University of Southern California Dissertations and Theses

Phospho-proteomic analysis of immune cell activation

(USC Thesis Other)

Phospho-proteomic analysis of immune cell activation

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Copyright 2022 Melanie A. MacMullan

PHOSPHO-PROTEOMIC ANALYSIS OF IMMUNE CELL ACTIVATION

by
Melanie A. MacMullan

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
(CHEMICAL ENGINEERING)

May 2022

ii
DEDICATION

This thesis is dedicated to the love of my life, Mason Montgomery, our dog, Ruby
MacMullan-Montgomery, and my family, Samuel, Donna, Sarah, and Ryan MacMullan, all of
whose constant support and encouragement have allowed me to dedicate my life to the study of
cancer biology.

iii
ACKNOWLEDGMENTS

I would like to thank Professor Graham and Professor Wang for providing the guidance
and resources to allow me to accomplish this work. I would also like to thank my lab mates from
the Graham lab, Dr. James Joly, Dr. Alireza Delfarah, Dr. Dongqing Zheng, Dr. Nicolas Hartel,
and Belinda Garana, as well as my lab mates from the Wang lab, Dr. John Mac, Dr. Xianhui
Chen, Dr. Yun Qu, and Zachary Dunn.

iv
TABLE OF CONTENTS
DEDICATION .............................................................................................................................. ii
ACKNOWLEDGMENTS ........................................................................................................... iii
LIST OF TABLES ...................................................................................................................... vii
LIST OF FIGURES ................................................................................................................... viii
ABSTRACT .................................................................................................................................. ix
INTRODUCTION ........................................................................................................................ 1
1. IMMUNE SYSTEM ...................................................................................................................... 1
2. IMMUNE CELL ENGINEERING .................................................................................................... 2
3. DEFICIENCIES IN ENGINEERED IMMUNE CELLS ......................................................................... 3
4. NATURAL KILLER CELLS AS AN ENGINEERED IMMUNE CELL ALTERNATIVE .............................. 4
CHAPTER 1. LC MS-BASED PHOSPHO-PROTEOMICS ................................................... 6
1.1. INTRODUCTION...................................................................................................................... 6
1.2. MATERIALS AND METHODS ................................................................................................... 7
1.3. CURRENT PHOSPHO-PROTEOMIC ANALYSES OF IMMUNE CELL SIGNALING .......................... 7
CHAPTER 2. PHOSPHO-PROTEOMICS OF NK-92 CELL STIMULATION ................... 9
2.1 NATURAL KILLER CELLS AND METHODS OF ACTIVATION AND PROLIFERATION ...................... 9
2.2. ESTABLISHED SIMILARITIES AND DIFFERENCES IN NK CELL STIMULATION
BY IL-2 OR IL-15 ....................................................................................................................... 10
2.3. IDENTIFICATION OF A NOVEL DEPENDENCE ON RSK SIGNALING FOR
NK-92 CELLS ............................................................................................................................ 12
2.3.1. Materials and Methods ............................................................................................... 12
2.3.2. IL-2 and IL-15 stimulate proliferation but not IFN-γ production
in NK-92 cells ...................................................................................................................... 19
v
2.3.3. IL-2 and IL-15 activate canonical signaling pathways
including JAK-STAT ............................................................................................................ 19
2.3.4. IL-2 and IL-15 induce highly similar signaling in NK-92 cells .................................. 21
2.3.5. Enrichment analyses identify kinases activated by IL-2 and IL-15 ............................ 23
2.3.6. Fuzzy C-means clustering identifies differing phospho-signaling kinetic profiles ..... 24
2.3.7. p90RSK signaling is required for IL-2- and IL-15-mediated proliferation ................ 25
2.4. DISCUSSION ........................................................................................................................ 26
CHAPTER 3. PHOSPHO-PROTEOMICS OF CAR-T SIGNALING.................................. 30
3.1. COMPLICATIONS WITH PHOSPHO-PROTEOMIC ANALYSIS OF CELL CO-CULTURES............... 30
3.2. CAR-T CELL ISOLATION FROM CELL-CELL CO-CULTURE................................................... 31
3.2.1. Materials and Methods ............................................................................................... 31
3.2.2. Anti-CD19 CAR-T cells kill CD19-expressing SKOV3 cells ...................................... 40
3.2.3. Formalin fixation does not impact phospho-peptide identification
or quantitation by LC-MS ..................................................................................................... 40
3.2.4. Magnet-activated cell sorting (MACS) enables purification of
co-cultured CAR-T cells to >90%......................................................................................... 41
3.3. DIRECT ANALYSIS OF ANTI-CD19 CAR-T CELLS STIMULATED BY
CD19-PRESENTING CANCER CELLS ........................................................................................... 42
3.3.1. Materials and Methods ............................................................................................... 42
3.3.2. Analysis of CAR-T cell signaling following stimulation by
antigen-presenting cancer cells ............................................................................................ 42
3.4. DISCUSSION ........................................................................................................................ 46
CHAPTER 4: FUTURE DIRECTIONS ................................................................................... 52
vi
4.1. EXPANDED DYNAMIC SIGNALING ANALYSES ..................................................................... 52
4.2. GENOMIC VALIDATION OF PHOSPHO-PROTEOMIC FINDINGS ............................................... 52
4.3. CHIMERIC ANTIGEN RECEPTOR SIGNALING EVENTS ........................................................... 53
4.4. CAR-NK CELLS ................................................................................................................. 53
REFERENCES............................................................................................................................ 55
APPENDICES ............................................................................................................................. 75
FIGURES AND FIGURE LEGENDS ................................................................................................. 75
SUPPLEMENTAL FIGURES AND FIGURE LEGENDS ....................................................................... 96

vii
List of Tables
All processed results tables as well as RAW LC-MS files are available on the MassIVE database
under the identifiers MSV000088281 and MSV000088632.

viii
List of Figures
Figure 1. Experimental design for phospho-proteomic analysis of IL-2 and IL-15
signaling in NK cells. .................................................................................................................... 75
Figure 2. IL-2 and IL-15 activate canonical signaling pathways including JAK-STAT. ............. 77
Figure 3. IL-2 and IL-15 induce quantitatively similar signaling in NK-92 cells. ....................... 80
Figure 4. Post-translational modification signature enrichment analysis (PTM-SEA)
reveals significant enrichment of p90RSK family kinases involved in propagating
NK-92 cell signaling. .................................................................................................................... 81
Figure 5. Fuzzy C-means clustering reveals distinct kinetic profiles of signaling
downstream of IL-2 and IL-15. ..................................................................................................... 83
Figure 6. p90RSK signaling is required for IL-2- and IL-15-mediated proliferation
of NK-92 cells. .............................................................................................................................. 85
Figure 7. Workflow for phospho-proteomic analysis of CAR signaling in CAR-T
cells stimulated with antigen-expressing cancer cells. ................................................................. 86
Figure 8. Anti-CD19 CAR-T cells kill CD19-expressing SKOV3 cells. .................................... 88
Figure 9. Formalin-fixation does not affect phospho-proteomics in T cells................................. 89
Figure 11. Phospho-proteomic analysis of the known CAR signaling network. .......................... 93
Figure 12. Analysis of CAR signaling following stimulation with antigen-expressing
cancer cells. ................................................................................................................................... 95
Supplementary Figure 1. Analysis of phospho-proteomic replicates identified outliers
that were excluded. ....................................................................................................................... 97
Supplemental Figure 2. IL-2 and IL-15 induce proliferation but not IFN 𝛾 production in
NK-92 cells. .................................................................................................................................. 98
Supplemental Figure 3. Reproducibility analysis of individual samples. ................................... 100
Supplemental Figure 4. Western blotting confirmed upregulation of STAT5A Y694
/ STAT5B Y699 phosphorylation by IL-2/IL-15........................................................................ 101
Supplemental Figure 5. Histograms of IL-2 and IL-15 phospho-proteomic AUC values. ......... 102
Supplemental Figure 6. Assessing the purity of CAR-T cell isolation by MACS following
co-culture with GFP-labeled cancer cells. .................................................................................. 103
Supplemental Figure 7. Histogram of log2 fold change in phosphorylation in CAR-T cells
stimulated with SKOV3.CD19 or SKOV3.NT cells. ................................................................. 104
ix
Abstract
The process of developing improved immunotherapies through the incorporation of
antigen-targeting chimeric antigen receptors (CARs) has been particularly successful in the
treatment of non-solid malignancies, such as leukemia and lymphoma (Melenhorst et al., 2022;
X. Zhao et al., 2020). These CARs incorporate an antigen-targeting small chain variable
fragment (scFv) derived from an antibody with T-cell signaling elements to enable a biomarker-
specific immune cell (Guedan et al., 2019). As such, CARs have been most successfully
implemented in T cells, with multiple CAR-T cells receiving FDA approval since 2017 (Mullard,
2021). Despite these successes, CAR-T cells remain poorly effective against solid tumors
(Stoiber et al., 2019; R. Zhao et al., 2019). Further, mechanisms of CAR activation have not been
fully characterized, complicating efforts to improve CAR designs for better targeting and
activation mechanisms. CAR-NK cells, which implement CARs designed for T cells into an
alternative immune cell platform, the natural killer (NK) cell, have seen some improvements
over CAR-T cell therapies (Albinger et al., 2021; Xie et al., 2020), but are also limited by a lack
of CAR optimization based on NK signaling events. While some studies have characterized CAR
signaling through stimulation of CAR-T cells by antigen-bound beads or through stable isotope
labeling of amino acids in culture (SILAC), these methods are either limited in the establishment
of a realistic cell-cell interaction or in terms of cost and time (Griffith et al., 2022a; Philipson et
al., 2020; Salter et al., 2019, 2021). As such, we sought to use phospho-proteomics to accomplish
two main objectives: 1) establish a better understanding of NK canonical signaling events to aid
in the development of CARs for this platform, and 2) develop an inexpensive and
straightforward method for analyzing CAR signaling from cell-cell interactions.

1
Introduction
1. Immune system
The immune system is comprised of a collection of cells whose main objective is to
protect the body from infection and “non-self” invaders (Kellie & Al-Mansour, 2017). The cell
types involved are broadly known as white blood cells or leukocytes and have a variety of
functions and roles in maintaining immunity (Tigner et al., 2020). The immune response can
generally be split into two categories: innate and adaptive immunity. The innate immune system
acts as a first responder and is responsible for the general elimination of pathogens, typically
either through the release of antimicrobial molecules which penetrate and destroy the infectious
agent or through pathogen ingestion (Kellie & Al-Mansour, 2017). Neutrophils, macrophages,
dendritic cells, natural killer cells, mast cells and basophils comprise the main cell types of the
innate immune system (Tigner et al., 2020). The adaptive immune system, made up primarily of
T cells and B cells, functions far more specifically, relying on contact with diseased or antigen-
presenting cells (APCs) to prompt activation and proliferation (Kellie & Al-Mansour, 2017).

In response to prolonged disease or infection, T cells play a particularly important and
interesting role. Mature T cells develop mainly into two different functional classes: those with
CD8
+
T-cell receptors (TCR) or those with CD4
+
TCRs (Janeway Jr et al., 2001). These markers
indicate whether the cell will differentiate into a cytotoxic (killer) T cell in the case of CD8
+

TCRs or induce an influx of immune cell activity as effector (helper) T cells in the case of CD4
+

TCRs. Helper T cells can be further differentiated into those which activate killing properties of
other immune cell types (TH1 cells) and those which activate the humoral immune response of
antibody development in B cells (TH2 cells) (Janeway Jr et al., 2001). All of this T cell activity is
2
stimulated through the interaction of the TCR with major histocompatibility complexes (MHC)
on the surfaces of infected cells or APCs. MHC class I molecules tend to be viral protein
subunits (peptides) which stimulate cytotoxic T cells to destroy and kill the APC. MHC class II
molecules are typically presented on APCs which have ingested infectious bacteria or fungi and
prompt the T cells to activate the cells they are interacting with through the release of
biomolecules called cytokines (Kellie & Al-Mansour, 2017). The specificity of the TCR to the
molecules which are being presented via the MHCs and the ability of T cells to both kill infected
cells and induce killing in other cell types make them an ideal target for cellular engineering
(Hedrick et al., 1984; D. Li et al., 2019; Yanagi et al., 1984).

2. Immune cell engineering
Immune cell engineering aims to improve immune cell function by amplifying diseased
cell specificity and killing without compromising safety. This broad field stemmed from the
concept of adoptive cell therapies (ACTs), in which tumor-specific T cells are expanded ex vivo
and injected into the tumor site of a cancerous patient (Perica et al., 2015). Since then, a variety
of therapeutics have been developed based on the concept of extracting immune cells from an
individual, engineering or modifying them, and injecting them back into that same individual
(Lim & June, 2017). Following a series of modifications and clinical trials, a specific type of
engineered T cell with a chimeric antigen receptor (CAR) was developed (Roybal & Lim, 2017).
Unlike normal T cells, the CAR-T cell recognizes biomarkers on a cancerous cell surface
without the requirement of presentation through MHCs, enabling a far greater affinity for cancer
cell recognition. This, in combination with the designed cell killing, allows these cells to
effectively and swiftly identify and kill malignant cells. The word chimeric is defined as “having
3
parts of different origins” and was originally used to describe a mythical animal formed from
parts of various animals. The CAR itself is designed through the incorporation of one or more
specific targeting regions called small-chain variable fragments (scFVs), derived from
antibodies, with transmembrane and intracellular signaling domains derived from T cells
(Guedan et al., 2019; D. Li et al., 2019). CAR-T cells have seen great clinical success, with the
first CAR-T cell FDA approval in 2017 and the most recent in February 2022, all for treatment
of hematological malignancies (Janssen Research & Development, LLC, 2022; Mullard, 2021).
While the FDA approval of CAR-T cells represents a breakthrough for cancer
immunotherapeutics, there are still a number of drawbacks and unknowns which prevent them
from having more widespread efficacy.

3. Deficiencies in engineered immune cells
Although CAR-T cell immunotherapeutics have seen promising clinical success, their
development has not been without limitations. These limitations include the lack of efficacy
against solid tumors, the co-development of cytokine release syndrome and neurotoxicity or
graft-versus-host disease (GVHD) with treatment, the loss of efficacy due to antigen loss or
modulation, and the evidence of leukemia relapse (Shah & Fry, 2019; Srivastava & Riddell,
2018; Stoiber et al., 2019; R. Zhao et al., 2019). Moreover, the development of these therapeutics
is labor-intensive and costly, with the average cost (exclusive of patient care cost) ranging from
$373,000 to $475,000 per infusion (Kansagra et al., 2020). Patients must receive a referral to be
considered for this type of treatment, and it has been found that only 48% of adults in the United
States live within 30 minutes of a hematopoietic cell transplantation center (Delamater & Uberti,
4
2016). Given this, there remains a well-defined need for an off-the-shelf immunotherapeutic with
minimal toxicity.

4. Natural killer cells as an engineered immune cell alternative
Natural killer (NK) cells are members of the innate immune system but have an inherent
killing activity similar to that of cytotoxic T cells which can enhance their efficacy as an
immunotherapy (Wu et al., 2020). Instead of relying on interactions with the MHC to prompt
activity as in T cells, NK cells operate based on feedback from a balance of stimulatory and
inhibitory receptors (Shimasaki et al., 2020). Also unlike T cells, NK cells actually suppress the
development of GVHD while maintaining graft-versus-tumor effect, suggesting that NK cells are
a good selection for off-the-shelf therapeutic development (Olson et al., 2010). Because of their
capacity to become activated based on a signaling imbalance from stimulatory and inhibitory
receptors, NK cells are often functional in a tumor microenvironment even without CAR-
engineering (Shimasaki et al., 2020). While the implementation of CARs in NK cells has been
successful in vitro and a number of CAR-NK cells are being evaluated in clinical trials, nearly all
of the CARs have been designed for T cells and not yet optimized for use in NK cells (Albinger
et al., 2021; Xie et al., 2020).

The survival and development of NK cells is known to be supported by co-stimulation
with cytokines, particularly IL-2 and IL-15, but the intracellular signaling events supporting
activation of NK cells following stimulation by these molecules has not been well characterized
(Wu et al., 2020). A better understanding of the canonical signaling patterns for proliferating
cells of this type could enable improved development of an optimal CAR for NK cells.
5
Additionally, because of the much larger number of stimulatory and inhibitory receptors on the
NK cell surface, a comparison of CAR signaling in NK cells to CAR signaling in T cells would
illuminate how well the CARs are functioning in NK cells and where there is room for
improvement.

6
Chapter 1. LC MS-Based Phospho-proteomics
1.1. Introduction
Intracellular communication is typically carried out through signal transduction. Protein
phosphorylation is an important signal transduction mechanism, as the process of
phosphorylation or dephosphorylation can result in the activation or inactivation of a number of
enzymes and receptors (Ardito et al., 2017). Protein phosphorylation is a reversible process
generally performed by protein kinases through the transfer of a phosphate group from ATP to a
protein, causing a conformational change that typically results in activation. Phosphate transfer is
typically carried out on serine (S), threonine (T), or tyrosine (Y) residues through an
esterification reaction between the hydroxyl group on these amino acid residues and the
phosphate group (Grisham & Garrett, 2013). Currently, there are 518 known protein kinases in
humans that control phosphorylation, and over 200,000 known phospho-sites (Ardito et al.,
2017; Day et al., 2016).

There are multiple ways to detect and measure this post-translational modification
(PTM), many of them involving the use of antibodies specific to phosphorylated protein forms.
These include Western blotting, flow cytometry, and antibody and protein arrays. Other methods
include fluorescence microscopy and mass spectrometry. Each of these has its own set of
advantages and limitations, but mass spectrometry in particular is becoming a more popular
approach for unbiased quantification of a large number of phospho-sites from whole cell and
tissue lysates (Day et al., 2016). Mass spectrometers are extremely sensitive analytical
instruments which can detect mass shifts that correspond to a phosphate group (80 Daltons),
making them ideal for detecting post-translational modifications. Coupled to a high-performance
7
liquid chromatographer (HPLC), tandem mass spectrometry (MS) enables high-throughput
peptide and phospho-peptide quantification.

1.2. Materials and methods
To characterize phosphorylation events in cellular material, the cells are first lysed and
the intracellular proteins are broken down into peptides via digestion with trypsin (Pierce
Trypsin Protease, 90058). Because of the negatively charged nature of a phosphate group,
phosphorylated peptides can be enriched from a sample using a positively charged medium. In
our workflows, we employ titanium dioxide (TiO2) labeled microbeads (GL sciences,
#1400B500) to tightly bind phosphorylated peptides. The peptide-bound beads are then washed
to ensure purity and eventually the peptides are eluted off the beads for analysis. Peptides are
passed through a Thermo EASY nanoLC coupled to a Thermo Q-Exactive Plus Mass
Spectrometer and the resulting spectra are analyzed using Proteome Discoverer (v2.2) software.

1.3. Current Phospho-proteomic Analyses of Immune Cell Signaling
As the field of engineered immune cells continues to expand, so does our understanding
of the intracellular signaling pathways that are involved in carrying out the signals prompted by
stimulation of chimeric antigen receptors engineered onto cell surfaces. In recent years, phospho-
proteomic analyses characterizing T cell receptor (TCR) and CAR-T cell signaling have been
reported (Álvarez-Salamero et al., 2017; Griffith et al., 2022a; Hwang et al., 2020; Navarro et al.,
2014; Philipson et al., 2020; Ramello et al., 2019; Salter et al., 2019, 2021; Tan et al., 2017). To
date, the only study characterizing immune signaling of NK cells has been an analysis of the
kinome which did not involve phospho-proteomics and was therefore not as high-throughput or
8
unbiased (König et al., 2012). Further, many studies characterizing CAR signaling in T cells
have been limited to stimulation by soluble ligands or antigen-bound microbeads, whereas we
propose to characterize this signaling in the context of cell-cell contact (Philipson et al., 2020;
Ramello et al., 2019; Salter et al., 2018, 2021). Phospho-proteomic analyses of NK cell canonical
signaling or CAR function in NK cells have not been performed, and by executing experiments
to characterize the endogenous and engineered signaling pathways involved in NK cell function
we hope to illuminate mechanisms that will aid in improved immunotherapeutic development.

9
Chapter 2. Phospho-proteomics of NK-92 Cell Stimulation
2.1 Natural killer cells and methods of activation and proliferation
Natural killer (NK) cells are members of the innate immune system and are known to
circulate the body in search of diseased cells or pathogens. The cell killing activity of NK cells
resembles that of CD8
+
cytotoxic T cells, but with the advantage that NK cells do not need to be
primed by a pathogenic antigen prior to killing initiation (Abel et al., 2018). NK cells rely on a
balance between activating and inhibitory receptors to initiate activity and are responsible for the
killing of any cell recognized as “non-self” (Sivori et al., 2019). Once a molecule binds to an
inhibitory or activating receptor on the NK cell surface, a signal propagates through the transfer
of phosphate groups to a number of proteins which indicates to the cell whether or not to activate
killing (Paul & Lal, 2017). An overwhelming activating signal is needed to prompt cell killing
and a number of subpopulations have been characterized which vary in killing mechanism based
on their relative expression of the CD56 ligand and the CD16 receptor (Lutz et al., 2011; Van
Acker et al., 2017). Four main steps are taken when a natural killer cell receives an
overwhelming activation signal in response to an invasive or diseased cell beginning with 1) the
formation of an immunological synapse mediated by receptor-ligand interactions between the
NK and target cell followed by 2) migration of the microtubule organizing center (MTOC) and
secretory lysosome towards the lytic synapse, enabling 3) the secretory lysosome to dock with
the NK cell plasma membrane and finally allowing 4) the secretory lysosome to fuse with the
target cell plasma membrane, prompting the release of cytotoxic molecules that result in target
cell destruction (Paul & Lal, 2017).

10
NK cell activity can be attributed to both the production of and response to cytokines,
biomolecules which act as inflammatory mediators for several functions including initiating and
sustaining their growth and recruiting and activating other immune cell types. Cytokine receptors
are expressed early on the NK cell surfaces and are specific to a number of interleukins (IL)
including IL-2, IL-3, IL-7, IL-9, IL-15, and IL-21 and well as other factors such as
Interferon(IFN)-gamma( 𝛾 ), tumor necrosis factor(TNF)-alpha( 𝛼 ) (Abel et al., 2018).Though
many of these cytokines share similar functional roles, they may be secreted by a variety of cell
types and their impact on the immune response is highly dependent on the cell surface
expression of receptors specific to them. While many of the functions of these cytokines have
been characterized under specific cellular conditions, few studies have studied isolated cellular
response to select cytokines.

2.2. Established similarities and differences in NK cell stimulation by IL-2 or IL-15
Despite low sequence homology, both IL-2 and IL-15 signal through heterotrimeric
receptor complexes that include the IL-2/15Rβ subunit (CD122) and the common γ chain (γc)
(CD132) (Meghnem et al., 2017; Rodella et al., 2001). The third member of the receptor
complex differs between the two cytokines, with IL-2 binding to complexes containing IL-2Rα
(CD25), and IL-15 binding to complexes containing IL-15Rα (CD215) (Hamid et al., 2006;
Tamzalit et al., 2014; Y. Yang & Lundqvist, 2020). The shared receptor has an intermediate
binding affinity to both IL-2 and IL-15 of (Kd ~ 10
-9
M for both IL-15 and IL-2), while the
isolated IL-2R 𝛼 binds IL-2 with a low affinity (Kd ~ 10
-8
M) and the isolated IL-15R 𝛼 binds IL-
15 with a relatively high affinity (Kd ~ 5x10
-11
M) (Hamid et al., 2006; Tamzalit et al., 2014; Y.
Yang & Lundqvist, 2020). The shared receptor is functionally bound to JAK1 and JAK3,
11
prompting downstream signaling events including tyrosine phosphorylation of STAT1, STAT3,
and STAT5 and activation of the RAS-MAP and PI-3 kinase pathways (Mitra et al., 2015).

Despite similarities in their shared receptor subunits and downstream signaling pathways,
IL-2 and IL-15 can elicit different functions in immune cells. For example, mice deficient in IL-2
or IL-15 have distinct phenotypes, with IL-2 deficient mice suffering from overaccumulation of
activated T and B cells while IL-15 deficient mice experience a lack of NK, NKT, and activated
cytotoxic T cells (Ma et al., 2000). Further receptor knockout experiments in mice have
demonstrated that IL-15/IL-15R signaling is essential for NK cell proliferation and development
while IL-2/IL-2R is not (Dunne et al., 2001). Thus, there is great interest in understanding IL-2
and IL-15 signaling in immune cells.

Comparisons of IL-2 and IL-15 signaling have generated conflicting results. Although
some studies have found that the cytokines produce indistinguishable signaling profiles
(Zambricki et al., 2005), others have demonstrated substantial differences (Waldmann, 2015).
Despite the functional relevance for NK cell function, studies of IL-2 and IL-15 signaling in NK
cells have been limited to approaches such as immunoblotting (Frank et al., 1995; Gilmour et al.,
2001) and phospho-flow cytometry (Pande et al., 2012) that can profile only a small number of
known signaling events for which validated antibodies exist. In contrast, liquid chromatography-
mass spectrometry (LC-MS)-based phospho-proteomics offers comprehensive, unbiased, and
quantitative characterization of protein phosphorylation without the need for validated antibodies
(MacMullan et al., 2019). As such, we sought to use LC-MS phospho-proteomics to compare the
12
early signaling dynamics induced by these cytokines in the immortalized NK cell line NK-92 and
then connect these signaling events to functional outcomes including proliferation (Fig. 1).

2.3. Identification of a Novel Dependence on RSK Signaling for NK-92 Cells

2.3.1. Materials and Methods
Cell Culture
The immortalized NK-92 cell line was a gift from Dr. Jihane Khalife of the Children’s
Hospital of Los Angeles. NK-92 cells were expanded in MEM Alpha (1x) + GlutaMAX media
(Gibco 32561037) containing 8% fetal bovine serum (Sigma F2442), 8% horse serum (Gibco
16050130), 1% non-essential amino acids (Lonza BE13-114E), 0.2 mM myoinositol (Sigma
I7508), 0.02 mM folic acid (Sigma F8758), 0.2 mM sodium pyruvate (Corning 25-000-Cl), 0.1
mM β-mercaptoethanol (Gibco 21985023), and 0.35% Pen Strep (Gibco 15140-122). Media was
filtered using a 0.2 μm bottle filter and warmed to 37°C before being added to cells. Interleukin-
2 (IL-2) (PeproTech 200-02) was added at a concentration of 100 U/mL during each media
refresh. Cells were initially recovered at a concentration of 0.5 million cells/mL but maintained
at a concentration of 0.25 million cells/mL for all subsequent passages. Cells were counted by
trypan blue exclusion with a hemocytometer. Cells were grown in a 5% CO2, 37°C and
humidified incubator and were used within 20 passages of thawing. For the experimental
conditions, cells were starved for 6 hours in fresh media lacking cytokines and subsequently
treated with 100 U/mL of either IL-2 or IL-15 (PeproTech 200-15). Each experimental condition
was carried out with four replicates (n=36 total), although 5 outlier samples were removed from
analysis (See Section on Data Analysis and Statistics).
13

Flow Cytometry Analysis
Following cell starvation, 0.5 million cells were treated with 100 U/mL IL-2 or IL-15 and
5 µg/mL Brefeldin A and allowed to incubate for 6 hours. Cells were then washed 1 time with
100 µL PBS and fixed and permeabilized in 100 µL Fixation and Permeabilizaton Solution from
the BD Cytofix/Cytoperm Fixation/Permeabilization Kit (BD BioSciences 554714) for 10
minutes at 4°C. Cells were washed 2 times using 100 µL 1X Perm/Wash buffer provided in the
Cytofix/Cytoperm Fixation/Permeabilization Kit and then labeled with a R-Phycoerythrin (PE)
labeled anti-human interferon (IFN)-γ antibody at a ratio of 1:100 in 100 µL 1X Perm/Wash
buffer and incubated for 30 min at 4°C. Cells were then washed 2 times using 100 µL 1X
Perm/Wash buffer and resuspended in 100 µL PBS for flow cytometry analysis using a Miltenyi
Biotec MACSQuant benchtop flow cytometer.

LC-MS Phospho-proteomics
Cell pellets were lysed in a modified urea-based lysis buffer (8 M urea, 100 mM Tris-HCl
pH 7, 1 mM activated sodium vanadate, 2.5 mM sodium pyrophosphate, 1 mM β-
glycerophosphate) by sonication in a cup-horn sonicator reservoir at 4°C. Protein concentration
was measured using a BCA assay (Pierce P123227) and 1 mg of each sample was used for
phospho-proteomics. Lysates were reduced with 5 mM DTT for 30 minutes at room temperature
(RT), alkylated with 25 mM iodacetamide for 30 minutes at RT in the dark, and quenched with
10 mM DTT for 10 minutes at RT. Proteins were then digested to peptides with trypsin (Thermo
Scientific 90057) using a 1:100 trypsin-to-lysate ratio by mass based on the manufacturer’s
recommendation. Tryptic peptides were then acidified with 5% trifluoroacetic acid (TFA),
14
desalted using HyperSep C18 Cartridges (Thermo Scientific 03-251-157), and eluted with 30%
acetonitrile. The eluents were vacuum dried and then resuspended in loading buffer (80%
acetonitrile (ACN), 6% TFA) to a concentration of 1 mg/mL for phospho-enrichment.

Phospho-peptide enrichment was adapted from the protocols developed by Matheron et al
and Zhou et al (Matheron et al., 2014; Zhou et al., 2013). Titanium dioxide beads (PolyLC
TT200C18) were weighed out at a ratio of 10 mg beads to 1 mg tryptic peptide and then
resuspended in 30 µL per sample of suspension buffer (30% ACN, 0.1% TFA) to prepare a bead
slurry. For each sample, 30 µL of bead slurry was aliquoted into a new tube. Reconstituted
tryptic peptides were added to the bead aliquots and samples were rotated at RT for 30 min. A
C8 StageTip (Sigma Aldrich 66882-U) was prepared and the bead-conjugated peptides were
added. Non phospho-peptides were washed away by three 150 µL additions of 50% ACN, 0.1%
TFA, and bead-bound phospho-peptides were eluted in two steps, first by 20 µL 10% ammonium
hydroxide and then by 30 µL 80% ACN, 2% formic acid (FA). Elutions were collected in a tube
containing 25 µL of a 10% FA neutralization buffer.

The eluents were vacuum dried and then reconstituted in 100 µL of 0.1% TFA for
desalting using C18 StageTips (Sigma Aldrich 66883-U) according to an adaptation of the
protocol developed by Rappsilber et al. (Rappsilber et al., 2007). Buffers and phospho-peptide
containing solutions were passed through StageTips by centrifugation at 2,000-3,000x g in a
bench top centrifuge. After conditioning the StageTips with two passes of 150 µL of 50% ACN
and by two passes of 150 µL of 0.1% TFA, the phospho-peptide solutions were added to the
StageTips. StageTip-bound phospho-peptides were washed three times with 150 µL of 0.1%
15
TFA and eluted with 100 µL of 60% ACN / 0.1% TFA. Eluents were vacuum dried and dried
phospho-peptides were resuspended in 6 µL of 0.1% FA.

The samples were randomized and 5 µL of each sample was injected onto an EASY-nLC
1200 ultra-high-performance liquid chromatography coupled to a Q Exactive Plus quadrupole-
Orbitrap mass spectrometer (Thermo Fisher Scientific). Peptides were separated by a reverse
phase analytical column (PepMap RSLC C18, 2 µm, 100 Å, 75 µm×25 cm). Flow rate was set to
300 nL/min at a gradient from 3% LC buffer B (0.1% formic acid, 80% acetonitrile) to 38% LC
buffer B in 110 min, followed by a 10-min washing step to 85% LC buffer B. The maximum
pressure was set to 1,180 bar, and column temperature was maintained at 50°C. Peptides
separated by the column were ionized at 2.4 kV in positive ion mode. MS1 survey scans were
acquired at the resolution of 70,000 from 350 to 1,800 m/z, with a maximum injection time of
100 ms and AGC target of 1e6. MS/MS fragmentation of the 14 most abundant ions were
analyzed at a resolution of 17,500, AGC target 5e4, maximum injection time 65 ms, and
normalized collision energy of 26. Dynamic exclusion was set to 30 sec, and ions with charge
+1, +7 and >+7 were excluded.

MS/MS fragmentation spectra were searched with Proteome Discoverer SEQUEST
(version 2.2, Thermo Scientific) against in silico tryptic digested Uniprot all-reviewed Homo
sapiens database (release June 2017, 42,140 entries) plus all recombinant protein sequences used
in this study. The maximum missed cleavages was set to two. Dynamic modifications were set to
phosphorylation on serine, threonine or tyrosine (+79.966 Da), oxidation on methionine
(+15.995 Da), and acetylation on protein N-terminus (+42.011 Da). Carbamidomethylation on
16
cysteine (+57.021 Da) was set as a fixed modification. The maximum parental mass error was set
to 10 ppm, and the MS/MS mass tolerance was set to 0.02 Da. The false discovery threshold was
set strictly to 0.01 using the Percolator Node validated by q-value. The relative abundance of
parental peptides was calculated by integration of the area under the curve of the MS1 peaks
using the Minora LFQ node. Individual phospho-site localization probabilities were determined
by the ptmRS node, and phospho-sites with <0.75 localization probability were removed. A
comparison of removal of phospho-sites with <0.25 localization probability was also performed
to evaluate loss of information by selecting only type 1 phospho-peptides. RAW LC-MS files
and processed result files are available at MassIVE database under identifier MSV000088281
(reviewer login: MSV000088281_reviewer; reviewer password: a).

Data analysis and Statistics
Proteome Discoverer peptide group abundance values of each sample were normalized to
the sample’s median value. No data imputation was performed for missing values. After
normalization, phospho-peptides were filtered in Perseus (Version 1.6.2.3) (Tyanova et al., 2016)
so that each condition had at least two quantified values. We then evaluated the relative
distribution of log2-transformed phospho-peptide intensities performed PCA to remove outliers.
Five samples were determined to be outliers based on low intensity signals, and PCA analysis
revealed that these samples did not cluster with their replicates (Supp. Fig 1). All five outliers
were removed from the analysis: replicate 2 for t = 0 min, replicate 4 for IL-15 at t = 5 min,
replicate 2 for IL-15 at t = 10 min, and replicates 2 and 4 for IL-2 at t = 10 min. With the
remaining data, the area under the curve (AUC) of each phospho-peptide time course was
calculated using the composite trapezoidal rule (Equation 1):
17

𝐴 𝑈𝐶 =
5
2
( 𝐼 0
+ 2 ∗ 𝐼 5
+ 2 ∗ 𝐼 10
+ 𝐼 15
) +
15
2
( 𝐼 15
+ 𝐼 30
) (1)

where 𝑰 𝒙 is the log2-transformed, median normalized intensity of the phospho-peptide signal at 𝒙
min. Next, we calculated the standard deviation of the AUC ( 𝒔 𝑨 𝑼 𝑪 ) by propagating the standard
deviations of the individual time points (e.g., 𝒔 𝑰 𝟓 for the standard deviation at t=5 min)
(Equation 2):

𝑠 𝐴 𝑈 𝐶 =
√
(
5
2
)
2
𝑠 𝐼 0
2
+ ( 5 )
2
𝑠 𝐼 5
2
+ ( 5 )
2
𝑠 𝐼 10
2
+ ( 10 )
2
𝑠 𝐼 15
2
+ (
15
2
)
2
𝑠 𝐼 30
2
(2)

Then, the t statistic was calculated (Equation 3):

𝑡 =
𝐴 𝑈 𝐶 𝑠 𝐴𝑈𝐶
√ 𝑛 ⁄
(3)

where n was 5 time points. Using this value, we calculated the p-value from the two-tailed, one-
sample t distribution with 4 degrees of freedom. Lastly, the resulting p-values were adjusted for
multiple hypothesis testing using the Benjamini-Hochberg method.

Kinase set enrichment analysis (KSEA) was performed using in-house Perl scripts and
post-translation modification signature enrichment analysis (PTM-SEA) performed using the
ssGSEA 2.0 R package available on Github (Krug et al., 2019). Fuzzy C-means clustering
18
analysis was performed using the Mfuzz R package from Bioconductor in R version 4.0.2
(Kumar & Futschik, 2007). Additional data processing and analyses were performed in
Microsoft Excel.

Drug Treatment
NK-92 cells were recovered in NK cell media for 24 hours before being counted by
trypan blue exclusion and distributed into separate populations for drug treatment in triplicate.
To inhibit RSK activity, we used the inhibitors LJI308 (MedChemExpress, HY-19713) at a
concentration of 20 µM and BI-D1870 (MedChemExpress, HY-10510) at concentrations of 2
µM and 10 µM based on published literature (Chae et al., 2020; Roffé et al., 2015; W. Wang et
al., 2019). Cell media and drug treatment were refreshed every 2 days, and cells were counted by
trypan blue exclusion after each passage.

Western Blotting
NK-92 cells were lysed in modified radioimmune precipitation assay buffer (50 mM Tris-
HCl (pH 7.5), 150 mM NaCl, 50 mM β-glycerophosphate, 0.5 mM Nonidet P-40, 0.25% sodium
deoxycholate, 10 mM sodium pyrophohsphate, 30 mM sodium fluoride, 2 mM EDTA, 1 mM
activated sodium vanadate, 20 µg/mL aprotinin, 10 µg/mL leupeptin, 1 mM DTT, and 1 mM
phenylmethylsulfonyl fluoride). Whole-cell lysates were resolved by SDS-PAGE on 4-15%
gradient gels and blotted onto nitrocellulose membranes (Bio-Rad). Membranes were blocked for
30 minutes and then incubated with primary antibodies at room temperature (RT) for 1 hour
following by secondary antibodies at RT in the dark for 1 hour. Blots were imaged using the
Odyssey IR Imaging System (LI-COR). Primary antibodies used for Western blot analysis were
19
as follows: pSTAT5 pY694 (4322, Cell Signaling) and anti-β-actin (66009-1-Ig, Proteintech).
IR-dye conjugated secondary antibodies used for Western blot analysis were as follows: 680RD
Goat anti-Rabbit (926-68071, Licor), 800RD Goat anti-Mouse (926-32210, Licor).

2.3.2. IL-2 and IL-15 stimulate proliferation but not IFN-γ production in NK-92 cells
To understand how NK cells respond to the cytokines IL-2 and IL-15, we stimulated the
immortalized NK cell line NK-92 with each cytokine individually. Notably, these cells are
typically grown in media supplemented with IL-2. First, we tested the effects of these cytokines
on proliferation and found that both cytokines induced robust cell growth (Supp. Fig. 2A). There
was no significant difference in the growth rate of NK-92 cells stimulated with either cytokine.
Next, we tested the effects of IL-2 and IL-15 on interferon (IFN)-gamma ( 𝛄 ) production, a
cytokine secreted by NK cells which can shape the immune response (Vivier et al., 2011), and
found that neither cytokine induced IFN-γ production (Supp. Fig. 2B). Taken together, this data
demonstrates that both IL-2 and IL-15 induce proliferation but not IFN-γ production in NK-92
cells.

2.3.3. IL-2 and IL-15 activate canonical signaling pathways including JAK-STAT
Next, we sought to identify the signaling pathways induced by IL-2 and IL-15 using LC-
MS-based phospho-proteomics. First, NK-92 cells were starved of cytokines for 6 h, and then
cells were stimulated with either IL-2 or IL-15 for 0, 5, 10, 15 or 30 min (Fig. 1). Phospho-
peptides were enriched using TiO2 beads and then identified and quantified using label-free LC-
MS. In total, we quantified 8,692 phospho-peptides on 3,023 proteins. Analysis of the correlation
between individual samples demonstrated high reproducibility (Supp. Fig. 3). For subsequent
20
quantitative analysis, we focused on the 3,619 phospho-peptides that were fully quantified across
all 5 time points. To capture the total phospho-signal over time, we calculated the area under the
curve (AUC) of the intensity for each phospho-peptide.

We then compared our phospho-proteomic data to the canonical IL-2 and IL-15 signaling
pathways (Fig. 2A-B). One canonical signal induced by IL signaling is activation of JAK3
(Ikemizu et al., 2012; Morris et al., 2018), which is functionally bound to the 𝛽 - and 𝛾 -chains of
the IL-2/15 receptor (IL2/15R) along with JAK1. Our data revealed that both IL-2 and IL-15
robustly increased phosphorylation of JAK3 on Y785, an autophosphorylation site required for
association with the adapter protein SH2B1 following IL-2 stimulation (Kurzer et al., 2004). We
also quantified JAK3 peptides with phosphorylated residues S17 and S17/S20, although these
phosphorylation sites had smaller AUC magnitudes than Y785 (average AUC values of 22.5 and
18.4 arbitrary units, respectively, compared to 47.8 arbitrary units for Y785) indicating less
induction of phosphorylation over the 30-minute time period following stimulation.

One of the primary downstream targets of JAKs is the signal transducer and activator of
transcription (STAT) family of transcription factors (Seif et al., 2017). Following IL-2/IL-15
stimulation, we observed increased phosphorylation on four phospho-peptides of STAT family
members including STAT1 S727, STAT3 S727, STAT5A Y694 / STAT5B Y699, and STAT5A
S780. Western blotting confirmed the induction of phosphorylation on STAT5A Y694 /
STAT5B Y699 in cells stimulated by IL-2 and IL-15 but not those starved of cytokine (Supp.
Fig. 4). The phosphorylation of STAT5A Y694 / STAT5B Y699 (these tryptic phospho-peptides
are indistinguishable) is crucial for activation and nuclear translocation of these transcription
21
factors (Fahrenkamp et al., 2015; Klejman et al., 2002). Phosphorylation of S727 on both STAT1
and STAT3 has also been shown to be required for promotion of transcription activity under
multiple cellular contexts (Decker & Kovarik, 2000). We also observed phosphorylation of
STAT4A S733, STAT5B S128, and STAT5B S193, although the regulation levels of these
phospho-peptides did not change in response to either IL-2 or IL-15 stimulation (i.e., AUC ≈ 0).

IL stimulation is also known to induce Ras/Raf/MAPK and PI3K/AKT signaling
pathways (Marzec et al., 2008; Mishra et al., 2014; Mitra et al., 2015). Examination of phospho-
signals from the Ras/Raf/MAPK pathway revealed one phosphorylation site on ARAF (S186),
two phosphorylation sites on BRAF (S365 and S729), and three phosphorylation sites on RAF1
(S29, S259, and S642). Among these sites, only ARAF S186 and RAF1 S642 were significantly
activated by IL-2 and IL-15 stimulation (average AUC of 38.5 and 24.9 arbitrary units,
respectively). Downstream of RAF, we measured slightly decreased phosphorylation of the
MAPK kinases MEK1 on T292/S304 and MEK2 on S23 and T394 in response to IL-2/15. In the
PI3K/AKT pathway, we quantified only one phosphorylation site on AKT2 (pT451), and its
phosphorylation was slightly decreased by IL-2 and IL-15 stimulation (average AUC of -17.1
arbitrary units). Taken together, these data support strong activation of the JAK/STAT signaling
pathway and suggest weaker activation of the PI3K/AKT and Ras/RAF/MAPK signaling
pathways in response to IL-2 and IL-15 in NK-92 cells.

2.3.4. IL-2 and IL-15 induce highly similar signaling in NK-92 cells
For all phospho-peptides in the canonical signaling pathways, the kinetic profile of
phosphorylation was qualitatively similar for IL-2 and IL-15-stimulated cells. We therefore
22
tested whether this similarity between IL-2 and IL-15 signaling held true across all measured
phospho-peptides. Plotting the AUC values for all phospho-peptides quantified in both IL-2 and
IL-15-stimulated cells revealed a strong correlation (r = 0.72, p < 0.00001) (Fig. 3A). No
phospho-peptides were significantly distant from the line of equality (i.e., where the AUC was
equal for IL-2 and IL-15). For both cytokines, we observed a roughly equal balance of positive
and negative AUC values (Supp. Fig. 5). Notably, by AUC, the most upregulated
phosphorylation site in both IL-2 and IL-15-stimulated cells was STAT5A Y694 / STAT5B
Y699 (average AUC of 115.3 arbitrary units). Interestingly, the serine/arginine repetitive matrix
protein 2 (SRRM2) had both one of the most upregulated phospho-peptides
(S1099/S1101/S1103, average AUC 97.1 arbitrary units) and one of the most downregulated
phospho-peptides (S440, average AUC -58.5 arbitrary units) (Fig. 3A-B).

We next searched the most up- and downregulated phospho-peptides for sites with known
regulatory function in the PhosphoSite Plus database (P. V. Hornbeck et al., 2015b). Among the
highly upregulated phospho-peptides, this analysis highlighted plastin-2 (LCP1) S5 which
regulates enzymatic activity and intracellular translocation (Koide et al., 2017) and the
serine/threonine-protein kinase 17B (STK17B/DRAK2) S12 which negatively regulates IL-2
signaling and T cell development (Mandarano et al., 2020). Other phospho-peptides with known
regulatory function included vimentin (VIM) pS56 (J. Li et al., 2016; Ratnayake et al., 2021),
the serine/threonine-protein kinase PAK 2 (PAK2) pS141 (T. Li et al., 2011; Zhan et al., 2003),
Wiskott-Aldrich syndrome protein (WAS) pS483 (Cory et al., 2003; C. Liu et al., 2013), 40S
ribosomal protein S6 (RPS6) pS236 & pS240 (Cerezo et al., 2021; Salizzato et al., 2016) and
stathmin (STMN1) pS25 (Alesi et al., 2016; Kuang et al., 2015, 2016). Among the most
23
downregulated phospho-peptides with known regulatory function, we found stromal interaction
molecule 1 (STIM1) pS608 (Casas-Rua et al., 2015; Pozo-Guisado et al., 2013; Tomas-Martin et
al., 2015), mitochondrial fission factor (MFF) pS146 (Ducommun et al., 2015), and
serine/threonine-protein phosphatase PP1γ catalytic subunit (PPP1CC) pT311 (Schmutz et al.,
2011). Taken together, these data demonstrate that IL-2 and IL-15 induce highly similar
signaling in NK-92 cells including many phospho-peptides with known regulatory function.

2.3.5. Enrichment analyses identify kinases activated by IL-2 and IL-15
Phospho-proteomics identifies the downstream phosphorylation sites but not the
upstream kinases that perform the phosphorylation. We therefore used bioinformatic approaches
to infer the kinases activated by IL-2 and IL-15 in NK-92 cells. First, we used post-translational
modification signature enrichment analysis (PTM-SEA) (Krug et al., 2019) which tests for
enrichment of kinase activity and signaling pathway signatures assembled from literature.
Ranking IL-2 and IL-15-stimulated phospho-peptides by AUC, PTM-SEA identified
p90RSK/RPS6KA1, PKCD/PRKCD, and RSK2/RPS6KA3 kinase signatures as significantly
positively enriched in both IL-2 and IL-15 stimulated cells (Fig. 4A). Phospho-peptides from the
IL-33 and EGFR1 signaling pathways were also significantly positively enriched in both IL-2
and IL-15 stimulated cells. Notably, although the signatures for p90RSK and RSK2 share four
overlapping phospho-peptides (LCP1 pS5, RPS6 pS236, RANBP3 pS126, and BAD pS75), each
kinase signature also had unique phospho-peptides including CARHSP1 pS52 and PDCD4
pS457 for p90RSK and STAT3 pS727 and TINF2 pS95 for RSK2 (Fig. 4B).

24
Next, we used kinase set enrichment analysis (KSEA), a technique that tests for statistical
enrichment of kinase substrates predicted by NetworKIN (Linding et al., 2008) or phospho-
peptides matching known kinase motifs (Keshava Prasad et al., 2009). In both IL-2 and IL-15-
stimulated NK-92 cells, RAF substrates were significantly positively enriched, consistent with
the known activation of Ras/RAF/MAPK signaling by IL-2/15 (Mishra et al., 2014; Nandagopal
et al., 2014; Ross & Cantrell, 2018; Turner et al., 1993). In addition, EGFR substrates were
enriched in both IL-2/15 stimulated cells, consistent with results from PTM-SEA. Among the
most downregulated motifs were multiple casein kinase (CK) II motifs (e.g., pSD.E where .
represents any amino acid) or GSK-3/ERK motifs (e.g., [pS|pT]P). Notably, nearly all kinase
substrate sets enriched in IL-2-stimulated cells were also enriched in IL-15-stimulated cells.
Taken together, the PTM-SEA and KSEA results suggest that IL-2 and IL-15 activate similar
kinases in NK-92 cells including the p90RSK family of kinases.

2.3.6. Fuzzy C-means clustering identifies differing phospho-signaling kinetic profiles
Because our analysis of AUC compared phospho-peptides by their magnitude of
signaling but not their kinetics, we next sought to identify different temporal patterns of
phospho-peptide signaling downstream of IL-2 and IL-15. Using the most up- and
downregulated phospho-peptides (i.e., top 10% by absolute value of AUC, n=724), we
performed fuzzy C-means clustering analysis (Kumar & Futschik, 2007), a soft clustering
approach which identifies common signaling trajectories while allowing phospho-peptides to
belong to multiple clusters if appropriate (Hu et al., 2015; Rahmatbakhsh et al., 2021; P. Yang et
al., 2015). This analysis revealed six distinct clusters with different kinetic profiles (Fig. 5).
Cluster 1 demonstrated rapid and sustained upregulation of phosphorylation, whereas Cluster 2
25
showed rapid and sustained decreased phosphorylation. Interestingly, both Clusters 1 and 2
contained primarily IL-15 phospho-signals (74.0% and 65.9%, respectively). Cluster 3 and
Cluster 5 also showed opposite kinetic profiles with initially decreased and increased
phosphorylation, respectively, followed by adaptation to basal levels. Finally, Cluster 4
demonstrated slow but sustained decreases in phosphorylation, whereas Cluster 6 represented
minimal initial signaling followed by increased phosphorylation at 10 and 15 min after
stimulation with IL-2 or IL-15. Interestingly, phospho-peptides stimulated by both IL-2 and IL-
15 which belong to the p90RSK signature (e.g., RPS6 pS236 & pS240) and the EGFR1 pathway
signature (e.g., STAT3 pS727) were identified in Cluster 1. This analysis further suggests that
both p90RSKs and the EGFR pathway are rapidly activated and sustained over 30-minute period
following IL-2/15 stimulation in NK-92 cells.

2.3.7. p90RSK signaling is required for IL-2- and IL-15-mediated proliferation
Following identification that p90RSK and RSK2 signaling was upregulated in IL-2 and
IL-15-stimulated NK-92 cells, we sought to test whether p90RSK signaling is required for
proliferation induced by IL-2 and IL-15. We therefore tested whether inhibiting RSK signaling
with two structurally distinct p90RSK inhibitors, LJI308 and BI-D1870, would inhibit
proliferation of NK-92 cells. Indeed, both inhibitors significantly reduced the growth of NK-92
cells treated with IL-2 or IL-15 (Fig. 6). This data supports that p90RSK signaling is required for
IL-2 and IL-15-mediated proliferation in NK-92 cells.

26
2.4. Discussion
Although it is widely recognized that interleukins including IL-2 and IL-15 regulate NK
cell function, the signaling pathways stimulated by these cytokines are still not fully understood.
To address this deficit, we present here the first LC-MS phospho-proteomic characterization of
IL-2 and IL-15 signaling in the immortalized NK cell line NK-92. By quantitative analysis of the
early signaling kinetics, we show that IL-2 and IL-15 activate highly similar intracellular
signaling in NK-92 cells, including both known (e.g., JAK/STAT (Gotthardt et al., 2019; Morris
et al., 2018)) and novel (e.g., p90RSK) signaling pathways. Remarkably, despite the low
homology of these cytokines, the magnitude of signaling in response to IL-2 and IL-15 was
extremely high across >3,600 phosphorylation sites (r=0.72) (Fig. 3A). Finally, using
pharmacological inhibitors, we then demonstrated that p90RSK signaling is required for IL-2/15-
mediated proliferation of NK-92 cells. Taken together, our results show the power of LC-MS
phospho-proteomics to elucidate NK cell function and outline a pipeline for the analysis of
signaling kinetics and downstream function in immune cells.

Previous studies of IL-2 and IL-15 signaling in NK cells were limited in that they require
validated, costly antibodies to detect phospho-signals (Frank et al., 1995; Gilmour et al., 2001;
Pande et al., 2012). In contrast, our LC-MS phospho-proteomic approach enables an unbiased
and quantitative characterization of phospho-signaling across thousands of phosphorylation sites.
Prior to our analysis, proteomic approaches to analyze NK cell signaling have been limited to a
kinase-selective phospho-proteomic approach that identified and quantified 313 phosphorylation
sites on 109 kinases at one time point (König et al., 2012). In contrast, our study identified
>8,000 phosphorylation sites and fully quantified >3,600 phosphorylation sites across five time
27
points. Thus, this work has greatly extended the landscape of known signaling events in NK
cells. Among the phospho-peptides we quantified downstream of IL-2 and IL-15, the most
upregulated phosphorylation site was STAT5A Y694 / STAT5B Y699 (these tryptic phospho-
peptides are indistinguishable). Remarkably, in terms of AUC, this phosphorylation site was
20% more upregulated than even the second most upregulated phospho-peptide (SRRM2
S1099/S1101/S1103). Notably, phosphorylation of these tyrosine residues on is crucial for
activation and nuclear translocation of STAT5A / STAT5B (Fahrenkamp et al., 2015; Klejman et
al., 2002). We also observed increased phosphorylation on STAT1 S727 and STAT3 S727, both
of which are required for promotion of transcriptional activity under multiple cellular contexts
(Decker & Kovarik, 2000). Taken together, this data demonstrates that IL-2 and IL-15
significantly upregulate the JAK/STAT pathway within 30 minutes of stimulation in NK-92
cells.

Our analysis also revealed that IL-2/15 activate the p90RSK family of kinases in NK-92
cells. Previous work has shown that IL-2 and IL-15 can activate p90RSK signaling in T cells
(Franklin et al., 1996; J. X. Lin et al., 2008), but we are unaware of previous reports showing this
connection in NK cells. Notably, our phospho-proteomic data showed only marginally increased
phosphorylation of p90RSK (RPS6KA1) on S380, a phosphorylation site often used as marker of
p90RSK activation (AUC values of 8.4 and 4.1 arbitrary units for IL-2 stimulation and IL-15,
respectively) (Salhi et al., 2015; Singh et al., 2019). However, our bioinformatic analysis (e.g.,
PTM-SEA) revealed the activation of p90RSK from the enrichment of known downstream
phosphorylation targets including the 40S ribosomal protein S6 (RPS6) S236, a downstream
target of p90RSK and RSK2 that induces enzymatic activity and cell growth (L. Lin et al., 2019;
28
Peng et al., 2010; Roux et al., 2003). In addition, the phosphorylation of plastin-2 (LCP1) S5 and
serine/threonine-protein kinase 17B (STK17B/DRAK2) S12, both targets of p90RSK and RSK2
(Machado et al., 2021; Schenk et al., 2017), was significantly upregulated in response to IL-2
and IL-15. Both LCP1 S5 and STK17B S12 have previously been linked to IL-2 signaling in
cytotoxic CD8+ T cells (Ross et al., 2016). Other notable p90RSK targets in our data set include
RANBP3 pS126 (Yoon et al., 2008) and BAD pS75 (Cronin et al., 2021).

While p90RSKs have been shown to be rapidly activated by IL-2, IL-15, and IL-7 in T
cells (J. X. Lin et al., 2008), their role in NK cells has not yet been elucidated. Previous research
has suggested that CREB activation might be a result of RSK-2 activation in NK cells stimulated
by IL-2, but this has not yet been conclusively demonstrated (Ponti et al., 2002). Importantly,
when we treated NK-92 cells with the RSK inhibitors BI-D1870 and LJI308, we observed a
complete suppression of cell growth and even cell death at higher doses, thus connecting our
phospho-proteomic findings to NK cell function. Future work will seek to identify whether
single or multiple members of the p90RSK family are required for proliferation of NK-92 cells.

In summary, this work presents quantitative insight into NK-92 cell signaling
downstream of IL-2 or IL-15 and offers the potential for improved understanding of NK cell
biology and immunoengineering. NK cells rely on a balance between activating and inhibitory
receptors to initiate activity, making it challenging to characterize the activation signals that can
be used for engineering NK cells for therapeutic usage (Paul & Lal, 2017; Sivori et al., 2019).
Because of their complexity, NK cells are typically engineered with tools that have been
optimized for T cells, resulting in suboptimal engineered cell designs (Albinger et al., 2021).
29
Regardless, given that the immortalized cell line NK-92 is being used clinically for tumor
immunotherapy (Cheng et al., 2012; X. Tang et al., 2018), our results may provide insight into
the essential components of IL-2/15 signaling to improve engineering of NK-92 cells for optimal
activity in vivo. As the field of immunotherapy continues to expand and to push into the clinic,
phospho-proteomic characterization of immune cell signaling will be critical to the improved
understanding of the complex mechanisms by which immune cell function is regulated and how
immune cells can be manipulated for therapeutic benefit.

30
Chapter 3. Phospho-proteomics of CAR-T Signaling
3.1. Complications with Phospho-proteomic Analysis of Cell Co-cultures
Phospho-proteomic studies of CAR signaling have been limited by the challenge of
stimulating CAR-expressing cells with their in vivo stimulus, antigen-presenting cancer cells. In
LC-MS-based proteomic studies, mixing two cell types such as CAR-T and target cancer cells
confounds analysis because both cell types express many of the same proteins. Additionally,
because phospho-signaling occurs on rapid time scales (Reddy et al., 2016; Torchia et al., 2018),
it is not feasible to use flow sorting to separate cell types before phospho-proteomic analysis. As
such, some studies have relied on antigen-presenting beads to activate CARs (Philipson et al.,
2020; Salter et al., 2018, 2021) but these methods fail to produce a realistic immunological
synapse with membrane interaction and ligand exchange. To overcome this limitation, others
have used stable isotope labeling with amino acids in cell culture (SILAC) to label CAR-T cells
and cancer cells before co-culture, thereby enabling deconvolution of peptides on the MS
(Griffith et al., 2022b; Ramello et al., 2019). However, SILAC-based methods remain expensive
and limited in the number of samples that can be quantified and the number of cell types that can
be co-cultured (X. Chen et al., 2015; X. Wang et al., 2018).

To overcome these limitations, we have developed a formalin fixation and cell sorting
method that enables label-free phospho-proteomic analysis of CAR-T cells stimulated with
antigen-expressing cancer cells. Specifically, after mixing CAR-T and cancer cells, we preserve
transient phospho-signals by formalin fixation, label CAR-T cells with magnetic beads coated
with an antibody that recognizes the CAR, and then purify CAR-T cells using magnet-activated
cell sorting (MACS) (S. Li et al., 2015) (Fig. 7). Isolated CAR-T cells are then lysed, and
31
formalin-fixed proteins are de-crosslinked and analyzed by phospho-proteomics. The method
presented here represents an easy and inexpensive method for phospho-proteomic analysis of
CAR signaling in immune cells stimulated by cancer cells expressing the target antigen.

3.2. CAR-T Cell Isolation from Cell-cell Co-culture

3.2.1. Materials and Methods
Plasmid Construction
The retroviral vector encoding anti-CD19 CAR was constructed by incorporating the
anti-CD19 ScFv derived from the anti-CD19 antibody FMC63 (Acro Biosystems) into the MP-
71 retroviral vector backbone kindly provided by Prof. Wolfgang Uckert (Max-Delbrück-Center
for Molecular Medicine) using methods previously described (Engels et al., 2003). The CAR
expression cassette within MP-71 backbone also contains a CD8 hinge, CD8 and CD3 𝜁
transmembrane domains, and CD28 and CD3 𝜁 signaling domains. The lentiviral vectors
encoding GFP and CD19 were constructed from the FUW backbone as previously described
(Han et al., 2017).

Vector Production
293T cells were obtained from ATCC (CRL-3216) and maintained in D10 medium
consisting of DMEM supplemented with 10% fetal bovine serum (FBS), 2 mM L-glutamine, and
0.35% Pen Strep (Gibco 15140-122). Media was filtered using a 0.2-μm bottle filter and warmed
to 37°C before being added to cells. Retroviral and lentiviral vectors were prepared by transient
transfection of 293T cells using a standard calcium phosphate precipitation method. Briefly, 18
32
million 293T cells were seeded in a 15-cm tissue-culture dish. When confluency reached 70%–
80%, retrovirus was produced by transfecting 293T cells with plasmids encoding the retroviral
backbone, the RD114 envelope protein and the packaging proteins (Gag-Pol). For lentivirus,
293T cells were transfected with plasmids encoding the lentiviral backbone, VSV-G envelope
protein, and packaging proteins (Gag-Pol). The viral supernatants were harvested 48 hours after
transfection and filtered through a 0.45-μm filter (Corning). Filtered viral supernatant was
ultracentrifuged at 25,000 rpm for 1.5 hours to increase viral titer.

Anti-CD19 CAR-T Cell Production
Frozen human peripheral blood mononuclear cells (PBMCs) (AllCells by
FisherScientific, 50-156-486) and were recovered in T-cell culture medium (TCM) consisting of
AIM-V Medium (ThermoFisher, 12055083), 5% human AB serum (GemCell, GeminiBio, 100-
512), 10 mM HEPES, 1% GlutaMax-100x, 12.25 mM N-acetylcysteine, and 0.35% Pen Strep
(Gibco, 15140-122). Media was filtered using a 0.2-μm bottle filter and warmed to 37°C before
being added to cells. Interleukin-2 (IL-2) (PeproTech, 200-02) was added at a concentration of
100 U/mL during each media refresh. PBMCs were recovered at a concentration of 1 ×
10
6
cells/mL overnight in a 5% CO2, 37°C and humidified incubator. T cells were activated by
adding Dynabeads Human T-Expander CD3/CD28 (Invitrogen, 11141D) at a bead:PBMC ratio
of 3:1 for 48 hours in a 5% CO2, 37°C and humidified incubator. For T cell transduction,
RetroNectin (TaKara, T202) was bound to a non-tissue culture treated plate overnight at 4°C.
Both T cells and viral supernatants were added to the RetroNectin and centrifuged for 90 minutes
at 300 g. Fresh media was added to the cells and transduced T cells were expanded for 2 weeks
in TCM, during which time the culture medium was replenished every 2 days and T cell density
33
was maintained between 0.5 and 1 × 10
6
cells/mL. Following expansion, T cells were then
reconstituted in freezing medium (90% FBS, 10% DMSO) and stored in the vapor phase of a
liquid nitrogen cell stock dewar until needed.

Cell Culture
Anti-CD19 CAR-T cells were expanded in TCM. Interleukin-2 (IL-2) was added at a
concentration of 100 U/mL during each media refresh. Cells were initially recovered at a
concentration of 1 × 10
6
cells/mL but maintained at a concentration between 0.5 and 1 ×
10
6
cells/mL for all subsequent passages. SKOV3 cells were obtained from ATCC (30-2007) and
transduced with either the lentiviral vector FUW-CD19 or the lentiviral vector FUW-GFP to
overexpress CD19 or GFP, respectively. Non-transduced SKOV3 cells were also used as
controls in these experiments. SKOV3 cells were expanded in RPMI media containing 15% FBS
(Sigma, F2442), 2 mM L-glutamine (Corning, MT25005CV), 20 mM HEPES (ThermoFisher
Scientific, 15630106), 1 mM β-mercaptoethanol (Gibco, 21985-023), and 0.35% Pen Strep
(Gibco, 15140-122). All cells were counted by Trypan Blue exclusion with a hemocytometer. All
cells were grown in a 5% CO2, 37°C and humidified incubator and were used within 20 passages
of thawing.

Flow Cytometry Analysis
To measure interferon gamma (IFNγ) production, cells were mixed at effector-to-target
ratios of 10:1, 5:1, 1:1, 0.5:1, and 0.2:1 to determine which ratio prompted the greatest response.
Cells were treated with 5 µg/mL of Brefeldin A (BioLegend, 420601) and allowed to incubate
for 6 hours in a 5% CO2, 37°C and humidified incubator. Cells were then washed 1 time with
34
100 µL PBS and then fixed and permeabilized in 100 µL Fixation and Permeabilization Solution
from the BD Cytofix/Cytoperm Fixation/Permeabilization Kit (BD BioSciences, 554714) for 10
minutes at 4°C. Cells were washed 2 times using 100 µL 1X Perm/Wash buffer provided in the
Cytofix/Cytoperm Fixation/Permeabilization Kit and then labeled with a R-Phycoerythrin (PE)
labeled anti-human interferon(IFN)-gamma( γ) antibody at a ratio of 1:100 in 100 µL 1X
Perm/Wash buffer and incubated for 30 minutes at 4°C. Cells were then washed 2 times using
100 µL 1X Perm/Wash buffer and resuspended in 100 µL PBS for flow cytometry analysis. Cells
were then analyzed using a Miltenyi Biotec MACSQuant benchtop flow cytometer.

To measure cytotoxicity, SKOV3 target cells were resuspended in 0.1% BSA in PBS to a
concentration of 1 × 10
6
cells/mL and stained with 1 µM CFSE for 10 min at 37°C. 1 equal
volume of FBS was added to stop the reaction, and cells were washed 3X with PBS to remove
excess CFSE before being reconstituted in TCM. Cells were then mixed at a target-to-effector
ratio of 1:1 and allowed to incubate for 24 hours in a 5% CO2, 37°C and humidified incubator.
Cells were then washed 1 time with 100 µL PBS and stained with 7AAD for 10 minutes at 4°C.
Cells were then analyzed using a Miltenyi Biotec MACSQuant benchtop flow cytometer.

Formalin Fixation Experiments
T cells expanded from the same PBMC donor as other experiments used in this study
were evaluated for the effect of protein crosslinking by formalin fixation on phospho-proteomic
results. The non-fixed population (NFF) was briefly centrifuged and washed once with PBS
before being immediately frozen at -80°C while the experimental population was fixed (FF) by
adding an equal volume of 20% neutral buffered formalin for a final formaldehyde concentration
35
of ~4%. Both FF and NFF populations were lysed according to the same protocol in a 5% SDS
lysis buffer as described below. Three biological replicates of each condition were collected and
analyzed to improve statistical power of detection.

Cell Co-culture and Magnet-associated Cell Sorting (MACS)
Co-cultures were conducted in biological triplicate with anti-CD19 CAR T cells
stimulated by CD19-expressing SKOV3 cells (SKOV3.CD19) as the experimental condition and
by non CD19-expressing SKOV3 cells (SKOV3.NT) as the control. Co-cultures were incubated
at a 1:1 target-to-effector ratio for 45 minutes in a 5% CO2, 37°C and humidified incubator.
Immediately following co-culture, cells were mixed with 20% neutral buffered formalin for a
final formaldehyde concentration of ~4%. For analysis of MACS purity, SKOV3.NT and
SKOV3.CD19 cells were first labeled with either CFSE or GFP. CFSE labeling was done as
described for the cytotoxicity assay. GFP labeling was performed via transfection with the FUW-
GFP vector. Labeled SKOV3.NT and SKOV3.CD19 cells were mixed with CAR-T cells for 45
min and then formalin fixed as described above. Next, the cell mixture was washed with PBS,
and CD19.CAR-T cells were labeled with biotinylated anti-Fab antibodies (ThermoFisher
Scientific, 31803). The co-culture was then mixed with CELLection Biotin Binder
(ThermoFisher Scientific, 11533D) Dynabeads for 20 minutes at 4°C with rotation. A magnetic
rack was used to pull Dynabead bounds cells out of solution, and the cell-bound beads were
washed 2x with 0.1% BSA in PBS and then treated with a DNase I-containing release buffer
provided by the CELLection kit to break the DNA linker and release the cells from the beads.
Isolated cells were then analyzed in technical duplicate by LC-MS proteomics or flow cytometry
for CFSE or GFP signal to evaluate purity.
36

Cell Lysis and Formaldehyde Removal
To de-cross link proteins following formalin fixation, cells were sonicated in a 5% SDS,
100 mM Tris buffer and then heated for 60 minutes at 80°C (Kast & Klockenbusch, 2010). This
process was repeated once more before proteins were reduced with 10 mM dithiothreitol (DTT)
for 30 minutes at room temperature (RT), alkylated with 20 mM iodacetamide (IAA) for 30
minutes at RT in the dark, and quenched with 10 mM DTT. Samples were then diluted 8x with
S-Trap buffer (90% methanol, 100 mM triethylammonium bicarbonate (TEAB) (ThermoFisher,
90114)) solution. The protein solution was then added directly to S-Trap mini columns (Protifi,
C02-mini-40) by centrifugation (400 µL at a time, 4,000 g spin for 30 seconds) and washed 3x
with S-Trap buffer to remove SDS (Marchione et al., 2020). 125 µL per column of Trypsin
solution was prepared by diluting Trypsin (Pierce Trypsin Protease, 90058) at a 1:20 protein
mass ratio in digestion buffer (50mM TEAB). Proteins were digested overnight in a 37°C water
bath. The following morning, the columns were placed in new collections. 80 µL of digestion
buffer was added to each column, and centrifugation for 1 minute at 1000x g was used to collect
the elution. 80 µL of 0.2% formic acid (FA) was added to the column and elution was collected
into the same tube following centrifugation for 1 minute at 1000x g. A final 80 µL of 50%
acetonitrile (ACN) / 0.2% FA was added to the column and the elution was collected into the
same tube following centrifugation for 1 minute at 1,000 g. Peptides were then vacuum dried
prior to phospho-enrichment or directly injected onto the LC-MS for proteomic analysis.

37
LC-MS Phospho-proteomics
300 µg of dried peptides were resuspended in 170 µL loading buffer (65% ACN, 2%
TFA, saturated with 8.6 mg/mL glutamic acid) for phospho-enrichment. Titanium dioxide beads
(GL sciences, #1400B500) were weighed out at a ratio of 10 mg beads to 1 mg tryptic peptide
for enrichment and resuspended in 30 µL per sample of loading buffer to prepare a bead slurry.
Samples and beads were vortexed (setting 6.0 on Vortex-Genie 2T, SIIE10013) for 15 minutes at
room temperature (RT) to ensure mixing. 30 µL of bead slurry was aliquoted into a new tube for
each sample. Reconstituted dried peptides were added to the bead aliquots and samples were
vortexed (setting 6.0 on Vortex-Genie 2T) at RT for 60 minutes. Samples were centrifuged and
the supernatant was discarded. 200 µL of loading buffer was added to each sample and vortexed
(setting 6.0 on Vortex-Genie 2T) at RT for 30 minutes followed by centrifugation and discard of
the supernatant. Samples were resuspended in 200 µL wash buffer 1 (WB1, 65% ACN, 0.5%
TFA) and vortexed (setting 6.0 on Vortex-Genie 2T) at RT for 30 minutes followed by
centrifugation and discard of the supernatant. Samples were resuspended in 200 µL wash buffer
2 (WB2, 65% ACN, 0.1% TFA) and vortexed (setting 6.0 on Vortex-Genie 2T) at RT for 30
minutes followed by centrifugation and discard of the supernatant. Samples were resuspended in
200 µL elution buffer 1 (EB1, 50% ACN, 0.3M NH4OH) and incubated for 1 hour on a
ThermoMixer at 45°C, 1,400 RPM. Following centrifugation, the supernatant was transferred to
a clean tube and the tube was placed in a speed vac to begin the drying down process. Beads
were then reconstituted in 200 µL elution buffer 2 (EB2, 5% ACN, 0.3M NH4OH) and incubated
for 1 hour on a ThermoMixer at 45°C, 1,400 RPM. Following centrifugation, the supernatant
was transferred to their corresponding tubes and samples were dried down.

38
Desalting was carried out on stage tips composed of empty pipette tips loaded with two
cores of C18. C18 stage tips were conditioned with two times 150 µL of 50% ACN followed by
two times 150 µL of 0.1% TFA. Reagents were passed through the stage tips by centrifugation
(2,000-3,000 g). Samples were resuspended in 100 µL of 0.1% TFA and transferred onto the
stage tips. Stage tips were placed in the original sample tubes to effectively “wash” the tube as
the sample eluted. Elutions were transferred to the stage tip for another passage. Stage tips were
then washed 3 times with 150 µL 0.1% TFA. Samples were eluted with 75 µL of 60% ACN,
0.1%TFA and dried down to completion. Dried phospho-peptides were resuspended in 6 µL of
LC buffer A (0.1% FA).

The samples were randomized and 5 µL of each sample was injected into an Easy 1200
nanoLC ultra high-performance liquid chromatography coupled with a Q Exactive Plus
quadrupole-Orbitrap mass spectrometer (Thermo Fisher Scientific). Peptides were separated by a
reverse-phase analytical column (PepMap RSLC C18, 2 µm, 100 Å, 75 µm×25 cm). Flow rate
was set to 300 nL/min at a gradient from 3% LC buffer B (0.1% formic acid, 80% acetonitrile) to
38% LC buffer B in 110 min, followed by a 10-min washing step to 85% LC buffer B. The
maximum pressure was set to 1,180 bar, and column temperature was maintained at 50°C.
Peptides separated by the column were ionized at 2.4 kV in positive ion mode. MS1 survey scans
were acquired at the resolution of 70,000 from 350 to 1,800 m/z, with a maximum injection time
of 100 ms and AGC target of 1e6. MS/MS fragmentation of the 14 most abundant ions were
analyzed at a resolution of 17,500, AGC target 5e4, maximum injection time 65 ms, and
normalized collision energy of 26. Dynamic exclusion was set to 30 sec, and ions with charge
+1, +7 and >+7 were excluded.
39

MS/MS fragmentation spectra were searched with Proteome Discoverer SEQUEST
(version 2.2, Thermo Scientific) against in silico tryptic digested Uniprot all-reviewed Homo
sapiens database (release August 2017, 42,140 entries) plus all recombinant protein sequences
used in this study. The maximum missed cleavages was set to two. Dynamic modifications were
set to phosphorylation on serine, threonine or tyrosine (+79.966 Da), oxidation on methionine
(+15.995 Da), and acetylation on protein N-terminus (+42.011 Da). Carbamidomethylation on
cysteine (+57.021 Da) was set as a fixed modification. The maximum parental mass error was set
to 10 ppm, and the MS/MS mass tolerance was set to 0.02 Da. The false discovery threshold was
set strictly to 0.01 using the Percolator Node validated by q-value. The relative abundance of
parental peptides was calculated by integration of the area under the curve of the MS1 peaks
using the Minora LFQ node. A signal-to-noise threshold of 1.5 was implemented. Individual
phospho-site localization probabilities were determined by the ptmRS node, and phospho-sites
with <0.75 localization probability were removed. Peak files were obtained through conversion
by the MSConvertGUI (v 3.0.21301-4175933). RAW LC-MS files and processed result files are
available at MassIVE database (M. Wang et al., 2018) under identifier MSV000088632
(username: MSV000088632_reviewer, password: MM2021).

Data Analysis and Statistics
Resulting peptide data was filtered for phospho-peptides having two or fewer missed
cleavages. No data was imputed. Phospho-peptide intensities were log 2 transformed and median
normalized. To analyze the effect of formalin fixation on the phospho-proteome, phospho-
peptide intensities and identities were compared directly.
40

3.2.2. Anti-CD19 CAR-T cells kill CD19-expressing SKOV3 cells
Human T cells were expanded from PBMCs and then engineered to express a second-
generation CAR that recognizes the tumor antigen CD19. Following recovery from freeze-thaw,
51% of the transduced T cells expressed the anti-CD19 CAR (Fig. 8A). We next tested whether
stimulation with SKOV3 ovarian cancer cells expressing either CD19 (SKOV3.CD19) or not
(SKOV3.NT) could stimulate CAR-T cells to produce IFNγ, a marker of activated T cells. We
tested several effector-to-target (E:T) cell ratios and found that stimulation with SKOV3.CD19
cells but not SKOV3.NT cells induced IFNγ production in CAR-T cells (Fig. 8B). The maximal
IFNγ response occurred with a 1:1 E:T cell ratio. We thus tested the extent of cell killing of
CAR-T cells co-cultured with SKOV3.NT or SKOV3.CD19 cells at a 1:1 E:T cell ratio and
found that anti-CD19 CAR-T cells mediated five-fold more killing of SKOV3.CD19 than
SKOV3.NT (Fig. 8C). Taken together, this data demonstrates that anti-CD19 CAR-T cells
respond to and kill cancer cells in a CD19-specific fashion.

3.2.3. Formalin fixation does not impact phospho-peptide identification or quantitation by
LC-MS
Next, we tested the effects of formalin fixation on the identification and quantitation of
phospho-peptides by LC-MS. We prepared parallel samples of T cells either by formalin fixation
or flash freezing, followed by lysis in 5% SDS buffer, and SDS removal and on-column protein
digestion using S-traps (Marchione et al., 2020). Phospho-peptides were then enriched by TiO2
pulldown and were analyzed by LC-MS. In both formalin-fixed and flash-frozen samples, we
identified 5,254 phospho-peptides from 2,110 proteins, of which 4,569 phospho-peptides from
41
1,931 proteins were quantified in at least 1 of 3 replicates per condition with 93.2% of phospho-
peptides identified in both conditions (Fig. 9A). Examining the phospho-peptides that were
quantified in all samples (i.e., no data imputation), we found that phospho-peptide intensities
were strongly correlated in formalin-fixed and flash-frozen conditions (Fig. 9B). This data
demonstrates that formalin-fixation does not alter phospho-proteomic identification or
quantitation by LC-MS.

3.2.4. Magnet-activated cell sorting (MACS) enables purification of co-cultured CAR-T
cells to >90%
Having demonstrated that formalin fixation does not affect phospho-proteomic data, we
next tested whether we could isolate CAR-T cells from formalin-fixed co-cultures of CAR-T and
SKOV3 cancer cells using magnet-activated cell sorting (MACS) (S. Li et al., 2015). To assess
the purity of CAR-T cell isolation, we first labeled SKOV3.CD19 and SKOV3.NT cells with the
fluorescent marker CFSE. Then, we used flow cytometry to measure the percentage of CFSE
+

SKOV3 cells during co-culture with anti-CD19 CAR-T cells and after MACS purification of
CAR-T cells (Fig. 10A). After labeling, the percentage of CFSE
+
SKOV3 cells reached 86.7-
92.6% (Fig. 10B). Following mixing of SKOV3.CD19/NT cells at a 1:1 ratio with unlabeled
anti-CD19 CAR-T cells, the percentage of CFSE
+
cells dropped to 40.9-43.9%. Lastly, after
formalin fixation and MACS purification, the percentage of CFSE
+
cells was reduced to 5.8-
7.6% of total live cells, supporting that the elution was >90% CAR-T cells. We also confirmed
this MACS sorting efficiency using GFP-labeled SKOV3 cells (Supp. Fig. 6). Together, this
data indicates that MACS can enrich CAR-T cells to high purity from formalin-fixed co-cultures
of SKOV3 cancer cells and CAR-T cells.
42

3.3. Direct Analysis of anti-CD19 CAR-T Cells Stimulated by CD19-presenting Cancer Cells

3.3.1. Materials and Methods
Everything was conducted as described in Section 3.2.1., with the addition of extended
data analysis described below.
Data Analysis and Statistics
For analysis of the CAR-T signaling phospho-proteomic dataset, we applied strict filters
to ensure that each phospho-peptide included in the analysis was 1) observed in at least one
technical replicate of every sample and 2) observed in at least 10/12 replicates. We then
performed an error of propagation on the technical replicates to calculate the standard deviation
of our biological replicates and used these values to perform a two-sample t-test. P-values were
corrected using the Benjamini-Hochberg correction with a false discovery rate of 0.05. Post-
translational modification signature enrichment analysis was performed according to the
established workflow (Krug et al., 2019).

3.3.2. Analysis of CAR-T cell signaling following stimulation by antigen-presenting cancer
cells
Next, we applied our formalin fixation and MACS purification protocol to analyze CAR-
T cell signaling. To measure activation of the CAR in response to its target antigen, we mixed
anti-CD19 CAR-T cells with either SKOV3.CD19 or SKOV3.NT cells (i.e., negative control) at
43
a 1:1 ratio. After 45 min, co-cultures were formalin fixed, and CAR-T cells were purified by
MACS, followed by lysis, protein de-crosslinking, tryptic digestion, TiO2 phospho-peptide
enrichment, and LC-MS proteomics. Across three biological replicates run in technical duplicate
in each condition, we identified 4,141 phospho-peptides from 1,964 proteins, of which 2,326
phospho-peptides from 1,293 proteins were quantified in at least 10 of 12 samples and one of
each technical replicate. Comparing CAR-T cells stimulated with SKOV3.CD19 cells to CAR-T
cells stimulated with SKOV3.NT cells, the log2 fold changes in phosphorylation across all
phospho-peptides were normally distributed, with mean and median log2 fold change slightly
greater than zero (Supp. Fig. 7). In addition, we found 171 phospho-peptides that were
quantified in all six samples of SKOV3.CD19-stimulated cells but zero samples of SKOV3.NT-
stimulated cells, suggesting that these phospho-peptides were strongly activated by CAR
signaling. Conversely, we found only 2 phospho-peptides that were quantified in all six
SKOV3.NT-stimulated samples but zero SKOV3.CD19-stimulated samples. Of the phospho-
peptides found exclusively in SKOV3.CD19 stimulated CAR-T cells, three phospho-peptides
were from MKI67 (pSer2105, pThr2692, and pSer538), a marker of cell proliferation (Sun &
Kaufman, 2018), one was from EGFR (pSer991), and one was from MEK1/2
(pSer226/pSer222). Taken together, these data support that anti-CD19 CAR-T cells stimulated
by CD19-expressing cancer cells have upregulated signaling relative to cells stimulated by non
CD19-expressing cancer cells.

We next compared our phospho-proteomic data to the known CAR signaling network
(Cell Signaling Technology, n.d.). In our dataset, we quantified two phospho-sites on the CAR
(CD3ζ pTyr142 and CD28 pTyr223) as well as 40 phospho-peptides from 33 proteins with
44
known roles in CAR signaling including LAT, GRB2, SOS, RAF, MEK1/2, and ERK1/2 (Fig.
11). Notably, these phospho-peptides included signals thought to propagate through both the
CD28 and CD3ζ activation domains (Benmebarek et al., 2019; Ramello et al., 2019; Salter et al.,
2018; Srivastava & Riddell, 2015; Stoiber et al., 2019; Van Der Stegen et al., 2015). Both CD3ζ
pTyr142 and CD28 pTyr209 were slightly downregulated by CAR activation. As expected, we
observed activation of the MAPK pathway, including the activation sites of both ERK1/MAPK3
(pThr202/pTyr204, log2 fold change = 1.2) and ERK2/MAPK1 (pThr185/pTyr187, log2 fold
change = 0.8). Interestingly, we observed >2-fold downregulation of pSer38 and pSer224 from
LAT (linker for activation of T-cells family member 1), a protein which couples CAR activation
to downstream kinases. Neither of these LAT phospho-sites has a known functional role,
although Ser224 is close to Tyr220, a site whose phosphorylation regulates LAT activity (Zhang
et al., 2000, p. 2). Lastly, we observed many upregulated phospho-sites on transcription factors
including JUND pSer90 (log2 fold change = 2.6), JUN pSer63 (log2 fold change = 2.3), and
ATF2 pSer112 (log2 fold change = 1.7). Together, this data show that our phospho-proteomic
analysis quantified many signaling nodes in the known CAR signaling network, further
validating the applicability of this approach.

Next, to identify the most significant differences between CAR-T cells stimulated
with SKOV3.CD19 cells and CAR-T cells stimulated with SKOV3.NT cells, we plotted our
phospho-proteomic data on a volcano plot (Fig. 12A). The most significantly different phospho-
peptides were also visualized on a heatmap to assess reproducibility (Fig. 12B). In total, 300
phospho-peptides were significantly upregulated and 67 were significantly downregulated in
CAR-T cells stimulated with SKOV3.CD19 cells (FDR-adjusted p-value < 0.05). Notably, the
45
significantly changing phosphorylation sites in our data set were significantly correlated with
changes induced by antibody-coated bead stimulation of a similar CAR (Supp. Fig. 8) (Salter et
al., 2018). We next annotated the significantly changing phospho-peptides with known
regulatory roles from the PhosphoSitePlus database (P. V. Hornbeck et al., 2015a) This analysis
identified notable phosphorylation sites including pSer251 of the ERBB receptor feedback
inhibitor 1 (ERRFI1), a phosphorylation site that promotes EGF signaling by inhibition of the
negative regulatory function of ERRFI1 (N. Liu et al., 2012), Thr693 of EGFR, whose
phosphorylation can affect EGFR activity and trafficking (Lan et al., 2019; Smith et al., 2021),
and Ser17 of SRC, which induces enzymatic activity (Schmitt & Stork, 2002) Among the
significantly downregulated phospho-peptides were Src-like-adapter (SLA) pSer190 and FYN-
binding protein 1 (FYB) pSer56. Although neither of these phospho-sites has a known function,
both SLA and FYB have known roles in T-cell receptor signaling (Da Silva et al., 1997; Krause
et al., 2000; J. Tang et al., 1999). Taken together, these results highlighted both novel and known
phosphorylation changes in anti-CD19 CAR-T cells stimulated by CD19-expressing SKOV3
cells.

Finally, to identify signaling pathways downstream of CAR activation, we
analyzed our phospho-proteomic data using post-translational modification-signature enrichment
analysis (PTM-SEA) (Krug et al., 2019). This analysis supported that CAR signaling activated
the ERK1/MAPK3 and ERK2/MAPK1 signaling pathways (Fig. 11C,D), corroborating our
observation of upregulated phosphorylation on the activation sites of these kinases. In addition,
PTM-SEA identified upregulated phospho-signatures including EGFR, EGF, protein kinase A
(PKACA), p38A/MAPK14, TIE2, and CHK1 as significantly upregulated in SKOV3.CD19-
46
stimulated CAR-T cells. The signature of p38A was also identified as significantly upregulated
in the PTM-SEA of CD19.CAR signaling data collected by Griffith et al, further corroborating a
role of this kinase in CAR signaling (Griffith et al., 2022b). PTM-SEA also identified phospho-
signatures of CDC7, a cyclin dependent kinase that regulates the cell cycle, and vorinostat, an
inhibitor of histone deacetylases, as significantly downregulated in SKOV3.CD19-stimulated
CAR-T cells. Taken together, these findings support that CARs activate both known and novel
signaling pathways in CAR-T cells.

3.4. Discussion
CAR immunotherapy has revolutionized cancer treatment. However, despite the rapid
growth in clinical and preclinical research, the analysis of CAR signaling by phospho-proteomics
(Griffith et al., 2022b; Ramello et al., 2019; Salter et al., 2018, 2021) and other methods (Drent
et al., 2019; Karlsson et al., 2015; G. Li et al., 2018; Philipson et al., 2020; Zhong et al., 2010)
has been limited. In addition, current methods for phospho-proteomic analysis of CAR signaling
require either bead-based antigen display to activate CAR signaling or stable isotope labeling
(i.e., SILAC) to distinguish phospho-peptides from target and immune cells on the MS. Here,
using formalin fixation and MACS, we have developed an efficient and cost-effective method for
label-free phospho-proteomic analysis of CAR immune cells stimulated with antigen-presenting
cancer cells. This method preserves transient phospho-signaling while recapitulating the
physiological stimulus for CAR immune cells without the need for stable isotope labels. Using
anti-CD19 CAR-T cells as an example, we have demonstrated the power of this method to reveal
downstream CAR signaling that is consistent with findings from other phospho-proteomic
47
approaches, and we expect that future applications will yield additional insight into how CARs
encode the cytotoxic function of immune cells.

Because of the time requirement for MACS (~2 h), formalin fixation is essential to
ensure that CAR-mediated phosphorylation events, which occur on rapid time scales (Chylek et
al., 2014), are not lost during cell sorting. In proteomic analysis by phospho-flow, fixation with
formalin or other agents is often used to preserve phosphorylation status (Drent et al., 2019;
Krutzik & Nolan, 2003, p.). However, for proteomic analysis by LC-MS, protein crosslinks must
be reversed. The detergent SDS is ideal for reversing formalin crosslinks (and for solubilizing
membrane proteins), but risks contaminating the LC-MS with a difficult to remove detergent.
Here, we have taken advantage of the recent commercial development of suspension trapping (S-
Trap) technology to remove SDS from cell lysates (Marchione et al., 2020). Indeed, we found
that formalin fixation and crosslink reversal resulted in no loss of phospho-proteomic
information (Fig. 9), suggesting that this technique could be implemented more broadly for other
rapidly changing post-translational modifications or protein-protein interactions.

Following formalin fixation, we used MACS to isolate CAR-T cells from SKOV3 cancer
cells. Although fluorescence-activated cell sorting (FACS) can isolate immune cells for
proteomic profiling (Myers et al., 2019), we found that FACS was unable to purify enough CAR-
T cells for phospho-proteomics (e.g., 300 μg, not shown). MACS, in contrast, isolated CAR-T
cells at >90% purity in an experimentally tractable amount of time (Fig. 10). Future work to
improve our method will focus on increasing the efficiency of CAR-T isolation by MACS,
potentially by using antibodies that more specifically recognize the CAR or by negative selection
48
against cancer cell-specific markers. The ~10% contamination of target cancer cells in our
phospho-proteomic data represents a limitation compared to SILAC labeling of CAR-T and
cancer cells (Griffith et al., 2022b; Ramello et al., 2019) or bead-based CAR activation
(Philipson et al., 2020; Salter et al., 2018, 2021) which analyze purer populations of CAR-T
cells. However, the advantages of formalin fixation and MACS isolation in terms of cost,
throughput, and realistic CAR activation by antigen-presenting cancer cells make this an
attractive method for phospho-proteomic analysis of CAR signaling.

In our phospho-proteomic analysis, CAR stimulation with CD19-expressing SKOV3
cancer cells activated a broad range of signals (Figs. 11,12). Many of the observed signaling
pathways are consistent with other analyses of CAR signaling, supporting the validity of our
approach (Benmebarek et al., 2019; Griffith et al., 2022b; Ramello et al., 2019; Salter et al.,
2018). One of the strongest signals we observed was the activation of the ERK/MAPK signaling
pathway, both at the level of individual ERK1/2 activation sites and PTM-SEA pathway
analysis. Notably, The ERK/MAPK pathway plays a role in both CAR and T cell signaling,
particularly in T cells that have not been primed with antigen (Adachi & Davisa, 2011; Fischer et
al., 2005; Helou et al., 2013; Rincón et al., 2001). However, the functional implications of ERK
signaling for downstream CAR-T function are currently unknown.

Beyond known CAR signaling pathways, our phospho-proteomic approach also
identified novel signaling pathways downstream of the CAR. First, PTM-SEA identified CHK1,
a serine/threonine protein kinase that regulates cell cycle processes (Bartek & Lukas, 2003), as
significantly upregulated by CAR activation. CHK1 has primarily been investigated for its role
49
in inhibiting replication stress in cancer cells (Qiu et al., 2018), but it can also positively regulate
EGF signaling in the absence of DNA damage by phosphorylating and inhibiting the negative
regulator ERRFI1 on pSer251 (N. Liu et al., 2012). In our data, ERRF1 pSer251 was
significantly upregulated by CAR activation (log2 fold change 2.60), supporting the possibility
that CHK1 functions to modulate signaling in CAR-T cells. However, the role of CHK1 in CAR
and TCR signaling has not yet been defined. In addition, our phospho-proteomic analysis
identified that two EGFR phosphorylation sites with known regulatory roles (e.g., pThr693 and
pSer991) (Lan et al., 2019; Smith et al., 2021; Tong et al., n.d.) and an EGFR kinase signature
were upregulated by CAR activation. Although a previous analysis of CAR-T cell signaling
suggested that activated CAR-T cells could phosphorylate EGFR (Karlsson et al., 2015), it has
traditionally been thought that hematopoietic cells do not express EGFR. However, sporadic
reports have demonstrated that EGFR is indeed expressed by human Treg (Zaiss et al., 2013) and
CD4
+
T cells (Zeboudj et al., 2018). Although we cannot exclude the possibility that these EGFR
phospho-peptides come from SKOV3 and not CAR-T cells, proteomic profiling of pure CAR-T
cells did support that CAR-T cells express EGFR (not shown). Thus, these data support the
possibility that EGFR signaling may play an unappreciated role in CAR-T signaling. Our PTM-
SEA also identified downregulation of phospho-signatures related to the kinase CDC7 and
histone deacetylase inhibitor vorinostat following CAR activation. Interestingly, in T cells,
CDC7 activity may be required for T cell activation by affecting ERK and NFκB signaling but
not proximal TCR signaling (E. W. Chen et al., 2019). Additionally, HDAC inhibitors have been
shown to enhance the proliferation and survival of adoptively transferred T cells in cancer
models (Ali et al., 2021; Lisiero et al., 2014). As such, our results point to interesting CAR-T
biology yet to be explored.
50

One limitation of our phospho-proteomic data is that we did not observe phosphorylation
sites on several proteins known to be involved in CAR signaling including ZAP70, PLCγ1, and
LCK. In addition, we did not capture some of the phosphorylation events known to occur on the
co-stimulatory and activation domains of the CAR (i.e., CD28 and CD3ζ), including those
present on the three immunoreceptor tyrosine-based activation motifs (ITAMs). Notably, many
of these phosphorylation events occur on tyrosine residues (e.g., ITAM domains, ZAP70
pTyr492, LCK pTyr192, etc.), suggesting that inclusion of a parallel phospho-tyrosine
enrichment (e.g., phospho-tyrosine antibody (Ramello et al., 2019), Src SH2 superbinder
(Griffith et al., 2022b)) would generate a more comprehensive view of the CAR phospho-
proteome. Nevertheless, among the CAR phosphorylation sites that we did quantify, it is
interesting that both were slightly downregulated (CD3ζ pTyr142 and CD28 pTyr209). This
contrasts with previous phospho-proteomic studies of second-generation CARs activated with
antibody-coated beads (Salter et al., 2018, 2021) and of a third-generation CAR activated with
cell displayed-antigen (Griffith et al., 2022b) where these phosphorylation sites were
significantly upregulated by CAR stimulation. Potentially, this discrepancy could reflect
differences in experimental design or that our CD19-expressing cancer cells have not maximally
activated the CAR. Regardless, even in the absence of differential phosphorylation levels at these
CAR sites, we observed robust activation of downstream signaling pathways (e.g., ERK) and
target-cell killing. Further research will be necessary to clarify how differences in experimental
design, CAR structure, and CAR activation methods regulate these important phosphorylation
sties.

51
Taken together, our work represents the first label-free phospho-proteomic analysis of
CAR signaling in CAR-T cells stimulated with antigen-presenting cancer cells. Our results
provide insight into the signaling mechanisms downstream of CAR activation and suggest how
CARs direct T cells to kill target cells. By coupling phospho-proteomic profiling with functional
analysis of CAR-T cell cytotoxicity, we envision that an improved understanding of CAR
signaling will enable engineering of next-generation CARs with increased potency, improved
persistence, and reduced toxic side effects.

52
Chapter 4: Future Directions
Though the application of phospho-proteomics for future analyses is boundless, we
present here a selection of future directions for this work which we feel would greatly enhance
the field of immunoproteomics.

4.1. Expanded Dynamic Signaling Analyses
While the work presented in Chapter 2 is specific to the NK cell system, the workflow we
have developed can be applied for the analysis of dynamic phospho-signaling in other cell
systems. Phospho-signaling analyses applied to one timepoint are useful for obtaining a snapshot
of cell signaling post stimulation, but time course analyses can elucidate mechanisms of
signaling that are essential for the activation of cellular activities. Through the combination of
phospho-signaling at various time points in succession with bioinformatic analyses, a better
understanding of the signaling events which enable immune cells to become sufficiently
activated against a target can be established. Through the development of this understanding,
decisions can be made about signaling elements which are essential to signaling that should be
incorporated in improved immunotherapeutics.

4.2. Genomic Validation of Phospho-proteomic Findings
Our phospho-proteomic characterization of dynamic signaling in NK-92 cells revealed a
dependence on RSK family kinases that had not previously been reported. Although we were
able to treat NK-92 cells with an RSK inhibitor to evaluate the validity of this dependence, we
did not have time to perform further experiments to identify which of the four RSK family
kinases is responsible for this dependence. Future experiments would knockout various genes for
53
these RSKs and evaluate the impact on NK-92 cell proliferation and expansion. Further, we
would perform stimulation experiments on these NK-92 knockout mutants using IL-2 and IL-15
and evaluate the phospho-signaling. These phospho-proteomic experiments would inform on the
signaling events required for the RSK role in NK-92 proliferation. Taken together, these
experiments would improve our understanding of the role of RSK family kinases in NK-92
proliferation as well as provide further evidence to support our phospho-proteomic findings.

4.3. Chimeric Antigen Receptor Signaling Events
The method we have described here has successfully enabled the phospho-proteomic
characterization of CAR-T cells in co-culture with antigen-presenting cells. Though we have
applied this method to the model of an anti-CD19 CAR-T cell co-cultured with CD19-expressing
SKOV3 cells, we believe that we have only scratched the surface of CAR signaling with this
application. To further understand the role of CAR signaling elements in the propagation of
activation signals, this method should be applied to T cells which are engineered to express
multiple different generations of CARs. The variation in phospho-signals observed via this
comparison will illuminate critical signaling events responsible for CAR activation, informing on
improved designs. By applying these methods to various CAR constructs or various co-cultures,
we believe we will be able to develop a better understanding of the signaling events attributable
to CAR stimulation and, as a result, develop improved immunotherapies.

4.4. CAR-NK Cells
Initially, we set out to develop improved CARs for implementation in NK cells. While
we fell short of that goal, we have now developed the tools to make that and other analyses
54
possible: 1) a better understanding of canonical and dynamic NK-92 cell signaling, and 2) a
method for analyzing CAR-engineered immune cells in co-culture. By combining our
understanding of essential activation mechanisms of the NK-92 cell with mechanisms for
analyzing different CAR constructs in NK cells, we believe that we will be able to develop a set
of criteria necessary for optimal CAR-NK cells. Through the identification and implementation
of these criteria, we hope that future studies will find improved immunotherapies with broader
therapeutic implications.

55
References
Abel, A. M., Yang, C., Thakar, M. S., & Malarkannan, S. (2018). Natural killer cells:
Development, maturation, and clinical utilization. Frontiers in Immunology, 9(AUG), 1–
23. https://doi.org/10.3389/fimmu.2018.01869
Adachi, K., & Davisa, M. M. (2011). T-cell receptor ligation induces distinct signaling pathways
in naïve vs. Antigen-experienced T cells. Proceedings of the National Academy of
Sciences of the United States of America, 108(4), 1549–1554.
https://doi.org/10.1073/pnas.1017340108
Albinger, N., Hartmann, J., & Ullrich, E. (2021). Current status and perspective of CAR-T and
CAR-NK cell therapy trials in Germany. Gene Therapy. https://doi.org/10.1038/s41434-
021-00246-w
Alesi, G., Jin, L., Magliocca, K., Kang, Y., Chen, Z., Shin, D., Khuri, F., & Kang, S. (2016).
RSK2 signals through stathmin to promote microtubule dynamics and tumor metastasis.
Oncogene, 35(41), 5412–5421. https://doi.org/10.1038/onc.2016.79.RSK2
Ali, A. I., Wang, M., von Scheidt, B., Dominguez, P. M., Harrison, A. J., Tantalo, D. G. M.,
Kang, J., Oliver, A. J., Chan, J. D., Du, X., Bai, Y., Lee, B., Johnstone, R. W., Darcy, P.
K., Kershaw, M. H., & Slaney, C. Y. (2021). A Histone Deacetylase Inhibitor,
Panobinostat, Enhances Chimeric Antigen Receptor T-cell Antitumor Effect Against
Pancreatic Cancer. Clinical Cancer Research, 27(22), 6222–6234.
https://doi.org/10.1158/1078-0432.CCR-21-1141
Álvarez-Salamero, C., Castillo-González, R., & Navarro, M. N. (2017). Lighting up T
lymphocyte signaling with quantitative phosphoproteomics. In Frontiers in Immunology
(Vol. 8, Issue AUG). https://doi.org/10.3389/fimmu.2017.00938
Ardito, F., Giuliani, M., Perrone, D., Troiano, G., & Muzio, L. Lo. (2017). The crucial role of
protein phosphorylation in cell signaling and its use as targeted therapy (Review).
International Journal of Molecular Medicine, 40(2), 271–280.
https://doi.org/10.3892/ijmm.2017.3036
Bartek, J., & Lukas, J. (2003). Chk1 and Chk2 kinases in checkpoint control and cancer. Cancer
Cell, 3(5), 421–429. https://doi.org/10.1016/S1535-6108(03)00110-7
56
Benmebarek, M. R., Karches, C. H., Cadilha, B. L., Lesch, S., Endres, S., & Kobold, S. (2019).
Killing mechanisms of chimeric antigen receptor (CAR) T cells. International Journal of
Molecular Sciences, 20(6). https://doi.org/10.3390/ijms20061283
Casas-Rua, V., Tomas-Martin, P., Lopez-Guerrero, A. M., Alvarez, I. S., Pozo-Guisado, E., &
Martin-Romero, F. J. (2015). STIM1 phosphorylation triggered by epidermal growth
factor mediates cell migration. Biochimica et Biophysica Acta - Molecular Cell Research,
1853(1), 233–243. https://doi.org/10.1016/j.bbamcr.2014.10.027
Cell Signaling Technology. (n.d.). CAR Signaling Networks Interactive Pathway.
https://www.cellsignal.com/pathways/car-signaling-networks
Cerezo, E. L., Houles, T., Lie, O., Sarthou, M. K., Audoynaud, C., Lavoie, G., Halladjian, M.,
Cantaloube, S., Froment, C., Burlet-Schiltz, O., Henry, Y., Roux, P. P., Henras, A. K., &
Romeo, Y. (2021). RIOK2 phosphorylation by RSK promotes synthesis of the human
small ribosomal subunit. In PLoS Genetics (Vol. 17, Issue 6).
https://doi.org/10.1371/journal.pgen.1009583
Chae, H. D., Dutta, R., Tiu, B., Hoff, F. W., Accordi, B., Serafin, V., Youn, M., Huang, M.,
Sumarsono, N., Davis, K. L., Lacayo, N. J., Pigazzi, M., Horton, T. M., Kornblau, S. M.,
& Sakamoto, K. M. (2020). RSK inhibitor BI-D1870 inhibits acute myeloid leukemia cell
proliferation by targeting mitotic exit. Oncotarget, 11(25), 2387–2403.
https://doi.org/10.18632/oncotarget.27630
Chen, E. W., Tay, N. Q., Brzostek, J., Gascoigne, N. R. J., & Rybakin, V. (2019). A Dual
Inhibitor of Cdc7/Cdk9 Potently Suppresses T Cell Activation. Frontiers in Immunology,
10, 1718. https://doi.org/10.3389/fimmu.2019.01718
Chen, X., Wei, S., Ji, Y., Guo, X., & Yang, F. (2015). Quantitative proteomics using SILAC:
Principles, applications, and developments. Proteomics, 15(18), 3175–3192.
https://doi.org/10.1002/pmic.201500108
Cheng, M., Zhang, J., Jiang, W., Chen, Y., & Tian, Z. (2012). Natural killer cell lines in tumor
immunotherapy. Frontiers of Medicine in China, 6(1), 56–66.
https://doi.org/10.1007/s11684-012-0177-7
Chylek, L. A., Akimov, V., Dengjel, J., Rigbolt, K. T. G., Hu, B., Hlavacek, W. S., & Blagoev,
B. (2014). Phosphorylation Site Dynamics of Early T-cell Receptor Signaling. PLoS
ONE, 9(8).
57
Cory, G. O. C., Cramer, R., Blanchoin, L., & Ridley, A. J. (2003). Phosphorylation of the
WASP-VCA domain increases its affinity for the Arp2/3 complex and enhances actin
polymerization by WASP. Molecular Cell, 11(5), 1229–1239.
https://doi.org/10.1016/S1097-2765(03)00172-2
Cronin, R., Brooke, G. N., & Prischi, F. (2021). The role of the p90 ribosomal S6 kinase family
in prostate cancer progression and therapy resistance. Oncogene, 40(22), 3775–3785.
https://doi.org/10.1038/s41388-021-01810-9
Da Silva, A. J., Li, Z., De Vera, C., Canto, E., Findell, P., & Rudd, C. E. (1997). Cloning of a
novel T-cell protein FYB that binds FYN and SH2-domain-containing leukocyte protein
76 and modulates interleukin 2 production. Proceedings of the National Academy of
Sciences of the United States of America, 94(14), 7493–7498.
https://doi.org/10.1073/pnas.94.14.7493
Day, E. K., Sosale, N. G., & Lazzara, M. J. (2016). Cell Signaling Regulation by Protein
Phosphorylation: A Multivariate, Heterogeneous, and Context-dependent Process.
Physiology & Behavior, 40(1), 185–192. https://doi.org/10.1016/j.copbio.2016.06.005
Decker, T., & Kovarik, P. (2000). Serine phosphorylation of STATs. Oncogene, 19(21), 2628–
2637. https://doi.org/10.1038/sj.onc.1203481
Delamater, P. L., & Uberti, J. P. (2016). Geographic access to hematopoietic cell transplantation
services in the United States. Bone Marrow Transplantation, 51(2), 241–248.
https://doi.org/10.1038/bmt.2015.246
Drent, E., Poels, R., Ruiter, R., van de Donk, N. W. C. J., Zweegman, S., Yuan, H., de Bruijn, J.,
Sadelain, M., Lokhorst, H. M., Groen, R. W. J., Mutis, T., & Themeli, M. (2019).
Combined CD28 and 4-1BB Costimulation Potentiates Affinity-tuned Chimeric Antigen
Receptor–engineered T Cells. Clinical Cancer Research, 25(13), 4014–4025.
https://doi.org/10.1158/1078-0432.CCR-18-2559
Ducommun, S., Deak, M., Sumpton, D., Ford, R. J., Núñez Galindo, A., Kussmann, M., Viollet,
B., Steinberg, G. R., Foretz, M., Dayon, L., Morrice, N. A., & Sakamoto, K. (2015).
Motif affinity and mass spectrometry proteomic approach for the discovery of cellular
AMPK targets: Identification of mitochondrial fission factor as a new AMPK substrate.
Cellular Signalling, 27(5), 978–988. https://doi.org/10.1016/j.cellsig.2015.02.008
58
Dunne, J., Lynch, S., O’Farrelly, C., Todryk, S., Hegarty, J. E., Feighery, C., & Doherty, D. G.
(2001). Selective Expansion and Partial Activation of Human NK Cells and NK
Receptor-Positive T Cells by IL-2 and IL-15. The Journal of Immunology, 167(6), 3129–
3138. https://doi.org/10.4049/jimmunol.167.6.3129
Engels, B., Cam, H., Schüler, T., Indraccolo, S., Gladow, M., Baum, C., Blankenstein, T., &
Uckert, W. (2003). Retroviral vectors for high-level transgene expression in T
lymphocytes. Human Gene Therapy, 14(12), 1155–1168.
https://doi.org/10.1089/104303403322167993
Fahrenkamp, D., Schmitz-Van De Leur, H., Küster, A., Chatain, N., & Müller-Newen, G.
(2015). Src family kinases interfere with dimerization of STAT5A through a
phosphotyrosine-SH2 domain interaction. Cell Communication and Signaling, 13(1), 1–
14. https://doi.org/10.1186/s12964-014-0081-7
Fischer, A. M., Katayama, C. D., Pagès, G., Pouysségur, J., & Hedrick, S. M. (2005). The role of
Erk1 and Erk2 in multiple stages of T cell development. Immunity, 23(4), 431–443.
https://doi.org/10.1016/j.immuni.2005.08.013
Frank, D. A., Robertson, M. J., Bonni, A., Ritz, J., & Greenberg, M. E. (1995). Interleukin 2
signaling involves the phosphorylation of Stat proteins. Proceedings of the National
Academy of Sciences of the United States of America, 92(17), 7779–7783.
https://doi.org/10.1073/pnas.92.17.7779
Franklin, R. A., Sawami, H., Terada, N., Lucas, J. J., & Gelfand, E. W. (1996). Activation of
p42erk2 MAPK and p90rsk by IL-2 occurs independently of protein kinase C. Clinical
Immunology and Immunopathology, 80(2), 116–122.
https://doi.org/10.1006/clin.1996.0104
Gilmour, K. C., Fujii, H., Cranston, T., Graham Davies, E., Kinnon, C., & Gaspar, H. B. (2001).
Defective expression of the interleukin-2/interleukin-15 receptor β subunit leads to a
natural killer cell-deficient form of severe combined immunodeficiency. Blood, 98(3),
877–879. https://doi.org/10.1182/blood.V98.3.877
Gotthardt, D., Trifinopoulos, J., Sexl, V., & Putz, E. M. (2019). JAK/STAT Cytokine Signaling
at the Crossroad of NK Cell Development and Maturation. Frontiers in Immunology,
10(November). https://doi.org/10.3389/fimmu.2019.02590
59
Griffith, A. A., Callahan, K. P., King, N. G., Xiao, Q., Su, X., & Salomon, A. R. (2022a). SILAC
Phosphoproteomics Reveals Unique Signaling Circuits in CAR-T Cells and the Inhibition
of B Cell-Activating Phosphorylation in Target Cells. Journal of Proteome Research.
https://doi.org/10.1021/acs.jproteome.1c00735
Griffith, A. A., Callahan, K. P., King, N. G., Xiao, Q., Su, X., & Salomon, A. R. (2022b). SILAC
Phosphoproteomics Reveals Unique Signaling Circuits in CAR-T Cells and the Inhibition
of B Cell-Activating Phosphorylation in Target Cells. Journal of Proteome Research.
https://doi.org/10.1021/acs.jproteome.1c00735
Grisham, R. H., & Garrett, C. M. (2013). Biochemistry (5th ed.). Cengage Learning.
Guedan, S., Calderon, H., Posey, A. D., & Maus, M. V. (2019). Engineering and Design of
Chimeric Antigen Receptors. Molecular Therapy - Methods and Clinical Development,
12(March), 145–156. https://doi.org/10.1016/j.omtm.2018.12.009
Hamid, Q., Ito, I., & Muro, S. (2006). Interleukins. In Encyclopedia of Respiratory Medicine (pp.
385–390).
Han, X., Bryson, P. D., Zhao, Y., Cinay, G. E., Li, S., Guo, Y., Siriwon, N., & Wang, P. (2017).
Masked Chimeric Antigen Receptor for Tumor-Specific Activation. Molecular Therapy,
25(1), 274–284. https://doi.org/10.1016/j.ymthe.2016.10.011
Hedrick, S. M., Cohen, D. I., Nielsen, E. A., & Davis, M. M. (1984). Isolation of cDNA clones
encoding T cell-specific membrane-associated proteins. Nature, 308(5955), 149–153.
https://doi.org/10.1038/308149a0
Helou, Y. A., Nguyen, V., Beik, S. P., & Salomon, A. R. (2013). ERK Positive Feedback
Regulates a Widespread Network of Tyrosine Phosphorylation Sites across Canonical T
Cell Signaling and Actin Cytoskeletal Proteins in Jurkat T Cells. PLoS ONE, 8(7).
https://doi.org/10.1371/journal.pone.0069641
Hornbeck, P. V., Zhang, B., Murray, B., Kornhauser, J. M., Latham, V., & Skrzypek, E. (2015a).
PhosphoSitePlus, 2014: Mutations, PTMs and recalibrations. Nucleic Acids Research,
43(D1), D512–D520. https://doi.org/10.1093/nar/gku1267
60
Hornbeck, P. V., Zhang, B., Murray, B., Kornhauser, J. M., Latham, V., & Skrzypek, E. (2015b).
PhosphoSitePlus, 2014: Mutations, PTMs and recalibrations. Nucleic Acids Research,
43(D1), D512–D520. https://doi.org/10.1093/nar/gku1267
Hu, C. W., Hsu, C. L., Wang, Y. C., Ishihama, Y., Ku, W. C., Huang, H. C., & Juan, H. F.
(2015). Temporal phosphoproteome dynamics induced by an ATP synthase inhibitor
citreoviridin. Molecular and Cellular Proteomics, 14(12), 3284–3298.
https://doi.org/10.1074/mcp.M115.051383
Hwang, J. R., Byeon, Y., Kim, D., & Park, S. G. (2020). Recent insights of T cell receptor-
mediated signaling pathways for T cell activation and development. Experimental and
Molecular Medicine, 52(5), 750–761. https://doi.org/10.1038/s12276-020-0435-8
Ikemizu, S., Chirifu, M., & Davis, S. J. (2012). IL-2 and IL-15 signaling complexes: Different
but the same. Nature Immunology, 13(12), 1141–1142. https://doi.org/10.1038/ni.2472
Janeway Jr, C., Travers, P., & Walport, M. (2001). T Cell-Mediated Immunity. In
Immunobiology: The Immune System in Health and Disease. Garland Science.
Janssen Research & Development, LLC. (2022). A Study of JNJ-68284528, a Chimeric Antigen
Receptor T Cell (CAR-T) Therapy Directed Against B-Cell Maturation Antigen (BCMA)
in Participants With Relapsed or Refractory Multiple Myeloma (CARTITUDE-1). US
National Library of Medicine. https://clinicaltrials.gov/ct2/show/NCT03548207
Kansagra, A., Farnia, S., & Majhail, N. (2020). Expanding Access to Chimeric Antigen Receptor
T-Cell Therapies: Challenges and Opportunities. American Society of Clinical Oncology
Educational Book, 40, e27–e34. https://doi.org/10.1200/edbk_279151
Karlsson, H., Svensson, E., Gigg, C., Jarvius, M., Olsson-Strömberg, U., Savoldo, B., Dotti, G.,
& Loskog, A. (2015). Evaluation of Intracellular Signaling Downstream Chimeric
Antigen Receptors. PLOS ONE, 10(12), e0144787.
https://doi.org/10.1371/journal.pone.0144787
Kast, J., & Klockenbusch, C. (2010). Optimization of formaldehyde cross-linking for protein
interaction analysis of non-tagged integrin β 1. Journal of Biomedicine and
Biotechnology, 2010. https://doi.org/10.1155/2010/927585
61
Kellie, S., & Al-Mansour, Z. (2017). Overview of the immune system. In Micro and
Nanotechnology in Vaccine Development. Elsevier Inc. https://doi.org/10.1201/b13424
Keshava Prasad, T. S., Goel, R., Kandasamy, K., Keerthikumar, S., Kumar, S., Mathivanan, S.,
Telikicherla, D., Raju, R., Shafreen, B., Venugopal, A., Balakrishnan, L., Marimuthu, A.,
Banerjee, S., Somanathan, D. S., Sebastian, A., Rani, S., Ray, S., Harrys Kishore, C. J.,
Kanth, S., … Pandey, A. (2009). Human Protein Reference Database—2009 update.
Nucleic Acids Research, 37(SUPPL. 1), 767–772. https://doi.org/10.1093/nar/gkn892
Klejman, A., Schreiner, S. J., Nieborowska-Skorska, M., Slupianek, A., Wilson, M., Smithgall,
T. E., & Skorski, T. (2002). The Src family kinase Hck couples BCR/ABL to STAT5
activation in myeloid leukemia cells. EMBO Journal, 21(21), 5766–5774.
https://doi.org/10.1093/emboj/cdf562
Koide, N., Kasamatsu, A., Endo-Sakamoto, Y., Ishida, S., Shimizu, T., Kimura, Y., Miyamoto,
I., Yoshimura, S., Shiiba, M., Tanzawa, H., & Uzawa, K. (2017). Evidence for Critical
Role of Lymphocyte Cytosolic Protein 1 in Oral Cancer. Scientific Reports, 7, 1–9.
https://doi.org/10.1038/srep43379
König, S., Nimtz, M., Scheiter, M., Ljunggren, H. G., Bryceson, Y. T., & Jänsch, L. (2012).
Kinome analysis of receptor-induced phosphorylation in human natural killer cells. PLoS
ONE, 7(1). https://doi.org/10.1371/journal.pone.0029672
Krause, M., Sechi, A. S., Konradt, M., Monner, D., Gertler, F. B., & Wehland, J. (2000). Fyn-
Binding Protein (Fyb)/Slp-76–Associated Protein (Slap), Ena/Vasodilator-Stimulated
Phosphoprotein (Vasp) Proteins and the Arp2/3 Complex Link T Cell Receptor (Tcr)
Signaling to the Actin Cytoskeleton. Journal of Cell Biology, 149(1), 181–194.
https://doi.org/10.1083/jcb.149.1.181
Krug, K., Mertins, P., Zhang, B., Hornbeck, P., Raju, R., Ahmad, R., Szucs, M., Mundt, F.,
Forestier, D., Jane-Valbuena, J., Keshishian, H., Gillette, M. A., Tamayo, P., Mesirov, J.
P., Jaffe, J. D., Carr, S. A., & Mani, D. R. (2019). A Curated Resource for Phosphosite-
specific Signature Analysis. Molecular and Cellular Proteomics, 18(3), 576–593.
https://doi.org/10.1074/mcp.TIR118.000943
Krutzik, P. O., & Nolan, G. P. (2003). Intracellular Phospho-protein Staining Technique for
Flow Cytometry: Monitoring Single Cell Signaling Events. Cytometry Part A, 55A, 61–
70.
62
Kuang, X. Y., Chen, L., Zhang, Z. J., Liu, Y. R., Zheng, Y. Z., Ling, H., Qiao, F., Li, S., Hu, X.,
& Shao, Z. M. (2015). Stathmin and phospho-stathmin protein signature is associated
with survival outcomes of breast cancer patients. Oncotarget, 6(26), 22227–22238.
https://doi.org/10.18632/oncotarget.4276
Kuang, X. Y., Jiang, H. S., Li, K., Zheng, Y. Z., Liu, Y. R., Qiao, F., Li, S., Hu, X., & Shao, Z.
M. (2016). The phosphorylation-specific association of STMN1 with GRP78 promotes
breast cancer metastasis. Cancer Letters, 377(1), 87–96.
https://doi.org/10.1016/j.canlet.2016.04.035
Kumar, L., & Futschik, M. E. (2007). Mfuzz: A software package for soft clustering of
microarray data. Bioinformation, 2(1), 5–7. https://doi.org/10.6026/97320630002005
Kurzer, J. H., Argetsinger, L. S., Zhou, Y.-J., Kouadio, J.-L. K., O’Shea, J. J., & Carter-Su, C.
(2004). Tyrosine 813 Is a Site of JAK2 Autophosphorylation Critical for Activation of
JAK2 by SH2-Bβ. Molecular and Cellular Biology, 24(10), 4557–4570.
https://doi.org/10.1128/mcb.24.10.4557-4570.2004
Lan, T., Pang, J., Wang, Z., Wang, Y., Qian, H., Chen, Y., & Wu, Y. (2019). Type II cGMP-
dependent protein kinase phosphorylates EGFR at threonine 669 and thereby inhibits its
activation. Biochemical and Biophysical Research Communications, 518(1), 14–18.
https://doi.org/10.1016/j.bbrc.2019.07.126
Li, D., Li, X., Zhou, W. L., Huang, Y., Liang, X., Jiang, L., Yang, X., Sun, J., Li, Z., Han, W. D.,
& Wang, W. (2019). Genetically engineered t cells for cancer immunotherapy. Signal
Transduction and Targeted Therapy, 4(1). https://doi.org/10.1038/s41392-019-0070-9
Li, G., Boucher, J. C., Kotani, H., Park, K., Zhang, Y., Shrestha, B., Wang, X., Guan, L., Beatty,
N., Abate-Daga, D., & Davila, M. L. (2018). 4-1BB enhancement of CAR T function
requires NF-κB and TRAFs. JCI Insight, 3(18), 121322.
https://doi.org/10.1172/jci.insight.121322
Li, J., Wang, R., Gannon, O. J., Rezey, A. C., Jiang, S., Gerlach, B. D., Liao, G., & Tang, D. D.
(2016). Polo-like kinase 1 regulates vimentin phosphorylation at Ser-56 and contraction
in smooth muscle. Journal of Biological Chemistry, 291(45), 23693–23703.
https://doi.org/10.1074/jbc.M116.749341
Li, S., Plouffe, B. D., Belov, A. M., Ray, S., Wang, X., Murthy, S. K., Karger, B. L., & Ivanov,
A. R. (2015). An integrated platform for isolation, processing, and mass spectrometry-
63
based proteomic profiling of rare cells in whole blood. Molecular and Cellular
Proteomics, 14(6), 1672–1683. https://doi.org/10.1074/mcp.M114.045724
Li, T., Zhang, J., Zhu, F., Wen, W., Zykova, T., Li, X., Liu, K., Peng, C., Ma, W., Shi, G., Dong,
Z., Bode, A. M., & Dong, Z. (2011). P21-activated protein kinase (PAK2)-mediated c-jun
phosphorylation at 5 threonine sites promotes cell transformation. Carcinogenesis, 32(5),
659–666. https://doi.org/10.1093/carcin/bgq271
Lim, W. A., & June, C. H. (2017). The Principles of Engineering Immune Cells to Treat Cancer.
Cell, 168(4), 724–740. https://doi.org/10.1016/j.cell.2017.01.016
Lin, J. X., Spolski, R., & Leonard, W. J. (2008). Critical role for Rsk2 in T-lymphocyte
activation. Blood, 111(2), 525–533. https://doi.org/10.1182/BLOOD-2007-02-072207
Lin, L., White, S. A., & Hu, K. (2019). Role of p90RSK in kidney and other diseases.
International Journal of Molecular Sciences, 20(4), 1–13.
https://doi.org/10.3390/ijms20040972
Linding, R., Jensen, L. J., Pasculescu, A., Olhovsky, M., Colwill, K., Bork, P., Yaffe, M. B., &
Pawson, T. (2008). NetworKIN: A resource for exploring cellular phosphorylation
networks. Nucleic Acids Research, 36(SUPPL. 1), 695–699.
https://doi.org/10.1093/nar/gkm902
Lisiero, D. N., Soto, H., Everson, R. G., Liau, L. M., & Prins, R. M. (2014). The histone
deacetylase inhibitor, LBH589, promotes the systemic cytokine and effector responses of
adoptively transferred CD8+ T cells. Journal for ImmunoTherapy of Cancer, 2(1), 8.
https://doi.org/10.1186/2051-1426-2-8
Liu, C., Bai, X., Wu, J., Sharma, S., Upadhyaya, A., Dahlberg, C. I. M., Westerberg, L. S.,
Snapper, S. B., Zhao, X., & Song, W. (2013). N-WASP Is Essential for the Negative
Regulation of B Cell Receptor Signaling. PLoS Biology, 11(11).
https://doi.org/10.1371/journal.pbio.1001704
Liu, N., Matsumoto, M., Kitagawa, K., Kotake, Y., Suzuki, S., Shirasawa, S., Nakayama, K. I.,
Nakanishi, M., Niida, H., & Kitagawa, M. (2012). Chk1 phosphorylates the tumour
suppressor Mig-6, regulating the activation of EGF signalling. EMBO Journal, 31(10),
2365–2377. https://doi.org/10.1038/emboj.2012.88
64
Lutz, C. T., Karapetyan, A., Al-Attar, A., Shelton, B. J., Holt, K. J., Tucker, J. H., & Presnell, S.
R. (2011). Human natural killer cells proliferate and die in vivo more rapidly than T cells
in healthy young and elderly adults. Journal of Immunology, 186(8), 4590–4598.
https://doi.org/10.1038/jid.2014.371
Ma, A., Boone, D. L., & Lodolce, J. P. (2000). The pleiotropic functions of interleukin 15: Not
so interleukin 2-like after all. Journal of Experimental Medicine, 191(5), 753–755.
https://doi.org/10.1084/jem.191.5.753
Machado, R. A. C., Stojevski, D., De Landtsheer, S., Lucarelli, P., Baron, A., Sauter, T., &
Schaffner-Reckinger, E. (2021). L-plastin Ser5 phosphorylation is modulated by the
PI3K/SGK pathway and promotes breast cancer cell invasiveness. Cell Communication
and Signaling, 19(1), 1–22. https://doi.org/10.1186/s12964-021-00710-5
MacMullan, M. A., Dunn, Z. S., Graham, N., Yang, L., & Wang, P. (2019). Quantitative
proteomics and metabolomics reveal biomarkers of disease as potential immunotherapy
targets and indicators of therapeutic efficacy. Theranostics, 9(25).
https://doi.org/10.7150/thno.37373
Mandarano, A. H., Harris, T. L., Creasy, B. M., Wehenkel, M., & McGargill, M. A. (2020).
Drak2 contributes to autoimmunity by negatively regulating IL-2 signaling and altering
regulatory T cell development. The Journal of Immunology, 204(1 Supplement), 228.20
LP-228.20.
Marchione, D. M., Ilieva, I., Devins, K., Sharpe, D., Pappin, D. J., Garcia, B. A., Wilson, J. P., &
Wojcik, J. B. (2020). HYPERsol: High-Quality Data from Archival FFPE Tissue for
Clinical Proteomics. Journal of Proteome Research, 19(2), 973–983.
https://doi.org/10.1021/acs.jproteome.9b00686
Marzec, M., Halasa, K., Kasprzycka, M., Wysocka, M., Liu, X., Tobias, J. W., Baldwin, D.,
Zhang, Q., Odum, N., Rook, A. H., & Wasik, M. A. (2008). Differential effects of
interleukin-2 and interleukin-15 versus interleukin-21 on CD4+ cutaneous T-cell
lymphoma cells. Cancer Research, 68(4), 1083–1091. https://doi.org/10.1158/0008-
5472.CAN-07-2403
Matheron, L., Van Den Toorn, H., Heck, A. J. R., & Mohammed, S. (2014). Characterization of
biases in phosphopeptide enrichment by Ti 4+-immobilized metal affinity
chromatography and TiO2 using a massive synthetic library and human cell digests.
Analytical Chemistry, 86(16), 8312–8320. https://doi.org/10.1021/ac501803z
65
Meghnem, D., Morisseau, S., Frutoso, M., Trillet, K., Maillasson, M., Barbieux, I., Khaddage,
S., Leray, I., Hildinger, M., Quéméner, A., Jacques, Y., & Mortier, E. (2017). Cutting
Edge: Differential Fine-Tuning of IL-2– and IL-15–Dependent Functions by Targeting
Their Common IL-2/15Rβ/γc Receptor. The Journal of Immunology, 198(12), 4563–
4568. https://doi.org/10.4049/jimmunol.1700046
Melenhorst, J. J., Chen, G. M., Wang, M., Porter, D. L., Chen, C., Collins, M. A., Gao, P.,
Bandyopadhyay, S., Sun, H., Zhao, Z., Lundh, S., Pruteanu-Malinici, I., Nobles, C. L.,
Maji, S., Frey, N. V., Gill, S. I., Tian, L., Kulikovskaya, I., Gupta, M., … June, C. H.
(2022). Decade-long leukaemia remissions with persistence of CD4+ CAR T cells.
Nature. https://doi.org/10.1038/s41586-021-04390-6
Mishra, A., Sullivan, L., & Caliguiri, M. A. (2014). Molecular Pathways: Interleukin-15
Signaling in Health and in Cancer. Clinical Cancer Research, 20(8), 2044–2050.
https://doi.org/10.1016/j.earlhumdev.2006.05.022
Mitra, S., Ring, A. M., Amarnath, S., Spangler, J. B., Li, P., Ju, W., Fischer, S., Oh, J., Spolski,
R., Weiskopf, K., Kohrt, H., Foley, J. E., Rajagopalan, S., Long, E. O., Fowler, D. H.,
Waldmann, T. A., Garcia, K. C., & Leonard, W. J. (2015). Interleukin-2 Activity Can Be
Fine Tuned with Engineered Receptor Signaling Clamps. Immunity, 42(5), 826–838.
https://doi.org/10.1016/j.immuni.2015.04.018
Morris, R., Kershaw, N. J., & Babon, J. J. (2018). The molecular details of cytokine signaling via
the JAK/STAT pathway. Protein Science, 27(12), 1984–2009.
https://doi.org/10.1002/pro.3519
Mullard, A. (2021). FDA approves fourth CAR-T cell therapy. Nature, 20(News and Analysis),
166.
Myers, S. A., Rhoads, A., Cocco, A. R., Peckner, R., Haber, A. L., Schweitzer, L. D., Krug, K.,
Mani, D. R., Clauser, K. R., Rozenblatt-Rosen, O., Hacohen, N., Regev, A., & Carr, S. A.
(2019). Streamlined Protocol for Deep Proteomic Profiling of FAC-sorted Cells and Its
Application to Freshly Isolated Murine Immune Cells. Molecular & Cellular Proteomics,
18(5), 995–1009. https://doi.org/10.1074/mcp.RA118.001259
Nandagopal, N., Ali, A. K., Komal, A. K., & Lee, S. H. (2014). The critical role of IL-15-PI3K-
mTOR pathway in natural killer cell effector functions. Frontiers in Immunology,
5(APR), 1–12. https://doi.org/10.3389/fimmu.2014.00187
66
Navarro, M. N., Goebel, J., Hukelmann, J. L., & Cantrell, D. A. (2014). Quantitative
phosphoproteomics of cytotoxic T cells to reveal protein kinase D 2 regulated networks.
Molecular and Cellular Proteomics, 13(12), 3544–3557.
https://doi.org/10.1074/mcp.M113.037242
Olson, J. A., Leveson-Gower, D. B., Gill, S., Baker, J., Beilhack, A., & Negrin, R. S. (2010). NK
Cells Mediate Reduction of GVHD by inhibiting activated, alloreactive T cells while
retaining GVT effects. Blood, 115(21), 4293–4301. https://doi.org/10.1182/blood-2009-
05-222190
Pande, V. S., Li, P., Moraga, I., Spolski, R., Özkan, E., Leonard, W. J., & Garcia, K. C. (2012).
Mechanistic and structural insight into the functional dichotomy between interleukin-2
and interleukin-15. Nature Immunology, 13(12), 1187–1195.
https://doi.org/10.1038/ni.2449.Mechanistic
Paul, S., & Lal, G. (2017). The molecular mechanism of natural killer cells function and its
importance in cancer immunotherapy. Frontiers in Immunology, 8(SEP).
https://doi.org/10.3389/fimmu.2017.01124
Peng, C., Cho, Y., Zhu, F., Xu, Y., Wen, W., Ma, W., Bode, A. M., & Dong, Z. (2010). RSK2
mediates NF‐κB activity through the phosphorylation of IκBα in the TNF‐R1 pathway.
The FASEB Journal, 24(9), 3490–3499. https://doi.org/10.1096/fj.09-151290
Perica, K., Varela, J. C., Oelke, M., & Schneck, J. (2015). Adoptive T Cell Immunotherapy For
Cancer. Rambam Maimonides Medical Journal, 6(1), e0004.
https://doi.org/10.5041/rmmj.10179
Philipson, B. I., O’Connor, R. S., May, M. J., June, C. H., Albelda, S. M., & Milone, M. C.
(2020). 4-1BB costimulation promotes CAR T cell survival through noncanonical NF-κB
signaling. Science Signaling, 13(625), eaay8248.
Ponti, C., Gibellini, D., Boin, F., Melloni, E., Manzoli, F. A., Cocco, L., Zauli, G., & Vitale, M.
(2002). Role of CREB transcription factor in c-fos activation in natural killer cells.
European Journal of Immunology, 32(12), 3358–3365. https://doi.org/10.1002/1521-
4141(200212)32:12<3358::AID-IMMU3358>3.0.CO;2-Q
Pozo-Guisado, E., Casas-Rua, V., Tomas-Martin, P., Lopez-Guerrero, A. M., Martin-Romero, F.
J., & Alvarez-Barrientos, A. (2013). Phosphorylation of stim1 at erk1/2 target sites
67
regulates interaction with the microtubule plus-end binding protein eb1. Journal of Cell
Science, 126(14), 3170–3180. https://doi.org/10.1242/jcs.125054
Qiu, Z., Oleinick, N. L., & Zhang, J. (2018). ATR/CHK1 inhibitors and cancer therapy.
Radiotherapy and Oncology : Journal of the European Society for Therapeutic Radiology
and Oncology, 126(3), 450–464. https://doi.org/10.1016/j.radonc.2017.09.043
Rahmatbakhsh, M., Gagarinova, A., & Babu, M. (2021). Bioinformatic Analysis of Temporal
and Spatial Proteome Alternations During Infections. Frontiers in Genetics, 12(July), 1–
29. https://doi.org/10.3389/fgene.2021.667936
Ramello, M. C., Benzaïd, I., Kuenzi, B. M., Lienlaf-Moreno, M., Kandell, W. M., Santiago, D.
N., Pabón-Saldaña, M., Darville, L., Fang, B., Rix, U., Yoder, S., Berglund, A., Koomen,
J. M., Haura, E. B., & Abate-Daga, D. (2019). An immunoproteomic approach to
characterize the CAR interactome and signalosome. Science Signaling, 12(568), 1–16.
https://doi.org/10.1126/scisignal.aap9777
Rappsilber, J., Mann, M., & Ishihama, Y. (2007). Protocol for micro-purification, enrichment,
pre-fractionation and storage of peptides for proteomics using StageTips. Nature
Protocols, 2(8), 1896–1906. https://doi.org/10.1038/nprot.2007.261
Ratnayake, W. S., Apostolatos, C. A., Breedy, S., Dennison, C. L., Hill, R., & Acevedo-Duncan,
M. (2021). Atypical PKCs activate Vimentin to facilitate prostate cancer cell motility and
invasion. Cell Adhesion and Migration, 15(1), 37–57.
https://doi.org/10.1080/19336918.2021.1882782
Reddy, R. J., Gajadhar, A. S., Swenson, E. J., Rothenberg, D. A., Curran, T. G., & White, F. M.
(2016). Early signaling dynamics of the epidermal growth factor receptor. Proceedings of
the National Academy of Sciences of the United States of America, 113(11), 3114–3119.
https://doi.org/10.1073/pnas.1521288113
Rincón, M., Flavell, R. A., & Davis, R. J. (2001). Signal transduction by MAP kinases in T
lymphocytes. Oncogene, 20(19 REV. ISS. 2), 2490–2497.
https://doi.org/10.1038/sj.onc.1204382
Rodella, L., Zamai, L., Rezzani, R., Artico, M., Peri, G., Falconi, M., Facchini, A., Pelusi, G., &
Vitale, M. (2001). Interleukin 2 and interleukin 15 differentially predispose natural killer
cells to apoptosis mediated by endothelial and tumour cells. British Journal of
Haematology, 115(2), 442–450. https://doi.org/10.1046/j.1365-2141.2001.03055.x
68
Roffé, M., Lupinacci, F. C., Soares, L. C., Hajj, G. N., & Martins, V. R. (2015). Two widely
used RSK inhibitors, BI-D1870 and SL0101, alter mTORC1 signaling in a RSK-
independent manner. Cellular Signalling, 27(8), 1630–1642.
https://doi.org/10.1016/j.cellsig.2015.04.004
Ross, S. H., & Cantrell, D. A. (2018). Signaling and Function of Interleukin-2 in T
Lymphocytes. Annual Review of Immunology, 36, 411–433.
https://doi.org/10.1146/annurev-immunol-042617-053352
Ross, S. H., Rollings, C., Anderson, K. E., Hawkins, P. T., Stephens, L. R., & Cantrell, D. A.
(2016). Phosphoproteomic Analyses of Interleukin 2 Signaling Reveal Integrated JAK
Kinase-Dependent and -Independent Networks in CD8+ T Cells. Immunity, 45(3), 685–
700. https://doi.org/10.1016/j.immuni.2016.07.022
Roux, P. P., Richards, S. A., & Blenis, J. (2003). Phosphorylation of p90 Ribosomal S6 Kinase
(RSK) Regulates Extracellular Signal-Regulated Kinase Docking and RSK Activity.
Molecular and Cellular Biology, 23(14), 4796–4804.
https://doi.org/10.1128/mcb.23.14.4796-4804.2003
Roybal, K. T., & Lim, W. A. (2017). Synthetic Immunology: Hacking Immune Cells to Expand
their Therapeutic Capabilities. Physiology & Behavior, 35, 139–148.
https://doi.org/10.1146/annurev-immunol-051116-052302.Synthetic
Salhi, A., Farhadian, J. A., Giles, K. M., Vega-Saenz De Miera, E., Silva, I. P., Bourque, C.,
Yeh, K., Chhangawala, S., Wang, J., Ye, F., Zhang, D. Y., Hernando-Monge, E.,
Houvras, Y., & Osman, I. (2015). RSK1 activation promotes invasion in nodular
melanoma. American Journal of Pathology, 185(3), 704–716.
https://doi.org/10.1016/j.ajpath.2014.11.021
Salizzato, V., Borgo, C., Cesaro, L., Pinna, L. A., & Donella-Deana, A. (2016). Inhibition of
protein kinase CK2 by CX-5011 counteracts imatinib-resistance preventing rpS6
phosphorylation in chronic myeloid leukaemia cells: New combined therapeutic
strategies. Oncotarget, 7(14), 18204–18218. https://doi.org/10.18632/oncotarget.7569
Salter, A. I., Ivey, R. G., Kennedy, J. J., Voillet, V., Alderman, E. J., Voytovich, U. J., Lin, C.,
Liu, L., Whiteaker, J. R., Gottardo, R., Amanda, G., & Riddell, S. R. (2019).
Phosphoproteomic analysis of chimeric antigen receptor signaling reveals kinetic and
quantitative differences that affect cell function. Science Signaling, 11(544).
https://doi.org/10.1126/scisignal.aat6753.Phosphoproteomic
69
Salter, A. I., Ivey, R. G., Kennedy, J. J., Voillet, V., Rajan, A., Alderman, E. J., Voytovich, U. J.,
Lin, C., Sommermeyer, D., Liu, L., Whiteaker, J. R., Gottardo, R., Paulovich, A. G., &
Riddell, S. R. (2018). Phosphoproteomic analysis of chimeric antigen receptor signaling
reveals kinetic and quantitative differences that affect cell function. Science Signaling,
11(544), 1–18. https://doi.org/10.1126/scisignal.aat6753
Salter, A. I., Rajan, A., Kennedy, J. J., Ivey, R. G., Shelby, S. A., Leung, I., Templeton, M. L.,
Muhunthan, V., Voillet, V., Sommermeyer, D., Whiteaker, J. R., Gottardo, R., Veatch, S.
L., Paulovich, A. G., & Riddell, S. R. (2021). Comparative analysis of TCR and CAR
signaling informs CAR designs with superior antigen sensitivity and in vivo function.
Science Signaling, 14(697), Article 697. https://doi.org/10.1126/scisignal.abe2606
Schenk, L. K., Möller-Kerutt, A., Klosowski, R., Wolters, D., Schaffner-Reckinger, E., Weide,
T., Pavenstädt, H., & Vollenbröker, B. (2017). Angiotensin II regulates phosphorylation
of actin-associated proteins in human podocytes. FASEB Journal, 31(11), 5019–5035.
https://doi.org/10.1096/fj.201700142R
Schmitt, J. M., & Stork, P. J. S. (2002). Gα and Gβγ Require Distinct Src-dependent Pathways to
Activate Rap1 and Ras. Journal of Biological Chemistry, 277(45), 43024–43032.
https://doi.org/10.1074/jbc.M204006200
Schmutz, I., Wendt, S., Schnell, A., Kramer, A., Mansuy, I. M., & Albrecht, U. (2011). Protein
Phosphatase 1 (PP1) is a post-translational regulator of the Mammalian Circadian Clock.
PLoS ONE, 6(6). https://doi.org/10.1371/journal.pone.0021325
Seif, F., Khoshmirsafa, M., Aazami, H., Mohsenzadegan, M., Sedighi, G., & Bahar, M. (2017).
The role of JAK-STAT signaling pathway and its regulators in the fate of T helper cells.
Cell Communication and Signaling, 15(1), 1–13. https://doi.org/10.1186/s12964-017-
0177-y
Shah, N. N., & Fry, T. J. (2019). Mechanisms of resistance to CAR T cell therapy. Nature
Reviews Clinical Oncology, 16(6), 372–385. https://doi.org/10.1038/s41571-019-0184-6
Shimasaki, N., Jain, A., & Campana, D. (2020). NK cells for cancer immunotherapy. Nature
Reviews Drug Discovery, 19(3), 200–218. https://doi.org/10.1038/s41573-019-0052-1
Singh, M. V., Kotla, S., Le, N. T., Ko, K. A., Heo, K. S., Wang, Y., Fujii, Y., Thi Vu, H.,
McBeath, E., Thomas, T. N., Gi, Y. J., Tao, Y., Medina, J. L., Taunton, J., Carson, N.,
Dogra, V., Doyley, M. M., Tyrell, A., Lu, W., … Abe, J. I. (2019). Senescent Phenotype
70
Induced by p90RSK-NRF2 Signaling Sensitizes Monocytes and Macrophages to
Oxidative Stress in HIV-Positive Individuals: Implications for Atherogenesis.
Circulation, 139(9), 1199–1216.
https://doi.org/10.1161/CIRCULATIONAHA.118.036232
Sivori, S., Vacca, P., Del Zotto, G., Munari, E., Mingari, M. C., & Moretta, L. (2019). Human
NK cells: Surface receptors, inhibitory checkpoints, and translational applications.
Cellular and Molecular Immunology, 16(5), 430–441. https://doi.org/10.1038/s41423-
019-0206-4
Smith, M. P., Ferguson, H. R., Ferguson, J., Zindy, E., Kowalczyk, K. M., Kedward, T., Bates,
C., Parsons, J., Watson, J., Chandler, S., Fullwood, P., Warwood, S., Knight, D., Clarke,
R. B., & Francavilla, C. (2021). Reciprocal priming between receptor tyrosine kinases at
recycling endosomes orchestrates cellular signalling outputs. The EMBO Journal, 40(14).
https://doi.org/10.15252/embj.2020107182
Srivastava, S., & Riddell, S. R. (2015). Engineering CAR-T cells: Design concepts. Trends in
Immunology, 36(8), 494–502. https://doi.org/10.1016/j.it.2015.06.004
Srivastava, S., & Riddell, S. R. (2018). CAR T Cell Therapy: Challenges to Bench-to-Bedside
Efficacy. Journal of Immunology, 200(2), 459–469.
https://doi.org/10.4049/jimmunol.1701155
Stoiber, S., Cadilha, B. L., Benmebarek, M.-R., Lesch, S., Endres, S., & Kobold, S. (2019).
Limitations in the Design of Chimeric Antigen Receptors for Cancer Therapy. Cells,
8(5), 472. https://doi.org/10.3390/cells8050472
Sun, X., & Kaufman, P. D. (2018). Ki-67: More than a proliferation marker. Chromosoma,
127(2), 175–186. https://doi.org/10.1007/s00412-018-0659-8
Tamzalit, F., Barbieux, I., Plet, A., Heim, J., Nedellec, S., Morisseau, S., Jacques, Y., & Mortier,
E. (2014). IL-15.IL-15Rα complex shedding following trans-presentation is essential for
the survival of IL-15 responding NK and T cells. Proceedings of the National Academy
of Sciences of the United States of America, 111(23), 8565–8570.
https://doi.org/10.1073/pnas.1405514111
Tan, H., Yang, K., Li, Y., Shaw, T. I., Wang, Y., Blanco, D. B., Wang, X., Cho, J. H., Wang, H.,
Rankin, S., Guy, C., Peng, J., & Chi, H. (2017). Integrative Proteomics and
Phosphoproteomics Profiling Reveals Dynamic Signaling Networks and Bioenergetics
71
Pathways Underlying T Cell Activation. Immunity, 46(3), 488–503.
https://doi.org/10.1016/j.immuni.2017.02.010
Tang, J., Sawasdikosol, S., Chang, J.-H., & Burakoff, S. J. (1999). SLAP, a dimeric adapter
protein, plays a functional role in T cell receptor signaling. Proceedings of the National
Academy of Sciences, 96(17), 9775–9780. https://doi.org/10.1073/pnas.96.17.9775
Tang, X., Yang, L., Li, Z., Nalin, A. P., Dai, H., Xu, T., Yin, J., You, F., Zhu, M., Shen, W.,
Chen, G., Zhu, X., Wu, D., & Yu, J. (2018). Erratum: First-in-man clinical trial of CAR
NK-92 cells: Safety test of CD33-CAR NK-92 cells in patients with relapsed and
refractory acute myeloid leukemia. American Journal of Cancer Research, 8(9), 1899.
Tigner, A., Ibrahim, S. A., & Murray, I. (2020). Histology, White Blood Cells. In NCBI
Bookshelf.
Tomas-Martin, P., Lopez-Guerrero, A. M., Casas-Rua, V., Pozo-Guisado, E., & Martin-Romero,
F. J. (2015). Phospho-STIM1 is a downstream effector that mediates the signaling
triggered by IGF-1 in HEK293 cells. Cellular Signalling, 27(3), 545–554.
https://doi.org/10.1016/j.cellsig.2014.12.017
Tong, J., Taylor, P., Peterman, S. M., Prakash, A., & Moran, M. F. (n.d.). Epidermal Growth
Factor Receptor Phosphorylation Sites Ser991 and Tyr998 Are Implicated in the
Regulation of Receptor Endocytosis and Phosphorylations at Ser1039 and Thr1041*□S.
14.
Torchia, M. L. G., Dutta, D., Mittelstadt, P. R., Guha, J., Gaida, M. M., Fish, K., Barr, V. A.,
Akpan, I. O., Samelson, L. E., Tagad, H. D., Debnath, S., Jenkins, L. M. M., Appella, E.,
& Ashwell, J. D. (2018). Intensity and duration of TCR signaling is limited by p38
phosphorylation of ZAP-70T293 and destabilization of the signalosome. Proceedings of
the National Academy of Sciences of the United States of America, 115(9), 2174–2179.
https://doi.org/10.1073/pnas.1713301115
Turner, B. C., Tonks, N. K., Rapp, U. R., & Reed, J. C. (1993). Interleukin 2 regulates Raf-1
kinase activity through a tyrosine phosphorylation-dependent mechanism in a T-cell line.
Proceedings of the National Academy of Sciences of the United States of America,
90(12), 5544–5548. https://doi.org/10.1073/pnas.90.12.5544
72
Tyanova, S., Temu, T., Sinitcyn, P., Carlson, A., Hein, M. Y., Geiger, T., Mann, M., & Cox, J.
(2016). The Perseus computational platform for comprehensive analysis of (prote)omics
data. Nature Methods, 13(9), 731–740. https://doi.org/10.1038/nmeth.3901
Van Acker, H. H., Capsomidis, A., Smits, E. L., & Van Tendeloo, V. F. (2017). CD56 in the
immune system: More than a marker for cytotoxicity? Frontiers in Immunology, 8(JUL),
1–9. https://doi.org/10.3389/fimmu.2017.00892
Van Der Stegen, S. J. C., Hamieh, M., & Sadelain, M. (2015). The pharmacology of second-
generation chimeric antigen receptors. Nature Reviews Drug Discovery, 14(7), 499–509.
https://doi.org/10.1038/nrd4597
Vivier, E., Raulet, D. H., Moretta, A., Caligiuri, M. A., Zitvogel, L., Lanier, L. L., Yokoyama,
W. M., & Ugolini, S. (2011). Innate or adaptive immunity? The example of natural killer
cells. Science, 331(6013), 44–49. https://doi.org/10.1126/science.1198687
Waldmann, T. A. (2015). The shared and contrasting roles of interleukin-2 (IL-2) and IL-15 in
the life and death of normal and neoplastic lymphocytes: Implications for cancer therapy.
Cancer Immunology Research, 3(3), 219–227.
Wang, M., Wang, J., Carver, J., Pullman, B. S., Cha, S. W., & Bandeira, N. (2018). Assembling
the Community-Scale Discoverable Human Proteome. Cell Systems, 7(4), 412-421.e5.
https://doi.org/10.1016/j.cels.2018.08.004
Wang, W., Marinis, J. M., Beal, A. M., Savadkar, S., Wu, Y., Taunk, P. S., Wu, N., Su, W., Wu,
J., Ahsan, A., Kurz, E., Chen, T., Yaboh, I., Li, F., Gutierrez, J., Diskin, B., Hundeyin,
M., Reilly, M., Lich, J. D., … Miller, G. (2019). RIP1 Kinase Drives Macrophage
Mediated Adaptive Immune Tolerance in Pancreatic Cancer. 34(5), 757–774.
https://doi.org/10.1016/j.ccell.2018.10.006.RIP1
Wang, X., He, Y., Ye, Y., Zhao, X., Deng, S., He, G., Zhu, H., Xu, N., & Liang, S. (2018).
SILAC-based quantitative MS approach for real-time recording protein-mediated cell-cell
interactions. Scientific Reports, 8(1), 1–9. https://doi.org/10.1038/s41598-018-26262-2
Wu, S. Y., Fu, T., Jiang, Y. Z., & Shao, Z. M. (2020). Natural killer cells in cancer biology and
therapy. Molecular Cancer, 19(1), 1–26. https://doi.org/10.1186/s12943-020-01238-x
73
Xie, G., Dong, H., Liang, Y., Ham, J. D., Rizwan, R., & Chen, J. (2020). CAR-NK cells: A
promising cellular immunotherapy for cancer. EBioMedicine, 59.
https://doi.org/10.1016/j.ebiom.2020.102975
Yanagi, Y., Yoshikai, Y., Leggett, K., Clark, S. P., Aleksander, I., & Mak, T. W. (1984). A
human T cell-specific cDNA clone encodes a protein having extensive homology to
immunoglobulin chains. Nature, 308(5955), 145–149. https://doi.org/10.1038/308145a0
Yang, P., Zheng, X., Jayaswal, V., Hu, G., Yang, J. Y. H., & Jothi, R. (2015). Knowledge-Based
Analysis for Detecting Key Signaling Events from Time-Series Phosphoproteomics Data.
PLoS Computational Biology, 11(8), 1–18. https://doi.org/10.1371/journal.pcbi.1004403
Yang, Y., & Lundqvist, A. (2020). Immunomodulatory Effects of IL-2 and IL-15; Implications
for Cancer Immunotherapy 1 2. 2, 1–20.
Yoon, S. O., Shin, S., Liu, Y., Ballif, B. A., Woo, M. S., Gygi, S. P., & Blenis, J. (2008). Ran-
Binding Protein 3 Phosphorylation Links the Ras and PI3-Kinase Pathways to
Nucleocytoplasmic Transport. Molecular Cell, 29(3), 362–375.
https://doi.org/10.1016/j.molcel.2007.12.024
Zaiss, D. M. W., van Loosdregt, J., Gorlani, A., Bekker, C. P. J., Gröne, A., Sibilia, M., van
Bergen en Henegouwen, P. M. P., Roovers, R. C., Coffer, P. J., & Sijts, A. J. A. M.
(2013). Amphiregulin Enhances Regulatory T Cell-Suppressive Function via the
Epidermal Growth Factor Receptor. Immunity, 38(2), 275–284.
https://doi.org/10.1016/j.immuni.2012.09.023
Zambricki, E., Shigeoka, A., Kishimoto, H., Sprent, J., Burakoff, S., Carpenter, C., Milford, E.,
& McKay, D. (2005). Signaling T-cell survival and death by IL-2 and IL-15. American
Journal of Transplantation, 5(11), 2623–2631. https://doi.org/10.1111/j.1600-
6143.2005.01075.x
Zeboudj, L., Maître, M., Guyonnet, L., Laurans, L., Joffre, J., Lemarie, J., Bourcier, S., Nour-
Eldine, W., Guérin, C., Friard, J., Wakkach, A., Fabre, E., Tedgui, A., Mallat, Z.,
Tharaux, P.-L., & Ait-Oufella, H. (2018). Selective EGF-Receptor Inhibition
in CD4+ T Cells Induces Anergy and Limits Atherosclerosis. Journal of the American
College of Cardiology, 71(2), 160–172. https://doi.org/10.1016/j.jacc.2017.10.084
Zhan, Q., Ge, Q., Ohira, T., Van Dyke, T., & Badwey, J. A. (2003). P21-Activated Kinase 2 in
Neutrophils Can Be Regulated by Phosphorylation at Multiple Sites and by a Variety of
74
Protein Phosphatases. The Journal of Immunology, 171(7), 3785–3793.
https://doi.org/10.4049/jimmunol.171.7.3785
Zhang, W., Trible, R. P., Zhu, M., Liu, S. K., McGlade, C. J., & Samelson, L. E. (2000).
Association of Grb2, Gads, and Phospholipase C-γ1 with Phosphorylated LAT Tyrosine
Residues. Journal of Biological Chemistry, 275(30), 23355–23361.
https://doi.org/10.1074/jbc.M000404200
Zhao, R., Cui, Y., Li, S., Qin, L., & Li, P. (2019). Current status and hurdles for CAR-T cell
immune therapy. Blood Science, 1(2), 148–155.
https://doi.org/10.1097/bs9.0000000000000025
Zhao, X., Yang, J., Zhang, X., Lu, X.-A., Xiong, M., Zhang, J., Zhou, X., Qi, F., He, T., Ding,
Y., Hu, X., De Smet, F., Lu, P., & Huang, X. (2020). Efficacy and Safety of CD28- or 4-
1BB-Based CD19 CAR-T Cells in B Cell Acute Lymphoblastic Leukemia. Molecular
Therapy Oncolytics, 18, 272–281. https://doi.org/10.1016/j.omto.2020.06.016
Zhong, X.-S., Matsushita, M., Plotkin, J., Riviere, I., & Sadelain, M. (2010). Chimeric antigen
receptors combining 4-1BB and CD28 signaling domains augment PI3kinase/AKT/Bcl-
XL activation and CD8+ T cell-mediated tumor eradication. Molecular Therapy: The
Journal of the American Society of Gene Therapy, 18(2), 413–420.
https://doi.org/10.1038/mt.2009.210
Zhou, H., Ye, M., Dong, J., Corradini, E., Cristobal, A., Heck, A. J. R., Zou, H., & Mohammed,
S. (2013). Robust phosphoproteome enrichment using monodisperse microsphere-based
immobilized titanium (IV) ion affinity chromatography. Nature Protocols, 8(3), 461–480.
https://doi.org/10.1038/nprot.2013.010

75
Appendices
Figures and Figure Legends

Figure 1. Experimental design for phospho-proteomic analysis of IL-2 and IL-15 signaling in NK
cells. NK-92 cells were starved of cytokines for 6 h and then stimulated with 100 U/ml of either
IL-2 or IL-15 for 0, 5, 10, 15, and 30 min. Cell lysates were harvested, trypsinized, and then
enriched for phospho-peptides using TiO2 beads before analysis by LC-MS. The kinetic
phospho-peptide data was analyzed using kinase set enrichment analysis (KSEA), post-
translational modification signature enrichment analysis (PTM-SEA), and fuzzy C-means
clustering. Lastly, based on the finding that both IL-2 and IL-15 induced RSK signaling, we used
76
RSK inhibitors (RSKi) to validate that RSK signaling is required for proliferation of NK-92 cells
stimulated with either IL-2 or IL-15.
77

Figure 2. IL-2 and IL-15 activate canonical signaling pathways including JAK-STAT. A)
Schematic of canonical signaling pathways downstream of the receptor complex of IL-2Rα, IL-
78
2/IL-15Rβ, and the common γ chain. The small circles with P represent phosphorylation sites
quantified by LC-MS phospho-proteomics. Each circle is colored from red to blue corresponding
to positive and negative AUC values as shown. B) Kinetic profiles for selected phospho-peptides
in canonical IL-2 and IL-15 signaling pathways. Individual orange circles (IL-2) and blue
squares (IL-15) represent data from independent biological replicates, and the line represents the
average of the biological replicates. Orange and blue asterisks represent adjusted t-test p-value <
0.05 for IL-2 and IL-15 AUC, respectively.
79

80
Figure 3. IL-2 and IL-15 induce quantitatively similar signaling in NK-92 cells. A) A scatter plot
of AUC values reveals a strong correlation (r = 0.72, p < 0.00001) between IL-2 and IL-15
phospho-signaling. The p-value was calculated using a two-tailed t-test for the observed
correlation coefficient of 3,619 phospho-peptide AUC values. Individual phospho-peptides with
significant AUC values are colored as shown in the legend. Selected phospho-peptides with
regulatory function according to the PhosphoSite Plus database (P. V. Hornbeck et al., 2015b)
are highlighted. B) Kinetic profiles of selected phospho-peptides following IL-2 or IL-15
stimulation. Each orange circle (IL-2) and blue square (IL-15) represent data from independent
biological replicates, and the line represents the average of the biological replicates. Orange and
blue asterisks represent adjusted t-test p-value < 0.05 for IL-2 and IL-15 AUC, respectively.

81

Figure 4. Post-translational modification signature enrichment analysis (PTM-SEA) reveals
significant enrichment of p90RSK family kinases involved in propagating NK-92 cell signaling.
A) PTM-SEA identified the p90RSK, RSK2, and the EGFR1 pathways as significantly (FDR <
0.01) upregulated in cells stimulated with either IL-2 or IL-15. Kinase set enrichment analysis
(KSEA) corroborated significant upregulation of the RAF1 and EGFR kinases substrates and
82
suggested downregulation of casein kinase (CK) II and GSK3/ERK substrates (FDR < 0.01). B)
Rank-rank plots of both the RSK2 and p90RSK/RPS6KA1 kinase signatures reveal phospho-
peptides identified in this dataset which overlap with those upregulated in the signature.

83

Figure 5. Fuzzy C-means clustering reveals distinct kinetic profiles of signaling downstream of
IL-2 and IL-15. A) The top 10% most up- and downregulated phospho-peptides by absolute
value of AUC (n=724) were subjected to fuzzy C-means clustering. 516 phospho-peptide
trajectories were assigned into 6 distinct clusters (i.e., membership score > 0.7). Phospho-
84
peptides corresponding to p90RSK (RSP6 pS236 & pS240) and EGFR1 pathways (STAT3
pS727) fell into cluster 1. The table shows the number and distribution of IL-2 and IL-15
stimulated phospho-peptides that were assigned to each cluster.

85

Figure 6. p90RSK signaling is required for IL-2- and IL-15-mediated proliferation of NK-92
cells. NK-92 cells were treated with the RSK inhibitors LJI308 (20µM), BI-D1870 (2 and 10
µM), or DMSO control, and cell growth was monitored for 4 days. Both RSK inhibitors
significantly reduced cell growth compared to the DMSO control (***, Student’s t-test p-value <
0.001).

86

Figure 7. Workflow for phospho-proteomic analysis of CAR signaling in CAR-T cells stimulated
with antigen-expressing cancer cells. To activate CAR signaling, CAR-T cells are co-cultured
with antigen-expressing cancer cells. To preserve protein phosphorylation, the cell mixture is
then fixed with formalin. Next, CAR-T cells are separated from cancer cells by MACS after
labeling with a biotinylated antibody that recognizes the CAR and streptavidin-coated magnetic
Dynabeads. The MACS eluate is then lysed and formalin crosslinks are reversed using an SDS
lysis buffer. Using an S-trap, SDS is removed from the lysate, and proteins are digested to
peptides with trypsin. Lastly, phospho-peptides are purified with TiO2 beads and analyzed by
LC-MS.
87

88
Figure 8. Anti-CD19 CAR-T cells kill CD19-expressing SKOV3 cells. A) Following expansion
from human PBMCs, T cells were either not transduced or transduced with a retroviral vector
encoding a second-generation CAR against CD19. Following freeze-thaw, flow cytometry with
an antibody that recognizes the anti-Fab region of the CAR demonstrated that 48% of transduced
T cells expressed the CAR. B) CAR-T cells were mixed with non-transduced SKOV3 cells
(SKOV3.NT) or CD19-expressing SKOV3 cells (SKOV3.CD19) at the shown effector-to-target
(E:T) cell ratios. Interferon (IFN)-gamma(γ) production was measured by flow cytometry 6
hours later. C) SKOV3.NT and SKOV3.CD19 cells were stained with CFSE and then mixed or
not mixed with CAR-T cells at a 1:1 E:T ratio. 24 h later, the percentage of dead SKOV3 cells
was measured by 7AAD staining using flow cytometry.

89

Figure 9. Formalin-fixation does not affect phospho-proteomics in T cells. A) The number of
phospho-peptides identified by LC-MS in formalin fixed and non-formalin fixed samples is
shown on a Venn diagram. Phospho-peptide recovery following fixation exceeded 93% in T
cells. B) Phospho-peptide signal intensities were similar for formalin fixed and non-fixed
samples (r = 0.76, p-value < 0.00001). Two biological replicates were used in each condition.

90

Figure 10. Assessing the purity of CAR-T cell isolation by MACS following co-culture with
cancer cells. A) Workflow for assessing the purity of CAR-T isolation from formalin-fixed co-
cultures of CAR-T cells with either SKOV3.CD19 or SKOV3.NT cells. B,C) Flow cytometry
plots of SKOV3.CD19 (B) and SKOV3.NT (C) cells before labeling with CFSE (upper left),
after labeling with CFSE (upper right), after 1:1 mixing with unlabeled CAR-T cells (lower left),
91
and after MACS elution (lower right). Formalin fixation was performed after 45 min of co-
culture. The purity of the isolated CAR-T cell eluates was >90%.

92

93
Figure 11. Phospho-proteomic analysis of the known CAR signaling network. CAR-T cells were
stimulated for 45 min with cancer cells that either expressed CD19 (SKOV3.CD19) or not
(SKOV3.NT). Co-cultures were formalin fixed, and CAR-T cells were isolated by MACS before
phospho-proteomic analysis. Three biological replicates were analyzed in technical duplicate for
each condition. A) Phospho-peptides identified in our dataset were matched to known CAR
signaling pathways adapted from Cell Signaling Technology (Cell Signaling Technology, n.d.).
Log2 fold change in phosphorylation levels of each phospho-peptide is indicated by the color of
the circle enclosing a “P” according to the legend. B) Heatmap of individual replicate values of
phospho-peptide levels for sites shown in A.
94

95
Figure 12. Analysis of CAR signaling following stimulation with antigen-expressing cancer
cells. CAR-T cells were stimulated for 45 min with cancer cells that either expressed CD19
(SKOV3.CD19) or not (SKOV3.NT). Co-cultures were formalin fixed, and CAR-T cells were
isolated by MACS before phospho-proteomic analysis. Three biological replicates were analyzed
in technical duplicate for each condition. A) Phospho-proteomic data were plotted on a volcano
plot to identify the most significant changes in phosphorylation levels. Phospho-peptides in grey
represent FDR-adjusted p-value < 0.05 comparing SKOV3.CD19 stimulation with SKOV3.NT
stimulation. Red (upregulated) and blue (downregulated) phospho-peptides have known
regulatory roles according the PhosphoSite Plus database (P. V. Hornbeck et al., 2015a). B) The
most significantly changing phospho-peptides (FDR adjusted p-value < 0.05 and log2 fold
change >3 or <-2) were visualized on a heatmap to assess reproducibility. C) PTM-SEA
identified significantly upregulated kinase, perturbation, and pathway signatures in CAR-T cells
stimulated with CD19-expressing cancer cells. D) Rank-rank plots of significantly enriched
kinase signatures identified by PTM-SEA.

96
Supplemental Figures and Figure Legends

97

Supplementary Figure 1. Analysis of phospho-proteomic replicates identified outliers that were
excluded. A) Log2-transformed peptide intensities were evaluated for all phospho-peptides in the
dataset. Outliers, denoted by red asterisks, were removed. B) PCA scores plot of phospho-
peptide intensities confirmed that four of the five outliers identified clustered differently from
other replicates.

98

Supplemental Figure 2. IL-2 and IL-15 induce proliferation but not IFN 𝛾 production in NK-92
cells. A) NK-92 cells were grown in media containing either IL-2 or IL-15 at a concentration of
100 U/mL. Cells were counted every fourth day by trypan blue exclusion, and the cell number
was normalized to the initial cell number. Data from two independent biological replicate
experiments. B) IFNγ production was measured by flow cytometry following starved NK cell
99
treatment with either IL-2 or IL-15 in biological replicate. Treatment with PMA and ionomycin
was used as the positive control (posCTRL) while no stimulation was maintained as the negative
control (negCTRL).

100

Supplemental Figure 3. Reproducibility analysis of individual samples. The Pearson correlation
coefficient was calculated for all sample pairs using the log2-transformed, median normalized
phospho-peptide intensities, and then the correlation coefficients were plotted on a heatmap
according to the white-to-red color scale.
101

Supplemental Figure 4. Western blotting confirmed upregulation of STAT5A Y694 / STAT5B
Y699 phosphorylation by IL-2/IL-15. NK-92 cells were starved of cytokines for 6 h, then
stimulated with 100 U/mL of IL-2 or IL-15 before cell lysis at the indicated times.
Phosphorylation of STAT5A Y694 / STAT5B Y699 was strongly induced by cytokine
stimulation. Actin was used as an equal loading control.

102

Supplemental Figure 5. Histograms of IL-2 and IL-15 phospho-proteomic AUC values. For both
IL-2 and IL-15, positive and negative AUC values were evenly balanced. For IL-2, mean AUC =
-0.1, median AUC = -0.6, and 48.2% of phospho-peptides with AUC > 0. For IL-15, mean AUC
= 0.1, median AUC = -1.6, and 46.0% of phospho-peptides with AUC > 0.
103

Supplemental Figure 6. Assessing the purity of CAR-T cell isolation by MACS following co-
culture with GFP-labeled cancer cells. A) Workflow for assessing the purity of CAR-T isolation
from formalin-fixed co-cultures of CAR-T cells with either SKOV3.CD19 or SKOV3.NT cells.
B,C) Flow cytometry plots of SKOV3.CD19 (B) and SKOV3.NT (C) cells before labeling with
GFP (upper left), after labeling with GFP (upper right), after 1:1 mixing with unlabeled CAR-T
cells (lower left), and after MACS elution (lower right). Formalin fixation was performed after
45 min of co-culture. The purity of the isolated CAR-T cell eluates was >90%.

104

Supplemental Figure 7. Histogram of log2 fold change in phosphorylation in CAR-T cells
stimulated with SKOV3.CD19 or SKOV3.NT cells. Anti-CD19 CAR-T cells were mixed at a 1:1
ratio with either SKOV3.CD19 or SKOV3.NT cells. After 45 min, co-cultures were formalin
fixed, and CAR-T cells were purified by MACS, followed by lysis and protein de-crosslinking,
tryptic digestion, TiO2 phospho-peptide enrichment, and LC-MS proteomics (n = three
biological replicates in each condition). The log2 fold change in phosphorylation comparing
stimulation with SKOV3.CD19 to SKOV3.NT cells was calculated for 2,326 quantified
phospho-peptides and plotted on a histogram.

105
Supplemental Figure 8. Comparison of significantly changing phosphorylation sites with Salter
et al. Significantly changing phosphorylation sites from our data (Fig. 12, n = 367) were
compared with significantly changing phosphorylation sites from Salter et al. (n = 1,189) (Salter
et al., 2018). Both data sets were collected from T cells expressing a CD28/CD3ζ CAR after 45
min of stimulation with either CD19-expressing cancer cells (our data) or microbeads coated
with an antibody that recognizes a Strep-tag II (STII) domain on the extracellular hinge of the
CAR (Salter et al.). The overlap between the two data sets was significant (r = 0.38, n = 45, p =
0.01).

Abstract (if available)

Abstract The process of developing improved immunotherapies through the incorporation of antigen-targeting chimeric antigen receptors (CARs) has been particularly successful in the treatment of non-solid malignancies, such as leukemia and lymphoma. These CARs incorporate an antigen-targeting small chain variable fragment (scFv) derived from an antibody with T-cell signaling elements to enable a biomarker-specific immune cell. As such, CARs have been most successfully implemented in T cells, with multiple CAR-T cells receiving FDA approval since 2017. Despite these successes, CAR-T cells remain poorly effective against solid tumors. Further, mechanisms of CAR activation have not been fully characterized, complicating efforts to improve CAR designs for better targeting and activation mechanisms. CAR-NK cells, which implement CARs designed for T cells into an alternative immune cell platform, the natural killer (NK) cell, have seen some improvements over CAR-T cell therapies, but are also limited by a lack of CAR optimization based on NK signaling events. While some studies have characterized CAR signaling through stimulation of CAR-T cells by antigen-bound beads or through stable isotope labeling of amino acids in culture (SILAC), these methods are either limited in the establishment of a realistic cell-cell interaction or in terms of cost and time. As such, we sought to use phospho-proteomics to accomplish two main objectives: 1) establish a better understanding of NK canonical signaling events to aid in the development of CARs for this platform, and 2) develop an inexpensive and straightforward method for analyzing CAR signaling from cell-cell interactions.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Improving antitumor efficacy of chimeric antigen receptor-engineered immune cell therapy with synthetic biology and combination therapy approaches

PDF

Engineering chimeric antigen receptor-directed immune cells for enhanced antitumor efficacy in solid tumors

PDF

Engineering therapeutics for the improved antitumor efficacy of chimeric antigen receptor T cell therapy

PDF

Enhancing chimeric antigen receptor-engineered immune cell therapy with synthetic biology and nanomedicine

PDF

Mechanistic model of chimeric antigen receptor T cell activation

PDF

Dendritic cell-specific vaccine utilizing antibody-mimetic ligand and lentivector system

PDF

Enhancing the specificity and cytotoxicity of chimeric antigen receptor Natural Killer cells

PDF

T cell regulation of HLA-DR

PDF

Engineering lentiviral vectors for gene delivery and immunization

PDF

Insights from a computational approach to natural killer cell activation

PDF

Engineered CAR-T cells for treatment of solid cancers

PDF

Combination therapy for solid tumor

PDF

Self-secretion of checkpoint blockade enhances antitumor immunity by murine chimeric antigen receptor-engineered T cells

PDF

Novel design and combinatory therapy to enhance chimeric antigen receptor engineered T cells (CAR-T) efficacy against solid tumor

PDF

The effect of tumor-mediated immune suppression on prostate cancer immunotherapy

PDF

Development of TCR-like antibody and novel chimeric antigen receptor for cancer immunotherapy

PDF

Development of an NFAT-GFP Jurkat T cell reporter system for acceleration of T Cell receptor affinity screening

PDF

Phosphorylation kinetics of T-cell receptor signaling domain in vitro

PDF

Immunotherapy of cancer

PDF

Engineering chimeric antigen receptor (CAR) -modified T cells for enhanced cancer immunotherapy

Asset Metadata

Creator MacMullan, Melanie Amber (author)

Core Title Phospho-proteomic analysis of immune cell activation

School Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Chemical Engineering

Degree Conferral Date 2022-05

Publication Date 04/16/2022

Defense Date 03/10/2022

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

Tag CAR-T cells,cellular activation,cellular signaling,immunology,immunoproteomics,immunotherapies,OAI-PMH Harvest,phospho-proteomics

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Graham, Nicholas (committee chair), Lee, Jerry (committee member), Wang, Pin (committee member)

Creator Email macmulla@usc.edu,melanie.macmullan@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC110964827

Unique identifier UC110964827

Document Type Dissertation

Format application/pdf (imt)

Rights MacMullan, Melanie Amber

Type texts

Source 20220416-usctheses-batch-926 (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. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.

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