged residues at the transmembrane domain border affect
transmembrane domain stability and dimerization………………………………………………………..…………..48
2.1 INTRODUCTION………………………………………………………………………………………………………………48
2.2 MATERIALS AND METHODS…………………………………………………………………………………………….53
2.3 RESULTS………………………………………………………………………………………………………………………….57
2.4 DISCUSSION……………………………………………………………………………………………………………………66
2.5 CONCLUSIONS………………………………………………………………………………………………………………..67
FUTURE DIRECTIONS……………………………………………………………………………………………………………..……69
REFERENCES……………………………………………………………………………………………………………………………….72

vi
List of Tables
Table 1.1 Structural statistics for the human CD33 transmembrane domain………………………..24
Table 2.1 Mutant integrin αIIb(W968V)β3 TM Complex Stabilities……………………………………….59
Table 2.2 Integrin αXβ2 TM complex stabilities in large unilamellar vesicles (LUV)……………….65

vii
List of Figures
Fig. 1 CD33 regulates microglial phagocytosis……………………………………………………………………6
Fig. 2 Structure of the CD33 ectodomain…………………………………………………………………………..8
Fig. 1.1 Structure of the extracellular CD33 domain……………………………………………………………14
Fig. 1.2 CD33 domain organization and sequence properties……………………………………………..15
Fig. 1.3 2D TROSY spectrum of CD33(Q251-R291) with assigned residues…………………………..22
Fig. 1.4 Structural ensemble of the CD33 transmembrane domain in phospholipid
bicelles…………………………………………………………………………………………….……………….……23
Fig. 1.5 Structure of the transmembrane domain of CD33………………………………….……….……..25
Fig. 1.6 AlphaFold2 model of the human CD33 transmembrane domain……………………………26
Fig. 1.7 ClustalΩ multiple sequence alignment of the human Siglec transmembrane
domains…………………………………………………………………………………………………….…………..27
Fig. 1.8 2D TROSY spectrum of CD33(Q251-R291) and CD33(Q251-Q364 overlaid……….……..29
Fig. 1.9 ClustalΩ multiple sequence alignment of the depicted human CD33rSiglecs………….30
Fig. 1.10 Secondary 13Cα chemical shifts of the TM and TM-CS domains……………………………….31
Fig. 1.11 Secondary 13Cα chemical shifts of the TM-CS domains calculated using POTENCI and
XPLOR-NIH overlaid……………………………………………………………………………………………….32
Fig. 1.12 Characterization of the CD33 IgC1 domain expressed in P. pastoris………………………..33
Fig. 1.13 Biophysical characterization of the IgC domain of CD33………………………………………….34
Fig. 1.14 Surface area of IgV and IgC interfaces……………………………………………………………………..35
Fig. 1.15 β sheet organization of human Siglec IgC domains………………………………………………….36
Fig. 1.16 Domain organization of human Siglecs……………………………………………………………………37
Fig. 1.17 Structural basis of CD33 homodimerization and Siglec IgC(βC-βD-βE) sequence
alignment……………………………………………………………………………………………………………….38
Fig. 1.18 Superposition of CD33 IgC1 with the first IgC domains of human Siglecs………………..40
Fig. 1.19 Superposition of free and ligand-bound CD33 IgV-IgC1 domains…………………………….43
Fig. 1.20 CD33 structure and signaling hypothesis…………………………………………………………………45
Fig. 2.1 Structure of the αIIbβ3 integrin heterodimer signaling mechanism………………………..49
Fig. 2.2 ClustalΩ multiple sequence alignment of the transmembrane domains of human
integrin subunits…………………………………………………………………………………………………….51
Fig. 2.3 Integrin αX and β2 transmembrane structures………………………………………………………..58
Fig. 2.4 Comparison of {1H}-
15N NOE values for αX and αX mutants…………………………………….61
Fig. 2.5 Overlay of secondary 13Cα chemical shifts of αX, αX(P1088A), and αX(P1088W)……..61
Fig. 2.6 Overlay of secondary 13Cα chemical shifts of αX, αX(P1088A), αX(G1092L), and
αX(S1094I)……………………………………………………………………………………………………………..62
Fig. 2.7 Overlay of secondary 13Cα chemical shifts of αX in DMPC, POPC, and DEPC……………63
Fig. 2.8 Integrin αXβ2 transmembrane complex stability…………………………………………………….64

viii
Abstract
Representing just under 30% of the genome and ~50% of all drug targets, membrane
proteins are an important class of proteins to characterize structurally. However, they represent
less than 1% of structures obtained in the RCSB Protein Database. This is due to their hydrophobic
properties, as well as their propensity to contain dynamically unstructured regions, which makes
obtaining structure difficult using traditional methods. NMR has emerged as a tool that uniquely
addresses the shortcomings of those traditional methods.
Many membrane proteins are found within the immune system as signaling receptors.
Structural studies can elucidate the specific mechanisms of signaling used by these proteins.
Within the innate immune system, the CD33 receptor modulates microglial activity. It is a known
drug target for AML, and now shows promise in Alzheimer disease as a target for inhibition.
Developing such a drug requires knowledge of its signaling mechanism, which is as of yet unknown.
In this thesis, we finish the structure of human CD33 by refining the previously acquired crystal
structure of its extracellular IgV-IgC domains and combining it with NMR-based structures of its
transmembrane and cytosolic domains. We find that the IgC domain adopts a C1-type β-sheet
structure that homodimerizes in solution via βC-βC strand pairing and βC- βD packing. In
comparing computationally derived structures, while homodimerization at this particular interface
can be found in several human Siglec family receptors, only Siglec-6 is predicted to do so in a similar
non-covalent manner. Dimerization does not occur at the 21-residue helical transmembrane
domain, containing a conserved thin neck, thick belly appearance that transitions into a base of
positively charged residues that leads to the cytosolic tail. The cytosolic domain is dynamically
unstructured, but takes on a somewhat rigid extended conformation due to its proline content.
This suggests that CD33 does not utilize a monomer-dimer shift in signaling, like a number of other
membrane receptors. Instead, binding of the dimerized extracellular domains by multivalent
ligands may constrain movement of the monomeric transmembrane and unstructured cytosolic
domains to a degree that they induce co-localization with an activating kinase.
By contrast, a monomer-dimer shift has been found by the Ulmer lab to be utilized in the
signaling mechanism of integrins, a heterodimer family within the membrane protein class with
inside-out signaling capabilities that is involved in cell migration and adhesion. Inactive integrin
forms a dimerized at the transmembrane domains, while destabilization leads to activation. While
well-conserved across the family, differential functions have resulted in the evolution of 18
different α integrin subunits and 8 β subunits. Variety in function and sequence indicate variety in
regulation of α- and β-subunit dimerization, and thus of activation. αXβ2 is an integrin family
member that is implicated in immune function, and has been found in microglia. This thesis
suggests that in contrast to previously explored αIIbβ3, the transmembrane domain of αX is
destabilized in comparison to αIIb due to a motif at its N-terminal border that disrupt α-helix
backbone dynamics. Moreover, this motif interacts with lipid headgroups, signifying that αX is
sensitive to its lipid environment.
Overall, this thesis presents how structural features influence two different signaling
mechanisms available to membrane proteins that are active in the immune system.

1
Introduction
Structural biology is a subfield of biology that seeks to obtain an understanding of how
physics and chemistry dictate the arrangement and interplay of biomolecules, and thus explore
the relationship between structure and function at the molecular level. Protein chemistry in
particular draws from this subfield, especially so in the realm of drug development. Notable
achievements include many HIV-1-inhibiting drugs such as raltitrexed, which targets thymidylate
synthase, flurbiprofen, which targets cyclooxygenase-2 to treat rheumatoid and osteoarthritis,
and STX-0119, used to treat lymphoma by targeting STAT3 (Batool et al., 2019). In that sense,
unraveling the structure and understanding the structure-to-function relationship of membrane
proteins is of vital importance. Membrane proteins play a vital role in physiological events and
serve as regulators in numerous cellular processes, including cell signaling, metabolism, and
homeostasis. Defects in transmembrane proteins can lead to well-known diseases such as cystic
fibrosis (Lukasiak & Zajac, 2021), hypertrophic cardiomyopathy (Schlossarek et al., 2011), and
Alzheimer’s disease (AD) (Raj et al., 2014). Moreover, interactions between proteins and cell
membranes are crucial in signaling events, where disruptions in lipid-protein interactions can
impair functional regulation. Given their significance in both healthy and diseased physiological
systems, membrane proteins are often the target of pharmacological substances, constituting
approximately 50% of all drug targets despite representing less than 30% of the entire protein
coding genome (Ulmschneider & Ulmschneider, 2008).
Despite this, our understanding of membrane protein structure and function is vastly
inferior to that of water-soluble proteins. Gaining structural insights into membrane proteins is
essential for comprehending and appreciating their roles in physiological and pathological

2
signaling events. It provides valuable information on the initiation and transmission of molecular
signaling across the membrane, the influence of environmental factors, and the interactions
between transmembrane proteins, as well as protein-lipid interactions. Of over 200,000+ solved
unique protein structures, and excluding 1,000,000+ computationally derived structures, less
than 1% of structures in the Protein Data Bank are for membrane proteins (Bank, n.d.-c;
Membrane Proteins of Known Structure, n.d.). Even within the 8,000 known membrane proteins
in the human genome, only about 1,711 unique proteins have an elucidated structure (Martin &
Sawyer, 2019; Membrane Proteins of Known Structure).
This highlights the difficulties faced in the structure determination process for membrane
proteins. For example, conventional methods like X-ray crystallography struggle to overcome two
main factors: obtaining a high-resolution crystal, and physiological relevance. Crystal formation
is difficult due to the hydrophobic nature of membrane proteins, which necessitates the presence
of lipids or detergents to prevent aggregation, wherein the lipids themselves interfere with
crystal formation by competing for packing interactions. The hydrophobic portions also often
result in impurities that compromise the crystal lattice. Even after those challenges are
overcome, only a small fraction of crystal structures includes embedded lipids in the crystal unit's
cell, raising questions about their true relevance in a biological setting. Moreover, in the last
decade, scientists are increasingly aware of a new class of proteins that are intrinsically
disordered (IDPs) or have intrinsically disordered regions (IDRs). By definition, IDPs resist the
formation of organized crystal structures, and if they do, the resulting crystal structure presents
limited biological information. This is a substantial weakness, given that many membrane

3
proteins have IDRs, particularly at membrane borders, that are significant to their function (J.
Cornish et al., 2020).
Nuclear magnetic resonance (NMR) is one alternative to X-ray crystallography that
mitigates some of its weaknesses. Protein NMR can calculate protein structures at atomic
resolution by measuring the response of nuclei to a strong magnetic field. By virtue of the unique
chemical environment of each nucleus in the protein molecule, depending on the experiment
performed, the response of each nucleus reveals physical properties such as relative position,
conformation, interacting partners, interaction kinetics, etc. With regard to membrane proteins,
its strongest advantage is its ability to function under near physiological conditions, which
includes a membrane environment. Detergent bicelles are used to mimic the cell bilayer,
ensuring structures are obtained in physiologically relevant conditions. These bicelles are made
up of a ratio of short- to long-chain lipids, wherein the long-chain lipids form a central bilayer disc
that is commonly stabilized by the short-chain lipids (Sanders & Schwonek, 1992). Another
advantage is that because the sample is in solution, NMR is able to sample dynamic
conformational and topological changes of the protein and the kinetics associated with those
changes. This is especially important for membrane proteins with IDRs, which may sample many
conformations. For membrane proteins and their interacting partners, this captures the nature
of their biological activity.
Structure is not the end all be all of understanding membrane protein activity. LUV
fluorescence spectroscopy is a method that combines fluorescence spectroscopy with the use
of large unilamellar vesicles (LUV) to observe kinetics of membrane proteins in physiologically
relevant conditions. Fluorescence spectroscopy excites proteins at specified wavelength(s), and

4
then captures an emission spectrum, either at one specific wavelength or within a specified
range. Changes in emission spectra, depending on experimental conditions, can provide
association constants, folding energies, and more. Isothermal titration calorimetry (ITC) is
another technique used in protein chemistry to measure the heat changes associated with
biological interactions occurring in a sample. In ITC, a protein of interest resides in a sample cell
and a reference cell, and a binding partner is titrated into the sample cell. The heat flow used to
maintain a consistent temperature is measured by the calorimeter, and the data obtained can
be used to calculate the heat of reaction, heat capacity, or other thermodynamic parameters.
ITC is particularly useful in the field to explore binding interactions and phase transitions. Given
that membrane proteins often undergo complex restructuring and binding interactions, it
provides valuable information about the thermodynamics, kinetics, and energetics of these
processes, helping to understand their underlying mechanisms. Like LUV fluorescence
spectroscopy, ITC can also be performed in physiologically relevant conditions using bicelles,
and have been used with integrins to elucidate their binding energies and the properties that
control their activation (A. J. Situ et al., 2014).
As mentioned previously, membrane proteins are significant in Alzheimer’s disease. AD is
a widespread neurodegenerative disease that affects 20-30% of the global population over 65
years old (Masters et al., 2015). It is estimated that by 2050 AD will afflict up to 14 million people
in the United States alone and cost more 1.1 trillion dollars to manage (Winston Wong, 2020).
This would severely strain health-care and social service systems, and it is thus of vital importance
to develop a therapy for AD. AD is a chronic disease, and in the absence of therapy is progressive
and fatal. The pathology of AD is currently hypothesized to be characterized by the inability to

5
clear extracellular amyloid-β (Aβ) peptides and of tau neurofibrillary tangles, which aggregate
and provoke an immune response and neuronal death (Mucke, 2009). AD is frequently divided
into early-onset (EO, <60 years) familial AD, which represents <5% of AD and is linked to dominant
mutations in Aβ processing genes, and into late-onset (LO) sporadic AD, the majority at >95% of
cases (Hollingworth et al., 2011; Naj et al., 2011; Tanzi, 2013). Aβ peptide is the cleavage product
of amyloid precursor protein (APP) by β- and γ-secretases, and is normally important for neuronal
growth and development. In healthy individuals, Aβ is rapidly cleared from the brain, but when
its concentration is increased by excess production or faulty clearance it can self-aggregate and
form the oligomers, protofibrils, fibrils, and plaques that drive AD pathogenesis (Mucke, 2009).
However, despite clear clinical association between level of Aβ aggregation and stage of
AD progression in afflicted individuals, Aβ aggregation is estimated to be present in ~30% of
healthy individuals with no clinically apparent cognitive impairment, and there is also increasing
evidence that neural immune response plays a role in AD pathology (Perez-Nievas et al., 2013).
With regard to LOAD, this has led to a paradigm shift in how researchers approach AD etiology,
and it presently includes a variety of factors ranging from changes in neuronal plasticity to innate
immune activation (Tijms et al., 2024). Specifically, neuroinflammation is estimated to lead to 10-
100 times more neuronal death than the plaques themselves (Heneka et al., 2015). Thus, drugs
targeting aggregated Aβ may be too late. Current AD therapies utilize monoclonal antibodies
targeting Aβ to reduce plaque burden, but their efficacy is hotly debated, and many past Aβtargeting drugs have failed clinical trial (Bachurin et al., 2017; Zhang et al., 2023).
Genome-wide association studies (GWAS) found approximately 30 genes linked to the
immune system that are associated with AD (Hollingworth et al., 2011; Naj et al., 2011; Sims et

6
al., 2017). Five of the aforementioned genes (CD33, TREM2, CR1, MS4A6A and INPP5D) are
associated with microglia. Microglia are the macrophages of the CNS, and act as the primary
active and innate immune response (Kettenmann et al., 2011). They are distributed throughout
the brain and spinal cord, and make up 5-10% of brain cells present depending on the region
(Lawson et al., 1990). Depending on the local environment and cell-cell interactions, microglia
exhibit context-dependent phenotypes that can be pro-inflammatory or anti-inflammatory. For
example, when microglia are converted into an M1-like state they release pro-inflammatory
cytokines and free radicals leading to neuroinflammation that aggravates AD. When microglia
are converted to a phagocytic mode they clear Aβ better and display a neuroprotective
phenotype (Griciuc & Tanzi, 2021; Naj et al., 2011). Microglia are also responsible for neuronal
plasticity through synaptic pruning, and depending on context, both hyper- and hypo-plasticity
could contribute to AD pathology (Rajendran & Paolicelli, 2018; Tijms et al., 2024). In
Figure 1. CD33 regulates microglial phagocytosis. Mouse studies show wild-type microglia express
a low level of CD33, while AD risk allele rs3865444 is associated with higher levels of CD33 that
suppress the microglial phagocytosis. Knockdown of CD33 significantly improves the phagocytosis
capability of microglia (Hollingworth et al., 2011; Naj et al., 2011; Sims et al., 2017). Figure from
(Siew & Chern, 2018).

7
hyperplasticity, microglia inappropriately phagocytose synapses and result in neuronal death as
seen above (Rajendran & Paolicelli, 2018). In hypoplasticity, lack of pruning results in lack of
ability for the brain to adapt to synapse loss as the patient progresses (Pascual-Leone et al.,
2011).
AD risk allele rs3865444 in particular is related to increased CD33 expression, while the
minor allele is associated with lower CD33 expression and is protective (OR = 0.89) (Hollingworth
et al., 2011; Naj et al., 2011). CD33 is a member of the family of sialic acid-binding
immunoglobulin-like receptors (Siglecs) that help distinguish between self and non-self, and that
regulate the function of cells in the innate and adaptive immune systems through the recognition
of sialylated glycan ligands (Macauley et al., 2014). They are commonly expressed in leukocytes
such as microglia, but the structural mechanics of Siglec signaling are little studied. Mouse studies
have shown that CD33 knockout improves microglial phagocytosis and mitigates Aβ pathology
(Fig. 1) (Siew & Chern, 2018). CD33 inhibition may thus represent a novel therapy for AD, and so
this study aims to elucidate the structural basis of CD33 signaling.
There are structural features known about CD33 and the Siglec family receptors. The Siglec
receptor family contains four different domains: two are extracellular immunoglobulin domains,
one IgV that binds its specific sialylated glycan ligand and a variable number and type of IgC
immunoglobulin domains, and then there is a transmembrane (TM) domain and a cytosolic tail
(Varki & Angata, 2006). Some are known to dimerize, and most contain some number of
immunoreceptor tyrosine-based inhibitory motifs (ITIM) with sequence S/I/V/LxYxxI/V/L and
ITIM-like motifs that when phosphorylated trigger an inhibitory downstream signaling cascade (S.
Siddiqui et al., 2017). Human CD33 has one IgV domain, one IgC domain, a transmembrane

8
domain, and a cytosolic tail with one ITIM and one
ITIM-like motif. A crystal structure has been found for
the IgV and IgC domains (Fig. 2) (Dodd, 2017) which
suggests that CD33 dimerizes at the IgC domain, but
it remains to be seen if this holds true under
physiological condition. Moreover, the
transmembrane and cytosolic tail remain to be
elucidated, and TM domains often contain significant
structural features to membrane protein function.
(Martin & Sawyer, 2019).
Aside from CD33, integrins are also
implicated in AD. Integrins are cell surface receptors
that are involved in cell adhesion and signaling
processes, and are crucial for maintaining the integrity and selective permeability of the bloodbrain barrier (BBB). Integrins are also implicated in Alzheimer's disease (AD) neuroinflammation
itself (Hogg et al., 2011). Leukocyte surface integrins play a crucial role in antigen presentation
within the central nervous system (CNS). Similarly, activated microglia express various integrins,
which can be upregulated in response to inflammation and injury. β2 has also been found to be
constitutively expressed in microglia, and an increase in expression of αXβ2 was observed in
reactive microglia, early evidence that AD can trigger an inflammatory response (Akiyama &
McGeer, 1990) and serving as biomarkers of endothelial activation and neuroinflammation (Rossi
et al., 2011).
Figure 2. Structure of the human CD33
ectodomain. The X-ray structure of CD33
indicates a non-covalent, homodimeric
assembly. Disulfide bonds are shown in
yellow (PDB entry 5ihb).

9
The integrin family exhibits a remarkable conservation in both sequence and structure
among its members. Each integrin is composed of one α and one β subunit forming a heterodimer,
featuring multiple globular extracellular domains, a single transmembrane (TM) domain, and a
relatively short cytoplasmic tail at the C-terminal (Arnaout et al., 2005). In humans, there are 18
different α subunits and 8 different β subunits (Hughes et al., 1996). While the extracellular
regions of the α and β subunits make up the majority of each protein and vary in length and
domain composition, the β integrin subunit, approximately 700 residues long, comprises a PSI
domain, a hybrid domain, four I-EGF domains, and a tail domain (Arnaout et al., 2005; Springer &
Wang, 2004; Zhu et al., 2008). In contrast, the α subunit, ranging from 940 to 1,120 residues in
length, includes a Propeller, a Thigh, and Calf1 and Calf2 domains (Zhu et al., 2008). Their
cytoplasmic domains are relatively small, with the α and β tails consisting of 15-78 and 46-70
amino acids, respectively (Wegener et al., 2007). The β tail, predominantly unstructured, shows
high sequence similarity across the family, with a membrane-proximal helix and a membranedistal unstructured region, while the α tails contain a highly conserved GFFKR motif in the
membrane-proximal portion, and the β tails feature two NPxY or NPxY-like motifs (Rocco et al.,
2008; Wegener et al., 2007). Notably, the β tail serves as a central hub for interactions between
integrins and cytoplasmic proteins, thereby facilitating integrin-related signaling pathways. As the
integrin family expanded its repertoire of functions, it evolved to encompass multiple members.
This diversity results in 24 unique integrin proteins, each fulfilling its own distinct yet nonredundant physiological function, as demonstrated by knockout (KO) mice studies (Hodivala-Dilke
et al., 1999; Tsakiris et al., 1999). Given the non-redundant physiological roles of these 24
integrins, there is considerable interest in elucidating their function-to-structure relationship.

10
The Ulmer lab has previously identified the regulatory mechanism of activation for the
αIIbβ3 integrin heterodimer through the work of Dr. Thomas Schmidt. Integrins are well known
for having bidirectional inside-out or outside-in allosteric signaling (Biose et al., 2023; Hogg et al.,
2011). During an outside-in signaling event, an extracellular ligand triggers allosteric changes and
integrin clustering, which leads to the recruitment and activation of tyrosine kinases and focal
adhesion kinases (Hynes et al, 2002). Inside-out signaling utilizes a far more complex mechanism,
where talin binding disrupts the interaction between the cytoplasmic and TM portion of the α
and β subunits, triggering structural changes to the extracellular domains which will eventually
result in an increased binding to the ECM(Kim et al., 2009; Wegener et al., 2007). Due to its
complex strategy, the insight-out signaling process is quite unique and cannot be found in other
physiological systems in nature, making an understanding of the molecular underpinnings of this
process highly important.
Using solution state NMR in order to obtain structural and kinetic data, this thesis looks
to explore the moleclar mechanisms behind the function of two transmembrane proteins
involved in AD and general immune function and signaling: CD33, and integrins. With a dearth of
information on the structural basis of CD33 signaling, and of the Siglec family in general, the
research aims to establish a hypothesis for CD33 signalling and how structural features may
contribute to its function. Prior work in the Ulmer lab with integrins has established a
dimerization activation model for the αIIbβ3 TM heterodimer, and further understanding of
function in the integrin family of heterodimers is sought by expanding into the αX and β2 subunits
and their interactions. Research of both contributes further to understanding of membrane
protein signaling mechanisms at a biophysical level.

11
CHAPTER 1 – The CD33 receptor dimerizes and utilizes lipid raft
clustering as a signaling mechanism, with implications for the wider
Siglec receptor family
1.1 – Introduction
Alzheimer’s disease (AD) is a neurodegenerative disorder affecting 20-30% of the senior
population. The current hypothesis of pathology is that extracellular amyloid-β (Aβ) and tau
aggregate into plaques and neurofibrillary tangles that provoke neuroinflammation and neuronal
death. Consequently, present therapies target Aβ peptide burden through antibody-based drugs
(van Dyck Christopher H. et al., 2023), but with limited efficacy. However, the exact relationship
between Aβ aggregation and AD is not clearly understood. Aβ aggregates can be present in
healthy individuals (Perez-Nievas et al., 2013), and researchers increasingly view AD as a
multifactorial disease (Tijms et al., 2024). Studies indicate that the immune response to the
plaques initiated by the central nervous system (CNS) can either exacerbate or alleviate AD
pathology (Bhattacherjee et al., 2021; Butovsky & Weiner, 2018). Microglia, the resident
phagocytes of the CNS, play a major role in the immune response of the CNS and the clearance
of Aβ aggregates (Butovsky & Weiner, 2018; Siew & Chern, 2018). It also plays a large role in
synaptic pruning, and thus plasticity of the brain during and following inflammatory damage due
to AD (Pascual-Leone et al., 2011). Microglia can take on a pro- or anti-inflammatory phenotype
depending on cellular context (Heneka et al., 2015), and represent a viable drug target to enhance
Aβ phagocytosis and reduce neuroinflammation.
Genome wide association studies have implicated microglial risk allele rs3865444 in AD,
which increases the expression of CD33 (Naj et al., 2011), a Siglec-type receptor that has been

12
shown to modulate activation and phagocytic capability. CD33 activation lowers microglial uptake
of Aβ plaques, thus exacerbating AD pathology, while inactivation improves AD pathology
(Bradshaw et al., 2013; Griciuc & Tanzi, 2021). CD33 is a member of the family of Siglec receptors
that help distinguish between self and non-self, and that regulate the function of cells in the
innate and adaptive immune systems through the recognition of self-associated sialylated glycan
ligands expressed on cell surfaces (Gonzalez-Gil et al., 2023; Gonzalez-Gil & Schnaar, 2021). In
humans, the Siglec receptor family has 15 members, 4 of which are classified as conserved Siglecs,
and 11 of which are classified as CD33-related Siglecs (CD33rSiglecs) (Angata & Varki, 2023). The
CD33rSiglecs diverged from the conserved Siglecs some 180 million years ago, developing in
mammals and mostly containing inhibitory motifs (Cao & Crocker, 2011), and thus form an
attractive target for immune inhibitory drugs (Crocker et al., 2012). They are commonly expressed
in leukocytes such as microglia, and although immunomodulation by Siglec receptors has been
extensively studied, the relationship between structure and function is poorly understood in this
receptor class. Siglecs are type I transmembrane proteins that generally consist of several
extracellular immunoglobulin domains: an N-terminal IgV-set domain that binds a unique
sialylated glycan ligand and a varying number of IgC-set domains, a single-pass transmembrane
(TM) domain, and a cytoplasmic tail (Gonzalez-Gil et al., 2023; S. Siddiqui et al., 2017). Different
Siglec receptors are believed to exist either as monomers, non-covalent dimers or disulfide-linked
dimers (S. Siddiqui et al., 2017).
With regard to AD, CD33 activation lowers microglial uptake of Aβ plaques, thus
exacerbating AD pathology, while inactivation improves AD pathology (Bradshaw et al., 2013;
Griciuc & Tanzi, 2021). It has 4 main domains: a terminal V-set domain, a singular C2-set

13
domain, a single pass transmembrane domain, and an intracellular domain containing
immunoreceptor tyrosine-based inhibitory motifs (ITIMs) that are crucial for its immune
function (S. S. Siddiqui et al., 2019). The IgV domain contains a sialoglycan-binding site centered
on a conserved arginine residue that ligates the sialic acid carboxylate (Fig. 1.1a)(Angata & Varki,
2023). CD33 binds the sialylated sulfated sequence Neu5Acα2-3[6SO3]Galβ1-4GlcNAc (Büll et
al., 2021) and its physiological ligands include keratan sulfate proteoglycans in the brain
(Gonzalez-Gil et al., 2022). With only one IgC domain, CD33, together with Siglec-15, is the
smallest Siglec receptor (Angata & Varki, 2023; Crocker et al., 2007). Studies cross-linking CD33
to antibodies shows it localizes to lipid raft domains in the membrane (Pérez-Oliva et al., 2011),
suggesting that lipid raft clustering of the receptors is part of its signaling mechanisms.
Following ligand binding, the ITIM can be found by Src family kinases that phosphorylate the
ITIM and consequently inhibit downstream pathways (Paul et al., 2000). ITIMs are a motif with
sequence S/I/V/LxYxxI/V/L that is phosphorylated at the tyrosine residue, and then
dephosphorylated to inhibit downstream pathways (Paul et al., 2000; Walter, Häusermann, et
al., 2008). It's noteworthy that the IgC domains of several Siglecs have been observed to form
dimers through disulfide bonds (A. L. Cornish et al., 1998; Floyd et al., 2000; S. Siddiqui et al.,

14
2017; Yu et al., 2001) and the third and fourth IgC domains of MAG (Siglec-4) have been
reported to homodimerize non-covalently (Pronker et al., 2016). While CD33 does not exhibit an
unpaired Cys, X-ray crystallography was used to obtain a structure for the V-set and C2-set
ectodomain (Fig. 2), and it suggests they dimerize at the interface of the IgC domains (Fig. 1.1)
(Dodd, 2017). However, crystal structures often utilize repeating contacts in order to generate
organized, packed crystals and unit cells that contain oligomeric assemblies that may not be
biologically relevant. The TM and cytosolic domain also remain to be found, regions which
constitute significant biological function in membrane proteins (Martin & Sawyer, 2019).
Figure 1.1. Structure of the extracellular CD33 domain. A, Crystal structure of the IgV-IgC domains
of human CD33 in complex with 3′-sialyllactose (PDB ID 5j06; chains A+D). Of the four molecules in
the unit cell, two are shown in blue and red. B, IgC dimerization in the single crystal. In the absence of
3′-sialyllactose, the crystallographic unit cell also contains four IgV-IgC molecules (PDB ID 5ihb). All IgV
domains are shown in light blue. Two IgC molecules in the unit cell, shown in red and blue, make
homodimeric contacts with each other, whereas the remaining two, also shown in red, pack the crystal
with such contacts as revealed by the adjacent IgC domain shown in light red.

15
Computational modelling of these unknown TM and cytosolic domain predicts a helical
TM and an almost entirely dynamically unstructured cytosolic tail with a possible small region of
secondary structure (Fig. 1.2). Although dynamically unstructured, cytosolic receptor tails serve
as hubs for the assembly of signaling complexes receptor tails and are often stabilized by
ligand/agonist binding (Ulmer et al., 2001). The ITIMs of the cytosolic tail are centered on Y340
and Y358. Upon ligand binding, these ITIMs become phosphorylated by Src family kinases, leading
to intracellular signaling pathways involving the recruitment of phosphatases like Src homology
region 2 domain-containing phosphatase-1 (SHP-1) and Src homology region 2 domain-containing
phosphatase-2 (SHP-2) (S. S. Siddiqui et al., 2019). These signaling events can have a variety of
phenotypes. In AML, phosphorylation of the ITIMs is followed by suppression of cytokine signaling
3 (SOCS3) recruitment, and consequently E3 ubiquitin ligase marking for proteasomal
degradation of both CD33 and SOCS3 (Orr et al., 2007). By contrast, in AD, CD33 downregulation
improves amyloid-β uptake and AD pathology, and it is hypothesized that CD33 recruitment of
Figure 1.2. CD33 domain organization and sequence properties. Data in green originates from
UniProtKB. Variation data (sourced from UniProt) shows non-genetic variation from the ExPASy and
dbSNP websites. Data in yellow originates from Pfam. Data in purple originates from Phosphosite.
Protein disorder predictions are based on RONN (Red: potentially disorderd region, Blue: probably
ordered region.) Hydropathy has been calculated using a sliding window of 15 residues and summing
up scores from standard hydrophobicity tables (Red: hydrophobic, Blue: hydrophilic).

16
SHP1/2 leads to downregulation of spleen tyrosine kinase (Syk) signalling to the PI3K pathways
triggering phagocytosis (Huang & Xu, 2019). As such, it is already a known drug target for acute
myeloid leukemia, and is a proposed drug target for AD (J. Liu et al., 2022; Malik et al., 2015). It
has also been suggested that SHP2 can modulate CD33 internalization depending on the cell line
(Walter, Raden, et al., 2008), but no structural studies have been performed. Obtaining the
structure of the cytosolic domains will often allow us to predict the structure of the ligand boundstate.
Modelling further predicts that the connection between the ectodomain and TM domain
is flexible. The backbone flexibility of this linker likely affects the allosteric properties of the
receptor. While the role of the TM domain is unknown for CD33, changes in TM domain structure
allow cytosolic ligands to sense the altered signaling state and trigger signaling (Schmidt, Ye, et
al., 2016), which for virtually any receptor is incompletely understood. By studying the structural
and thermodynamic parameters behind CD33 activation and interactions, we can better
understand how CD33 contributes to microglial activity, and thus how it can generate a microgliabased immune response that exacerbates or ameliorates AD pathology.
Here, we report the CD33 TM domain structure, characterize the structural preferences
of the cytosolic domain, and establish the oligomerization state of the IgC2 and TM domains. As
introduced, in order to obtain structures that are as close to physiologically relevant as possible,
we use solution state NMR within a bicelle membrane mimic. For an α-helical TM domain that is
estimated to be approximately 20-25 amino acids (approximately 30- 38 Å), having a long-chain
lipid such as DMPC, with its 14 carbon length chain is appropriate to span the length of the TM
domain (Cheng et al., 2009). Various NMR experiments are performed in order to generate an

17
ensemble of structures and association data. Using this information, the corresponding properties
for the human Siglecs family are predicted and the implications for Siglec signaling are discussed.
1.2 - Materials and Methods
Expression and purification of transmembrane CD33 peptides—
Using the cDNA of human CD33 (Mammalian Gene Collection Clone ID 5217182, UniProt entry
P20138) as template, an insert encompassing Gln251-Gln364 was generated by PCR and subcloned
into the pET-44 expression vector (Novagen, Inc.) with maltose binding protein (MBP) as Nterminal fusion protein and an intervening TEV protease cleavage site. A shorter variant,
encompassing Gln251-Arg291, was subsequently constructed using standard techniques.
Expression was induced in E. coli BL21(DE3)pLysS,T1R cells (Sigma-Aldrich, Inc.) cultured at 37 °C in
M9 minimal medium, containing combinations of 99% 13C-D glucose, 99% 15NH4Cl, and 99% D2O,
at an OD600 of 1.0 by adding IPTG to 1.0 mM. Cells were lysed by sonication in 50 mM
NaH2PO4/Na2HPO4, pH 7.4, 300 mM NaCl, 100 mM SDS, 20 mM imidazole, 2 mM βmercaptoethanol. The clarified lysate was applied on a HiTrap IMAC HP column (GE Amersham,
Inc.) charged with Ni2+ for immobilized metal affinity chromatography (IMAC). The column was
washed with 50 mM NaH2PO4/Na2HPO4, pH 7.4, 300 mM NaCl, 25 mM SDS solution, followed by
50 mM NaH2PO4/Na2HPO4, pH 7.4, 300 mM NaCl, 8 M urea, 20 mM imidazole, and 50 mM
NaH2PO4/Na2HPO4, pH 7.4, 300 mM NaCl, 8 M urea, 50 mM imidazole. Then, bound protein was
eluted in 50 mM NaH2PO4/Na2HPO4, pH 7.4, 300 mM NaCl, 8 M urea, 300 mM imidazole solution.
The peptide was cleaved from the fusion protein using TEV protease at a molar ratio of 1:50

18
overnight at 30 °C in 50 mM Tris·HCl, pH 8.0, 1 mM DTT solution, leaving a Gly as the N-terminal
residue preceding Gln251. Uncleaved protein and fusion protein were removed by IMAC and the
column flow-through was applied on a Hamilton PRP-3 reverse-phase HPLC column. The peptides
were eluted using a linear gradient from 70%/30% buffer A (H2O, 0.1% TFA) / buffer B (70%
acetonitrile, 30% 1-propanol, 0.1% TFA) to 10%/90% in 30 min. Subsequent to freeze-drying,
peptide purity was verified by SDS-PAGE and NMR.
Expression and purification of CD33 IgC1 domain—
An insert encompassing Asp140-Thr232 of the human CD33 gene was generated by PCR and
subcloned into the pPIC9K yeast expression vector (Invitrogen) with the α-mating signal sequence,
and a N-terminal Flag-tag for detection. Additionally, the C169S substitution was implemented to
avoid unspecific disulfide bond formation (Fig. 1.1). Electrocompetent histidine auxotrophic Pichia
(Komagataella phaffii) cells (his4-∆1; BioGrammatics, Inc. BG12) were transformed. Single colonies
were picked and cultured in 4 ml of BMGY medium (1% yeast extract, 2% peptone, 4·10-5% biotin,
100 mM KH2PO4/K2HPO4, pH 7.5, 1.34% yeast nitrogen base, and 1% glycerol) for 24 hours (30 °C
and 200 rpm). The culture was spun down and the cell pellet was resuspended in 4ml of BMMY
medium (1% yeast extract, 2% peptone, 4·10-5% biotin, 100 mM KH2PO4/K2HPO4, pH 7.5, 1.34%
yeast nitrogen base, and 0.5% methanol) to a OD600 of 1 and grew for 24 hours (30 °C and 200
rpm). After 24 hours, fresh additional methanol was added (0.5%), and cells grown for another 24
hours. Subsequently, cells were pelleted by centrifugation and the supernatant analyzed by
Western blot using the monoclonal anti-Flag antibody conjugated with M2-alkaline phosphatase
(Sigma Aldrich, cat. no. A9469). The best expressing clone was grown in BMGY medium for 16

19
hours at 30°C and 250 rpm. This culture was spun down and the cell pellet was resuspended in 1
liter of BMMY medium to a OD600 of 2. Cells were grown for 72 hours at 25 °C and 250 rpm, and
methanol was replenished every 12 hours to a concentration of 0.5%. The supernatant was
collected by pelleting the cells by centrifugation (4 °C and 4000g for 20 min.), the pH was adjusted
to 7.5 using 8 M NaOH, stirred on ice for 2 hours, and filtered (0.2 µm). The supernatant was
applied on a HiTrap IMAC HP column (GE Amersham, Inc.) charged with Ni2+ for immobilized metal
affinity chromatography (IMAC). The column was washed with 50 mM KH2PO4/K2HPO4, pH 7.5, 300
mM NaCl, 20 mM imidazole. Then, the bound protein was eluted in 50 mM KH2PO4/K2HPO4, pH
7.5, 300 mM NaCl, 300 mM imidazole. For deglycosylation, Endo H (UniProt entry P04067)
(Robbins et al., 1984) was added to a final concentration of 0.04 µM and incubated overnight at
37 oC. The eluate was concentrated to 0.2 ml and applied on a Superdex 75 Increase 10 300 GL
column equilibrated with 10 mM NaH2PO4/Na2HPO4, pH 7.4, 140 mM NaCl, 3mM KCl.

NMR sample preparation—
TM peptide concentrations were measured in acetonitrile-water solution by UV spectroscopy
(ε205nm= 149,580 M-1cm-1 and ε280nm= 3,105 M-1cm-1 for Gln251-Arg291 and Gln251-Gln364,
respectively) and defined amounts of peptide were freeze-dried. Peptides were taken up in 320 μL
of 350 mM 1,2-dihexanoyl-sn-glycero-3-phosphocholine (DHPC), 105 mM 1,2-dimyristoyl-snglycero-3-phosphocholine (DMPC), 6% D2O, 0.02% w/v NaN3 buffered by 25 mM HEPES·NaOH, pH
7.4 to yield concentrations of 1.1 mM and 0.8 mM for Gln251-Arg291 and Gln251-Gln364,
respectively. The Gln251-Arg291 peptide-bicelle complex was aligned relative to the magnetic field
by a stretched, negatively charged polyacrylamide gel of 320 μL volume, which was polymerized

20
from a 4.2% w/v solution of acrylamide (AA), 2-acrylamido-2-methyl-1-propanesulfonate (AMPS)
and bisacrylamide (BIS) with a monomer-to-crosslinker ratio of 49:1 (w/w) and a molar ratio of
95:5 of AA to AMPS (Ulmer et al., 2003). For IgC1(D140-T232 with C169S) a 0.4 mM sample (ε280nm=
12,615 M-1cm-1) in 10 mM NaH2PO4/Na2HPO4, pH 7.4, 140 mM NaCl, 3mM KCl was prepared.
NMR spectroscopy—
NMR experiments were carried out on a cryoprobe-equipped Bruker Avance 700 spectrometer at
35 °C. Data were processed and analyzed with the nmrPipe package and CARA. 2H/13C/15N-labeled
N peptide and TROSY-type H-N detection (Pervushin et al., 1997) was used for HNCA, HNCACB,
HN(CA)CO and HNCO-based backbone assignments, the measurement of 3
JC’Cγ and 3JNCγ coupling
(Hu et al., 1997), and the detection of 1JNH, 1JCαC’, 1JC’N as well as 1JNH+1DNH, 1JCαC’+1DCαC’, 1JC’N+1DC’N
couplings (Chou et al., 2000; Fitzkee & Bax, 2010; Jaroniec et al., 2004; Kontaxis et al., 2000) using
isotropic and aligned samples, respectively.
Structure calculation of the CD33 TM domain—
The bicelle-embedded CD33 TM peptide structure of the well-folded Ala265-His285 residues was
calculated by simulated annealing, starting at 3000 K using the program XPLOR-NIH (Schwieters et
al., 2003). The peptide termini were represented by random-coil conformations. Backbone
dihedral angle constraints were derived from N, Hα Cα, Cβ and C’ chemical shifts using the program
TALOS-N (Shen & Bax, 2015). χ1 side-chain angle restraints were derived from the 3JC’Cγ and 3JNCγ
coupling constants(Hu et al., 1997). In addition to standard force field terms for covalent geometry
(bonds, angles, and improper dihedrals) and nonbonded contacts (Van der Waals repulsion),

21
dihedral angle and interproton distance restraints were implemented using quadratic square-well
potentials, and a backbone-backbone hydrogen-bonding potential and torsion angle potential of
mean force were employed (Grishaev & Bax, 2004; Kuszewski et al., 1997). The difference between
predicted and experimental residual dipolar couplings (RDC; Δ1D) was described by a quadratic
harmonic potential. The final values for the force constants of the different terms in the simulated
annealing target function are as follows: 1,000 kcal·mol-1·Å-2 for bond lengths; 500 kcal·mol-1·rad-2
for angles and improper dihedrals, which serve to maintain planarity and chirality; 4 kcal·mol-1·Å-4
for the quartic Van der Waals repulsion term; 30 kcal·mol-1·Å-2 for interproton distance restraints;
500 kcal·mol-1·rad-2 for dihedral angle restraints; 0.3 kcal·mol-1·Hz-2 for 1DNH RDC restraints and
1DC’N and 1DCαC’ scaled relative to 1DNH according to their dipolar interaction constants; 1.0 for the
torsion angle potential; and a directional force of 0.20 and a linearity force of 0.05 for the
hydrogen-bonding potential. A total of 20 structures were calculated (Fig. 1.4 and Table 1). Analysis
of the surface area of human CD33 ectodomains (PDB: 5j06) was performed using the SASA
function of GROMACS v5.8 (Abraham et al., 2015; Dodd et al., 2017)

22
1.3 – Results
CD33 forms a monomeric, 21-residue transmembrane helix
We first determined the structure of the CD33 TM domain in isotropic phospholipid
bicelles by expressing and purifying the CD33 Gln251-Arg291 construct and performing
multidimensional heteronuclear NMR spectroscopy. Given the absence of long-range
interproton distance restraints (NOEs) for a linear α-helix, structural restraints relied on H-N, C-C
and C′-N residual dipolar couplings. A 2D HN-N correlation TROSY experiment was obtained,
allowing a fingerprint of the CD33 TM domain (Fig. 1.3). Further experimentation generated an
ensemble of 20 structures calculated by simulated annealing, with a coordinate precision of 0.15
Å for backbone heavy atoms (Table 1 and Fig. 1.4). Residues Ala265-His285 adopted α-helical
Figure 1.3. 2D
TROSY spectrum of
CD33(Q251-R291)
with assigned
residues. HN-N
correlation NMR
spectrum of
2H/13C/15N-labeled
CD33(Q251-R291) in
350 mM DHPC, 105
mM DMPC, 6% D2O,
0.02% w/v NaN3, 25
mM HEPES•NaOH,
pH 7.4 at 35° C and
700 MHz.

23
conformation (Fig. 1.5), revealing a TM helix length of 21 residues. This helix differs from the
AlphaFold2 model (Fig. 1.6), highlighting the need for an experimental approach. The first
charged residue on the intracellular side, Lys283, establishes the cytosolic helix orientation
(positive-inside rule) (von Heijne, 1992) and, apart from this residue, no canonical membrane
anchors such as Trp or Tyr are present (Fig. 1.3 and 1.5a). The side-chain distribution along the
TM helix shows an interesting pattern. Within the extracellular leaflet, residues with small
sidechains, Ala265, Gly266, Ala269 and Ala272, constitute a relatively thin helix (Fig. 1.5a-b). This
Figure 1.4. Structural ensemble of the CD33 transmembrane domain in phospholipid
bicelles. Superposition of all 20 calculated simulated annealing structures of the CD33 TM
domain. The backbone and side-chains are shown in red and blue, respectively. Structural
statistics are summarized in Table 1.

24
Table 1. Structural statistics for the human CD33 transmembrane domaina
R.m.s. deviations from experimental dihedral restraints (deg) 0.01 ± 0.04
All (43)b
R.m.s. deviations from experimental residual dipolar couplings (Hz)c
1
DNH (19) 0.98 ± 0.08
1DNC´ (21) 0.66 ± 0.04
1DCαC´ (21) 0.49 ± 0.04
R.m.s. deviations from experimental distance restraints (Å)
All (5) 0.04 ± 0.02
Interresidue short range (1 < |i - j| < 5) (5) 0.04 ± 0.02
Deviations from idealized covalent geometry
Bonds (Å) 0.02 ± 0.01
Angles (deg) 0.47 ± 0.40
Impropers (deg) 0.42 ± 0.05
Coordinate precision (Å)d
Backbone non-hydrogen atoms 0.15
All non-hydrogen atoms 0.82
Measures of structural quality
ELJ (kcal mol-1)
3 e -128.1
Residues in most favorable region of Ramachandran plotf 100%
a
Statistics for all 20 calculated simulated annealing structures, encompassing structured residues
CD33(G264-H285).
bTorsion angle restraints included 21 φ, 21 ψ, and 1 determined χ1 angles. c
R.m.s. deviations are normalized to an alignment tensor magnitude of 10 Hz.
dDefined as the average r.m.s. difference between the 20 simulated annealing structures and the
mean coordinates.
e
The Lennard–Jones van der Waals energy was calculated with the CHARMM PARAM 19/20
parameters and not included in the simulated annealing target function.
f
Calculated using PROCHECK V3.5.4 (Laskowski et al., 1993).

25
contrasts the following relatively bulky residues; two Phe residues are particularly noteworthy.
On the intracellular membrane-proximal side, a cluster of positive charges formed by R286-R287-
K288 and R291 gives the membrane-proximal region a distinctively positive charge (Fig. 1.3 and
1.5b). Thus, the CD33 TM helix is composed of a thin neck and a thicker belly, which is followed
by a positively charged, cytosolic patch.
To convey a signal across the membrane, receptors may change the association state of
their TM domain(s) (Arkhipov et al., 2013; Bocharov et al., 2008; Lau et al., 2009). In case of CD33,
Figure 1.5. Structure of the transmembrane domain of CD33. A-B, Cartoon representation of the
average CD33 TM domain structure and its surface charge distribution color-coded by electrostatic
potential. C-D, Structural model including electrostatic potential of the TM domain of human Siglec-7.
Electrostatic potentials were calculated using APBS (Baker et al., 2001).
(a) (b)
(c) (d)

26
such a scenario would require TM helix homodimerization. To assess the default oligomerization
state of the TM domain, we have determined the rotational correlation time, termed τc, of the
bicelle-TM helix complex, which informs on particle size. With reference to the values of bicelleimmersed monomeric and dimeric integrin αIIbβ3 TM peptides of approximately 20 ns and 30 ns,
respectively (Suk et al., 2012), the determined τc of 16.9 ± 0.2 ns bicelles identifies the CD33 TM
helix as monomeric. We also view its structural features not to be conducive for homodimeric helixhelix interactions. Packing in the “neck” region would lead to a relatively large crossing angle, which
may require a longer TM helix length than 21 residues to remain membrane immersed (Lau, Dua,
et al., 2008). Likewise, the “belly”, centered on consecutive Phe residues (Phe279-Phe280), seems
unsuited to establish defined contacts and its “back” is also characterized by relatively featureless
Figure 1.6. AlphaFold2 model of the human CD33 transmembrane domain. Helical
conformation is predicted for Ala265-Val294. The model is color-coded according to the per-residue
model confidence score (pLDDT).

27
hydrophobic surface (Fig. 1.5a-b). Accordingly, the CD33 receptor is unlikely to convey its ligand
engagement via a monomer-dimer equilibrium of its TM domain.
A neck-belly-positive patch is universal to TM domain of CD33rSiglecs
Human Siglecs are divided into four conserved receptors (SN, CD22, MAG and Siglec-15) and
eleven CD33rSiglecs that have evolved relatively rapidly (Angata & Varki, 2023; Crocker et al.,
2007). To extend the finding of the CD33 TM domain structure and to assess conserved structural
features, we have aligned all human TM domain sequences (Fig. 1.7). The conserved receptors only
contain the cytosolic positively charged patch of CD33 but otherwise appear relatively featureless.
Siglec-14, 15 and 16 associate intramembranously with adaptor proteins such as DAP12 (Angata &
Varki, 2023; Crocker et al., 2007). Siglec-14 appears most distinct from all other Siglecs; it lacks the
positive patch and a high incidence of Tyr and Trp anchors in the intracellular membrane leaflet
Figure 1.7. ClustalΩ multiple sequence alignment of the human Siglec transmembrane domains.
Conserved amino acids are colored by the Jalview multiple alignment editor (Clamp et al., 2004) using
the ClustalX color scheme with manual modifications.

28
make this TM helix border ambiguous. In contrast, the Arg reported to make intramembrane
contact with DAP12 (Angata et al., 2006) is relatively far N-terminal compared to Siglec-15 and -16.
Siglec-16 exhibits the positive patch and shares a C/S/YxxC sequence motif with all other
CD33rSiglecs except Siglec-10 and -14. Indeed, Siglec-10 is the greatest outlier among
CD33rSiglecs, with thin neck-thick belly characteristics least pronounced. Besides C/S/YxxC, all
remaining CD33rSiglecs also exhibit at least two Phe residues. Interestingly, for a number of
receptors the Phe is shifted to center of the helix relative to CD33 and even increased to three Phe
residues (Fig. 1.7). To obtain structural insight into this subclass, we constructed a model of the
Siglec-7 TM domain based on the CD33 structure (Fig. 1.5c-d). The Phe residues line one helix face,
the belly, whereas its back is again relatively inconspicuous and the neck region also “thin”. The
neck-belly characteristic together with the positively charged parch is therefore a hallmark of most
CD33rSiglecs TM domains and referred to as NBP-helix. We also note that a high incidence of
backbone-sidechain hydrogen bonds through the C/S/YxxC motif and oftentimes additional such
residues (Fig. 1.5a and 1.5c). Finally, the variation of the first charged residue on the intracellular
side among CD33rSiglecs is noted. For Siglec-5, this suggests the TM helix to be longer by one
residue compared to CD33. For Siglec-6, -11, and -14, an additional Arg appears two positions
before the first Lys of CD33, which may also affect TM helix length. In addition to the discussed
structural variations, differences in membrane immersion are likely to separate the CD33rSiglecs
into two groups at the level of their TM domains (Fig. 1.7).

29
The CD33 cytosolic region is unstructured and rich in proline-induced extended conformationsThe
TM helix terminates at His285 and our TM construct ended at Arg291. Next, we examined the
consequences of extending this construct to the C-terminus (Gln364) and characterized the
structure of the cytosolic region. With an anticipated random-coil conformation, NMR
spectroscopy is again well suited to characterize its average conformation (Fig. 1.8). Although not
a prerequisite to the binding of intracellular ligands, structural propensities away from random-coil
conformation may indicate an anticipation of a bound conformation and a prominent binding
region (Garcıá -Alvarez et al., 2003). As noted earlier, CD33 contains ITIM and ITIM-like motifs,
predicted for residues Leu338-Leu343 and Thr356-Val361, respectively (Fig. 1.9). The Src
Figure 1.8. 2D TROSY
spectrum of CD33(Q251-
R291) and CD33(Q251-
Q364 overlaid.
Superposition of the HN-N
correlation NMR spectra of
CD33(Q251-R291) and
CD33(Q251-Q364),
corresponding to the TM
and TM-CS domains,
respectively, at 35° C and
700 MHz. Proteins were
2H/13C/15N-labeled and
reconstituted in 350 mM
DHPC, 105 mM DMPC,
6% D2O, 0.02% w/v NaN3,
25 mM HEPES•NaOH, pH
7.4.

30
homology-2-containing tyrosine phosphatase 1 and 2 (Shp1 and 2) bind the ITIM motifs of CD33
via tandem SH2-SH2 domains (Paul et al., 2000; Walter, Raden, et al., 2008). The structure of these
domains in complex with unrelated peptides, shows these ligands to be bound in extended
conformation (Y. Liu et al., 2016). In terms of NMR parameter, secondary 13Cα chemical shifts
provide a readily interpretable parameter of secondary structure propensity (Spera & Bax, 1991;
Wishart & Case, 2002). To determine these shifts, we have extended our TM domain backbone
assignment to the cytosolic (CS) domain.
First, we noted that extending the construct mitigated fraying of the TM helix; Lys283-
Ala289 exhibited modestly higher, positive 13Cα shifts indicative of increased helical content (Fig.
1.10). That is, two helix turns were somewhat stabilized. However, extending the TM construct
invariably reduces the mobility of its terminal residues and the relatively small magnitude of 13Cα
shift increases indicated that the influence of the cytosolic region on the TM domain structure was
Figure 1.9. ClustalΩ multiple sequence alignment of the depicted human CD33rSiglecs.
Conserved amino acids are colored by the Jalview multiple alignment editor (Clamp et al., 2004)
using the ClustalX color scheme. The immunoreceptor tyrosine-based inhibitory motifs (ITIM) are
indicated.

31
small. Starting with Ala290, secondary 13Cα shifts were close to zero,
albeit with several distinct outliers (Fig. 1.10). Shifts near zero represent random coil conformations
(Spera & Bax, 1991; Wishart & Case, 2002), revealing the absence of a predefined ITIM binding
region. Phosphorylation of Tyr, which takes place away far from the backbone, is not expected to
change this disposition. The “outliers” in 13Cα shifts occurred in residues preceding proline residues,
present at positions 301, 308, 319, 330 and 350 (Fig. 1.9-1.10). Proline generally coerces preceding
residues into extended conformations (Schimmel & Flory, 1968) and the observed negative
secondary 13Cα shifts reflect such conformations. For Pro319, this effect is mitigated by the
conformational flexibility of preceding Gly318. However, because this effect is unaccounted for in
13Cα random coil chemical shifts (RCCSs) used to calculate secondary structure in most programs,
we used POTENCI to generate 13Cα the RCCSs for CD33 (Q251-Q364) and used them in place of
those used by XPLOR-NIH. POTENCI is a new method for predicting RCCSs is based on a large,
curated database of chemical shifts for protein segments with confirmed disorder. It explicitly
considers pH and temperature, includes sequence-dependent nearest and next-nearest neighbor
Figure 1.10. Secondary
13Cα chemical shifts of
the TM and TM-CS
domains. For random
coil conformations, shifts
are close to zero,
whereas positive and
negative shifts denote
helical and extended
backbone propensities,
respectively (Spera &
Bax, 1991; Wishart &
Case, 2002).

32
corrections, and incorporates second-order corrections (Nielsen & Mulder, 2018). There are
differences obtained using POTENCI, notably that the residues preceding P308 and P330 no longer
show negative secondary chemical shift values (Fig. 1.11). However, the region in general still
displays increased negativity in the residues preceding prolines 301, 319, and 350. Notably, the
mitigating effects of Gly318 as an extremely flexible residue are quite strong, yet POTENCI still
indicates a tendency toward secondary structure. This suggests that to an extent, this effect is not
an artefact of data analysis. In sum, the intracellular portion of CD33 is dynamically unstructured
but conspicuously influenced by its proline residues whereas ITIM-SH2 binding is not anticipated
by deviating from random-coil conformation.
The prominent influence of proline residues in the CS domain prompted us to examine their
conservation for the most similar CD33rSiglecs. Prolines are abundant for all examined Siglecs, but
a clear conservation pattern was not evident (Fig. 1.9). Interestingly, Siglecs-5, -8, -9, -11, and -12 are
more Pro-rich than CD33 and Siglec-6 and -7. They also exhibit at least one stretch of consecutive
prolines preceding ITIM1. The most conserved sequences in the cytosolic domains are the ITIM(-
Figure 1.11. Secondary
13Cα chemical shifts of
the TM-CS domains
calculated using
POTENCI and XPLORNIH overlaid. Random
coil chemical shifts for
CD33 (Q251-Q364) were
generated using XPLORNIH, as in Fig.. 1.10, and
POTENCI (Neilsen &
Mulder, 2018), then
subtracted from
experimentally obtained
13Cα chemical shifts and
overlaid for comparison.

33
like) motifs with a consensus sequence (I/V/L/S)-X-Y-X-X-(L/V)(Crocker et al., 2007). A high amino
acid conservation preceding and succeeding the ITIMs is also noted.
The IgC1 domain dimerizes in solution, rendering CD33 dimeric
Figure 1.12. Characterization of the CD33 IgC1 domain expressed in P. pastoris. A,
Electrospray mass spectrum of IgC1(D140-T232/C169S). Three molecular species with
differently processed N-termini were identified as indicated. B, NMR spectrum of 0.42 mM IgC1
in 10 mM NaH2PO4/Na2HPO4, pH 7.4, 140 mM NaCl, 3 mM KCl and 6% D2O. The spectrum
was recorded at a 1H frequency of 700 MHz and 35 °C.

34
With the TM domain exhibiting no dimerization tendencies, the possible dimerization of the IgC
domain (Fig. 1) will define the structure of CD33. To examine its aggregation state in solution, we
have expressed IgC(Cys169Ser) in P. pastoris. To render it suitable for molecular weight
determination by gel filtration, we trimmed its glycosylation to single N-Acetylglucosamine
Figure 1.13. Biophysical characterization of the IgC domain of CD33. A, 1H NMR spectrum of 0.4
mM CD33(D140-T232/C169S) in 10 mM NaH2PO4/Na2HPO4, pH 7.4, 140 mM NaCl, 3mM KCl at 35° C
and 700 MHz. B, CD spectrum of 20 M CD33(D140-T232/C169S) in 10 mM KH2PO4/K2HPO4, pH 7.4.
C, Gel filtration of CD33(D140-T232/C169S) as part of its last purification step in 10 mM
NaH2PO4/Na2HPO4, pH 7.4, 140 mM NaCl, 3mM KCl using a Superdex 75 Increase 10 300 GL
column.

35
(GlcNAc) moieties. On Sephacryl S100 and Superdex 75 columns apparent molecular weights of
31.5 and 34.3 kDa were obtained, respectively (Fig. 1.13). Mass spectrometry confirmed a
molecular weight of 12.7 kDa for our construct (Fig. 1.13) and, accordingly, 25.4 kDa for the dimer.
While a trimer could also be considered (38.1 kDa), it is likely that its non-spherical shape (Fig. 1)
caused it to travel faster than expected in the gel filtration matrix. Thus, we conclude that IgC is a
dimer in solution.
The IgV-IgC crystal structure reveals dimerization to be mediated by the parallel pairing of
βC strands (Fig. 1.14). Sidechain packing between intermolecular βC-βD* and also βC contacts to
βF*
stabilize the interface that buries a surface area of 808 Å2 (Fig. 1.14) (Abraham et al., 2015).
Figure 1.14. Surface area of IgV
and IgC interfaces. Crystal structure
of the IgV-IgC domains of human
CD33 in complex with 3′-sialyllactose
(PDB ID 5j06). The IgV domain is
shown in yellow and the IgC domain
in teal, and together shown in green.
The buried surface area (BSA) was
found for the IgV-IgV interface (84 Å2),
the IgV-IgC interface (402 Å2), and
IgC-IgC interface (808 Å2) using
molecular modelling program
GROMACS, v.5.8 (Abraham et al.,
2015)

36
Interestingly, βD exhibits a conspicuous change in strand registry effected by Gly188-Pro189 that
allows Arg190 to point towards βC and engage Ile176 in packing interactions (Fig. 1.15b). Further
notable packing interactions are mediated by Leu180-Leu187 and Ser181-Thr185 (Fig. 1.15b and
1.16). Non-covalent dimerization was also reported for MAG where the third and fourth IgC
domains form hydrophobic contacts over an area of 2,037 Å2 (Pronker et al., 2016). Nonetheless,
MAG dimerization is too weak to be detected by gel filtration (Pronker et al., 2016), highlighting
the well-defined nature of the IgC1 dimer interface of CD33. On a more formal note, IgC domains
Figure 1.15. β sheet organization of human Siglec IgC domains. A, Comparison of Siglec IgC
topologies. IgC1 of CD33 dimerizes via βC pairing. For IgC2, two topologies are differentiated. In
one βA pairs solely with the B-E sheet, whereas in the other an additional strand βA' also pairs
with the G-F-C sheet. This latter topology is designated IgC2* in contrast to the simpler IgC2
version. Occasionally, a β-strand (βC') in the βC-βE linker is formed but helical conformation is
also possible (Fig. 1.16d). B, Structural examples of IgC topologies in comparison to IgC1 of CD33
shown in red. For IgC1 of Siglec-5 and -11, dimerization via βD appears to be blocked by the βCβE linkers. However, IgC1 of Siglec-5 contains an unpaired Cys in βC that may dimerize (A. L.
Cornish et al., 1998). The second IgC2 domain of Siglec-8 exhibits an unpaired Cys in the βC-βE
linker that may also dimerize (Floyd et al., 2000). The second IgC domains of Sigled-5 and -11
exemplify C2*-type topology. For CD33, the crystal structure is shown (PDB ID 5ihb), whereas for
the other Siglecs AlphaFold2 models are displayed. These models are color-coded according to
their per-residue model confidence score (pLDDT).

37
are categorized as either C1- or C2-type folds, for which the existence of the βD sheet is a
principal differentiator (Fig. 1.15a) (Williams, 1987; Williams & Barclay, 1988). The existence of
βD for IgC of CD33 classifies this domain to adopt a C1-type fold. In sum, the IgC1 crystal
structure in combination with gel filtration identify IgC1 to dimerize via a specific packing
interactions, thereby effecting the non-covalent association of CD33 into a dimer with separated
TM helices and dynamically unstructured cytosolic domain (Fig. 1.20a).

1.4 - Discussion
Relevance of CD33 homodimerization mechanism to other Siglecs
Figure 1.16. Domain organization of human Siglecs. C, Siglec domain organization
according to AlphaFold2 models and available crystal structures (CD22 5vkj, CD33 5ihb,
Siglec5 2zg2). In contrast to AlphaFold2 predictions, the first IgC domains of Siglec-6 and
Siglec-15 were classified as C1-type based on sequence alignment (Fig. 5D-E). For SN,
d12(IgC2) may, in contrast to prediction, also be of C1-type.

38
The availability of AlphaFold2 structural models of all human Siglecs (Jumper et al., 2021) allowed
Fig. 1.17. Structural basis of CD33 homodimerization and Siglec IgC(βC-βD-βE) sequence
alignment. A, Cartoon representation of the IgC homodimeric assembly found in the IgV-IgC
crystal structure of CD33 (PDB ID 5ihb). With β strands A-B-E-D and C-F-G present, a C1-type Ig
fold is adopted (Fig. 1.14a). B, Overview of the dimerization interface. C, Superposition of the first
IgC domains of human CD33 and CD22. The crystal structure of CD22 (PDB ID 5vkj) was used.
D, ClustalΩ multiple sequence alignment of the βC-βD-βE region of the first IgC domains of
human CD33rSiglecs. Selected interdomain interactions are indicated. E, Analogous alignment of
CD33 with the conserved human Siglecs. Conserved amino acids are colored by the Jalview
multiple alignment editor (Clamp et al., 2004) using the ClustalX color scheme.

39
us to screen for the CD33 dimerization motif and to examine the nature of their IgC domains.
Structural superpositions showed that IgC1 of CD33 most closely resembles the first IgC domain of
other Siglecs (Fig. 1.18, Fig. Fig. 1.15b, and data not shown). This also applied to Siglec-15, which
lacks the IgV-IgC disulfide bond (Fig. 1.1a) found in all other Siglecs. Intriguingly, none of the first
IgC domains models of CD33rSiglecs exhibited βD that stabilizes dimerization in CD33 (Fig. 1.18).
It was replaced by either short helices or loops that would block any pairing at this location.
However, for many Siglecs, the region between βC and βE was modeled with low confidence (Fig.
1.18). Aligning the actual sequences of this region revealed that Siglec-6 is almost identical to CD33
and reproduces the features that underpin the existence of βD (Fig. 1.17d). The remaining
CD33rSiglecs, however, lack the hallmark of the relatively short βC-βD and βD-βE connectors, and
the characteristic Gly-Pro pair of βD. The available crystal structure of the first IgC domain of Siglec5 indeed confirms the absence of βD (Zhuravleva et al., 2008). As such, those CD33rSiglecs are
unlikely to form βD and represent C2-type Ig folds. For the conserved Siglecs, βD is present in the
crystal structure of CD22 and MAG (Ereño-Orbea et al., 2017; Pronker et al., 2016) and predicted
for SN and Siglec-15 (Fig. 1.17c and Fig. 1.18), classifying them as IgC1 domains. Sequence
alignment illustrates longer βD sheets and longer βC-βD linkers than found in CD33 (Fig. 1.17e).
The longer βC-βD linkers shield the dimerization interface and mismatch βC-βD heights (Fig. 1.17c).
It therefore appears that, for the first IgC domains, homodimerization via parallel βC pairing is only
possible for CD33 and Siglec-6.
For the remaining IgC domains of CD33rSiglecs, C1-type folds were only found for the
membrane-proximal Ig domains of Siglec-5 and -11 (Fig. Fig. 1.15b-1.16). In both cases, the βC-βD

40
Figure 1.18. Superposition of CD33 IgC1 with the first IgC domains of human Siglecs. CD33 is
shown in red. AlphaFold2 models of the indicated Siglecs are color-coded according to the perresidue model confidence score (pLDDT). Backbone Cα, Hα, N, C and O coordinates were aligned
using PyMOL.

41
linkers appear to block the canonical dimerization interface (Fig. Fig. 1.15b), making dimerization
unlikely. The other IgC domains lack βD, which classifies them as C2-type and makes their noncovalent dimerization via βC pairing improbable. They exhibit notable variability in the βC-βE
intervening sequence but also in strand βA (Fig. 1.15a). While this strand pairs with βB, oftentimes
it forms an extension that pairs with βG of the second β-sheet (Fig. 1.15a-b), a feature that was
first observed in an IgC1 domain (“Many of the Immunoglobulin Superfamily Domains in Cell
Adhesion Molecules and Surface Receptors Belong to a New Structural Set Which Is Close to That
Containing Variable Domains,” 1994) and is common among IgC2 domains of Siglecs where we
indicate it with an asterisk as IgC2* domain (Fig. 1.15a). In the conserved Siglecs, IgC1 domains are
more common and Sialoadhesin alternates IgC1 with IgC2 domains (Fig. 1.16). However, none of
the IgC1 domains appear capable of dimerizing in a manner similar to CD33. As dimerization of the
IgC2*-IgC1 pair of MAG (Fig. 1.16) shows, it is nonetheless difficult to predict dimerization ab initio
especially for relatively weak interactions. Thus, Siglecs exhibit a rich variety of IgC domain
topologies with the non-covalent dimerization of IgC1 domains appearing to be the exception and
not the rule (Fig. 1.16).
As noted in Introduction, disulfide-linked dimerization was reported for Siglec-5 and -8 (A.
L. Cornish et al., 1998; Floyd et al., 2000). For Siglec-8, an unpaired Cys is present in the βC-βE
linker of its second IgC2 domain (position 303; Fig. Fig. 1.15b-1.16). This position is equivalent to
the dimerization interface of CD33. In contrast, for Siglec-5, the unpaired Cys residue localizes to
strand βC of the third IgC2 domain (position 364; Fig. Fig. 1.15b-1.16). This dimerization site would
therefore be different from CD33.

42
Functional aspects of the CD33 structure
With the structural characterization of the TM and CS domains of CD33 and the IgV-IgC1 crystal
structure, the first complete structure of Siglec receptor is achieved (Fig. 1.20a). CD33 introduces
the element of non-covalent IgC1 dimerization into Siglec structures. Dimerization via βC strand
pairing is only shared with Siglec 6, which is evolutionary most closely related to CD33 (“Siglecs,”
2018). The sophisticated dimerization interface contrasts with the simple disulfide-mediated
dimerization of Siglec-5 and -8, and extensive but weak dimerization of MAG (A. L. Cornish et al.,
1998; Floyd et al., 2000; Pronker et al., 2016). Evidently, only one point mutation is required to
introduce Cys, whereas βC pairing is stabilized by several interactions that seem to rely on the
presence of strand βD (Fig. 1.17a, b, and d). Initially, a common ancestor to CD33 and Siglec-6
precursor likely achieved partial dimer formation. As more stabilizing mutations accumulated,
permanent dimerization was achieved attesting to an evolutionary advantage of dimeric over
monomeric CD33. Principally, this advantage may lie at the level of ligand binding or receptor
signaling.
As opposed to disulfide-linked receptors, the many contacts achieving non-covalent IgC1
dimerization lock the relative IgC1-IgC1 domain-domain orientation more decisively. This will aid
the initial assembly of the dimeric receptor, and confer a preferred domain-domain orientation for
the IgV pair. IgV connects to IgC1 via a 33-residue linker, a buried surface area of 402 Å2 (Abraham
et al., 2015), and is restrained by the IgV-IgC1 disulfide bond (Fig. 1). Moreover, Siglec-6 dimerizes
via its first IgC domain in contrast to the disulfide linkages of Siglec-5 and -8 in later domains (Fig.
1.16). For some Siglecs, instances of unexpectedly selective protein/ligand targeting are
documented (Blixt et al., 2003; Büll et al., 2021; Wisnovsky et al., 2021) suggesting that not only

43
are ligand affinities important but also the arrangement of binding sites on multivalent ligands. In
general, the number, coupling and properties of the IgC domains may encode such spatial
preference. Non-covalent dimerization may be another element to encode a spatial binding
preference even more precisely. As such, non-covalent dimerization of CD33 and Siglec-6 may
encode a spatial binding configuration that is important to their specific functional contexts.
CD33 signaling hypothesis
Receptors signal ligand binding into the cytosol. With many receptors consisting of more than one
subunit, the question arises whether a change in subunit monomer-dimer equilibrium could
Figure 1.19. Superposition of free and ligand-bound CD33 IgV-IgC1 domains. A, Crystal
structures of the IgVIgC1 domains of human CD33 in complex with 3′-sialyllactose (PDB ID 5j06;
chains A+D) and free (PDB ID 5ihb; chains A+D) were aligned using PyMOL. Backbone Cα, N, C and
O coordinates of residues 22-139 (IgV) of chain A were superimposed with a root-mean-square
deviation of 0.23 Å. For residues 142-232 (IgC1) of chain A, a corresponding r.m.s.d. of 0. 55 Å
resulted. B, The analogous evaluation for CD33 in complex 6′-sialyllactose (PDB ID 5j0b; chains A+D)
yielded r.m.s.d. of 0.14 Å and 0.26 Å for IgV and IgC1, respectively.

44
contribute to CD33 signaling. A signal is often transduced by shifting the monomer-dimer
equilibrium of the TM helices following ligand binding (Arkhipov et al., 2013; Lau et al., 2009). For
CD33, IgV-IgC1 crystal structures are available free and bound with 3′-sialyllactose, 6′-sialyllactose
and P-22. However, ligand binding does not change the IgV backbone structure or IgV-IgC1
orientation significantly (Fig. 1.19). It is therefore unlikely that ligand binding to IgV leads to a
structural rearrangement that is transmitted to or beyond IgC1. Unlike integrins, which also exhibit
non-covalently associated extracellular subunits (Hynes, 2002; Xiong et al., 2002), the machinery
for rearranging the extracellular domains appears to be missing in CD33. The TM domain of CD33
was found to be monomeric. Together with its characteristic NBP appearance (Fig. 1.5a-b), it
appears unsuited to dimerize. Moreover, a relatively long linker of 33 residues connects IgC1 to the
TM domain (Fig. 1.20a), which would uncouple any TM helix monomer-dimer equilibrium from the
corresponding IgC1 equilibrium or a change in IgV-IgC1 domain-domain orientation (Schmidt, Ye,
et al., 2016). The long linker may facilitate ligand binding by extending IgV beyond the glycocalyx
coat of the membrane. A shift in the monomer-dimer equilibrium of CD33, away from the IgC1
dimer or away from the TM monomer, is therefore unlikely to contribute to its activation.
If a multivalent ligand engages IgV, it can nonetheless oligomerize these receptors. For
CD33, P22 monomers had no effect on basal cellular phagocytosis whereas P22 conjugated to
microparticles increased phagocytosis (Miles et al., 2019). If ligand-binding sites are clustered,
receptors will, in turn, also cluster in the membrane. In other words, clustering concentrates the
receptor TM and CS domains in the membrane and cytosol, respectively. It has been reported that
ligand-bound CD33 and other activated Siglecs localize to membrane microdomains (lipid rafts)
(Ando et al., 2015, p. 3; Ha et al., 2020). Lipid raft clustering has also been demonstrated to

45
predicate CD33-induced megakaryocyte differentiation and apoptosis through activation of the
ERK signalling cascade (Ha et al., 2020). Accordingly, clustered TM domains may display an altered
lipid preference, promoting their transfer to lipid rafts (Fig. 1.19c). The CS domain is not
anticipating a bound conformation and, outside of the ITIMs, prolines distinctly modulate its
dynamically unstructured ensemble. Clustering restricts the lateral diffusion of CD33 and also
concentrates the dynamically disordered cytosolic domain. In analogy to other immune receptors
(Xiao et al., 2022), the CS domains may undergo a liquid–liquid phase separation (LLPS) once a
Figure 1.20. CD33 structure and signaling hypothesis. A-B, Structural model of the full-length
human CD33 receptor. The IgV-IgC1 crystal structure (PDB ID 5ihb, chains A+D) is connected to the
TM domain structure via a dynamically unstructured linker. The cytosolic domain is represented by
unstructured conformations. C-E, Proposed signaling mechanisms of CD33. First, a multivalent ligand
clusters CD33. The ensuing proximity of TM and CS domains may attract the TM domains to lipid rafts
and/or trigger a liquid-liquid phase separation of the CS domains. This may appose CD33 with a Srcfamily kinase in the raft area and/or the separated phase, allowing phosphorylating of the ITIMs to
render the receptor active.

46
critical concentration is locally achieved. LLPS could occur alongside CD33 localization into lipid
rafts (Fig. 1.19d), or independent of it (Fig. 1.19e), and also allow a productive co-localization with
a Src family kinase. In addition to impacting ligand binding, dimerization of CD33 and Siglec-6,
would reduce the entropic cost of clustering the receptor and, thus, lower the threshold of
receptor activation. In conclusion, the structure of CD33 establishes the first complete structure of
a Siglec, clarifies its signaling possibilities, and suggests an activating clustering mechanisms. The
dimeric CD33 structure constitutes a spatial restraint on multivalent ligand binding, impacts
signaling thermodynamics, and illustrates that CD33 drug targeting applications may achieve
greater efficiency when targeting the dimeric receptor structure.
1.6– Conclusions
In conclusion, the molecular basis of signal transduction for CD33, and perhaps other
dimerizing CD33-related Siglec receptors, is likely to originate from lipid raft clustering. Here we
use NMR spectroscopy and mass spectrometry to show that CD33 dimerizes at the βC-βD interface
within its IgC domain, confirming previous X-ray crystallography results. Unlike the X-ray structure,
the IgC domain is classified as an IgC1 domain rather than IgC2 domain, containing a We were able
to generate an ensemble TM domain structure, obtaining an α-helix spanning 21 residues from
A265 to H285. Inserting the TM domain in bicelle mimics of using various lipid lengths allowed us
to look at TM domain boundaries. Helical exposure at the N-terminus, i.e. the extracellular domain,
is clearly demarcated at G264, while the C-terminus in the cytosol can be buried. NMR also shows
that the cytosolic tail is unstructured, mirroring computational predictions, and that SHP2
interactions with it induce structural changes. These data show that the IgC domain is significant

47
to CD33 stability, and that structural features at the TM domain and cytosolic tail can impact
context-dependent microglial activity.

48
CHAPTER 2 – Prolines and charged residues at the transmembrane
domain border affect transmembrane domain stability and dimerization

2.1 – Introduction
Integrins are another membrane protein that play a role in disease, with a unique insideout signaling mechanism. They are implicated in a variety of diseases, but here we focus again on
AD. Integrins play an important role in AD both through its regulation of the blood-brain barrier
(BBB) and through aiding in activation and movement of innate immune cells. They interact with
extracellular matrix (ECM) proteins such as collagen, laminin, fibronectin and vitronectin (Hynes,
2002), which are connected to the actin cytoskeleton by integrin clusters known as large transient
complexes (focal adhesion). In doing so, their adhesive interaction contributes to the structural
integrity of the BBB by stabilizing endothelial cell junctions and preventing their disruption,
preventing movement of aβ plaques and tau neurofibrillary tangles in and around neuronal cells
(Engelhardt & Sorokin, 2009; Halder et al., 2023). Integrins are also implicated in Alzheimer's
disease (AD) neuroinflammation itself (Hogg et al., 2011). Leukocyte surface integrins play a
crucial role in antigen presentation within the central nervous system (CNS). Similarly, activated
microglia express various integrins, which can be upregulated in response to inflammation and
injury. Recent research revealed that microglia can engage with fibrillary Aβ via a complex
involving α6β1 integrin and the B-class scavenger receptor CD36. This interaction triggers a
proinflammatory response in microglia, resulting in elevated production of cytokines and
chemokines (Bamberger et al., 2003). Furthermore, studies on mixed neuron-glia cell cultures
from the cerebellum demonstrated that neurons exposed to Aβ prompt a phagocytic response in
microglia mediated by the binding of phosphatidylserine (PS) on neuron surfaces to αvβ3/5

49
integrin receptors on microglia (Neniskyte et al., 2011). β2 has also been found to be
constitutively expressed in microglia, and an increase in expression of αXβ2 was observed in
reactive microglia, early evidence that AD can trigger an inflammatory response (Akiyama &
McGeer, 1990) and serving as biomarkers of endothelial activation and neuroinflammation (Rossi
et al., 2011).
Integrins are a family of type I heterodimeric transmembrane receptor proteins
composed of a variety of combinations of α and β type I transmembrane subunits. Association of
Fig. 2.1. Structure of the αIIbβ3 integrin heterodimer signaling mechanism. The TM domains of
the αIIb and β3 subunits are dimerized in the inactive conformation. When the subunits overcome
ΔG°TM and dissociate, integrin takes on an active conformation. Adapted from (Lau et al., 2009).

50
the α and β TM domains regulates the unique inside-out signaling transduction for which
integrins are known (Arnaout et al, 2005). Disruption of that association leads to allosteric
structural rearrangement that activates the integrin heterodimer and triggers bidirectional
signaling (Hughes et al., 1996; Partridge et al., 2005). The specific side chains present in the TM
domain affect dimerization by altering hydrogen bonding between side chains, backbone
interactions within the α-helix(ces) and β-sheets, and with lipids (Kim et al., 2012; Moon &
Fleming, 2011; Wagner et al., 2022). Thus, the overall stability of the TM complex in competition
with the protein environment is fundamental to establishing the threshold for receptor
activation, commonly expressed as the free energy of TM complex formation, termed ΔG°TM (Fig.
2.1). For example, in the family of integrin cell adhesion receptors, the association of the α and β
TM helices maintains the inactive state and ΔG°TM dominates the activation threshold of the
integrins studied thus far (Kim et al., 2009; Partridge et al., 2005; Schmidt, Situ, et al., 2016).
In previous work, the Ulmer lab has shown that dimerization capacity of αIIbβ3 integrin
is linked to N-terminal prolines in the TM domain (Fig. 2.2) (Schmidt, Situ, et al., 2016). With its
unique structural property of having its sidechain linked back to its amide group, prolines confer
a great deal of significance in structure-to-function relationships where they occur (Wedemeyer
et al., 2002). Prolines are often found at the border of transmembrane helices, with a likelihood
of occurrence of ~25% (Ulmschneider & Sansom, 2001). This is true for the human integrin
subunits as well, where proline is highly conserved at the TM border (Fig. 2.2) ((Schmidt, Situ, et
al., 2016). Residues at the TM border are significant in and of themselves because they interact
with lipid headgroups both during protein folding and as the protein functions (Moon & Fleming,

51
2011). The thickness of the lipid bilayer, determined by acyl chain length and motility, can affect
the topology of protein TM domains in this way. Changes in thickness can induce changes in tilt
and conformation of amino acid sidechains in order to find the least energetic arrangement of
the TM domain (S. H. Park & Opella, 2005). An example of this type of rearrangement occurs in
“snorkeling” by lysine and arginine residues, whereine their positive charge allows them to peek
to the surface of the lipid bilayer (Kim et al., 2012). In this way, beyond the structure of the TM
Fig. 2.2. ClustalΩ multiple sequence alignment of the transmembrane domains of human
integrin subunits. A, Sequence alignment of the transmembrane domains of integrin α-subunits and
B, Sequence alignment of the transmembrane domains of integrin β-subunits. Amino acid numbering
follows the αIIb and β3 subunits respectively. Conserved amino acids are colored by the Jalview
multiple alignment editor1 using the ClustalX color scheme (Clamp et al., 2004).

52
domains themselves, lipid makeup of the membrane can contribute to integrin signaling
activation.
While highly conserved, mammalian integrins derive 24 unique proteins made from a
combination of 18 α and 8 β subunits. These 24 integrins each fulfull a specific physiological
function, as demonstrated by knockout (KO) mice studies (Hodivala-Dilke et al., 1999; Tsakiris et
al., 1999). Given the non-redundant physiological roles of these 24 integrins, there is considerable
interest in elucidating their function-to-structure relationship. With two prolines at the TM
domain border (both outside in the linker region and within the domain itself), and its relevance
to immune function, αXβ2 was chosen for comparative study against αIIbβ3. For β2 integrins, an
additional interhelical Ser-Thr hydrogen bond was also reported. This is indicative of an ancient
evolutionary trend towards the gradual incorporation of polar contacts in interactions among
membrane proteins that is thought to expand the structural diversity of membrane protein
configurations. (A. Situ & Ulmer, 2019). However, non-β2 integrins are able to modify ΔG°TM solely
by adjusting packing variations, leaving the functional advantage of interhelical hydrogen bonding
unclear. In principle, a hydrogen bond could stabilize ΔG°TM by ~0.6 kcal/mol, even up to 1.7
kcal/mol (Joh et al., 2008). However, the introduction of an equivalent hydrogen bond in αIIbβ3
produced only modest stabilization at an improvement of ~0.3 kcal/mol, up to 0.5 kcal/mol
(Schmidt, Situ, et al., 2016). In this chapter, we demonstrate that αXβ2 does not have a
homologous signaling mechanism to αIIbβ3, and discuss the role of proline and α-helix backbone
hydrogen bonding dynamics in TM domain stabilization, and thus regulation of αXβ2 activity in a
context dependent manner.

53
2.2 – Materials and Methods
Expression and purification of human Integrin αX and β2 transmembrane —
Using the cDNA of human integrin αX Uniprot P20702 (ITAX_HUMAN, Integrin alpha-X) and β2
Uniprot P05107 (ITB2_HUMAN, Integrin beta-2) respectively, inserts encompassing Glu1079-
Lys1119 for αX and Val674-Tyr713 for β2 were generated by PCR and subcloned into the pET-44
expression vector (Novagen, Inc.) with the third IgG-binding domain of protein G (GB3) as Nterminal fusion protein and an intervening tobacco etch virus protease cleavage site. Expression
was induced in E. coli BL21(DE3)pLysS,T1R cells (Sigma-Aldrich, Inc.) cultured at 37 °C in M9
minimal medium, containing combinations of 99% 13C-D glucose, 99% 15NH4Cl, and 99% D2O, at
an OD600 of 1.0 by adding IPTG to 1.0 mM. Cells were harvested by centrifugation 5 h after
induction and lysed by sonication in 50 mM Tris·HCl, pH 8.0, 300 mM NaCl, 100 mM SDS, 20 mM
imidazole, and 2 mM β-mercaptoethanol. The clarified lysate was applied on a HiTrap Chelating
HP column (GE Amersham, Inc.) charged with Ni2+ for immobilized metal affinity chromatography
(IMAC). The column was washed with 50 mM Tris·HCl, pH 8.0, 300 mM NaCl, and 25 mM SDS,
followed by buffer exchange into 50 mM Tris·HCl, pH 8.0, 300 mM NaCl, 8 M urea, and 20 mM
imidazole and additional washing with an elevated imidazole concentration of 50 mM. After
another buffer exchange into 50 mM Tris·HCl, pH 8.0, 300 mM NaCl, 6 M guanidine·HCl, and 20
mM imidazole, bound protein was eluted by raising the imidazole concentration to 300 mM. The
peptide was cleaved from the fusion protein using TEV protease at a molar ratio of 1:25 overnight
at 30 °C in 50 mM Tris·HCl, pH 8.0, 1 mM DTT solution, leaving a Gly as the N-terminal residue
preceding Glu1089 and Val674. Uncleaved protein and fusion protein were removed by IMAC and
the column flow-through was applied on a Hamilton PRP-3 reverse-phase HPLC column. The

54
peptides were eluted using a linear gradient from 70%/30% buffer A (H2O, 0.1% TFA) / buffer B
(70% acetonitrile, 30% 1-propanol, 0.1% TFA) to 10%/90% in 30 min. The 〈X(P1086A), 〈X(P1088A),
〈X(P1088W), 〈X(G1092L), and 〈X(S1094I) mutants were prepared via site-specific mutagenesis
using the QuikChange protocol (Novagen). Expression and Isolation preceded as previously
described. Subsequent to freeze-drying, peptide purity was verified by SDS-PAGE and NMR.
NMR sample preparation—
TM peptide concentrations were measured in acetonitrile-water solution by UV spectroscopy
(ε280nm= 4470 M-1cm-1 and ε280nm= 6990 M-1cm-1 for αX and β2, respectively) and defined amounts
of peptide were freeze-dried. The resulting fractions were dissolved in 320 μl of aqueous bicelle
solution during a heating (42°)/cooling cycle (-20°) to yield concentrations of 1.2 mM and 0.6 mM
for Gln251-Arg291 and Gln251-Gln364, respectively. The bicelle solution was comprised of longchain lipids, 1-palmitoyl-2-oleoyl-sn-glycero-3-phospho-choline (POPC), and short chain lipids,
1,2- dihexanoyl-sn-glycero-3-phosphocholine (DHPC), at a q-factor of 0.3 (or otherwise
mentioned), 25 mM HEPES∘NaOH, pH 7.4, 120 mM NaCl, 6% D2O, 0.02% w/v NaN3. To study
changes in lipid length, POPC (16:0-18:1 carbon chain length) was substituted with 14:0-14:0 1,2-
dimyristoyl-sn-glycero-3-phosphocholine (DMPC) and 20:1-20:1 1,2-dieicosenoyl-sn-glycero-3-
phosphocholine (DEPC) at a q-factor of 0.5.
LUV fluorescence spectroscopy—

55
Large unilamellar vesicles (LUVs) were prepared by mixing a 2 concentration of 3 µΜ and the αX
concentrations indicated in 25 mM HEPES pH 7.4, 120 NaCl, 6 mM POPC solution. The solution was
extruded to yield large unilamellar vesicles (LUV). All spectra were recorded at 28 °C. To extract the
association constant KXY of the resulting TM complex, the change in fluorescence intensity at 320 nm was
used to linearly monitor the fractional change in the amount of heterodimer.
NMR spectroscopy—
NMR experiments were carried out on a cryoprobe-equipped Bruker Avance 700 spectrometer at
40 °C. Data were processed and analyzed with the nmrPipe package and CARA. 2H/13C/15N-labeled
N peptide and TROSY-type H-N detection (Pervushin et al., 1997) was used for HNCA, HNCACB,
HN(CA)CO and HNCO-based backbone assignments, the measurement of 3JC’Cγ and 3JNCγ coupling
(Hu et al., 1997), and the detection of 1JNH, 1JCαC’, 1JC’N as well as 1JNH+1DNH, 1JCαC’+1DCαC’, 1JC’N+1DC’N
couplings (Chou et al., 2000; Fitzkee & Bax, 2010; Jaroniec et al., 2004; Kontaxis et al., 2000) using
isotropic and aligned samples, respectively. [1H]-15N NOE measurements were carried over from
prior experimentation with αIIb and β3.
ITC measurements—
Employing a Microcal VP-ITC calorimeter, 10 μM of β2 peptide in the 1.425 ml sample cell was
titrated with αX peptide. We injected αX in 9 μl aliquots over a period of 10 s. Unless otherwise
specified, measurements were carried out at 28 °C in 25 mM NaH2PO4/Na2HPO4, pH 7.4, 43
mM DHPC and 17 mM of long-chain lipid. Prior to data analysis, the measurements were
corrected for the heat of dilutions of the αX and β2 peptides. The reaction enthalpy (ΔH°) and

56
equilibrium constant (KXY), but not the αX β2 complex stoichiometry, were calculated from the
measured heat changes, ΔHi, as described previously (A. J. Situ et al., 2014; Vu et al., 2021).
Structure calculation of αX and β2integrins—
The bicelle-embedded structures of the well-folded αX Thr1087-Phe113 and β2 Asn678-Asp709
residues were calculated by simulated annealing, starting at 3000 K using the program XPLORNIH (Schwieters et al., 2003). The peptide termini were represented by random-coil
conformations. Backbone dihedral angle constraints were derived from N, Hα Cα, Cβ and C’
chemical shifts using the program TALOS-N (Shen & Bax, 2015). χ1 side-chain angle restraints were
derived from the 3JC’Cγ and 3JNCγ coupling constants (Hu et al., 1997). In addition to standard force
field terms for covalent geometry (bonds, angles, and improper dihedrals) and nonbonded
contacts (Van der Waals repulsion), dihedral angle and interproton distance restraints were
implemented using quadratic square-well potentials, and a backbone-backbone hydrogenbonding potential and torsion angle potential of mean force were employed (Grishaev & Bax,
2004; Kuszewski et al., 1997). The difference between predicted and experimental residual
dipolar couplings (RDC; Δ1D) was described by a quadratic harmonic potential. The final values
for the force constants of the different terms in the simulated annealing target function are as
follows: 1,000 kcal·mol-1·Å-2 for bond lengths; 500 kcal·mol-1·rad-2 for angles and improper
dihedrals, which serve to maintain planarity and chirality; 4 kcal·mol-1·Å-4 for the quartic Van der
Waals repulsion term; 30 kcal·mol-1·Å-2 for interproton distance restraints; 500 kcal·mol-1·rad-2 for
dihedral angle restraints; 0.3 kcal·mol-1·Hz-2 for 1
DNH RDC restraints and 1DC’N and 1DCαC’ scaled
relative to 1DNH according to their dipolar interaction constants; 1.0 for the torsion angle potential;

57
and a directional force of 0.20 and a linearity force of 0.05 for the hydrogen-bonding potential. A
total of 20 structures were calculated for both.
2.3 – Results
The structures of the αX and/or β2 TM domains cannot be homologous to αIIb and β3
To obtain a thermodynamic reference for β2 TM complex stabilities, we first examined the TM
association of αX and β2 TM peptides in large unilamellar vesicles (LUV) by fluorescence
spectroscopy. To our surprise, we did not detect any interaction between the peptides (Fig.
2.3a). Between the β2 and β3 TM domains, there are no significant amino acid differences in
the dimerization interface (Fig. 2.3). Previously, we studied the
αIIb(W967Y/W968V/G972S/W988Y) mutant, which implements all but one of the relevant
sequence differences to αX (Fig. 2.2), to assess the possible strength of the αX(S1093)-β2(T686)
hydrogen bond. Its ΔG°TM with corresponding β3(V700T) was -4.4 ± 0.1 kcal/mol (A. Situ &
Ulmer, 2023). The remaining, potentially important sequence difference is αIIb(W967P). We
evaluated this mutation in phospholipid bicelles by isothermal titration calorimetry, which
offers highest accuracy (Schmidt et al., 2015; A. J. Situ et al., 2014). The substitution
destabilized the αIIb(W968V)β3 TM complex by 0.43 ± 0.04 kcal/mol (Table 2.1). Accordingly,
ΔG°TM of αXβ2 was expected at approximately -4.0 kcal/mol. To fall short of this expectation
implies that, despite high sequence identities (Fig. 2.2), the structures of the αX and/or β2 TM
domains cannot be fully homologous to their αIIb and β3 counterparts.

58
Figure 2.3. Integrin αX and β2 transmembrane structures. (a) The structure of the integrin β2 was
determined and is shown in blue overlayed with the corresponding β3 structure in light blue. (b) The
structure of the integrin 〈X was determined and is shown in red overlayed with the corresponding 〈IIb
structure in light red. (c) The monomeric 〈X and β2 structures, shown in light red and light blue, were
overlayed onto a model of the 〈Xβ2 TM complex, colored in red and blue, respectively. (d) In the model
of the 〈Xβ2 TM complex, the destabilized 〈X helix segment (Thr1087-Val1095) is colored in light red.

59
Table 2.1. Mutant integrin αIIb(W968V)β3 TM Complex Stabilities
Peptides KXYa ΔHo
[kcal/mol]
TΔSo
[kcal/mol]
ΔGo
[kcal/mol]
ΔΔGo
[kcal/mol]b
αIIb(W968V) + β3d 13200 ±
600 -14.3 ± 0.2 -8.6 ± 0.2 -5.68 ± 0.03 –
αIIb(W967P/W968V) + β3 6400 ± 300 -14.1 ± 0.2 -8.9 ± 0.2 -5.25 ± 0.02 0.43 ± 0.04
a
Measurements in 43 mM DHPC, 17 mM POPC, 25 mM NaH2PO4/Na2HPO4 pH 7.4 at 28 °C
bΔΔGo = ΔGo
mutant - ΔGo
αIIb(W968V)β3
c
In bicelles, competition for Thr from water molecules unduly diminishes ΔGo
mutant (Schmidt,
Situ, et al., 2016)
dValues taken from (Lau et al., 2009; Lau, Partridge, et al., 2008).
The difference between αX and αIIb mainly lies in backbone dynamics
To identify structural differences between the subunits, we determined the structures of the αX
and β2 TM domains in phospholipid bicelles by multidimensional NMR spectroscopy (Fig. 2.3) in
analogy to the previously determined αIIb and β3 structures(Lau, Dua, et al., 2008; Lau, Partridge,
et al., 2008). The backbone structures of β2 and β3 were essentially indistinguishable; the
relatively long TM helix, estimated at 29 residues (Lau, Partridge, et al., 2008), closely matched
(Fig. 2.3a). Small deviations occurred at both helix termini, at the N-terminus between β2(I679)
and β3(I693), and at the C-terminus between β2(D709) and β3(I721), with the β2 helix deviating
slightly from linear conformation. However, these relatively small differences appear unlikely to
be responsible for the lower-than-expected ΔG°TM of αXβ2.
The structure of αX and αIIb were indistinguishable for approximately two thirds of the
TM domain, αX(G1096-F114) or αIIb(G975-F993) (Fig. 2.3b). At the N-terminus, however, the

60
backbone structures of αX and αIIb consistently differed for at least two helix turns (Fig. 2.3d).
The distortion of the αX helix misplaced the hydrogen bonding residue Ser1093 relative to its
αIIb(G972) counterpart. To better understand the consequence of the structural differences
between αX and αIIb, we assembled the αXβ2 TM complex based on the prototpyical
αIIb(W968V)β3 TM complex (A. J. Situ et al., 2021). The intersubunit contacts, including the
implemented αX(S1093)-β2(T686) hydrogen bond, nudged the N-terminal helix turns of αX but
not of β2 towards a more linear helix. In other words, there is no hindrance to a well-packed αXβ2
complex but the monomeric αX structure has to change.
The xPxxxGSS motif couples hydrogen bond strength to the membrane environment
The destabilization of the two N-terminal TM helix turns in αX is quite unusual and, to our
knowledge, unprecedented in MP. Moreover, the stark difference between αX and αIIb properties
seems surprising given only few sequence differences. To identify the origin of the αX
destabilization, we screened for amino acid substitution(s) that can revert it. With
αIIb(W967)/αX(P1088) perhaps the most conspicuous of the sequence differences, we studied its

61
Fig. 2.4. Comparison of {1H}-
15N NOE values for αX and αX mutants. Backbone dynamics in the
form of {1H}-15N NOE values are compared between of αX, αX(P1088A), αX(P1088W) and
αX(S1094I).
Fig. 2.5. Overlay of secondary 13Cα chemical shifts of αX, αX(P1088A), and αX(P1088W). For
random coil conformations, shifts are close to zero, whereas positive and negative shifts denote
helical and extended backbone propensities, respectively (Spera & Bax, 1991; Wishart & Case,
2002).

62
reversal first. For αX(P1088W), helical content for the entirety of the two N-terminal helix turns
was restored and {1H}-
15N NOE values increased (Fig. 2.5 and 2.6). Trp is the most prominent
membrane anchor of all amino acids (A. J. Situ et al., 2018) and we wondered if lipid anchoring
contributed to this observation. However, αX(P1088A) was indistinguishable from the Trp
mutation (Fig. 2.6), negating this possibility. For the sake of completeness, we also examined
individual αX(G1092L) and αX(S1094I) point mutations. To our surprise, each substitution also
removed the destabilization of the N-terminal helix turns (Fig. 2.7). Thus, the destabilized Nterminal turns arose from a cooperative behavior of the xPxxxGSS sequence where x denotes a
hydrophobic residue. We hypothesize two factors underpin this behavior. First, Pro and Gly are
not preferred in a-helices, making their a-helical conformation less stable. Second, membrane
immersion of Gly and Ser is mildly unfavorable. Consequently, side chain immersion may not
Fig. 2.6. Overlay of secondary 13Cα chemical shifts of αX, αX(P1088A), αX(G1092L), and
αX(S1094I). For random coil conformations, shifts are close to zero, whereas positive and negative
shifts denote helical and extended backbone propensities, respectively (Spera & Bax, 1991; Wishart &
Case, 2002).

63
provide enough driving force to stabilize the xPxxxGSS backbone hydrogen bonds in the
membrane. However, this consideration usually extends over all TM helix residues and the
division of the αX helix is unusual. TM helices are assumed to behave as a unit. This suggests that
the motif distorts membrane lipids in turn.
For xPxxxGSS to exhibit reduced helical content implies that hydrogen bonds are
destabilized. As these bonds must form, interactions with lipid headgroups or water molecules
have to exist. As a corollary, xPxxxGSS stability may be coupled to membrane properties. For
example, shorter-chain lipids should destabilize xPxxxGSS more than longer-chain lipids by
allowing more proximal hydrogen bond partners and exerting less penalties for not forming
intrahelical bonds. To test this hypothesis, we monitored helical content by NMR spectroscopy
Fig. 2.7. Overlay of secondary 13Cα chemical shifts of αX in DMPC, POPC, and DEPC. For
random coil conformations, shifts are close to zero, whereas positive and negative shifts denote
helical and extended backbone propensities, respectively (Spera & Bax, 1991; Wishart & Case,
2002).

64
for lipids with hydrocarbon chain length of 14:0-14:0 (DMPC), 16:0-18:1 (POPC) and 20:1-20:1
(DEPC). Increased helical content will incrementally increase 13Cα shifts relative to their residuespecific random coil values. When reducing the hydrocarbon chain length, helical content
decreased accordingly (Fig. 2.7). It is thus evident that the helical content of the two N-terminal
helix turns of αX is coupled to lipid hydrocarbon length (membrane thickness).
Figure 2.8. Integrin αXβ2 transmembrane complex stability. Fluorescence spectra of the depicted
β2 TM peptides at the indicated αX peptide concentrations. All spectra were recorded at a β2
concentration of 3 mM and the αX concentrations indicated in 25 mM HEPES pH 7.4, 120 NaCl, 6 mM
POPC solution at 28 °C. The solution was extruded to yield large unilamellar vesicles (LUV).

65
To verify the functionality of the xPxxxGSS motif in the sealed bilayer environment of
LUVs, we next determined ΔG°TM for αX(P1088A)β2 (Table 2.2 and Fig. 2.8). For this pair, an
interaction with ΔG°TM of 3.7 ± 0.4 kcal/mol was indeed detected (Table 2.2). Within
experimental uncertainties, this value is within the expected value of 4.0 kcal/mol based on the
substituted αIIbβ3 TM complex. If the αX(S1093)-β2(T686) hydrogen bond is instrumental in
this stabilization, the β2(T686V) should decrease ΔG°TM again. We also note that β2(T686) is not
in the dimerization interface and not relevant for packing outside of hydrogen bonding. In
confirmation of this expectation, a stable αX(P1088A)β2(T686V) TM complex was not
detectable (Fig. 2.8 and Table 2.2). Thus, the membrane environment is sensed by xPxxxGSS,
encoded in the interhelical αXβ2 hydrogen bond strength, and integrated into ΔG°TM.
Table 2.2. Integrin αXβ2 TM complex stabilities in large unilamellar vesicles (LUV)
Peptides a KXY ΔGo [kcal/mol] b
αX + β2 – c –
αX(P1088A) + β2 570 ± 450 -3.7 ± 0.4
αX(P1088A) + β2(T686V) – c –
αIIb(W967Y/W968V/G972S/W988Y) +
β3(V700T) d
1450 ± 140 -4.4 ± 0.1
a
Measurements in extruded 6 mM POPC, 25 mM NaH2PO4/Na2HPO4 pH 7.4 at 28 °C
bΔG° = ΔRT ln KXY = ΔH° - TΔS° where T denotes the absolute temperature and R the gas constant
c
no binding detected
dTaken from (Lau et al., 2009; Lau, Partridge, et al., 2008)

66
2.4 – Discussion
αX has evolved a xPxxxGSS motif that destabilizes a third of its TM domain
The problem of membrane protein folding is two-fold. First of all, how does selective
peptide transport to opposite sides of the cellular membrane contribute to the establishment of
protein topology? Secondly, how is the membrane protein inserted, folded, and integrated into
the cell membrane? For membrane proteins with oligomeric α-helices such as the integrin
family, the TM α-helix is considered the smallest unit of assembly, and is usually folded first
before insertion into the lipid bilayer (Phillips & Miller, 2021; Skach, 2009). This makes the
structural features of the xPxxxGSS motif found for αX highly unusual for its class of protein. A
full third of the αX TM domain is destabilized by this motif, that is eight of the 23 TM residues
have unfavorable features for α-helical conformation (Fig. 2.6). The motif contains sidechains
that are unfavorable for a single-pass α-helix protein such as αX. While the kink in proline is
generally favorable for tight packing of multimeric proteins, they otherwise introduce a host of
destabilizing interhelical interactions and open the TM helix to interactions from the
environment. (Schimmel & Flory, 1968; Wedemeyer et al., 2002). Similarly, glycine is often
found within the membrane to aid packing of multi-pass proteins, (Dong et al., 2012), and
serines and threonines are not usually found buried in the membrane, rather on the surface to
aid in turns (Ballesteros et al., 2000). Secondary structures are usually stabilized by
intramolecular hydrogen bonds and the presence of unfavorable hydrophobic interactions in the
immediate environment (Hubbard & Kamran Haider, 2010; Rehman et al., 2024). αXβ2 exists
with destabilized backbone hydrogen bonds that couple hydrogen bond strength with backbone

67
dynamics and the membrane environment. This motif shows that membrane domains can
utilize opposing principles as a tool to aid in their function.
The xPxxxGSS motif acts as an activation switch for αXβ2 integrin in its function
Paradoxically, the xPxxxGSS motif is beneficial to αXβ2 function. The helical content of
αXβ2 changes with changes in lipid length in membrane mimics (Fig. 2.7). Mutations of residues
within the xPxxxGSS motif rescue helix content, demonstrating that the motif serves as a sensor
for the membrane environment. Previously explored αIIbβ3 integrin utilizes a corresponding
αIIbβ3(W968xxxG972) motif identified in the TM complex of integrin aIIbb3 with αIIb(G972) and
αX(S972) at homologous sites that safequards αIIbβ3 stabilization (A. J. Situ et al., 2021). By
contrast, engagement of the xPxxxGSS motif maintains the active state, and disengagement
stabilizes the dimeric, inactive state. It can thus be considered an activation motif.
2.5 - Conclusion
In conclusion, despite a similar base mechanism of signal transduction through monomerdimer shift, αXβ2 does not make use of the same regulatory mechanisms as αIIβb. Here we use
NMR spectroscopy and LUV fluorescence spectroscopy to show that sequence differences in αX
against αIIb results in destabilization of a third of the TM α helix at the N-terminus, thus leading to
stabilization of the disassociated, active conformation. This difference, the αX(xPxxxGSS) motif,
resembles the αIIb(W968xxxG972) safeguarding motif identified in the TM complex of integrin
αIIbβ3 with αIIb(G972) and aX(S972) at homologous sites. Both motifs dynamically control ΔG°TM.

68
However, while in case of aIIbb3 the inactive state is destabilized by the motif, for b2 integrins we
regard the motif to stabilize the inactive state. In other words, xPxxxGSS is expected to constitute
an activation switch for aXb2. This motif causes αXβ2 to be more sensitive to changes in the
membrane environment, as demonstrated by sensitivity to different lipid lengths in the membrane
mimic. This may aid in the function of b2 integrins, which are present in immune cells and
additionally can mediate phagocytosis. Phagocytosis is a special form of cell endocytosis, whereby
cells ingest microbes, dead cells and debris solid through vesicles. Integrin activation occurs
through intracellular binding. However, membrane perturbation then adds a mechanical aspect
to integrin activation. Sensitivity to the membrane environment conferred by the xPxxxGSS motif
may lower the threshold of b2 integrin activation when the lipid environment shifts, thus
maintaining an active state could be beneficial for phagocytosis.

69
Future Directions
There is additional work that can be done to further explore the structural basis of both
CD33 and integrin function. In order to further investigate the signaling mechanism of CD33, the
following can be done:
ITIM phosphorylation and ligand binding of the cytosolic tail
The cytosolic tail has two potential ITIM sites that can be phosphorylated by the Src kinase
at Tyr340 and Tyr358, which are consequently dephosphorylated by SH2-domain containing
phosphatases such as SHP1 and SHP2 (Paul et al., 2000; Walter, Raden, et al., 2008). Binding
assays have revealed numerous other binding partners with a large variety of functions. COMMD3
is one, and it interacts with E3 ubiquitin ligase and downregulates pro-inflammatory cytokines
like NF-ΚB in cancer (Burstein et al., 2005; Huttlin et al., 2021; Y. F. Liu et al., 2013). Cbl is an E3
ubiquitin ligase that potentially marks CD33 for downregulation (Meng et al., 1999; Walter,
Häusermann, et al., 2008). GIMAP5 is a final example, a GTPase shown to moderate immune
function and promote an anti-inflammatory response (Luck et al., 2020; A. Y. Park et al., 2024;
Patterson et al., 2018). These variable binding partners induce a variety of functions. Given that
we hypothesize CD33 undergoes lipid raft clustering, and the extended conformation conferred
upon the cytosolic tail by its many prolines, it is possible that ligand binding at the cytosolic tail
involves large complex formation that induce a signaling cascade (Simons & Toomre, 2000;
Varshney et al., 2016). However, binding of the mostly unstructured cytosolic tail may also result
in restructuring that stabilizes local or global structures, perhaps inhibiting further ligand binding,
as is the case for many proteins (Alberts et al., 2002). Obtaining the bound structure and binding

70
energy of the CD33 cytosolic tail and its various ligands would provide valuable insight into how
CD33 responds to binding sialylated glycan ligands in the extracellular space to induce a contextdependent microglial response.
Membrane mimic with cholesterol and glycosphingolipids
We hypothesized that the structure of CD33 suggests utilization of lipid raft clustering to
transmit signal. The lipid rafts are enriched in sphingolipids, often sphingomyelin, and cholesterol
by ~50% and 300-500% respectively as compared to the surrounding bilayer (Anchisi et al., 2013).
It would thus be useful to study the interaction of the TM domains within the context of a
membrane mimic containing cholesterol and/orsphingolipids. This has been done before in solidstate NMR using a variety of cholesterol and sphingolipid mixes for membrane proteins
mammalian translocator protein TPSO and influenza A M2 protein (Cady et al., 2011; Della Ripa
et al., 2018; Jaipuria et al., 2018). These studies can elucidate how the CD33 TM domain interacts
with its lipid environment, whether that differs between a lipid raft or free bilayer, and if the TMTM association state is altered upon lipid raft clustering.
In order to further explore the activation dynamics of the integrin family subunits, the
following can be done:
Comparison of αXβ2 with αMβ2
Given that αXβ2 is non-comparable to αIIbβ3, it warrants interest in comparing other
integrin α subunits that could have greater similarities in the structural basis of function. Of the
other human integrins, αM has the greatest homology to αX (Fig. 2.2). It contains the same
xPxxxGSS motif found in αX, and could utilize congruent signal transduction mechanisms. There

71
is great overlap in protein ligands for both integrins, but each retains specificity in function - in
cases such as iC3b, αXβ2 binds at two moieties distinct from the moieties bound by αMβ2 (VorupJensen et al., 2003). They are also both expressed on immune cells, but of different types. αXβ2
is the primary β2 integrin on dendritic cell sand tissue-specific macrophages, while αMβ2 is the
major β2 integrin on neutrophils. While minor, these differences could translate to observable
structural differences that can elucidate the molecular basis of differences in signaling
mechanisms within the integrin family.

72
References

Abraham, M. J., Murtola, T., Schulz, R., Páll, S., Smith, J. C., Hess, B., & Lindahl, E. (2015). GROMACS: High
performance molecular simulations through multi-level parallelism from laptops to supercomputers.
SoftwareX, 1–2, 19–25. https://doi.org/10.1016/j.softx.2015.06.001
Akiyama, H., & McGeer, P. L. (1990). Brain microglia constitutively express beta-2 integrins. Journal of
Neuroimmunology, 30(1), 81–93. https://doi.org/10.1016/0165-5728(90)90055-r
Alberts, B., Johnson, A., Lewis, J., Raff, M., Roberts, K., & Walter, P. (2002). Signaling through EnzymeLinked Cell-Surface Receptors. In Molecular Biology of the Cell. 4th edition. Garland Science.
https://www.ncbi.nlm.nih.gov/books/NBK26822/
Anchisi, L., Dessì, S., Pani, A., & Mandas, A. (2013). Cholesterol homeostasis: A key to prevent or slow
down neurodegeneration. Frontiers in Physiology, 3, 486. https://doi.org/10.3389/fphys.2012.00486
Ando, M., Shoji, T., Tu, W., Higuchi, H., Nishijima, K., & Iijima, S. (2015). Lectin-dependent localization of
cell surface sialic acid-binding lectin Siglec-9. Cytotechnology, 67(4), 601–608.
https://doi.org/10.1007/s10616-014-9691-6
Angata, T., Hayakawa, T., Yamanaka, M., Varki, A., & Nakamura, M. (2006). Discovery of Siglec-14, a novel
sialic acid receptor undergoing concerted evolution with Siglec-5 in primates. The FASEB Journal, 20(12),
1964–1973. https://doi.org/10.1096/fj.06-5800com
Angata, T., & Varki, A. (2023). Discovery, classification, evolution and diversity of Siglecs. Molecular
Aspects of Medicine, 90, 101117. https://doi.org/10.1016/j.mam.2022.101117
Arkhipov, A., Shan, Y., Das, R., Endres, N. F., Eastwood, M. P., Wemmer, D. E., Kuriyan, J., & Shaw, D. E.
(2013). Architecture and Membrane Interactions of the EGF Receptor. Cell, 152(3), 557–569.
https://doi.org/10.1016/j.cell.2012.12.030
Bachurin, S. O., Bovina, E. V., & Ustyugov, A. A. (2017). Drugs in Clinical Trials for Alzheimer’s Disease: The
Major Trends. Medicinal Research Reviews, 37(5), 1186–1225. https://doi.org/10.1002/med.21434
Ballesteros, J. A., Deupi, X., Olivella, M., Haaksma, E. E., & Pardo, L. (2000). Serine and threonine residues
bend alpha-helices in the chi(1) = g(-) conformation. Biophysical Journal, 79(5), 2754–2760.
Bamberger, M. E., Harris, M. E., McDonald, D. R., Husemann, J., & Landreth, G. E. (2003). A Cell Surface
Receptor Complex for Fibrillar β-Amyloid Mediates Microglial Activation. The Journal of Neuroscience,
23(7), 2665–2674. https://doi.org/10.1523/JNEUROSCI.23-07-02665.2003
Bank, R. P. D. (n.d.-a). RCSB PDB - 5IHB: Structure of the immune receptor CD33. Retrieved May 9, 2024,
from https://www.rcsb.org/structure/5IHB
Bank, R. P. D. (n.d.-b). RCSB PDB - 5J0B: Structure of the immune receptor CD33 in complex with 6’-
sialyllactose. Retrieved May 10, 2024, from https://www.rcsb.org/structure/5J0B
Bank, R. P. D. (n.d.-c). RCSB PDB: Homepage. Retrieved June 7, 2024, from https://www.rcsb.org/

73
Batool, M., Ahmad, B., & Choi, S. (2019). A Structure-Based Drug Discovery Paradigm. International
Journal of Molecular Sciences, 20(11), 2783. https://doi.org/10.3390/ijms20112783
Bhattacherjee, A., Jung, J., Zia, S., Ho, M., Eskandari-Sedighi, G., St. Laurent, C. D., McCord, K. A., Bains,
A., Sidhu, G., Sarkar, S., Plemel, J. R., & Macauley, M. S. (2021). The CD33 short isoform is a gain-offunction variant that enhances Aβ1–42 phagocytosis in microglia. Molecular Neurodegeneration, 16(1),
19. https://doi.org/10.1186/s13024-021-00443-6
Biose, I. J., Ismael, S., Ouvrier, B., White, A. L., & Bix, G. J. (2023). The Potential Role of Integrin Signaling
in Memory and Cognitive Impairment. Biomolecules, 13(1), 108. https://doi.org/10.3390/biom13010108
Blixt, O., Collins, B. E., Nieuwenhof, I. M. van den, Crocker, P. R., & Paulson, J. C. (2003). Sialoside
Specificity of the Siglec Family Assessed Using Novel Multivalent Probes: IDENTIFICATION OF POTENT
INHIBITORS OF MYELIN-ASSOCIATED GLYCOPROTEIN *. Journal of Biological Chemistry, 278(33), 31007–
31019. https://doi.org/10.1074/jbc.M304331200
Bocharov, E. V., Mayzel, M. L., Volynsky, P. E., Goncharuk, M. V., Ermolyuk, Y. S., Schulga, A. A.,
Artemenko, E. O., Efremov, R. G., & Arseniev, A. S. (2008). Spatial Structure and pH-dependent
Conformational Diversity of Dimeric Transmembrane Domain of the Receptor Tyrosine Kinase EphA1∗.
Journal of Biological Chemistry, 283(43), 29385–29395. https://doi.org/10.1074/jbc.M803089200
Bradshaw, E. M., Chibnik, L. B., Keenan, B. T., Ottoboni, L., Raj, T., Tang, A., Rosenkrantz, L. L., Imboywa,
S., Lee, M., Korff, A. V., Morris, M. C., Evans, D. A., Johnson, K., Sperling, R. A., Schneider, J. A., Bennett, D.
A., & Jager, P. L. D. (2013). CD33 Alzheimer’s disease locus: Altered monocyte function and amyloid
biology. Nature Neuroscience, 16(7), 848–850. https://doi.org/10.1038/nn.3435
Büll, C., Nason, R., Sun, L., Van Coillie, J., Madriz Sørensen, D., Moons, S. J., Yang, Z., Arbitman, S.,
Fernandes, S. M., Furukawa, S., McBride, R., Nycholat, C. M., Adema, G. J., Paulson, J. C., Schnaar, R. L.,
Boltje, T. J., Clausen, H., & Narimatsu, Y. (2021). Probing the binding specificities of human Siglecs by cellbased glycan arrays. Proceedings of the National Academy of Sciences, 118(17), e2026102118.
https://doi.org/10.1073/pnas.2026102118
Burstein, E., Hoberg, J. E., Wilkinson, A. S., Rumble, J. M., Csomos, R. A., Komarck, C. M., Maine, G. N.,
Wilkinson, J. C., Mayo, M. W., & Duckett, C. S. (2005). COMMD Proteins, a Novel Family of Structural and
Functional Homologs of MURR1. Journal of Biological Chemistry, 280(23), 22222–22232.
https://doi.org/10.1074/jbc.M501928200
Butovsky, O., & Weiner, H. L. (2018). Microglial signatures and their role in health and disease. Nature
Reviews Neuroscience, 19(10), Article 10. https://doi.org/10.1038/s41583-018-0057-5
Cady, S., Wang, T., & Hong, M. (2011). Membrane-Dependent Effects of a Cytoplasmic Helix on the
Structure and Drug Binding of the Influenza Virus M2 Protein. Journal of the American Chemical Society,
133(30), 11572–11579. https://doi.org/10.1021/ja202051n
Cao, H., & Crocker, P. R. (2011). Evolution of CD33-related siglecs: Regulating host immune functions and
escaping pathogen exploitation?: Evolution of CD33-related siglecs. Immunology, 132(1), 18–26.
https://doi.org/10.1111/j.1365-2567.2010.03368.x

74
Cheng, H.-T., Megha, & London, E. (2009). Preparation and Properties of Asymmetric Vesicles That Mimic
Cell Membranes*. Journal of Biological Chemistry, 284(10), 6079–6092.
https://doi.org/10.1074/jbc.M806077200
Chou, J. J., Delaglio, F., & Bax, A. (2000). Measurement of one-bond 15N-13C’ dipolar couplings in
medium sized proteins. Journal of Biomolecular NMR, 18(2), 101–105.
https://doi.org/10.1023/a:1008358318863
Cornish, A. L., Freeman, S., Forbes, G., Ni, J., Zhang, M., Cepeda, M., Gentz, R., Augustus, M., Carter, K. C.,
& Crocker, P. R. (1998). Characterization of Siglec-5, a Novel Glycoprotein Expressed on Myeloid Cells
Related to CD33. Blood, 92(6), 2123–2132. https://doi.org/10.1182/blood.V92.6.2123
Cornish, J., Chamberlain, S. G., Owen, D., & Mott, H. R. (2020). Intrinsically disordered proteins and
membranes: A marriage of convenience for cell signalling? Biochemical Society Transactions, 48(6),
2669–2689. https://doi.org/10.1042/BST20200467
Crocker, P. R., McMillan, S. J., & Richards, H. E. (2012). CD33-related siglecs as potential modulators of
inflammatory responses: CD33-related siglecs and inflammatory responses. Annals of the New York
Academy of Sciences, 1253(1), 102–111. https://doi.org/10.1111/j.1749-6632.2011.06449.x
Crocker, P. R., Paulson, J. C., & Varki, A. (2007). Siglecs and their roles in the immune system. Nature
Reviews Immunology, 7(4), 255–266. https://doi.org/10.1038/nri2056
Della Ripa, L. A., Petros, Z. A., Cioffi, A. G., Piehl, D. W., Courtney, J. M., Burke, M. D., & Rienstra, C. M.
(2018). Solid-State NMR of highly 13C-enriched cholesterol in lipid bilayers. Methods, 138–139, 47–53.
https://doi.org/10.1016/j.ymeth.2018.01.008
Dong, H., Sharma, M., Zhou, H.-X., & Cross, T. A. (2012). Glycines: Role in α-Helical Membrane Protein
Structures and a Potential Indicator for Native Conformation. Biochemistry, 51(24), 4779–4789.
https://doi.org/10.1021/bi300090x
Engelhardt, B., & Sorokin, L. (2009). The blood-brain and the blood-cerebrospinal fluid barriers: Function
and dysfunction. Seminars in Immunopathology, 31(4), 497–511. https://doi.org/10.1007/s00281-009-
0177-0
Ereño-Orbea, J., Sicard, T., Cui, H., Mazhab-Jafari, M. T., Benlekbir, S., Guarné, A., Rubinstein, J. L., &
Julien, J.-P. (2017). Molecular basis of human CD22 function and therapeutic targeting. Nature
Communications, 8(1), 764. https://doi.org/10.1038/s41467-017-00836-6
Fitzkee, N. C., & Bax, A. (2010). Facile measurement of 1H–15N residual dipolar couplings in larger
perdeuterated proteins. Journal of Biomolecular NMR, 48(2), 65–70. https://doi.org/10.1007/s10858-
010-9441-9
Floyd, H., Ni, J., Cornish, A. L., Zeng, Z., Liu, D., Carter, K. C., Steel, J., & Crocker, P. R. (2000). Siglec-8.
Journal of Biological Chemistry, 275(2), 861–866. https://doi.org/10.1074/jbc.275.2.861
Garcıá-Alvarez, B., de Pereda, J. M., Calderwood, D. A., Ulmer, T. S., Critchley, D., Campbell, I. D.,
Ginsberg, M. H., & Liddington, R. C. (2003). Structural Determinants of Integrin Recognition by Talin.
Molecular Cell, 11(1), 49–58. https://doi.org/10.1016/S1097-2765(02)00823-7

75
Gonzalez-Gil, A., Li, T. A., Kim, J., & Schnaar, R. L. (2023). Human sialoglycan ligands for immune
inhibitory Siglecs. Molecular Aspects of Medicine, 90, 101110.
https://doi.org/10.1016/j.mam.2022.101110
Gonzalez-Gil, A., Porell, R. N., Fernandes, S. M., Maenpaa, E., Li, T. A., Li, T., Wong, P. C., Aoki, K.,
Tiemeyer, M., Yu, Z. J., Orsburn, B. C., Bumpus, N. N., Matthews, R. T., & Schnaar, R. L. (2022). Human
brain sialoglycan ligand for CD33, a microglial inhibitory Siglec implicated in Alzheimer’s disease. The
Journal of Biological Chemistry, 298(6), 101960. https://doi.org/10.1016/j.jbc.2022.101960
Gonzalez-Gil, A., & Schnaar, R. L. (2021). Siglec Ligands. Cells, 10(5), 1260.
https://doi.org/10.3390/cells10051260
Griciuc, A., & Tanzi, R. E. (2021). The role of innate immune genes in Alzheimer’s disease. Current
Opinion in Neurology, 34(2), 228. https://doi.org/10.1097/WCO.0000000000000911
Grishaev, A., & Bax, A. (2004). An Empirical Backbone−Backbone Hydrogen-Bonding Potential in Proteins
and Its Applications to NMR Structure Refinement and Validation. Journal of the American Chemical
Society, 126(23), 7281–7292. https://doi.org/10.1021/ja0319994
Ha, S.-H., Kwak, C.-H., Park, J.-Y., Abekura, F., Lee, Y.-C., Kim, J., Chung, T.-W., & Kim, C.-H. (2020). 3ʹsialyllactose targets cell surface protein, SIGLEC-3, and induces megakaryocyte differentiation and
apoptosis by lipid raft-dependent endocytosis. Glycoconjugate Journal, 37(2), 187–200.
https://doi.org/10.1007/s10719-019-09902-1
Halder, S. K., Delorme-Walker, V. D., & Milner, R. (2023). Β1 integrin is essential for blood–brain barrier
integrity under stable and vascular remodelling conditions; effects differ with age. Fluids and Barriers of
the CNS, 20(1), 52. https://doi.org/10.1186/s12987-023-00453-0
Heneka, M. T., Carson, M. J., El Khoury, J., Landreth, G. E., Brosseron, F., Feinstein, D. L., Jacobs, A. H.,
Wyss-Coray, T., Vitorica, J., Ransohoff, R. M., Herrup, K., Frautschy, S. A., Finsen, B., Brown, G. C.,
Verkhratsky, A., Yamanaka, K., Koistinaho, J., Latz, E., Halle, A., … Kummer, M. P. (2015).
Neuroinflammation in Alzheimer’s Disease. The Lancet. Neurology, 14(4), 388–405.
https://doi.org/10.1016/S1474-4422(15)70016-5
Hodivala-Dilke, K. M., McHugh, K. P., Tsakiris, D. A., Rayburn, H., Crowley, D., Ullman-Culleré, M., Ross, F.
P., Coller, B. S., Teitelbaum, S., & Hynes, R. O. (1999). Β3-integrin–deficient mice are a model for
Glanzmann thrombasthenia showing placental defects and reduced survival. Journal of Clinical
Investigation, 103(2), 229–238.
Hogg, N., Patzak, I., & Willenbrock, F. (2011). The insider’s guide to leukocyte integrin signalling and
function. Nature Reviews. Immunology, 11(6), 416–426. https://doi.org/10.1038/nri2986
Hollingworth, P., Harold, D., Sims, R., Gerrish, A., Lambert, J.-C., Carrasquillo, M. M., Abraham, R.,
Hamshere, M. L., Pahwa, J. S., Moskvina, V., Dowzell, K., Jones, N., Stretton, A., Thomas, C., Richards, A.,
Ivanov, D., Widdowson, C., Chapman, J., Lovestone, S., … Williams, J. (2011). Common variants at ABCA7 ,
MS4A6A/MS4A4E , EPHA1 , CD33 and CD2AP are associated with Alzheimer’s disease. Nature Genetics,
43(5), 429–435. https://doi.org/10.1038/ng.803

76
Hu, J.-S., Grzesiek, S., & Bax, A. (1997). Two-Dimensional NMR Methods for Determining χ1 Angles of
Aromatic Residues in Proteins from Three-Bond JC‘Cγ and JNCγ Couplings. Journal of the American
Chemical Society, 119(7), 1803–1804. https://doi.org/10.1021/ja963625z
Huang, T. Y., & Xu, H. (2019). Bringing Order out of Chaos: Establishing an Epistatic Relationship between
CD33 and TREM2. Neuron, 103(5), 747–749. https://doi.org/10.1016/j.neuron.2019.08.019
Hubbard, R. E., & Kamran Haider, M. (2010). Hydrogen Bonds in Proteins: Role and Strength. In
Encyclopedia of Life Sciences. John Wiley & Sons, Ltd.
https://doi.org/10.1002/9780470015902.a0003011.pub2
Hughes, P. E., Diaz-Gonzalez, F., Leong, L., Wu, C., McDonald, J. A., Shattil, S. J., & Ginsberg, M. H. (1996).
Breaking the Integrin Hinge: A DEFINED STRUCTURAL CONSTRAINT REGULATES INTEGRIN SIGNALING (∗).
Journal of Biological Chemistry, 271(12), 6571–6574. https://doi.org/10.1074/jbc.271.12.6571
Huttlin, E. L., Bruckner, R. J., Navarrete-Perea, J., Cannon, J. R., Baltier, K., Gebreab, F., Gygi, M. P.,
Thornock, A., Zarraga, G., Tam, S., Szpyt, J., Gassaway, B. M., Panov, A., Parzen, H., Fu, S., Golbazi, A.,
Maenpaa, E., Stricker, K., Guha Thakurta, S., … Gygi, S. P. (2021). Dual proteome-scale networks reveal
cell-specific remodeling of the human interactome. Cell, 184(11), 3022-3040.e28.
https://doi.org/10.1016/j.cell.2021.04.011
Hynes, R. O. (2002). Integrins: Bidirectional, Allosteric Signaling Machines. Cell, 110(6), 673–687.
https://doi.org/10.1016/S0092-8674(02)00971-6
Jaipuria, G., Giller, K., Leonov, A., Becker, S., & Zweckstetter, M. (2018). Insights into
Cholesterol/Membrane Protein Interactions Using Paramagnetic Solid-State NMR. Chemistry (Weinheim
an Der Bergstrasse, Germany), 24(66), 17606–17611. https://doi.org/10.1002/chem.201804550
Jaroniec, C. P., Ulmer, T. S., & Bax, A. (2004). Quantitative J correlation methods for the accurate
measurement of 13 C - 13 C α dipolar couplings in proteins. Journal of Biomolecular NMR, 30(2), 181–194.
https://doi.org/10.1023/B:JNMR.0000048946.71249.2f
Joh, N. H., Min, A., Faham, S., Whitelegge, J. P., Yang, D., Woods, V. L., & Bowie, J. U. (2008). Modest
stabilization by most hydrogen-bonded side-chain interactions in membrane proteins. Nature,
453(7199), 1266–1270. https://doi.org/10.1038/nature06977
Jumper, J., Evans, R., Pritzel, A., Green, T., Figurnov, M., Ronneberger, O., Tunyasuvunakool, K., Bates, R.,
Žídek, A., Potapenko, A., Bridgland, A., Meyer, C., Kohl, S. A. A., Ballard, A. J., Cowie, A., Romera-Paredes,
B., Nikolov, S., Jain, R., Adler, J., … Hassabis, D. (2021). Highly accurate protein structure prediction with
AlphaFold. Nature, 596(7873), 583–589. https://doi.org/10.1038/s41586-021-03819-2
Kettenmann, H., Hanisch, U.-K., Noda, M., & Verkhratsky, A. (2011). Physiology of Microglia. Physiological
Reviews, 91(2), 461–553. https://doi.org/10.1152/physrev.00011.2010
Kim, C., Lau, T.-L., Ulmer, T. S., & Ginsberg, M. H. (2009). Interactions of platelet integrin αΙΙb and β3
transmembrane domains in mammalian cell membranes and their role in integrin activation. Blood,
113(19), 4747–4753. https://doi.org/10.1182/blood-2008-10-186551

77
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.
https://doi.org/10.1038/nature10697
Kontaxis, G., Clore, G. M., & Bax, A. (2000). Evaluation of Cross-Correlation Effects and Measurement of
One-Bond Couplings in Proteins with Short Transverse Relaxation Times. Journal of Magnetic Resonance,
143(1), 184–196. https://doi.org/10.1006/jmre.1999.1979
Kuszewski, J., Gronenborn, A. M., & Clore, G. M. (1997). Improvements and Extensions in the
Conformational Database Potential for the Refinement of NMR and X-ray Structures of Proteins and
Nucleic Acids. Journal of Magnetic Resonance, 125(1), 171–177.
https://doi.org/10.1006/jmre.1997.1116
Laskowski, R. A., MacArthur, M. W., Moss, D. S., & Thornton, J. M. (1993). PROCHECK: A program to
check the stereochemical quality of protein structures. Journal of Applied Crystallography, 26(2), 283–
291. https://doi.org/10.1107/S0021889892009944
Lau, T.-L., Dua, V., & Ulmer, T. S. (2008). Structure of the Integrin αIIb Transmembrane Segment. Journal
of Biological Chemistry, 283(23), 16162–16168. https://doi.org/10.1074/jbc.M801748200
Lau, T.-L., Kim, C., Ginsberg, M. H., & Ulmer, T. S. (2009). The structure of the integrin αIIbβ3
transmembrane complex explains integrin transmembrane signalling. The EMBO Journal, 28(9), 1351–
1361. https://doi.org/10.1038/emboj.2009.63
Lau, T.-L., Partridge, A. W., Ginsberg, M. H., & Ulmer, T. S. (2008). Structure of the Integrin β3
Transmembrane Segment in Phospholipid Bicelles and Detergent Micelles. Biochemistry, 47(13), 4008–
4016. https://doi.org/10.1021/bi800107a
Lawson, L., Perry, V. H., Dri, P., & Gordon, S. (1990). Heterogeneity in the distribution and morphology of
microglia in the normal adult mouse brain—ScienceDirect. https://www-sciencedirectcom.libproxy2.usc.edu/science/article/pii/030645229090229W?via%3Dihub
Liu, J., Tong, J., & Yang, H. (2022). Targeting CD33 for acute myeloid leukemia therapy. BMC Cancer, 22(1),
24. https://doi.org/10.1186/s12885-021-09116-5
Liu, Y. F., Swart, M., Ke, Y., Ly, K., & McDonald, F. J. (2013). Functional interaction of COMMD3 and
COMMD9 with the epithelial sodium channel. American Journal of Physiology-Renal Physiology, 305(1),
F80–F89. https://doi.org/10.1152/ajprenal.00158.2013
Liu, Y., Lau, J., Li, W., Tempel, W., Li, L., Dong, A., Narula, A., Qin, S., & Min, J. (2016). Structural basis for
the regulatory role of the PPxY motifs in the thioredoxin-interacting protein TXNIP. Biochemical Journal,
473(2), 179–187. https://doi.org/10.1042/BJ20150830
Luck, K., Kim, D.-K., Lambourne, L., Spirohn, K., Begg, B. E., Bian, W., Brignall, R., Cafarelli, T., CamposLaborie, F. J., Charloteaux, B., Choi, D., Coté, A. G., Daley, M., Deimling, S., Desbuleux, A., Dricot, A.,
Gebbia, M., Hardy, M. F., Kishore, N., … Calderwood, M. A. (2020). A reference map of the human binary
protein interactome. Nature, 580(7803), 402–408. https://doi.org/10.1038/s41586-020-2188-x
Lukasiak, A., & Zajac, M. (2021). The Distribution and Role of the CFTR Protein in the Intracellular
Compartments. Membranes, 11(11), 804. https://doi.org/10.3390/membranes11110804

78
Macauley, M. S., Crocker, P. R., & Paulson, J. C. (2014). Siglec-mediated regulation of immune cell
function in disease. Nature Reviews Immunology, 14(10), 653–666. https://doi.org/10.1038/nri3737
Malik, M., Chiles, J., Xi, H. S., Medway, C., Simpson, J., Potluri, S., Howard, D., Liang, Y., Paumi, C. M.,
Mukherjee, S., Crane, P., Younkin, S., Fardo, D. W., & Estus, S. (2015). Genetics of CD33 in Alzheimer’s
disease and acute myeloid leukemia. Human Molecular Genetics, 24(12), 3557–3570.
https://doi.org/10.1093/hmg/ddv092
Many of the Immunoglobulin Superfamily Domains in Cell Adhesion Molecules and Surface Receptors
Belong to a New Structural Set Which is close to That Containing Variable Domains. (1994). Journal of
Molecular Biology, 238(4), 528–539. https://doi.org/10.1006/jmbi.1994.1312
Martin, J., & Sawyer, A. (2019). Elucidating the structure of membrane proteins. BioTechniques, 66(4),
167–170. https://doi.org/10.2144/btn-2019-0030
Masters, C. L., Bateman, R., Blennow, K., Rowe, C. C., Sperling, R. A., & Cummings, J. L. (2015).
Alzheimer’s disease. Nature Reviews Disease Primers, 1(1), 15056. https://doi.org/10.1038/nrdp.2015.56
Membrane Proteins of Known Structure. (n.d.). Retrieved June 7, 2024, from
https://blanco.biomol.uci.edu/mpstruc/
Meng, W., Sawasdikosol, S., Burakoff, S. J., & Eck, M. J. (1999). Structure of the amino-terminal domain of
Cbl complexed to its binding site on ZAP-70 kinase. Nature, 398(6722), 84–90.
https://doi.org/10.1038/18050
Miles, L. A., Hermans, S. J., Crespi, G. A. N., Gooi, J. H., Doughty, L., Nero, T. L., Markulić, J., Ebneth, A.,
Wroblowski, B., Oehlrich, D., Trabanco, A. A., Rives, M.-L., Royaux, I., Hancock, N. C., & Parker, M. W.
(2019). Small Molecule Binding to Alzheimer Risk Factor CD33 Promotes Aβ Phagocytosis. iScience, 19,
110–118. https://doi.org/10.1016/j.isci.2019.07.023
Moon, C. P., & Fleming, K. G. (2011). Side-chain hydrophobicity scale derived from transmembrane
protein folding into lipid bilayers. Proceedings of the National Academy of Sciences, 108(25), 10174–
10177. https://doi.org/10.1073/pnas.1103979108
Mucke, L. (2009). Alzheimer’s disease. Nature, 461(7266), 895–897. https://doi.org/10.1038/461895a
Naj, A. C., Jun, G., Beecham, G. W., Wang, L.-S., Vardarajan, B. N., Buros, J., Gallins, P. J., Buxbaum, J. D.,
Jarvik, G. P., Crane, P. K., Larson, E. B., Bird, T. D., Boeve, B. F., Graff-Radford, N. R., De Jager, P. L., Evans,
D., Schneider, J. A., Carrasquillo, M. M., Ertekin-Taner, N., … Schellenberg, G. D. (2011). Common variants
in MS4A4/MS4A6E, CD2uAP, CD33, and EPHA1 are associated with late-onset Alzheimer’s disease.
Nature Genetics, 43(5), 436–441. https://doi.org/10.1038/ng.801
Neniskyte, U., Neher, J. J., & Brown, G. C. (2011). Neuronal death induced by nanomolar amyloid β is
mediated by primary phagocytosis of neurons by microglia. The Journal of Biological Chemistry, 286(46),
39904–39913. https://doi.org/10.1074/jbc.M111.267583
Nielsen, J. T., & Mulder, F. A. A. (2018). POTENCI: Prediction of temperature, neighbor and pH-corrected
chemical shifts for intrinsically disordered proteins. Journal of Biomolecular NMR, 70(3), 141–165.
https://doi.org/10.1007/s10858-018-0166-5

79
Orr, S. J., Morgan, N. M., Elliott, J., Burrows, J. F., Scott, C. J., McVicar, D. W., & Johnston, J. A. (2007).
CD33 responses are blocked by SOCS3 through accelerated proteasomal-mediated turnover. Blood,
109(3), 1061–1068. https://doi.org/10.1182/blood-2006-05-023556
Park, A. Y., Leney-Greene, M., Lynberg, M., Gabrielski, J. Q., Xu, X., Schwarz, B., Zheng, L.,
Balasubramaniyam, A., Ham, H., Chao, B., Zhang, Y., Matthews, H. F., Cui, J., Yao, Y., Kubo, S., Chanchu, J.
M., Morawski, A. R., Cook, S. A., Jiang, P., … Lenardo, M. J. (2024). GIMAP5 deficiency reveals a
mammalian ceramide-driven longevity assurance pathway. Nature Immunology, 25(2), 282–293.
https://doi.org/10.1038/s41590-023-01691-y
Park, S. H., & Opella, S. J. (2005). Tilt Angle of a Trans-membrane Helix is Determined by Hydrophobic
Mismatch. Journal of Molecular Biology, 350(2), 310–318. https://doi.org/10.1016/j.jmb.2005.05.004
Partridge, A. W., Liu, S., Kim, S., Bowie, J. U., & Ginsberg, M. H. (2005). Transmembrane Domain Helix
Packing Stabilizes Integrin αIIbβ3 in the Low Affinity State *. Journal of Biological Chemistry, 280(8),
7294–7300. https://doi.org/10.1074/jbc.M412701200
Pascual-Leone, A., Freitas, C., Oberman, L., Horvath, J. C., Halko, M., Eldaief, M., Bashir, S., Vernet, M.,
Shafi, M., Westover, B., Vahabzadeh-Hagh, A. M., & Rotenberg, A. (2011). Characterizing Brain Cortical
Plasticity and Network Dynamics Across the Age-Span in Health and Disease with TMS-EEG and TMSfMRI. Brain Topography, 24(3), 302–315. https://doi.org/10.1007/s10548-011-0196-8
Patterson, A. R., Endale, M., Lampe, K., Aksoylar, H. I., Flagg, A., Woodgett, J. R., Hildeman, D., Jordan, M.
B., Singh, H., Kucuk, Z., Bleesing, J., & Hoebe, K. (2018). Gimap5-dependent inactivation of GSK3β is
required for CD4+ T cell homeostasis and prevention of immune pathology. Nature Communications,
9(1), 430. https://doi.org/10.1038/s41467-018-02897-7
Paul, S. P., Taylor, L. S., Stansbury, E. K., & McVicar, D. W. (2000). Myeloid specific human CD33 is an
inhibitory receptor with differential ITIM function in recruiting the phosphatases SHP-1 and SHP-2. 96(2),
8.
Perez-Nievas, B. G., Stein, T. D., Tai, H.-C., Dols-Icardo, O., Scotton, T. C., Barroeta-Espar, I., FernandezCarballo, L., de Munain, E. L., Perez, J., Marquie, M., Serrano-Pozo, A., Frosch, M. P., Lowe, V., Parisi, J. E.,
Petersen, R. C., Ikonomovic, M. D., López, O. L., Klunk, W., Hyman, B. T., & Gómez-Isla, T. (2013).
Dissecting phenotypic traits linked to human resilience to Alzheimer’s pathology. Brain, 136(8), 2510–
2526. https://doi.org/10.1093/brain/awt171
Pérez-Oliva, A. B., Martínez-Esparza, M., Vicente-Fernández, J. J., Corral-San Miguel, R., GarcíaPeñarrubia, P., & Hernández-Caselles, T. (2011). Epitope mapping, expression and post-translational
modifications of two isoforms of CD33 (CD33M and CD33m) on lymphoid and myeloid human cells.
Glycobiology, 21(6), 757–770. https://doi.org/10.1093/glycob/cwq220
Pervushin, K., Riek, R., Wider, G., & Wüthrich, K. (1997). Attenuated T2 relaxation by mutual cancellation
of dipole–dipole coupling and chemical shift anisotropy indicates an avenue to NMR structures of very
large biological macromolecules in solution. Proceedings of the National Academy of Sciences, 94(23),
12366–12371. https://doi.org/10.1073/pnas.94.23.12366
Phillips, B. P., & Miller, E. A. (2021). Membrane protein folding and quality control. Current Opinion in
Structural Biology, 69, 50–54. https://doi.org/10.1016/j.sbi.2021.03.003

80
Pronker, M. F., Lemstra, S., Snijder, J., Heck, A. J. R., Thies-Weesie, D. M. E., Pasterkamp, R. J., & Janssen,
B. J. C. (2016). Structural basis of myelin-associated glycoprotein adhesion and signalling. Nature
Communications, 7(1), 13584. https://doi.org/10.1038/ncomms13584
Raj, T., Ryan, K. J., Replogle, J. M., Chibnik, L. B., Rosenkrantz, L., Tang, A., Rothamel, K., Stranger, B. E.,
Bennett, D. A., Evans, D. A., De Jager, P. L., & Bradshaw, E. M. (2014). CD33: Increased inclusion of exon 2
implicates the Ig V-set domain in Alzheimer’s disease susceptibility. Human Molecular Genetics, 23(10),
2729–2736. https://doi.org/10.1093/hmg/ddt666
Rajendran, L., & Paolicelli, R. C. (2018). Microglia-Mediated Synapse Loss in Alzheimer’s Disease. The
Journal of Neuroscience, 38(12), 2911–2919. https://doi.org/10.1523/JNEUROSCI.1136-17.2017
Rehman, I., Farooq, M., & Botelho, S. (2024). Biochemistry, Secondary Protein Structure. In StatPearls.
StatPearls Publishing. http://www.ncbi.nlm.nih.gov/books/NBK470235/
Robbins, P. W., Trimble, R. B., Wirth, D. F., Hering, C., Maley, F., Maley, G. F., Das, R., Gibson, B. W., Royal,
N., & Biemann, K. (1984). Primary structure of the Streptomyces enzyme endo-beta-Nacetylglucosaminidase H. Journal of Biological Chemistry, 259(12), 7577–7583.
https://doi.org/10.1016/S0021-9258(17)42829-8
Rossi, B., Angiari, S., Zenaro, E., Budui, S. L., & Constantin, G. (2011). Vascular inflammation in central
nervous system diseases: Adhesion receptors controlling leukocyte–endothelial interactions. Journal of
Leukocyte Biology, 89(4), 539–556. https://doi.org/10.1189/jlb.0710432
Sanders, C. R., & Schwonek, J. P. (1992). Characterization of magnetically orientable bilayers in mixtures
of dihexanoylphosphatidylcholine and dimyristoylphosphatidylcholine by solid-state NMR. Biochemistry,
31(37), 8898–8905. https://doi.org/10.1021/bi00152a029
Schimmel, P. R., & Flory, P. J. (1968). Conformational energies and configurational statistics of
copolypeptides containing l-proline. Journal of Molecular Biology, 34(1), 105–120.
https://doi.org/10.1016/0022-2836(68)90237-4
Schlossarek, S., Mearini, G., & Carrier, L. (2011). Cardiac myosin-binding protein C in hypertrophic
cardiomyopathy: Mechanisms and therapeutic opportunities. Journal of Molecular and Cellular
Cardiology, 50(4), 613–620. https://doi.org/10.1016/j.yjmcc.2011.01.014
Schmidt, T., Situ, A. J., & Ulmer, T. S. (2016). Structural and thermodynamic basis of proline-induced
transmembrane complex stabilization. Scientific Reports, 6(1), 1–7. https://doi.org/10.1038/srep29809
Schmidt, T., Suk, J.-E., Ye, F., Situ, A. J., Mazumder, P., Ginsberg, M. H., & Ulmer, T. S. (2015). Annular
Anionic Lipids Stabilize the Integrin αIIbβ3 Transmembrane Complex *. Journal of Biological Chemistry,
290(13), 8283–8293. https://doi.org/10.1074/jbc.M114.623504
Schmidt, T., Ye, F., Situ, A. J., An, W., Ginsberg, M. H., & Ulmer, T. S. (2016). A Conserved EctodomainTransmembrane Domain Linker Motif Tunes the Allosteric Regulation of Cell Surface Receptors. Journal
of Biological Chemistry, 291(34), 17536–17546. https://doi.org/10.1074/jbc.M116.733683
Schwieters, C. D., Kuszewski, J. J., Tjandra, N., & Marius Clore, G. (2003). The Xplor-NIH NMR molecular
structure determination package. Journal of Magnetic Resonance, 160(1), 65–73.
https://doi.org/10.1016/S1090-7807(02)00014-9

81
Shen, Y., & Bax, A. (2015). Protein Structural Information Derived from NMR Chemical Shift with the
Neural Network Program TALOS-N. In H. Cartwright (Ed.), Artificial Neural Networks (pp. 17–32).
Springer. https://doi.org/10.1007/978-1-4939-2239-0_2
Siddiqui, S. S., Matar, R., Merheb, M., Hodeify, R., Vazhappilly, C. G., Marton, J., Shamsuddin, S. A., & Al
Zouabi, H. (2019). Siglecs in Brain Function and Neurological Disorders. Cells, 8(10), 1125.
https://doi.org/10.3390/cells8101125
Siddiqui, S., Schwarz, F., Springer, S., Khedri, Z., Yu, H., Deng, L., Verhagen, A., Naito-Matsui, Y., Jiang, W.,
Kim, D., Zhou, J., Ding, B., Chen, X., Varki, N., & Varki, A. (2017). Studies on the Detection, Expression,
Glycosylation, Dimerization, and Ligand Binding Properties of Mouse Siglec-E. Journal of Biological
Chemistry, 292(3), 1029–1037. https://doi.org/10.1074/jbc.M116.738351
Siew, J. J., & Chern, Y. (2018). Microglial Lectins in Health and Neurological Diseases. Frontiers in
Molecular Neuroscience, 11. https://doi.org/10.3389/fnmol.2018.00158
Siglecs: A journey through the evolution of sialic acid-binding immunoglobulin-type lectins. (2018).
Developmental & Comparative Immunology, 86, 219–231. https://doi.org/10.1016/j.dci.2018.05.008
Simons, K., & Toomre, D. (2000). Lipid rafts and signal transduction. Nature Reviews Molecular Cell
Biology, 1(1), Article 1. https://doi.org/10.1038/35036052
Sims, R., Lee, S. J. van der, Naj, A. C., Bellenguez, C., Badarinarayan, N., Jakobsdottir, J., Kunkle, B. W.,
Boland, A., Raybould, R., Bis, J. C., Martin, E. R., Grenier-Boley, B., Heilmann-Heimbach, S., Chouraki, V.,
Kuzma, A. B., Sleegers, K., Vronskaya, M., Ruiz, A., Graham, R. R., … Schellenberg, G. D. (2017). Rare
coding variants in PLCG2 , ABI3 , and TREM2 implicate microglial-mediated innate immunity in
Alzheimer’s disease. Nature Genetics, 49(9), 1373–1384. https://doi.org/10.1038/ng.3916
Situ, A. J., Kang, S.-M., Frey, B. B., An, W., Kim, C., & Ulmer, T. S. (2018). Membrane Anchoring of α-Helical
Proteins: Role of Tryptophan. The Journal of Physical Chemistry B, 122(3), 1185–1194.
https://doi.org/10.1021/acs.jpcb.7b11227
Situ, A. J., Kim, J., An, W., Kim, C., & Ulmer, T. S. (2021). Insight Into Pathological Integrin αIIbβ3
Activation From Safeguarding The Inactive State. Journal of Molecular Biology, 433(7), 166832.
https://doi.org/10.1016/j.jmb.2021.166832
Situ, A. J., Schmidt, T., Mazumder, P., & Ulmer, T. S. (2014). Characterization of Membrane Protein
Interactions by Isothermal Titration Calorimetry. Journal of Molecular Biology, 426(21), 3670–3680.
https://doi.org/10.1016/j.jmb.2014.08.020
Situ, A., & Ulmer, T. (2019). Universal principles of membrane protein assembly, composition and
evolution. PLOS ONE, 14, e0221372. https://doi.org/10.1371/journal.pone.0221372
Situ, A., & Ulmer, T. (2023). Comparison of Integrin αIIbβ3 Transmembrane Association in Vesicles and
Bicelles. Biochemistry, 62. https://doi.org/10.1021/acs.biochem.3c00177
Skach, W. R. (2009). Cellular mechanisms of membrane protein folding. Nature Structural & Molecular
Biology, 16(6), 606–612. https://doi.org/10.1038/nsmb.1600

82
Spera, S., & Bax, A. (1991). Empirical correlation between protein backbone conformation and C.alpha.
And C.beta. 13C nuclear magnetic resonance chemical shifts. Journal of the American Chemical Society,
113(14), 5490–5492. https://doi.org/10.1021/ja00014a071
Suk, J.-E., Situ, A. J., & Ulmer, T. S. (2012). Construction of Covalent Membrane Protein Complexes and
High-Throughput Selection of Membrane Mimics. Journal of the American Chemical Society, 134(22),
9030–9033. https://doi.org/10.1021/ja304247f
Tanzi, R. E. (2013). A Brief History of Alzheimer’s Disease Gene Discovery. Journal of Alzheimer’s Disease,
33(s1), S5–S13. https://doi.org/10.3233/JAD-2012-129044
Tijms, B. M., Vromen, E. M., Mjaavatten, O., Holstege, H., Reus, L. M., Van Der Lee, S., Wesenhagen, K. E.
J., Lorenzini, L., Vermunt, L., Venkatraghavan, V., Tesi, N., Tomassen, J., Den Braber, A., Goossens, J.,
Vanmechelen, E., Barkhof, F., Pijnenburg, Y. A. L., Van Der Flier, W. M., Teunissen, C. E., … Visser, P. J.
(2024). Cerebrospinal fluid proteomics in patients with Alzheimer’s disease reveals five molecular
subtypes with distinct genetic risk profiles. Nature Aging, 4(1), 33–47. https://doi.org/10.1038/s43587-
023-00550-7
Tsakiris, D. A., Scudder, L., Hodivala-Dilke, K., Hynes, R. O., & Coller, B. S. (1999). Hemostasis in the mouse
(Mus musculus): A review. Thrombosis and Haemostasis, 81(2), 177–188.
Ulmer, T. S., Ramirez, B. E., Delaglio, F., & Bax, A. (2003). Evaluation of Backbone Proton Positions and
Dynamics in a Small Protein by Liquid Crystal NMR Spectroscopy. Journal of the American Chemical
Society, 125(30), 9179–9191. https://doi.org/10.1021/ja0350684
Ulmer, T. S., Yaspan, B., Ginsberg, M. H., & Campbell, I. D. (2001). NMR Analysis of Structure and
Dynamics of the Cytosolic Tails of Integrin αIIbβ3 in Aqueous Solution. Biochemistry, 40(25), 7498–7508.
https://doi.org/10.1021/bi010338l
Ulmschneider, M. B., & Sansom, M. S. P. (2001). Amino acid distributions in integral membrane protein
structures. Biochimica et Biophysica Acta (BBA) - Biomembranes, 1512(1), 1–14.
https://doi.org/10.1016/S0005-2736(01)00299-1
Ulmschneider, M. B., & Ulmschneider, J. P. (2008). Membrane adsorption, folding, insertion and
translocation of synthetic trans-membrane peptides. Molecular Membrane Biology, 25(3), 245–257.
https://doi.org/10.1080/09687680802020313
van Dyck Christopher H., Swanson Chad J., Aisen Paul, Bateman Randall J., Chen Christopher, Gee
Michelle, Kanekiyo Michio, Li David, Reyderman Larisa, Cohen Sharon, Froelich Lutz, Katayama Sadao,
Sabbagh Marwan, Vellas Bruno, Watson David, Dhadda Shobha, Irizarry Michael, Kramer Lynn D., &
Iwatsubo Takeshi. (2023). Lecanemab in Early Alzheimer’s Disease. New England Journal of Medicine,
388(1), 9–21. https://doi.org/10.1056/NEJMoa2212948
Varki, A., & Angata, T. (2006). Siglecs—The major subfamily of I-type lectins. Glycobiology, 16(1), 1R-27R.
https://doi.org/10.1093/glycob/cwj008
Varshney, P., Yadav, V., & Saini, N. (2016). Lipid rafts in immune signalling: Current progress and future
perspective. Immunology, 149(1), 13–24. https://doi.org/10.1111/imm.12617

83
von Heijne, G. (1992). Membrane protein structure prediction: Hydrophobicity analysis and the positiveinside rule. Journal of Molecular Biology, 225(2), 487–494. https://doi.org/10.1016/0022-
2836(92)90934-C
Vorup-Jensen, T., Ostermeier, C., Shimaoka, M., Hommel, U., & Springer, T. A. (2003). Structure and
allosteric regulation of the αXβ2 integrin I domain. Proceedings of the National Academy of Sciences,
100(4), 1873–1878. https://doi.org/10.1073/pnas.0237387100
Vu, H. N., Situ, A. J., & Ulmer, T. S. (2021). Isothermal Titration Calorimetry of Membrane Proteins. In I.
Schmidt-Krey & J. C. Gumbart (Eds.), Structure and Function of Membrane Proteins (Vol. 2302, pp. 69–
79). Springer US. https://doi.org/10.1007/978-1-0716-1394-8_5
Wagner, A., Galicia-Andrés, E., Teufl, M., Gold, L., Obinger, C., Sykacek, P., Oostenbrink, C., & Traxlmayr,
M. W. (2022). Identification of Activating Mutations in the Transmembrane and Extracellular Domains of
EGFR. Biochemistry, 61(19), 2049–2062. https://doi.org/10.1021/acs.biochem.2c00384
Walter, R. B., Häusermann, P., Raden, B. W., Teckchandani, A. M., Kamikura, D. M., Bernstein, I. D., &
Cooper, J. A. (2008). Phosphorylated ITIMs Enable Ubiquitylation of an Inhibitory Cell Surface Receptor.
Traffic, 9(2), 267–279. https://doi.org/10.1111/j.1600-0854.2007.00682.x
Walter, R. B., Raden, B. W., Zeng, R., Häusermann, P., Bernstein, I. D., & Cooper, J. A. (2008). ITIMdependent endocytosis of CD33-related Siglecs: Role of intracellular domain, tyrosine phosphorylation,
and the tyrosine phosphatases, Shp1 and Shp2. Journal of Leukocyte Biology, 83(1), 200–211.
https://doi.org/10.1189/jlb.0607388
Wedemeyer, W. J., Welker, E., & Scheraga, H. A. (2002). Proline cis-trans isomerization and protein
folding. Biochemistry, 41(50), 14637–14644.
Wegener, K. L., Partridge, A. W., Han, J., Pickford, A. R., Liddington, R. C., Ginsberg, M. H., & Campbell, I.
D. (2007). Structural basis of integrin activation by talin. Cell, 128(1), 171–182.
https://doi.org/10.1016/j.cell.2006.10.048
Williams, A. F. (1987). A year in the life of the immunoglobulin superfamily. Immunology Today, 8(10),
298–303. https://doi.org/10.1016/0167-5699(87)90016-8
Williams, A. F., & Barclay, A. N. (n.d.). THE IMMUNOGLOBULIN SUPERFAMILY-DOMAINS FOR CELL
SURFACE RECOGNITION1,2.
Winston Wong, P. (2020). Economic Burden of Alzheimer Disease and Managed Care Considerations. 26.
https://www.ajmc.com/view/economic-burden-of-alzheimer-disease-and-managed-care-considerations
Wishart, D. S., & Case, D. A. (2002). [1]—Use of Chemical Shifts in Macromolecular Structure
Determination. In T. L. James, V. Dötsch, & U. Schmitz (Eds.), Methods in Enzymology (Vol. 338, pp. 3–34).
Academic Press. https://doi.org/10.1016/S0076-6879(02)38214-4
Wisnovsky, S., Möckl, L., Malaker, S. A., Pedram, K., Hess, G. T., Riley, N. M., Gray, M. A., Smith, B. A. H.,
Bassik, M. C., Moerner, W. E., & Bertozzi, C. R. (2021). Genome-wide CRISPR screens reveal a specific
ligand for the glycan-binding immune checkpoint receptor Siglec-7. Proceedings of the National Academy
of Sciences of the United States of America, 118(5), e2015024118.
https://doi.org/10.1073/pnas.2015024118

84
Xiao, Q., McAtee, C. K., & Su, X. (2022). Phase separation in immune signalling. Nature Reviews
Immunology, 22(3), 188–199. https://doi.org/10.1038/s41577-021-00572-5
Xiong, J.-P., Stehle, T., Zhang, R., Joachimiak, A., Frech, M., Goodman, S. L., & Arnaout, M. A. (2002).
Crystal Structure of the Extracellular Segment of Integrin αVβ3 in Complex with an Arg-Gly-Asp Ligand.
Science, 296(5565), 151–155. https://doi.org/10.1126/science.1069040
Yu, Z., Maoui, M., Wu, L., Banville, D., & Shen, S. (2001). mSiglec-E, a novel mouse CD33-related siglec
(sialic acid-binding immunoglobulin-like lectin) that recruits Src homology 2 (SH2)-domain-containing
protein tyrosine phosphatases SHP-1 and SHP-2. Biochemical Journal, 353(Pt 3), 483–492.
Zhang, Y., Chen, H., Li, R., Sterling, K., & Song, W. (2023). Amyloid β-based therapy for Alzheimer’s
disease: Challenges, successes and future. Signal Transduction and Targeted Therapy, 8(1), 1–26.
https://doi.org/10.1038/s41392-023-01484-7
Zhuravleva, M. A., Trandem, K., & Sun, P. D. (2008). STRUCTURAL IMPLICATIONS OF SIGLEC-5 MEDIATED
SIALO-GLYCAN RECOGNITION. Journal of Molecular Biology, 375(2), 437–447.
https://doi.org/10.1016/j.jmb.2007.10.009

Abstract (if available)

Abstract Membrane proteins play crucial roles in the immune system as signaling receptors. Structural studies can reveal the specific signaling mechanisms these proteins use. In the innate immune system, the CD33 receptor modulates microglial activity and shows promise as an Alzheimer’s disease drug target. Understanding CD33’s signaling mechanism is essential for drug development, but it remains unknown. This thesis completes the structure of human CD33 by refining the previously obtained crystal structure of its extracellular IgV-IgC domains and integrating NMR-based structures of its transmembrane and cytosolic domains. It reveals that CD33 does not signal via a monomer-dimer shift like other receptors. Instead, binding of dimerized extracellular domains by multivalent ligands restricts movement of the monomeric transmembrane and unstructured cytosolic domains, potentially leading to co-localization with an activating kinase.

By contrast, the Ulmer lab discovered that the integrin αIIbβ3 signals via a monomer-dimer shift. Integrins, involved in cell migration and adhesion, form inactive dimers at their transmembrane domains, with destabilization triggering activation. Functional and sequence diversity suggests varied regulation of α- and β-subunit dimerization and activation. This thesis suggests that in contrast to αIIbβ3, the transmembrane domain of αX is destabilized in comparison to αIIb due to an xPxxxGSS motif at its N-terminal border that disrupt α-helix backbone dynamics., disrupting α-helix backbone dynamics and interacting with lipid headgroups, indicating sensitivity to its lipid environment.

This thesis demonstrates how structural features influence distinct signaling mechanisms in membrane proteins active in the immune system.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Role of CD33 structure and function in Alzheimer’s disease

PDF

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

PDF

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

PDF

Dual effects of transmembrane proline residues on integrin function

PDF

Integrin β3TM structural investigations using phospholipid nanodiscs

PDF

Structural studies on functional amyloids and the mechanism of aggregation and disaggregation of Huntingtin Exon 1

PDF

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

PDF

Protein phosphatase 2A and annexin A5: modulators of cellular functions

PDF

Phosphorylation kinetics of T-cell receptor signaling domain in vitro

PDF

Modulation of transcription and receptor function with synthetic small molecules and multi-finctional integrin ligands

PDF

The interaction of SARS-CoV-2 with CD1d/NKT antigen presentation pathway

PDF

Genetic and molecular analysis of a rhythmic behavior in C. elegans: how neuropeptide signaling conveys temporal information

PDF

Human papillomavirus type 16 entry via the annexin A2 heterotetramer leads to infection and immune evasion

PDF

The modulation of dynamin and receptor endocytosis machinery using elastin-like polypeptides

Asset Metadata

Creator Vu, Han (author)

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

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Medical Biophysics

Degree Conferral Date 2024-08

Publication Date 07/18/2024

Defense Date 05/23/2024

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

Tag Alzheimer's disease,CD33,integrin,membrane protein,microglia,NMR,OAI-PMH Harvest,structural biology

Format theses (aat)

Language English

Contributor Electronically uploaded by the author (provenance)

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

Creator Email han.vu.hlc@gmail.com,hannvu@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC113998FYP

Unique identifier UC113998FYP

Identifier etd-VuHan-13267.pdf (filename)

Legacy Identifier etd-VuHan-13267

Document Type Dissertation

Format theses (aat)

Rights Vu, Han

Internet Media Type application/pdf

Type texts

Source 20240730-usctheses-batch-1186 (batch), 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 author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright.

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

Repository Email cisadmin@lib.usc.edu