Phosphorylation Site Dynamics of Early T-cell Receptor
Signaling
Lily A. Chylek1,2,3., Vyacheslav Akimov4., Jörn Dengjel5., Kristoffer T. G. Rigbolt5¤a, Bin Hu1,6¤b,
William S. Hlavacek1,2,6*, Blagoy Blagoev4*
1 Theoretical Division, Los Alamos National Laboratory, Los Alamos, New Mexico, United States of America, 2 Center for Nonlinear Studies, Los Alamos National
Laboratory, Los Alamos, New Mexico, United States of America, 3 Department of Chemistry and Chemical Biology, Cornell University, Ithaca, New York, United States of
America, 4 Department of Biochemistry and Molecular Biology, University of Southern Denmark, Odense M, Denmark, 5 Department of Dermatology, Medical Center;
Freiburg Institute for Advanced Studies (FRIAS); BIOSS Centre for Biological Signalling Studies; ZBSA Center for Biological Systems Analysis, University of Freiburg,
Freiburg, Germany, 6 Department of Biology, University of New Mexico, Albuquerque, New Mexico, United States of America
Abstract
In adaptive immune responses, T-cell receptor (TCR) signaling impacts multiple cellular processes and results in T-cell
differentiation, proliferation, and cytokine production. Although individual protein–protein interactions and phosphorylation events have been studied extensively, we lack a systems-level understanding of how these components cooperate to
control signaling dynamics, especially during the crucial first seconds of stimulation. Here, we used quantitative proteomics
to characterize reshaping of the T-cell phosphoproteome in response to TCR/CD28 co-stimulation, and found that diverse
dynamic patterns emerge within seconds. We detected phosphorylation dynamics as early as 5 s and observed widespread
regulation of key TCR signaling proteins by 30 s. Development of a computational model pointed to the presence of novel
regulatory mechanisms controlling phosphorylation of sites with central roles in TCR signaling. The model was used to
generate predictions suggesting unexpected roles for the phosphatase PTPN6 (SHP-1) and shortcut recruitment of the actin
regulator WAS. Predictions were validated experimentally. This integration of proteomics and modeling illustrates a novel,
generalizable framework for solidifying quantitative understanding of a signaling network and for elucidating missing links.
Citation: Chylek LA, Akimov V, Dengjel J, Rigbolt KTG, Hu B, et al. (2014) Phosphorylation Site Dynamics of Early T-cell Receptor Signaling. PLoS ONE 9(8):
e104240. doi:10.1371/journal.pone.0104240
Editor: Jon C. D. Houtman, University of Iowa, United States of America
Received May 22, 2014; Accepted July 7, 2014; Published August 22, 2014
Copyright: ß 2014 Chylek et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The authors confirm that all data underlying the findings are fully available without restriction. All relevant data are within the paper and its
Supporting Information files.
Funding: This work was supported by National Institutes of Health/National Institute of General Medical Sciences (NIH/NIGMS) grants P50GM085273 and
R01GM076570; US Department of Energy Contract DE-AC52-06NA25396 through the Los Alamos Center for Nonlinear Studies and the Laboratory-directed
Research and Development (LDRD) Program; the Excellence Initiative of the German Federal and State Governments through the Freiburg Institute for Advanced
Studies (FRIAS) in the School of Life Sciences–LifeNet, and the Center for Biological Signalling Studies (BIOSS); the Danish Council for Independent Research/
Natural Sciences; and the Lundbeck Foundation. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the
manuscript.
Competing Interests: The authors have declared that no competing interests exist.
* Email: wish@lanl.gov (WSH); bab@bmb.sdu.dk (BB)
. These authors contributed equally to this work.
¤a Current address: Gubra, Hørsholm, Denmark
¤b Current address: SRA International, Atlanta, Georgia, United States of America
A useful reasoning aid is a mechanistic model, meaning a model
in which information about molecular interactions is cast in a form
that enables simulations consistent with physicochemical principles. Simulation of such a model reveals the logical consequences
of the collected knowledge upon which the model is based.
Comparisons of model simulations to experimental measurements
can drive discovery through generation of hypotheses and
identification of knowledge gaps [7].
Successful integration of modeling and experimentation depends on both approaches having compatible and relevant levels
of resolution. Phosphorylation dynamics can be elucidated using
several high-throughput techniques, including reverse-phase protein arrays [8], micro-western arrays [9], and mass spectrometry
(MS) [10]. MS-based techniques can yield quantitative information about the abundance of proteins phosphorylated at specific
amino acid residues without reliance on availability of phospho-
Introduction
Protein phosphorylation is a fundamental part of cellular
information processing, with a role in controlling numerous
physiological functions, including immune defenses [1]. Links
between dysfunctional regulation of phosphorylation and disease
underscore the need to elucidate underlying regulatory mechanisms [2]. To this end, phosphorylation-dependent signaling
networks have been investigated extensively, largely in studies
targeting individual proteins and interactions. However, cell
signaling is marked by features, such as feedback and feedforward
loops [3,4], parallel pathways [5], and crosstalk [6], which may
only be apparent when a network is studied as a whole. For this
reason, multiplexed measurements of phosphorylation dynamics
are needed, paired with reasoning aids for interpretation of these
data.
PLOS ONE | www.plosone.org
1
August 2014 | Volume 9 | Issue 8 | e104240
Phosphorylation Site Dynamics of Early TCR Signaling
changes in the levels of phosphorylation at individual tyrosine
residues in response to TCR stimulation with anti-CD3, antiCD28, and secondary antibodies (Fig. 1, A to C; see Materials and
Methods). Three independent experiments were performed,
resulting in the identification of over 700 unique pTyr sites, of
which over 500 were detected in multiple experiments, with
significant correlation across experiments (Fig. 1, D and E; Table
S1 in File S1). Possible sources of variability in quantification and
detection are mentioned in the Materials and Methods section.
The level of phosphorylation for each site was quantified at 5, 15,
30 and 60 s of TCR/CD28 stimulation, relative to the
corresponding level in unstimulated cells. These experiments
targeted a period of signaling that has thus far been largely
uncharacterized using MS-based proteomics or traditional biochemical assays, which have mostly been used at later timepoints.
Our measurements map in unprecedented detail the earliest
intracellular events and reveal that even within the first minute of
TCR/CD28 co-stimulation, dramatic and diverse biochemical
changes occur within the cell, preparing the ground for later
events. To analyze these data, we took a knowledge-based/modelguided approach, which is summarized in Fig. S1A in File S2.
Regulated changes in phosphorylation ($2-fold increases or
decreases) occurred as early as 5 s after stimulation, with the
number of regulated sites increasing to 138 after 60 s of
stimulation (Fig. 1F). Time courses of phosphorylation fall into
four distinct clusters, which reveal that the abundance of some
phosphopeptides increase, others decrease, and some changes
occur earlier than others (Fig. 1G). These results clearly demonstrate that even within the first 60 s of TCR stimulation there are
diverse patterns of phosphorylation dynamics. Regulated sites map
to proteins with various cellular functions, including pivotal
signaling factors such as receptors, adapter proteins, phospholipases, phosphatases and kinases from multiple distinct kinase
families. In the group of sites showing rapid dynamics we find wellestablished TCR signaling proteins such as LCK, LAT, PLCG1
among many others (Fig. 1H; Figs. S2 and S3 in File S2; Table S1
in File S1). These results attest to rapid, multi-functional signaling
downstream of the TCR, consistent with the known diversity of
pathways that emanate from the receptor [23].
Indeed, subsequent enrichment analysis (Fig. S2 in File S2)
revealed that among the proteins with detected phosphorylation
changes, the most frequent pathway association was with the TCR
pathway. At the same time, other pathways, such as those
influencing metabolism and protein synthesis, were also detected.
These results suggest that TCR signaling may influence these
general cellular functions quickly, consistent with evidence that T
cells make committed decisions within 60 s of antigen contact
[28].
site-specific antibodies [11], and measurements can be made with
fine time resolution [12], which is needed to decipher the order of
phosphorylation events. Thus, MS-based proteomics has the
potential to make unique contributions to systems biology
modeling [13].
However, modeling and proteomics have not yet become tightly
integrated, in part because of the technical challenges of
constructing and parameterizing a model with sufficient detail
and scope to be used for analysis of proteomic data. Proteomic
measurements give information about phosphorylation levels at
specific amino acid residues (sites); thus, a compatible model
requires similar site-specific resolution. For this task, traditional
modeling approaches (e.g., ordinary differential equations) can be
difficult or impossible to apply [14], which has catalyzed
development of the specialized techniques of rule-based modeling
[15]. Rule-based models make it possible to simulate site-specific
biomolecular interactions in a manner consistent with physicochemical principles.
Rule-based modeling has been used to study several immunoreceptor signaling systems [16,17,18,19,20], although in each case,
the scope of the model has been restricted to a handful of signaling
readouts. Development of models with sufficient scope to connect
to proteomic data has faced additional challenges; large models
can be costly to simulate and the complexity of the model can
hinder communication of the model’s content. To overcome these
obstacles, simulation techniques for large models [21] and
methods for model annotation and visualization [22] have recently
been developed. Although these modeling capabilities have been
demonstrated to a limited extent, use of large models to decode
high-content data, generate hypotheses, and drive the discovery of
biological insights has thus far remained uncharted territory.
We have developed a new approach for characterizing signal
initiation using a rule-based model to interpret temporal
phosphoproteomic data. We have applied this approach to study
initiation of T-cell receptor (TCR) signaling, which is an essential
process in the adaptive immune response [23]. The TCR and
related antigen recognition receptors transmit signals that are
dependent on site-specific details. These receptors are characterized by immunoreceptor tyrosine-based activation motifs (ITAMs),
which each contain two tyrosine residues that can be phosphorylated. It has been found that the specific phosphoform of an
ITAM can determine whether activating or inhibitory signals are
transmitted [24]. Additionally, TCR signal initiation relies on the
kinase LCK, which can be phosphorylated at a minimum of three
sites: phosphorylation of two of these sites (Y394 and Y505) have
opposing influences in regulating kinase activity [25], and
phosphorylation of the third site (Y192) regulates the affinity of
the SH2 domain [26]. These examples underscore the need to
investigate the site-specific dynamics of immunoreceptor signaling
[27].
Our results 1) characterize early TCR signaling with finer time
resolution than previous proteomic studies of this system, 2) reveal
mechanisms primarily operative on short timescales immediately
after stimulation, and 3) demonstrate how mechanistic modeling
and high-content measurements can be combined to develop a
predictive understanding of cellular information processing.
Dynamical modeling drives identification of knowledge
gaps
To investigate regulation of pTyr sites with well-characterized
roles in TCR signaling, we developed a computational model
based on principles of chemical kinetics and known proteinprotein interactions (Fig. 2, Supplementary Text S1 in File S3,
Fig. S4 in File S2, Table S2 in File S1, and Files A and B in File
S1). The model, formulated in terms of local rules for interactions
[15], accounts for 10 proteins containing 16 pTyr sites detected
experimentally (Table S3 in File S2) and seven additional proteins
linked to their regulation. The pTyr sites included in the model
belong to three classes (Fig. S5 in File S2): 1) sites phosphorylated
without dependence on prior receptor phosphorylation, 2) sites
phosphorylated after receptor phosphorylation, and 3) sites that
are dephosphorylated. Model-guided analysis of these pTyr site
Results
Immediate and extensive reshaping of the T-cell
phosphoproteome
To characterize the first minute of TCR signaling, we
performed time-resolved quantitative phosphoproteomic experiments, which allowed direct and accurate measurements of the
PLOS ONE | www.plosone.org
2
August 2014 | Volume 9 | Issue 8 | e104240
Phosphorylation Site Dynamics of Early TCR Signaling
Figure 1. Phosphoproteomic analysis of TCR signal initiation. (A) T cells grown in SILAC media were stimulated with antibodies that crosslink
CD3 and CD28. Lysates from differentially labeled cells were processed as indicated and relative abundances of phosphopeptides were quantified. (B)
Flow system used to stimulate cells. (C) For each detected phosphopeptide, peak intensities in MS spectra were quantified to determine
phosphorylation levels at 5, 15, 30 and 60 s after stimulation relative to the phosphorylation level in unstimulated cells. Results from paired spectra
(top) were combined (bottom right). Peptides were identified by tandem MS (bottom left). (D) Measurement reproducibility. For each point, the y-axis
indicates the relative phosphopeptide abundance measured in one of three replicate experiments; the x-axis indicates the corresponding average. R
is Pearson’s correlation coefficient. (E) Venn diagram indicating the numbers of phosphopeptides detected in individual and different combinations
of replicate experiments. (F) Number of regulated pTyr sites (.2-fold change) at each indicated time point. (G) Results from clustering of time
courses. (H) Diverse proteins undergo regulated phosphorylation. Boxes represent proteins; each oval and residue number next to a box identifies a
regulated pTyr site and its cluster membership.
doi:10.1371/journal.pone.0104240.g001
pY163, pY181, and pY417 in PAG1, which interact with SFKs to
bring them into proximity of PAG1-bound CSK [31]. We also
detected increased phosphorylation of Y566 in PTPN6, which is a
substrate of LCK and is associated with positive regulation of
phosphatase activity [32]. PTPN6 Y566 is phosphorylated as
rapidly as ZAP70 Y493 (Fig. 2; cf. Fig. S6, G and K in File S2)
and PTPN6 is the only protein tyrosine phosphatase that we
observed to undergo regulated phosphorylation (Fig. 1H and
Table S1 in File S1), suggesting a role in signal initiation.
Incorporating PTPN6-mediated dephosphorylation of the sites
listed above into the model enabled the model to reproduce
measured time courses for these sites (Fig. 2 and Fig. S6 in File S2).
dynamics suggested that the phosphatase PTPN6 plays a positive
role in TCR signaling (Fig. S1B in File S2) and that WAS is
initially activated via a previously unappreciated pathway (Fig.
S1C in File S2).
Dephosphorylation of inhibitory pTyr sites
Involvement of a phosphatase in initiating TCR signaling was
suggested by rapid dephosphorylation of six potentially inhibitory
pTyr sites (Fig. 2 and Fig. S6 in File S2): 1) pY192 in the LCK
SH2 domain, which reduces SH2-pTyr affinity [26]; 2) pY299 in
DOK2, which binds RASA1 (p120 RasGAP), a negative regulator
of RAS [29]; pY449 in DOK1, which binds CSK, which
phosphorylates LCK and other SRC-family kinases (SFKs) at a
C-terminal tyrosine residue to facilitate autoinhibition [30]; and 4)
PLOS ONE | www.plosone.org
3
August 2014 | Volume 9 | Issue 8 | e104240
Phosphorylation Site Dynamics of Early TCR Signaling
Figure 2. Model for TCR signaling. Proteins considered in a rule-based model for TCR signaling are represented by rounded boxes. Separate
boxes indicate the phosphosites considered in the model. Sites detected in phosphoproteomic experiments are each associated with a pair of
heatmaps, in which the upper heatmap reflects averaged experimental measurements of relative pTyr abundance and the lower heatmap reflects
simulated phosphorylation levels at matching time points. The color scale of each heatmap is unique: black represents the lowest and green
represents the highest level of phosphorylation for that site. Interactions are represented by arrows according to the conventions illustrated at
bottom. The number in the lower right corner of a protein box represents the number of components of the protein (domains, motifs, and/or pTyr
sites) considered in the model.
doi:10.1371/journal.pone.0104240.g002
sustained, as LAT phosphorylation in PTPN6 KD cells at 60 s is
less than in normal cells.
To further test our model, we performed an in vitro
phosphatase assay in which LCK was immunoprecipitated from
PTPN6 KD cells and then incubated with purified recombinant
PTPN6. We found that both Y192 and Y505 sites on LCK
became dramatically less phosphorylated when incubated with
PTPN6 compared to the mock treated sample (Fig. S8 in File S2).
This finding is consistent with our model, in which pTyr sites
observed to undergo net loss of phosphorylation, such as LCK
Y192, are assumed to be substrates of PTPN6.
Prediction and tests of a positive role for PTPN6 in early
signaling
We simulated the effect of lowered PTPN6 abundance. The
model predicted dampening of stimulatory phosphorylation and
enhancement of inhibitory phosphorylation, including increased
phosphorylation of LCK Y192 (Fig. 3B), sustained phosphorylation of the C-terminal tyrosine (Y505) of LCK (Fig. 3C), decreased
phosphorylation of Y493 in the activation loop of ZAP70
(Fig. 3D), and decreased LAT phosphorylation (Fig. 3E). According to the model, these simulation results arise from disruption of
PTPN6-mediated positive feedback loops (Fig. 3K and Fig. S1, D
and E in File S2).
To test the prediction that PTPN6 positively regulates TCR
signaling, we used RNAi to knockdown PTPN6 (Fig. 3F, bottom).
Expression of PTPN11 (SHP-2), a phosphatase that is closely
related to PTPN6, was unaffected by PTPN6 knockdown, attesting
to the specificity of the knockdown (Fig. S7 in File S2). We then
used phosphosite-specific antibodies to monitor phosphorylation of
LCK, ZAP70, and LAT in normal and PTPN6 KD cells after
TCR/CD28 co-stimulation (Fig. 3, G and H). Sustained phosphorylation of LCK Y192 and Y505 (Fig. 3, G and H, bottom)
and decreased phosphorylation of ZAP70 Y493 and LAT Y191
(Fig. 3, I and J, bottom) were observed, in qualitative agreement
with model predictions (Fig. 3, B to E and G to J, top). At 60 s
after stimulation, ZAP70 phosphorylation is similar in normal and
PTPN6 KD cells (Fig. 3I, bottom), indicating that the positive
early effect of PTPN6 on ZAP70 phosphorylation is transient. In
contrast, the effect on LAT phosphorylation is evidently more
PLOS ONE | www.plosone.org
A shortcut pathway connects the TCR to WAS activation
A second novel mechanism of TCR signal initiation was
suggested by fast phosphorylation of WAS Y291 (cf. Fig. S6, G and
L in File S2). WAS can be recruited to the plasma membrane by
the adaptor protein NCK1 through a pathway dependent on LAT
and LCP2 (SLP-76) [33], which are substrates of ZAP70 [23]
(Fig. 4A). However, we observed that WAS is phosphorylated
before ZAP70 (Fig. 4C): the fold-change in WAS pY291 at 5 s is
significantly greater than the fold change in ZAP70 pY493
(p = 0.019, one-tailed t-test). Thus, we reasoned that NCK1 may
be present at the plasma membrane prior to ZAP70 activation,
presumably through binding of its N-terminal SH3 domain to a
proline-rich sequence (PRS) in CD3E, which takes place in the
absence of TCR phosphorylation [34]. Adding this interaction to
the model created a shortcut pathway to WAS activation (Fig. 4B)
4
August 2014 | Volume 9 | Issue 8 | e104240
Phosphorylation Site Dynamics of Early TCR Signaling
Figure 3. PTPN6 mediates positive feedbacks. (A to E) Model-predicted cumulative phosphorylation of the indicated pTyr sites in normal (WT)
and PTPN6 KD cells. The cumulative phosphorylation of a site was calculated as the area under the corresponding time course of phosphorylation (0
to 60 s). Area is normalized to WT cells. (F to J) Simulation results (top) and immunoblots (bottom) showing the predicted and measured effects of
PTPN6 KD on pTyr site dynamics. PTPN6 KD was modeled by setting the copy number of PTPN6 to 0. Simulated time courses are visualized as series
of dots whose areas are proportional to relative phosphorylation levels. For each pTyr site, phosphorylation levels are normalized by the level of
phosphorylation in unstimulated WT cells. Note that WT time courses present results shown previously in Fig. 2. IB, immunoblot; Quant.,
quantification; WCL, whole-cell lysate; Sim., simulation. (K) Hypothesized positive feedback loops involving PTPN6 incorporated in the model for TCR
signaling. In these loops, LCK phosphorylates and activates PTPN6, and PTPN6 dephosphorylates sites that contribute to negative regulation of LCK.
Thus, PTPN6 has a positive effect on phosphorylation events downstream of LCK, including LCK-mediated phosphorylation of ZAP70 and ZAP70mediated phosphorylation of LAT. Blots are representative of the results from multiple (at least two) experiments. Each repeated immunoblot
measurement is characterized by a coefficient of variation (CV) below 0.25, where CV is estimated as the ratio of the sample standard deviation to the
sample mean.
doi:10.1371/journal.pone.0104240.g003
and enabled simulated phosphorylation of WAS to precede
phosphorylation of ZAP70 (Fig. 2).
To confirm shortcut activation of WAS, we used RNAi to
knockdown LCP2 (Fig. 4D), which mediates WAS activation as
part of the well-characterized pathway of Fig. 4A. We found that
the early association of NCK1 with pTyr sites is unaffected by
LCP2 KD (Fig. 4E). Moreover, WAS phosphorylation is not
substantially reduced in LCP2 KD cells (Fig. 4F), consistent with
LCP2-independent phosphorylation of WAS. In contrast, phosphorylation of Y783 in PLCG1, which is dependent on LCP2
[23], is ablated in LCP2 KD cells (Fig. 4G).
WAS phosphorylation is dependent on LCP2 at times beyond
the first minute of signaling [35], so we assume that the shortcut
pathway to WAS activation is transient, which is consistent with
the unusual signaling role of the CD3E PRS. Because this PRS
overlaps Y188 in the CD3E immunoreceptor tyrosine-based
activation motif (ITAM), phosphorylation of Y188 inhibits SH3PRS binding and enables SH2-pTyr binding [36]. Model
simulations indicate that the shortcut pathway to WAS activation
is deactivated by ITAM phosphorylation (Fig. S9A in File S2) as
the LCP2-dependent pathway is coordinately engaged (Fig. S9B in
PLOS ONE | www.plosone.org
File S2). Engagement of LCP2-dependent pathways for WAS and
PLCG1 activation is supported by immunoblot measurements of
LCP2 phosphorylation (Fig. S9, C and D in File S2).
Discussion
This study of pTyr site dynamics has revealed processes that
have been systematically overlooked in the past because of the
speed with which they occur. We have monitored the phosphosite
dynamics of early TCR signaling with finer temporal resolution
than in previous proteomic studies of TCR signaling (see Table S4
in File S2 and references cited therein) and with greater breadth
than earlier studies of early TCR signaling events employing
relatively low-throughput assays [37,38], and we developed a
mechanistic model for TCR signaling that reproduces measured
time courses of phosphorylation for a greater number of specific
sites than previously developed models for immunoreceptor
signaling (see Table S5 in File S2 and references cited therein).
We detected over 100 pTyr sites that undergo greater than twofold changes in abundance during the first minute of TCR
signaling. Even on these short timescales, time courses show
5
August 2014 | Volume 9 | Issue 8 | e104240
Phosphorylation Site Dynamics of Early TCR Signaling
Figure 4. WAS activation. (A) Long pathway for WAS recruitment to the plasma membrane. Phosphorylated CD247 recruits ZAP70, which
phosphorylates LAT. Phosphorylated LAT binds the GRAP2 SH2 domain. The GRAP2 SH3 domain binds LCP2. Phosphorylated LCP2, a substrate of
ZAP70, binds the SH2 domain of NCK1/2. The C-terminal SH3 domain in NCK1/2 binds a proline-rich sequence (PRS) in WAS. (B) Short pathway for
WAS recruitment. The N-terminal SH3 domain in NCK1/2 binds a PRS in CD3E, and the NCK1/2 C-terminal SH3 domain binds WAS. (C) Comparison of
measured time courses of phosphorylation for WAS Y291 and ZAP70 Y493. Data is scaled such that the phosphorylation level of each site is 1 at 60 s.
Error bars indicate standard deviations. (D) Efficiency of LCP2 KD. (E) Inducible association of NCK1 with pTyr-containing proteins in normal (WT) and
LCP2 KD cells. (F and G) Simulations (top) and immunoblots (bottom) showing the predicted and measured effects of LCP2 KD on phosphorylation of
WAS Y291 and PLCG1 Y783 upon TCR/CD28 co-stimulation for the indicated times. Simulation results are plotted as in Fig. 3, F to J. WT time courses
present results shown previously in Fig. 2. Abbreviations are as in Fig. 3. Blots are representative of the results from multiple (at least two)
experiments. As for Fig. 3, the estimated CV is less than 0.25 for each repeated immunoblot measurement.
doi:10.1371/journal.pone.0104240.g004
studies. For example, in vitro, PTPN6 has previously been found
to be capable of dephosphorylating the inhibitory C-terminal
tyrosine of LCK when the SH2 domain is deleted [39].
The view of PTPN6 as an overall negative regulator of TCR
signaling [40] has been based mostly on studies of the motheaten
mouse, which is deficient in Ptpn6 and suffers from severe
autoimmunity [41]. Recent work has hinted at a more nuanced
role. Studies of mice with a T-cell specific Ptpn6 deletion indicate
that loss of Ptpn6 in T cells does not lead to overt autoimmunity
[42], nor does it affect the number of memory precursor cells [43].
It has also been found that mechanisms controlling PTPN6
expression are distinct from those controlling other negative
regulators of TCR signaling [44]: levels of PTPN6 mRNA and
protein are not affected by the miR-181a microRNA, which
negatively regulates expression of multiple phosphatases linked to
suppression of TCR signaling. Thus, PTPN6 appears to be
somewhat enigmatic. Contributing to uncertainty about the
function of PTPN6 is an incomplete catalog of its substrates,
which is incomplete partly because known substrates do not match
an obvious consensus sequence [45]. Our findings, together with
those mentioned above, point to a need to identify signaling
proteins whose phosphorylation states are regulated by PTPN6,
and to characterize the function of this phosphatase in TCR
signaling under precisely controlled conditions.
distinct patterns: the abundances of some pTyr sites increase,
others decrease, and some changes occur sooner than others. The
proteins containing these sites map to diverse cellular functions
and include kinases, phospholipases, actin regulators, and transcription factors, many of which are known players in T-cell
activation. The significance of these results is that by 60 s, which in
many studies is taken as an early time point for measurement,
significant changes have already occurred within the cell.
We found that multiple putatively negative regulatory sites
(including sites in PAG1 and LCK) were rapidly dephosphorylated
as the PTPN6 phosphatase was phosphorylated at an activating
site. Inclusion of a mechanism causally linking these events allowed
our model to reproduce measured time course data and to
generate testable predictions. These predictions were validated
experimentally, giving credence to the hypothesized link between
PTPN6 activation and dephosphorylation of putatively inhibitory
pTyr sites. Our model predicted that loss of PTPN6 would result
in sustained phosphorylation of these pTyr sites, and reduction of
phosphorylation at other, activating sites (including sites in ZAP70
and LAT). These predictions were confirmed through RNAimediated knockdown of PTPN6 expression and immunoblot
measurements with phosphosite-specific antibodies. These results
provide strong motivation for future studies of the possible early
positive role of PTPN6, ideally in primary cells. We note that a
positive role for PTPN6 has been suggested by the results of earlier
PLOS ONE | www.plosone.org
6
August 2014 | Volume 9 | Issue 8 | e104240
Phosphorylation Site Dynamics of Early TCR Signaling
The results presented here suggest that, in Jurkat T cells,
PTPN6 (the human ortholog of Ptpn6) has an early positive effect
that accelerates signaling, before its negative effects become
dominant. The negative effects of PTPN6, such as dephosphorylation of the LCK activation loop [25], may serve to prevent
deleterious overshoots that would otherwise be caused by its
positive effects, in addition to setting the baseline level of TCR
signaling. As a participant in positive feedback loops, which can
act as amplifiers, PTPN6 may also contribute to regulation of Tcell sensitivity. Such a role has been suggested in earlier studies of
PTPN6 [32].
Several caveats are worth noting. Firstly, although we have
demonstrated that PTPN6 acts directly on LCK Y192 and Y505
in vitro, we have not conclusively determined if PTPN6 directly
acts on the sites that are observed to undergo dephosphorylation in
our proteomics experiments, or if instead PTPN6 influences some
or all of these sites in an indirect manner. Nonetheless, our
knockdown results indicate that PTPN6 positively influences
specific events in early signaling, and evidence for a more indirect
mechanism would not alter this finding. Secondly, the dephosphorylated sites may have roles that are multifaceted, rather than
strictly inhibitory. For example, phosphorylation of Y192 in LCK
may enhance kinase activity by limiting SH2 association with the
C-terminal phosphotyrosine that mediates autoinhibition, which is
a regulatory mechanism that may be operative in the case of Src
[46]. However, phosphorylation of LCK Y192 has been found to
have an overall negative effect on important readouts of TCR
signaling [26], indicating that impairing LCK’s ability to associate
with its binding partners outweighs potential enhancement of
kinase activity through relief of autoinhibition. Thus, it is apparent
that categorization of a protein or site as ‘‘positive’’ or ‘‘negative’’
is dependent on context and such categorization must be made
with caution. Finally, the results presented here are based on the
Jurkat T cell line, which has been a source of much of our current
knowledge of TCR signaling mechanisms and is amenable to MS
measurements. Use of a cell line was required to obtain the
quantities of proteins required for MS-based assays of protein
phosphorylation and to obtain the fine time resolution desired.
However, Jurkat T cells do not perfectly recapitulate the behavior
of T cells in vivo. Characterization of very early signaling
mechanisms in primary T cells poses significant technical
challenges and is beyond the intended scope of the present study.
The breadth and fine time resolution of our proteomic data
allowed us to determine the order in which events occur. One of
the fastest events observed was phosphorylation of the actin
regulator WAS, which surprisingly preceded activating phosphorylation of the kinase ZAP70. It has previously been reported that
WAS is recruited to the plasma membrane via a pathway
involving LAT and LCP2 (SLP-76) [33], which are activated
through ZAP70-dependent phosphorylation [23]. This mechanism
of WAS activation did not allow our model to reproduce the
observed WAS phosphorylation dynamics. In contrast, a previously unappreciated shortcut pathway, which is apparently active
only transiently, allowed the model to reproduce the data.
Experimentally, knockdown of LCP2 expression did not attenuate
the early WAS phosphorylation, consistent with model predictions
and the presence of an alternative pathway. These results indicate
that the flow of information through different pathways may shift
as signaling progresses. Furthermore, the shortcut pathway may
explain how the PRS in CD3E contributes to the ability of the
TCR to respond to a range of agonist molecules. The PRS in
CD3E and its interaction with NCK1 are known to be more
consequential for responses to weak agonists [47] than strong
agonists [48]. This difference may arise because weak agonists
PLOS ONE | www.plosone.org
tend to induce only partial TCR phosphorylation [49], allowing
longer-lasting NCK1-CD3E association. Although the interactions
forming the shortcut pathway have been characterized individually, their combined role in facilitating rapid WAS activation has
not hitherto been investigated. Thus, the results presented here
complement past work by suggesting a potential mechanism by
which the PRS of CD3E enables responses to weak agonists.
Our findings suggest that TCR signaling is initiated by proteins
that transition from positive to negative roles. This strategy
resembles bang-bang control [50], in which a controller assumes
extreme values. PTPN6 appears to switch TCR signaling ‘‘on’’
upon signal detection and ‘‘off’’ after a period of signal
transmission. Another apparent mediator of bang-bang control is
CD3E, which is initially ‘‘on’’ and provides a shortcut pathway to
WAS activation by recruiting NCK1 prior to receptor phosphorylation, but later is turned ‘‘off’’ as the CD3E ITAM is
phosphorylated. The advantage of bang-bang control, or mode
switching (a transition from positive to negative signaling by a
protein with both on and off functions), can be appreciated by
considering that a superior brake system decided the winner of the
1921 French Grand Prix by enabling fast approaches to turns [51].
The apparent on/off functions of PTPN6 and CD3E may allow a
T cell to initiate signaling events with maximal acceleration and
then avoid deleterious overshoots through application of a
molecular brake.
Physiologically, we speculate that bang-bang control of TCR
signal initiation may allow a T cell to launch rapid but controlled
responses to infection. T cells scan antigen-presenting cells quickly
and have been shown to decide if foreign antigen is present in
under 1 min [28]. Thus, the effect of a fairly short delay in
phosphorylation of LAT or WAS, for example, could potentially
have a major impact on the number of antigen-specific T cells
responding to an infection. Bang-bang control is operative in gene
circuits with negative autoregulation [52] and in stem cell
population dynamics [53] and may represent a widely used design
principle of cellular regulatory systems.
Materials and Methods
Cell culture
Jurkat T cells, clone E6-1 (ATCC TIB-152), were grown in
RPMI (Invitrogen) supplemented with penicillin/streptomycin
(100 U/ml, 100 mg/ml), 10% dialyzed fetal calf serum (Invitrogen), and one of three SILAC labels (Sigma-Aldrich, Denmark):
13
14
L-arginine and L-lysine (Arg0/Lys0); L-arginine- C6- N4 and
2
13
15
13
L-lysine- H4 (Arg6/Lys4); or L-arginine- C6- N4 and L-lysine15
C6- N2 (Arg10/Lys8) (Fig. 1A). Before stimulation of TCR
signaling, cells were serum starved for 16 h. Starved cells were
diluted with medium (RPMI supplemented with 10 mM HEPES)
to a density of 0.9–1.06108 cells/ml and stimulated with a 1:1
mixture of pre-crosslinked anti-CD3 antibody (clone HIT3a,
Santa Cruz) and anti-CD28 antibody (clone CD28.2, Santa Cruz)
(4 mg/ml in RPMI supplemented with 10 mM HEPES). The antiCD3 and anti-CD28 antibodies were crosslinked by incubation
with anti-mouse IgG (Dako) at 4uC overnight. Differentially
labeled cells (Fig. 1A) were stimulated for 0 s (Lys0/Arg0), 5 s
(Lys4/Arg6), and 30 s (Lys6/Arg10) using a qPACE setup
(Fig. 1B) as described earlier [12]. A second set of differentially
labeled cells were stimulated for 0 s (Arg0/Lys0), 15 s (Arg6/
Lys4), and 60 s (Arg10/Lys6) (2.7–3.06108 cells per condition).
Because of the large number of cells required to obtain sufficient
protein for LC-MS/MS analysis, we performed the three replicate
temporal phosphoproteomic experiments of Fig. 1 serially over a
7
August 2014 | Volume 9 | Issue 8 | e104240
Phosphorylation Site Dynamics of Early TCR Signaling
time frame of months, which likely contributed to the measurement variability illustrated in Fig. 1D.
nature, which is likely to explain the variability in detection of
peptides summarized in Fig. 1E. Detection of non-overlapping sets
of peptides from experiment to experiment is expected unless
coverage of the phosphoproteome is complete, which is difficult to
achieve.
Proteomics sample preparation
Cells were lysed in 8 M urea with 25 mM Tris, pH 8.0. Lysates
from differentially labeled and stimulated cells (three conditions, as
indicated in Fig. 1A) were mixed in a 1:1:1 ratio, centrifuged, and
reduced with 1 mM DTT for 40 min at 25uC, followed by
alkylation with 5.5 mM iodoacetamide for 40 min. Cell lysates
were subjected to Lys-C (Wako) digestion at a 1:100 enzyme/
protein ratio for 5 h at room temperature. The lysates were diluted
(4X) using 25 mM Tris, pH 8.0, and then digested with trypsin
(Promega) overnight at room temperature. Digested cell lysates
were acidified with TFA and desalted using Sep-Pak (Waters) in
accordance with the manufacturer’s instructions, followed by
lyophilization of the tryptic peptides.
Data analysis
Raw data files from three biological replicates were processed
using MaxQuant (version 1.0.13.13) as described earlier [56].
Briefly, peak lists were generated by the MaxQuant program using
the following search parameters: triple SILAC with heavy labels
Arg6/Lys4 and Arg10/Lys8; a maximum of two missed trypsin
cleavages; use of the six most intense peaks per 100 Da interval for
generation of MS/MS peak lists; and a mass tolerance of 7 ppm
on precursors and 0.5 Da (CID) for fragment ions. A fixed
modification was carbamidomethylation of cysteine (Cys, +
57.021464 Da) and variable modifications were oxidation of
methionine (+15.994915 Da), N-terminal protein acetylation (Nterminal, +42.010565 Da), and phosphorylation of serine, threonine and tyrosine (Ser/Thr/Tyr, +79.966331 Da). We used the
Mascot engine (v.2.3) (http://www.matrixscience.com) to search
the generated peak-lists files (*.msm) against the IPI database
(version 3.69) [57], which contains a list of frequently observed
contaminants, concatenated with reverse copies of all entries. The
acquired Mascot DAT files (*.dat) together with the raw files were
processed and quantified by MaxQuant using the following
parameters: the false discovery rate (FDR) for peptides, proteins
and sites of modifications was required to be below 1% as assessed
by the number of positive hits searched in the reverse database;
and minimum peptide length was set at 6. MaxQuant automatically calculated the localization probabilities of all tyrosine
phosphorylation sites as described earlier [10] and quantified
intensity/peptide abundance ratios for each individual phosphosite. All tyrosine phosphorylated peptides (MS scan spectra) were
manually inspected for arginine-proline conversion and each
peptide abundance ratio was normalized in accordance with the
number of proline residues in the corresponding peptide sequence.
The peptide containing Y394 in LCK was matched to a
miscleaved peptide that is unique to this protein.
Immunoprecipitation and purification of
phosphopeptides
The lyophilized peptides were subjected to immunoprecipitation using a PhosphoScan Kit (P-Tyr-100, catalog number 7900,
Cell Signaling Technology) and anti-phosphotyrosine antibody
(clone 4G10, catalog number 16-101, Millipore). Briefly, peptides
were dissolved in 2 ml of IAP buffer per experiment (50 mM
MOPS, pH 7.2; 10 mM sodium phosphate; and 50 mM NaCl),
refrigerated, centrifuged to remove undissolved peptides in the
pellet, and immunoprecipitated with 80–100 ml of the antiphosphotyrosine antibody resin for 3–4 h at 4uC. Beads were
washed three times with the IAP buffer and two times with a salt
solution (50 mM NaCl), and phosphopeptides were eluted using
0.15% TFA solution. Three sequential elutions were performed;
each time, the volume of the 0.15% TFA solution used was equal
to that of the bead volume. Eluted phosphopeptides were further
purified by using TiO2 spheres as described earlier [10]. We note
that it was not possible to use antibodies having the same lot
numbers for each of the replicate temporal phosphoproteomic
experiments of Fig. 1, which likely contributed to the measurement variability illustrated in Fig. 1D.
LC-MS/MS analysis
Bioinformatics analysis
Peptides were analyzed using LC-MS/MS as previously
described [54]. Eluted samples were dried almost to completeness
in a SpeedVac and analyzed using an LTQ-Orbitrap XL
instrument (Thermo Scientific), which was interfaced with an
Agilent 1100 nanoflow system (Agilent Technologies) and
equipped with a nano-electrospray ion source (Proxeon Biosystems). Phosphopeptides were injected into a fused silica column
packed in-house with 3 mm C18 beads (Reprosil, Dr. Maisch
HPLC) applying a 120 min gradient from 8 to 64% acetonitrile in
0.5% acetic acid at a flow rate of 250 nl/min. We operated the
Orbitrap XL in the data-dependent mode. The five most intense
ions after full scan survey (MS spectra for m/z from 350 to 1,600)
were subjected to MS/MS fragmentation using the CID activation
technique with the following settings: R = 60,000 (MS resolution),
a normalized collision energy of 35%, and an isolation window of
2.0 Th. In MS/MS acquisition, we used q = 0.25 (collision
endothermicity) and an activation time of 30 ms. Slightly different
settings were used for the third biological replicate: the range of m/
z was 300–2000 and multistage activation (MSA) was used. For all
MS runs, ions selected for fragmentation were dynamically
excluded for 45 s and lock mass ions were used for internal mass
calibration [55] to obtain constant mass accuracy during analysis.
We note that sampling of ions for MS/MS analysis is stochastic in
PLOS ONE | www.plosone.org
For bioinformatics analysis, the peptide abundance ratios
obtained from the three biological replicates were averaged. For
cases where a pTyr site was detected in multiple peptides, we
considered the least modified peptide when evaluating pTyr site
dynamics. Clustering of pTyr sites showing $2-fold dynamics at a
minimum of one time-point was performed using the fuzzy
c-means algorithm as implemented in GProX with default
parameters [58]. Enrichment analysis was also performed using
GProX by retrieving Gene Ontology (GO) [59] and Pfam [60]
annotations from the UniProt database [61] and testing for overrepresentation of terms in each cluster, which was assessed using
Fisher’s exact test. Only terms occurring at least two times in a
cluster and attaining a p-value less than 0.05 after correction for
multiple testing using the Benjamini and Hochberg algorithm
were regarded as significant. For identification of enriched
pathways, UniProt accession keys were uploaded to DAVID
[62] and analyzed with default parameters. Sequences of kinase
domains were aligned using Clustal W2 [63] and the resulting tree
file was visualized using the iTOL tool [64]. The ‘‘princomp’’
function of MATLAB was used for principal component analysis.
8
August 2014 | Volume 9 | Issue 8 | e104240
Phosphorylation Site Dynamics of Early TCR Signaling
are given in Table S2 in File S1. Finally, some parameters were
determined through fitting. These parameters were constrained
during fitting in one of two ways. Some were simply constrained to
have positive values. Others were constrained to have values
between specified lower and upper bounds, which were set on the
basis of various empirical considerations, which are noted in Table
S2 in File S1. Fitting was performed initially using a brute force
approach (coarse grid search), followed by targeted parameter
refinement using the variable metric method [68]. PTPN6 and
LCP2 knockdowns were modeled by setting the copy numbers of
PTPN6 and LCP2 to zero. The model visualization in Fig. S4 in
File S2 is drawn in accordance with established conventions [22].
Simulation results were visualized using MATLAB, version 7.10.0
(R2010a) (MathWorks, Natick, MA).
Modeling and simulation
A chemical-kinetics model for TCR signaling in a single cell was
formulated using a rule-based approach, which enabled concise
representation of biomolecular interactions and efficient simulation of multi-site phosphorylation [15]. The goal of model building
was to leverage available mechanistic knowledge to construct a
model that includes as many observed pTyr sites as possible. The
knowledge base of the model was developed through a dataguided literature search. Phosphorylation sites and the proteins
containing these sites were selected for inclusion in the model if
they were detected in the phosphoproteomic experiments, were
known to be involved in TCR signaling based on information in
the primary literature, and if they had a known kinase,
phosphatase, and/or binding partner. A second set of proteins
and sites were included if, based on published findings, they were
necessary for regulation of the sites detected in experiments.
Residue numbers were assigned for naming purposes in accordance with standard UniProt numbering. The proteins and
interactions included in the model are identified and discussed
in the Supplementary Text S1 in File S3. The initial model that we
constructed on the basis of available mechanistic knowledge was
deemed deficient because it was unable to reproduce the observed
dephosphorylation dynamics of four putatively inhibitory pTyr
sites (Fig. S6I, M–O in File S2) and also because it was unable to
reproduce the observed fast phosphorylation dynamics of WAS
(Fig. S6L in File S2). To address these shortcomings, we extended
the model to include the hypothetical mechanisms of Fig. 3K (Fig.
S1B, D, E in File S2) and Fig. 4B (Fig. S1C in File S2). Unlike
other aspects of the model, these mechanisms cannot be supported
by literature citations, which is why experimental tests of model
predictions focused on probing these aspects of the model. The
mechanism of Fig. 3K was initially suggested by detection of
activating phosphorylation of PTPN6 (Fig. 1H; Fig. S6K in File
S2). The mechanism of Fig. 4B was initially suggested by reports
in the literature about the interactions that comprise the shortcut
pathway to WAS activation, especially the interaction between
CD3E and NCK1 [34]. Rules for noncovalent interactions and
post-translational modifications (i.e., tyrosine phosphorylation and
dephosphorylation) were specified using BNGL, a domain-specific
language for formulating models of biochemical reaction kinetics
[65]. BioNetGen [65] was used to process File A in File S1 (a.bngl
file) to generate an XML encoding of the model, which served as
input for NFsim [21], together with File B in File S1 (a.rnf file). File
B in File S1 specifies simulation protocols, including an
equilibration procedure that served to find the unstimulated
steady state. NFsim implements a particle-based kinetic Monte
Carlo algorithm [66]. Thus, NFsim produces results that reflect
the stochastic nature of chemical reactions; it uses rules to generate
reaction events, which are selected to occur randomly. For this
reason and also because our model is formulated for a single cell,
whereas our experimental measurements correspond to averages
over a large population of cells, we performed multiple simulation
runs and the results were averaged to obtain smooth curves for
comparisons with the proteomic, population-averaged data.
NFsim simulation results were validated using a different
simulation tool, RuleMonkey [67]. Model parameters were
estimated in three ways, as indicated in Table S2 in File S1 (see
the footnotes). Some parameters were assigned values reported in
the literature, which were determined in one of two ways: in an
experimental study or in an earlier modeling study. Other
parameters were determined through simplifying assumptions or
physicochemical constraints. These parameters were assigned
values related to and determined by other parameter values; the
relationships between the independent and dependent parameters
PLOS ONE | www.plosone.org
DNA constructs for RNAi silencing and generation of
stable cell lines
A lentiviral vector pSicoR (Addgene plasmid 11579) [69] and
plasmids of the 3rd generation packaging system for producing
viral particles [70] were used: pMD2.G (Addgene plasmid 12259),
pMDLg/pRRE (Addgene plasmid 12251) and pRSV-Rev (Addgene plasmid 12253), which were obtained via Addgene’s
Material Transfer Agreement. These DNA plasmids were kindly
deposited in Addgene by Drs. Tyler Jacks and Didier Trono.
An EGFP cassette in the vector pSicoR was exchanged with a
puromycin resistance gene cassette, resulting in a modified
pSicoR-puro vector allowing puromycin resistance-based selection
of shRNA expressing cells. RNAi sequences potentially targeting
the PTPN6 and LCP2 transcripts were generated using available
Web resources (http://www.dharmacon.com) according to published recommendations for siRNA/shRNA design [71,72].
shRNA DNA constructs were designed using recommended
guidelines [69] and available Web resources. Briefly, the shRNA
sequences were synthesized (DNA Technology, Denmark) as two
complementary DNA oligonucleotides:
59-T(N19)TTCAAGAGA(rN19)TTTTTTC-39 and
59-TCGAGAAAAAA(N19)TCTCTTGAA(rN19)A-39
where N19 is the sense strand of the target sequence and rN19 is
the antisense strand. The oligonucleotides were annealed as
described earlier [72] and directly cloned into the vector pSicoRpuro. Clones were selected for verification by DNA sequencing.
We used the following targeting sequences for RNAi: 59GAGCATGACACAACCGAAT-39 for PTPN6 and 59-GGACCAGACAGAAGAGAGA-39 for LCP2. Verified DNA constructs
were used to produce lentiviral particles as described earlier [69]
with modifications. Briefly, 10 mg of lentiviral vector and 5 mg of
each packaging plasmid were co-transfected in one 15 cm dish of
HEK-293T cells using the transfection reagent METAFECTENE
(Biontex Laboratories) according to the manufacturer’s instructions. Supernatants were harvested 48 and 72 h after infection and
viral particles were concentrated by ultracentrifugation at 115,000
RCF for 2 h at 4uC. Viral stocks were diluted in cell culture media
and used for infection of Jurkat T cells to generate stable cell lines
expressing the described RNAi constructs. Cells stably expressing
shRNA sequences were grown in RPMI medium with 4 mg/ml of
puromycin for 5 days and used for qPACE-based co-stimulation of
TCR/CD28 signaling for immunoblot experiments. Depletion of
PTPN6 and LCP2 was confirmed by immunoblotting.
9
August 2014 | Volume 9 | Issue 8 | e104240
Phosphorylation Site Dynamics of Early TCR Signaling
Immunoblotting
Table S5 in File S2 includes citations of References [80,81,82,
83,84,85]. The Supplementary Text S1 (File S3) includes citations
of References [86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,
101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,
116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,
131,132,133,134,135,136,137,138,139,140,141,142,143,144,145,
146,147,148,149,150,151,152,153,154,155,156,157,158,159,160,
161,162,163,164,165,166,167,168,169,170,171,172,173,174,175,
176,177,178,179,180,181,182,183,184,185,186,187,188,189,190,
191,192,193,194,195,196,197,198,199,200,201,202,203,204,205,
206,207,208,209,210,211,212,213,214,215,216,217,218,219,220,
221,222].
Equal amounts of normal Jurkat T cells (WT) and Jurkat T cells
with stable knockdown of PTPN6 (PTPN6 KD) or LCP2 (LCP2
KD) were stimulated for either 0, 30, and 60 s or 0, 10, and 60 s
using our qPACE setup (Fig. 1B). Harvested cells were lysed using
ice-cold lysis buffer [modified RIPA buffer: 150 mM NaCl;
50 mM Tris, pH 7.5; 1% v/v NP-40; 1 mM EDTA; proteases
inhibitors (cOmplete Tablets, Roche); 1 mM sodium orthovanadate; 2 mM NaF; and 2 mM b-glycerophosphate]. The cell lysates
were centrifuged, mixed with 66Laemmli buffers and resolved on
Novex 4–12% Bis-Tris gradient gels (Invitrogen) using MES
running buffer followed by protein transfer to nitrocellulose
membrane, blocking with 5% BSA and incubation with primary
and HRP-conjugated secondary antibodies. To quantify western
blots, we used the Analyze Gels function in the ImageJ software
tool
(http://imagej.nih.gov/ij/docs/guide/user-guide.pdf).
Chemiluminescence was measured, and we considered different
exposure times to ensure that images were analyzed well before
saturation. Values for bands corresponding to site-specific
antibody staining were normalized using values for corresponding
total protein loading controls. The following antibodies were used
for western blots: Phospho-Lck (Tyr 505), Phospho-Zap-70 (Tyr
493), Phospho-LAT (Tyr 191), Zap-70, WASP, NCK1 (Cell
Signaling); LAT (Santa Cruz Biotechnology, Inc.); PhosphoWASp (Tyr 290) (Sigma-Aldrich); Phospho-Lck (Tyr 192)
(Abcam); and Shp-1, Shp-2, and Slp-76 mouse mAb (BD
Biosciences). Secondary anti-mouse and anti-rabbit antibodies
were obtained from GE Healthcare, UK.
Supporting Information
File S1 This Zip file combines Tables S1 and S2 (Excel
spreadsheets) and Files A and B (plain-text files). Table
S1 in File S1. Proteomic data. This Excel spreadsheet provides a
listing time courses and residue numbers of phosphopeptides
detected in each of three LC-MS/MS experiments. Table S2 in
File S1. Parameter estimates. This Excel spreadsheet provides a
listing parameter estimates used in the model for TCR signaling.
File A in File S1. Executable model specification. This plain-text
file provides an executable model specification, which can be
processed by BioNetGen. The filename extension should be
changed to ‘‘.bngl’’ for processing by BioNetGen. File B in File
S1. Simulation protocol. This plain-text file provides a definition
of a simulation protocol, which can be processed by NFsim. The
filename extension should be changed to ‘‘.rnf’’ for processing by
NFsim.
(ZIP)
In vitro dephosphorylation assay
Jurkat T cells with stable knockdown of PTPN6 were starved
for 16 hours and harvested by centrifugation. The cells were lysed
using ice-cold lysis buffer [modified RIPA buffer: 150 mM NaCl;
50 mM Tris, pH 7.5; 1% v/v NP-40; 1 mM EDTA; proteases
inhibitors (Complete tablets, Roche); 1 mM sodium orthovanadate; 2 mM NaF; and 2 mM b-glycerophosphate]. The cell lysate
was centrifuged; a supernatant was supplemented with the SDS up
to 0.5% and incubated for 30 minutes on ice. The cell lysate was
diluted with the lysis buffer up to 0.1% SDS. Mouse anti-LCK
antibody (6 mg) bound to Protein G beads was used for
immunoprecipitation of LCK for 5 hours at 4uC. The beads were
washed three times with the lysis buffer and five times with ice-cold
in vitro assay buffer (50 mM HEPES, pH 7.4, 2 mM DTT,
100 mM NaCl, 2 mM EDTA). Washed beads were divided into
two equal parts with 50 ml in vitro assay buffer. The first sample
was supplemented with 1 mg of an active human recombinant
protein PTPN6 (Millipore, cat. No. 14-591) and second sample
was vehicle treated. Both samples were incubated at 37uC for
30 minutes with gentle shaking. Thereafter, the samples were
mixed with 66 Laemmli buffers and resolved on Novex 4–12%
Bis-Tris gradient gels (Invitrogen) using MOPS running buffer
followed by protein transfer to nitrocellulose membrane, blocking
with 5% BSA and incubation with primary and HRP-conjugated
secondary antibodies.
For immunoprecipitation, an anti-LCK antibody (Mouse, clone
MOL 171, BD Pharmingen) was used. For immunoblotting,
antibodies specific for phosphorylated Y192 in LCK (Abcam) and
LCK (Rabbit, Cell Signaling) were used.
File S2 This PDF file combines Figures S1–S9 and
Tables S3–S5. Figure S1 in File S2. Overview of methodology
and summary of main results. (A) An integrated experimental and
model-based approach was used to characterize initial phosphorylation events in TCR signaling, generate non-trivial predictions,
and test these predictions. A model based solely on previously
elucidated mechanisms of TCR signaling did not reproduce the
phosphorylation dynamics observed for the following five sites:
LCK Y192, DOK1 Y449, DOK2 Y299, PAG1 Y417, and WAS
Y291. Incorporation of novel mechanisms enabled the dynamics
of these sites to be reproduced and led to generation of predictions
that were tested experimentally. (B) Proposed roles of PTPN6 in
early and late signaling. In early signaling (bold lines), PTPN6
plays a positive role by dephosphorylating negative regulatory
sites, including LCK Y192, PAG1 Y163, DOK1 Y449, and
DOK2 Y299. Later in signaling (thin lines), the negative
regulatory capabilities of PTPN6 become dominant. (C) Proposed
dual pathways for activation of WAS. In early signaling (bold
lines), WAS is recruited to the plasma membrane through
interaction with NCK1/2 in association with CD3E. As signaling
progresses over time, the longer pathway for WAS recruitment
(thin lines), which is dependent on LCP2, becomes dominant. (D)
A PTPN6-mediated positive feedback loop in which PTPN6
dephosphorylates LCK Y192, thereby enhancing the ability of the
LCK SH2 domain to interact with pTyr sites. LCK activates
PTPN6 through phosphorylation and direct interaction. (E) In a
second PTPN6-mediated positive feedback loop, LCK phosphorylates and activates PTPN6. PTPN6 dephosphorylates PAG1,
which reduces the ability of PAG1 to co-localize LCK and CSK,
which reduces phosphorylation of LCK at its inhibitory Cterminal tyrosine and relieves autoinhibition. Figure S2 in File
S2. Enrichment analysis. For proteins containing regulated pTyr
sites, we tested for enrichment of associated GO terms compared
Supporting Information
The Supporting Information consists of 17 items: Tables S1 and
S2 and Files A and B in File S1; and Figures S1–S9 and Tables
S3–S5 in File S2; and File S3 (Supplementary Text S1). Table S4
in File S2 includes citations of References [73,74,75,76,77,78,79].
PLOS ONE | www.plosone.org
10
August 2014 | Volume 9 | Issue 8 | e104240
Phosphorylation Site Dynamics of Early TCR Signaling
to proteins containing detected but unregulated pTyr sites (i.e.,
pTyr sites for which phosphorylation changed less than two-fold).
(A) Cluster-specific enrichment analysis based on ‘‘biological
process’’ terms. (B) Cluster-specific enrichment analysis based on
‘‘molecular function’’ terms. (C) Cluster-specific enrichment
analysis based on ‘‘cellular compartment’’ terms. (D) We also
tested for enrichment of associated Pfam domain names. In A
through D, color is used to indicate the negative logarithm (base
10) of the z-transformed p-value associated with each term. The
lighest shade of green corresponds to the highest level of
enrichment. Black corresponds to no enrichment. (E) Information
about detected pTyr sites was uploaded to the DAVID resource
and processed using default parameters to identify pathways
enriched for regulated pTyr sites. The y-axis reports the negative
logarithm (base 10) of the p-value for each of the indicated
pathways. Pathway enrichment scores are reported on the y-axis.
The most enriched pathway is the ‘‘T cell receptor signaling
pathway.’’ Figure S3 in File S2. Phylogenetic relationships of
protein kinases with detected pTyr sites. The tree shown was built
based on kinase domain sequences. Protein kinases that contain
regulated pTyr sites are indicated with red lettering; these kinases
are also represented in Fig. 1h. Kinase families are indicated by
background colors. The following abbreviations are used for
protein kinase family names: TK, tyrosine kinases; TKL, tyrosine
kinase-like; CMGC, the CDK/MAPK/GSK3/CLK group;
AGC, protein kinase A, G, and C families; CAMK, calcium and
calmodulin regulated kinases; and STE, homologs of the yeast
STE7, STE11, and STE20 genes. Figure S4 in File S2.
Visualization of model. Proteins, domains, and linear motifs are
represented as boxes, which are nested to indicate structural
relationships. Lines that begin and end with an arrowhead
represent direct binding interactions. Arrowheads point to
functional components that mediate protein-protein interactions.
Lines that originate at a box representing an enzyme (a kinase or
phosphatase) and end with an open circle, or open circle overlayed
with a diagonal bar, indicate enzyme-substrate relationships. An
open circle denotes phosphorylation and an open circle overlaid
with a diagonal bar denotes dephosphorylation. Flags (vertical
lines connected to a small square box at top and a text label at
bottom) represent sites of post-translational modification. All of
these sites are pTyr sites. Compartmental locations of proteins are
indicated by boxed labels near the lower left corners of protein
boxes. The following symbols are used to denote locations: E,
extracellular; M, membrane anchored; and C, cytosol. Locations
that can be inferred are not indicated. Protein boxes are organized
in layers, which are indicated by shading. Stimulating antibodies
are represented in the top layer, TCR/CD3 and CD28 are
represented in the next layer, their direct interaction partners are
indicated in the next layer, and so on. Elements of this map
directly related to elements of the underlying rule-based model
that it visualizes. A rule-based model is composed of molecule type
definitions and rules, as well as specifications of rate laws,
parameters, and initial conditions. Molecule type definitions of
the model are illustrated here by protein boxes. Rules are
illustrated by arrows. Each arrow corresponds to a single rule or a
set of related rules. Numbers next to arrows reference rules
presented in File A in File S1. Figure S5 in File S2. Principal
component analysis of time-course data used to guide model
specification and estimate model parameter values. Experimental
time courses for the 16 pTyr sites included in the model were
analyzed by principal component analysis and found to separate
into three classes, which are distinguished by different background
colors and labeled 1–3. Time courses in Class 1 correspond to
pTyr sites that were observed to undergo increases in phosphorPLOS ONE | www.plosone.org
ylation; according to the model, these increases occur through
mechanisms that do not require prior ITAM phosphorylation. An
example of a pTyr site in Class 1 is CD3G Y171, which in the
model can be phosphorylated by LCK bound to CD28 through a
constitutive interaction that does not require ITAM phosphorylation. Time courses in Class 2 correspond to pTyr sites that were
also observed to undergo increases in phosphorylation; however,
according to the model, these increases in phosphorylation occur
through mechanisms that require ITAM phosphorylation. For
example, ZAP70 must be recruited to a phosphorylated ITAM
before it can be phosphorylated by LCK at Y493. Time courses in
Class 3 correspond to pTyr sites were observed to undergo
decreases in phosphorylation. Figure S6 in File S2. Experimental and simulated time courses. (A through P) Phosphorylation
dynamics of the 16 pTyr sites used to guide model construction
and estimate model parameters are plotted. Points represent the
average of measurements from three independent phosphoproteomic experiments, with error bars representing standard
deviations. Simulation results are plotted as solid lines. Experimental measurements and simulation results are normalized to
baseline and log2 transformed. (Q and R) Measured phosphorylation dynamics of pTyr sites in PAG1, additional to the site shown
in Panel m. The dynamics of these sites are similar to the dynamics
of PAG1 pY163; these sites are not explicitly considered in the
model. Note that the results presented here were presented earlier
in Figs. 2–4 in different form. Figure S7 in File S2. PTPN11
levels in normal cells and in cells depleted of PTPN6.
Immunoblots of PTPN11 (SHP-2) in normal cells (WT) and in
cells depleted of PTPN6 (SHP-1 KD). Blots are representative of
the results from multiple (at least two) experiments. Figure S8 in
File S2. In vitro phosphatase activity of PTPN6. Immunoprecipitated LCK was treated or untreated with recombinant PTPN6
and immunoblotted using phospho-tyrosine specific antibodies as
indicated. LCK specific antibodies were used to show equal
amounts of immunoprecipitated LCK. Blots are representative of
the results from multiple (at least two) experiments. Figure S9 in
File S2. Disengagement of the shortcut pathway and engagement
of the longer LCP2-dependent pathway to WAS activation. (A)
Predicted association of NCK1/2 with CD3E at 0 and60 s of
stimulation. The y-axis indicates the number of NCK molecules
associated with TCR/CD3 complexes per cell. (B) Predicted
association of NCK1/2 with phosphorylated LCP2 at 0 and 60 s
of stimulation. The y-axis indicates the number of NCK molecules
associated with LCP2 per cell. (C) Immunoblot of LCP2
phosphorylation at Y145 in normal (WT) and LCP2 KD cells
stimulated for the indicated times. (D) Immunoblot of LCP2
phosphorylation at Y113 in normal (WT) and LCP2 KD cells
stimulated for the indicated times. Blots are representative of the
results from multiple (at least two) experiments. Table S3 in File
S2. Proteins and pTyr sites included in the model for TCR
signaling. Table S4 in File S2. Summary of earlier phosphoproteomic studies of TCR signaling. Table S5 in File S2.
Comparison of selected models for immunoreceptor signaling in
which phosphosite dynamics were considered.
(PDF)
File S3 This PDF file provides the Supplementary Text.
Supplementary Text S1 in File S3. This PDF file provides
extensive annotation of the model.
(PDF)
Acknowledgments
We thank J. R. Faeder, B. Goldstein, R. N. Gutenkunst, and R. G. Posner
for helpful discussions and technical assistance.
11
August 2014 | Volume 9 | Issue 8 | e104240
Phosphorylation Site Dynamics of Early TCR Signaling
the data: VA BB LAC JD WSH BH KTGR. Contributed to the writing of
the manuscript: VA BB LAC JD WSH BH KTGR.
Author Contributions
Conceived and designed the experiments: VA BB LAC JD WSH KTGR.
Performed the experiments: VA BB LAC JD WSH BH KTGR. Analyzed
References
27. Chylek LA (2013) Decoding the language of phosphorylation site dynamics.
Science Signaling 6: jc2.
28. Brodovitch A, Bongrand P, Pierres A (2013) T lymphocytes sense antigens
within seconds and make a decision within one minute. Journal of Immunology
191: 2064–2071.
29. Mashima R, Hishida Y, Tezuka T, Yamanashi Y (2009) The roles of Dok
family adapters in immunoreceptor signaling. Immunological Reviews 232:
273–285.
30. Schoenborn JR, Tan YX, Zhang C, Shokat KM, Weiss A (2011) Feedback
circuits monitor and adjust basal Lck-dependent events in T cell receptor
signaling. Science Signaling 4: ra59.
31. Solheim SA, Torgersen KM, Tasken K, Berge T (2008) Regulation of FynT
function by dual domain docking on PAG/Cbp. Journal of Biological
Chemistry 283: 2773–2783.
32. Stefanová I, Hemmer B, Vergelli M, Martin R, Biddison WE, et al. (2003)
TCR ligand discrimination is enforced by competing ERK positive and SHP-1
negative feedback pathways. Nature Immunology 4: 248–254.
33. Barda-Saad M, Braiman A, Titerence R, Bunnell SC, Barr VA, et al. (2005)
Dynamic molecular interactions linking the T cell antigen receptor to the actin
cytoskeleton. Nature Immunology 6: 80–89.
34. Gil D, Schamel WW, Montoya M, Sanchez-Madrid F, Alarcon B (2002)
Recruitment of Nck by CD3e reveals a ligand-induced conformational change
essential for T cell receptor signaling and synapse formation. Cell 109: 901–
912.
35. Cao L, Ding Y, Hung N, Yu K, Ritz A, et al. (2012) Quantitative
phosphoproteomics reveals SLP-76 dependent regulation of PAG and Src
family kinases in T cells. PLOS ONE 7: e46725.
36. Kesti T, Ruppelt A, Wang JH, Liss M, Wagner R, et al. (2007) Reciprocal
regulation of SH3 and SH2 domain binding via tyrosine phosphorylation of a
common site in CD3e. Journal of Immunology 179: 878–885.
37. Houtman JCD, Houghtling RA, Barda-Saad M, Toda Y, Samelson LE (2005)
Early phosphorylation kinetics of proteins involved in proximal TCR-mediated
signaling pathways. Journal of Immunology 175: 2449–2458.
38. Huse M, Klein LO, Girvin AT, Faraj JM, Li QJ, et al. (2007) Spatial and
temporal dynamics of T cell receptor signaling with a photoactivatable agonist.
Immunity 27: 76–88.
39. Chiang GG, Sefton BM (2001) Specific dephosphorylation of the Lck tyrosine
protein kinase at Tyr-394 by the SHP-1 protein-tyrosine phosphatase. Journal
of Biological Chemistry 276: 23173–23178.
40. Lorenz U (2009) SHP-1 and SHP-2 in T cells: two phosphatases functioning at
many levels. Immunological Reviews 228: 342–359.
41. Shultz LD, Schweitzer PA, Rajan TV, Yi T, Ihle JN, et al. (1993) Mutations at
the murine motheaten locus are within the hematopoietic cell protein-tyrosine
phosphatase (Hcph) gene. Cell 73: 1445–1454.
42. Johnson DJ, Pao LI, Dhanji S, Murakami K, Ohashi PS, et al. (2013) Shp1
regulates T cell homeostasis by limiting IL-4 signals. Journal of Experimental
Medicine 210: 1419–1431.
43. Fowler CC, Pao LI, Blattman JN, Greenberg PD (2010) SHP-1 in T cells limits
the production of CD8 effector cells without impacting the formation of longlived central memory cells. Journal of Immunology 185: 3256–3267.
44. Li QJ, Chau J, Ebert PJ, Sylvester G, Min H, et al. (2007) miR-181a is an
intrinsic modulator of T cell sensitivity and selection. Cell 129: 147–161.
45. Ren L, Chen X, Luechapanichkul R, Selner NG, Meyer TM, et al. (2011)
Substrate specificity of protein tyrosine phosphatases 1B, RPTPa, SHP-1, and
SHP-2. Biochemistry 50: 2339–2356.
46. Stover DR, Furet P, Lydon NB (1996) Modulation of the SH2 binding
specificity and kinase activity of Src by tyrosine phosphorylation within its SH2
domain. Journal of Biological Chemistry 271: 12481–12487.
47. Tailor P, Tsai S, Shameli A, Serra P, Wang J, et al. (2008) The proline-rich
sequence of CD3e as an amplifier of low-avidity TCR signaling. Journal of
Immunology 181: 243–255.
48. Szymczak AL, Workman CJ, Gil D, Dilioglou S, Vignali KM, et al. (2005) The
CD3e proline-rich sequence, and its interaction with Nck, is not required for T
cell development and function. Journal of Immunology 175: 270–275.
49. Morris GP, Allen PM (2012) How the TCR balances sensitivity and specificity
for the recognition of self and pathogens. Nature Immunology 13: 121–128.
50. Berkovitz LD, Medhin NG (2012) Nonlinear Optimal Control Theory. Boca
Raton, FL: Chapman and Hall/CRC.
51. Borgeson G (1998) The Golden Age of the American Racing Car, Second
Edition. Warrendale, PA: Society of Automotive Engineers.
52. Wall ME, Hlavacek WS, Savageau MA (2004) Design of gene circuits: lessons
from bacteria. Nature Reviews Genetics 5: 34–42.
53. Itzkovitz S, Blat IC, Jacks T, Clevers H, van Oudenaarden A (2012) Optimality
in the development of intestinal crypts. Cell 148: 608–619.
54. Rigbolt KTG, Prokhorova TA, Akimov V, Henningsen J, Johansen PT, et al.
(2011) System-wide temporal characterization of the proteome and phospho-
1. Lim WA, Pawson T (2010) Phosphotyrosine signaling: evolving a new cellular
communication system. Cell 142: 661–667.
2. Hunter T (2009) Tyrosine phosphorylation: thirty years and counting. Current
Opinion in Cell Biology 21: 140–146.
3. Tyson JJ, Chen KC, Novak B (2003) Sniffers, buzzers, toggles and blinkers:
dynamics of regulatory and signaling pathways in the cell. Current Opinion in
Cell Biology 15: 221–231.
4. Ma W, Trusina A, El-Samad H, Lim WA, Tang C (2009) Defining network
topologies that can achieve biochemical adaptation. Cell 138: 760–773.
5. Logue JS, Morrison DK (2012) Complexity in the signaling network: insights
from the use of targeted inhibitors in cancer therapy. Genes & Development
26: 641–650.
6. Lowell CA (2011) Src-family and Syk kinases in activating and inhibitory
pathways in innate immune cells: signaling cross talk. Cold Spring Harbor
Perspectives in Biology 3: a002352.
7. Di Ventura B, Lemerle C, Michalodimitrakis K, Serrano L (2006) From in vivo
to in silico biology and back. Nature 443: 527–533.
8. Tibes R, Qiu Y, Lu Y, Hennessy B, Andreeff M, et al. (2006) Reverse phase
protein array: validation of a novel proteomic technology and utility for analysis
of primary leukemia specimens and hematopoietic stem cells. Molecular
Cancer Therapeutics 5: 2512–2521.
9. Ciaccio MF, Wagner JP, Chuu CP, Lauffenburger DA, Jones RB (2010)
Systems analysis of EGF receptor signaling dynamics with microwestern arrays.
Nature Methods 7: 148–155.
10. Olsen JV, Blagoev B, Gnad F, Macek B, Kumar C, et al. (2006) Global, in vivo,
and site-specific phosphorylation dynamics in signaling networks. Cell 127:
635–648.
11. Rigbolt KTG, Blagoev B (2012) Quantitative phosphoproteomics to characterize signaling networks. Seminars in Cell & Developmental Biology 23: 863–
871.
12. Dengjel J, Akimov V, Olsen JV, Bunkenborg J, Mann M, et al. (2007)
Quantitative proteomic assessment of very early cellular signaling events.
Nature Biotechnology 25: 566–568.
13. Cox J, Mann M (2011) Quantitative, high-resolution proteomics for datadriven systems biology. Annual Review of Biochemistry 80: 273–299.
14. Hlavacek WS, Faeder JR, Blinov ML, Posner RG, Hucka M, et al. (2006)
Rules for modeling signal-transduction systems. Science’s STKE 2006: re6.
15. Chylek LA, Harris LA, Tung CS, Faeder JR, Lopez CF, et al. (2014) Rulebased modeling: a computational approach for studying biomolecular site
dynamics in cell signaling systems. Wiley Interdisciplinary Reviews Systems
Biology and Medicine 6: 13–36.
16. Germain RN, Meier-Schellersheim M, Nita-Lazar A, Fraser ID (2011) Systems
biology in immunology: a computational modeling perspective. Annual Review
of Immunology 29: 527–585.
17. Feinerman O, Jentsch G, Tkach KE, Coward JW, Hathorn MM, et al. (2010)
Single-cell quantification of IL-2 response by effector and regulatory T cells
reveals critical plasticity in immune response. Molecular Systems Biology 6:
437.
18. Barua D, Hlavacek WS, Lipniacki T (2012) A computational model for early
events in B cell antigen receptor signaling: analysis of the roles of Lyn and Fyn.
Journal of Immunology 189: 646–658.
19. Mukherjee S, Zhu J, Zikherman J, Parameswaran R, Kadlecek TA, et al.
(2013) Monovalent and multivalent ligation of the B cell receptor exhibit
differential dependence upon Syk and Src family kinases. Science Signaling 6:
ra1.
20. Mukhopadhyay H, Cordoba SP, Maini PK, van der Merwe PA, Dushek O
(2013) Systems model of T cell receptor proximal signaling reveals emergent
ultrasensitivity. PLOS Computational Biology 9: e1003004.
21. Sneddon MW, Faeder JR, Emonet T (2011) Efficient modeling, simulation and
coarse-graining of biological complexity with NFsim. Nature Methods 8: 177–
183.
22. Chylek LA, Hu B, Blinov ML, Emonet T, Faeder JR, et al. (2011) Guidelines
for visualizing and annotating rule-based models. Molecular BioSystems 7:
2779–2795.
23. Smith-Garvin JE, Koretzky GA, Jordan MS (2009) T cell activation. Annual
Review of Immunology 27: 591–619.
24. O’Neill SK, Getahun A, Gauld SB, Merrell KT, Tamir I, et al. (2011)
Monophosphorylation of CD79a and CD79b ITAM motifs initiates a SHIP-1
phosphatase-mediated inhibitory signaling cascade required for B cell anergy.
Immunity 35: 746–756.
25. Ingley E (2008) Src family kinases: regulation of their activities, levels and
identification of new pathways. Biochimica et Biophysica Acta 1784: 56–65.
26. Couture C, Songyang Z, Jascur T, Williams S, Tailor P, et al. (1996)
Regulation of the Lck SH2 domain by tyrosine phosphorylation. Journal of
Biological Chemistry 271: 24880–24884.
PLOS ONE | www.plosone.org
12
August 2014 | Volume 9 | Issue 8 | e104240
Phosphorylation Site Dynamics of Early TCR Signaling
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
79.
80.
81.
82.
83. Lipniacki T, Hat B, Faeder JR, Hlavacek WS (2008) Stochastic effects and
bistability in T cell receptor signaling. Journal of Theoretical Biology 254: 110–
122.
84. Li QJ, Dinner AR, Qi S, Irvine DJ, Huppa JB, et al. (2004) CD4 enhances T
cell sensitivity to antigen by coordinating Lck accumulation at the
immunological synapse. Nature Immunology 5: 791–799.
85. Lee KH, Dinner AR, Tu C, Campi G, Raychaudhuri S, et al. (2003) The
immunological synapse balances T cell receptor signaling and degradation.
Science 302: 1218–1222.
86. Chylek LA, Stites EC, Posner RG, Hlavacek WS (2013) Innovations of the
rule-based modeling approach. In: Prokop A, Csukás B, editors. Systems
Biology: Integrative Biology and Simulation Tools. Dordrecht: Springer. pp.
273–300.
87. Chylek LA, Wilson BS, Hlavacek WS (2014) Modeling biomolecular site
dynamics in immunoreceptor signaling systems. In: Corey SJ, Kimmel M,
Leonard JN, editors. A Systems Biology Approach to Blood. Dordrecht:
Springer. In press.
88. Sekar JA, Faeder JR (2012) Rule-based modeling of signal transduction: a
primer. Methods in Molecular Biology 880: 139–218.
89. Danos V, Feret J, Fontana W, Harmer R, Krivine J (2007) Rule-based
modelling of cellular signalling. Lecture Notes in Computer Science 4703: 17–
41.
90. Krivine J, Danos V, Benecke A (2009) Modelling epigenetic information
maintenance: a Kappa tutorial. Lecture Notes in Computer Science 5643: 17–
32.
91. Abraham RT, Weiss A (2004) Jurkat T cells and development of the T-cell
receptor signalling paradigm. Nature Reviews Immunology 4: 301–308.
92. Weissman AM, Frank SJ, Orloff DG, Mercep M, Ashwell JD, et al. (1989) Role
of the zeta chain in the expression of the T cell antigen receptor: genetic
reconstitution studies. EMBO Journal 8: 3651–3656.
93. Alarcon B, Berkhout B, Breitmeyer J, Terhorst C (1988) Assembly of the
human T cell receptor-CD3 complex takes place in the endoplasmic reticulum
and involves intermediary complexes between the CD3-c?d?e core and single T
cell receptor a or b chains. Journal of Biological Chemistry 263: 2953–2961.
94. Alarcon B, Gil D, Delgado P, Schamel WW (2003) Initiation of TCR signaling:
regulation within CD3 dimers. Immunological Reviews 191: 38–46.
95. Reth M (1989) Antigen receptor tail clue. Nature 338: 383–384.
96. Iwashima M, Irving BA, van Oers NSC, Chan AC, Weiss A (1994) Sequential
interactions of the TCR with two distinct cytoplasmic tyrosine kinases. Science
263: 1136–1139.
97. Arnett KL, Harrison SC, Wiley DC (2004) Crystal structure of a human CD3e/d dimer in complex with a UCHT1 single-chain antibody fragment.
Proceedings of the National Academy of Sciences of the United States of
America 101: 16268–16273.
98. Risueno RM, Schamel WW, Alarcon B (2008) T cell receptor engagement
triggers its CD3e and CD3f subunits to adopt a compact, locked conformation.
PLoS ONE 3: e1747.
99. Xu C, Gagnon E, Call ME, Schnell JR, Schwieters CD, et al. (2008)
Regulation of T cell receptor activation by dynamic membrane binding of the
CD3e cytoplasmic tyrosine-based motif. Cell 135: 702–713.
100. Gil D, Gutierrez D, Alarcon B (2001) Intracellular redistribution of nucleolin
upon interaction with the CD3e chain of the T cell receptor complex. Journal
of Biological Chemistry 276: 11174–11179.
101. Hara T, Fu SM, Hansen JA (1985) Human T cell activation. II. A new
activation pathway used by a major T cell population via a disulfide-bonded
dimer of a 44 kilodalton polypeptide (9.3 antigen). Journal of Experimental
Medicine 161: 1513–1524.
102. Linsley PS, Clark EA, Ledbetter JA (1990) T-cell antigen CD28 mediates
adhesion with B cells by interacting with activation antigen B7/BB-1.
Proceedings of the National Academy of Sciences of the United States of
America 87: 5031–5035.
103. Azuma M, Ito D, Yagita H, Okumura K, Phillips JH, et al. (1993) B70 antigen
is a second ligand for CTLA-4 and CD28. Nature 366: 76–79.
104. Freeman GJ, Gribben JG, Boussiotis VA, Ng JW, Restivo VA Jr, et al. (1993)
Cloning of B7-2: a CTLA-4 counter-receptor that costimulates human T cell
proliferation. Science 262: 909–911.
105. Gonfloni S, Weijland A, Kretzschmar J, Superti-Furga G (2000) Crosstalk
between the catalytic and regulatory domains allows bidirectional regulation of
Src. Nature Structural Biology 7: 281–286.
106. Bergman M, Mustelin T, Oetken C, Partanen J, Flint NA, et al. (1992) The
human p50csk tyrosine kinase phosphorylates p56lck at Tyr-505 and down
regulates its catalytic activity. EMBO Journal 11: 2919–2924.
107. Sicheri F, Kuriyan J (1997) Structures of Src-family tyrosine kinases. Current
Opinion in Structural Biology 7: 777–785.
108. Vang T, Abrahamsen H, Myklebust S, Enserink J, Prydz H, et al. (2004)
Knockdown of C-terminal Src kinase by siRNA-mediated RNA interference
augments T cell receptor signaling in mature T cells. European Journal of
Immunology 34: 2191–2199.
109. Xu W, Harrison SC, Eck MJ (1997) Three-dimensional structure of the
tyrosine kinase c-Src. Nature 385: 595–602.
110. Xu W, Doshi A, Lei M, Eck MJ, Harrison SC (1999) Crystal structures of c-Src
reveal features of its autoinhibitory mechanism. Molecular Cell 3: 629–638.
111. Williams JC, Weijland A, Gonfloni S, Thompson A, Courtneidge SA, et al.
(1997) The 2.35 Å crystal structure of the inactivated form of chicken Src: a
proteome of human embryonic stem cell differentiation. Science Signaling 4:
rs3.
Olsen JV, de Godoy LM, Li G, Macek B, Mortensen P, et al. (2005) Parts per
million mass accuracy on an Orbitrap mass spectrometer via lock mass
injection into a C-trap. Molecular & Cellular Proteomics 4: 2010–2021.
Cox J, Mann M (2008) MaxQuant enables high peptide identification rates,
individualized p.p.b.-range mass accuracies and proteome-wide protein
quantification. Nature Biotechnology 26: 1367–1372.
Kersey PJ, Duarte J, Williams A, Karavidopoulou Y, Birney E, et al. (2004)
The International Protein Index: an integrated database for proteomics
experiments. Proteomics 4: 1985–1988.
Rigbolt KTG, Vanselow JT, Blagoev B (2011) GProX, a user-friendly platform
for bioinformatics analysis and visualization of quantitative proteomics data.
Molecular & Cellular Proteomics 10: 10.1074/mcp.O1110.0075450, 1–10.
Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, et al. (2000) Gene
Ontology: tool for the unification of biology. Nature Genetics 25: 25–29.
Punta M, Coggill PC, Eberhardt RY, Mistry J, Tate J, et al. (2012) The Pfam
protein families database. Nucleic Acids Research 40: D290–D301.
UniProt Consortium (2012) Reorganizing the protein space at the Universal
Protein Resource (UniProt). Nucleic Acids Research 40: D71–D75.
Huang DW, Sherman BT, Lempicki RA (2009) Bioinformatics enrichment
tools: paths toward the comprehensive functional analysis of large gene lists.
Nucleic Acids Research 37: 1–13.
Larkin MA, Blackshields G, Brown NP, Chenna R, McGettigan PA, et al.
(2007) Clustal W and Clustal X version 2.0. Bioinformatics 23: 2947–2948.
Letunic I, Bork P (2007) Interactive Tree Of Life (iTOL): an online tool for
phylogenetic tree display and annotation. Bioinformatics 23: 127–128.
Faeder JR, Blinov ML, Hlavacek WS (2009) Rule-based modeling of
biochemical systems with BioNetGen. Methods in Molecular Biology 500:
113–167.
Yang J, Monine MI, Faeder JR, Hlavacek WS (2008) Kinetic Monte Carlo
method for rule-based modeling of biochemical networks. Physical Review E
78: 031910.
Colvin J, Monine MI, Gutenkunst RN, Hlavacek WS, Von Hoff DD, et al.
(2010) RuleMonkey: software for stochastic simulation of rule-based models.
BMC Bioinformatics 11: 404.
Davidon WC (1991) Variable metric method for minimization. SIAM Journal
on Optimization 1: 1–17.
Ventura A, Meissner A, Dillon CP, McManus M, Sharp PA, et al. (2004) Crelox-regulated conditional RNA interference from transgenes. Proceedings of
the National Academy of Sciences of the United States of America 101: 10380–
10385.
Dull T, Zufferey R, Kelly M, Mandel RJ, Nguyen M, et al. (1998) A thirdgeneration lentivirus vector with a conditional packaging system. Journal of
Virology 72: 8463–8471.
Birmingham A, Anderson E, Sullivan K, Reynolds A, Boese Q, et al. (2007) A
protocol for designing siRNAs with high functionality and specificity. Nature
Protocols 2: 2068–2078.
Taxman DJ, Livingstone LR, Zhang J, Conti BJ, Iocca HA, et al. (2006)
Criteria for effective design, construction, and gene knockdown by shRNA
vectors. BMC Biotechnology 6: 7.
Nguyen V, Cao L, Lin JT, Hung N, Ritz A, et al. (2009) A new approach for
quantitative phosphoproteomic dissection of signaling pathways applied to T
cell receptor activation. Molecular & Cellular Proteomics 8: 2418–2431.
Brockmeyer C, Paster W, Pepper D, Tan CP, Trudgian DC, et al. (2011) T cell
receptor (TCR)-induced tyrosine phosphorylation dynamics identifies THEMIS as a new TCR signalosome component. Journal of Biological Chemistry
286: 7535–7547.
Mayya V, Lundgren DH, Hwang SI, Rezaul K, Wu L, et al. (2009)
Quantitative phosphoproteomic analysis of T cell receptor signaling reveals
system-wide modulation of protein-protein interactions. Science Signaling 2:
ra46.
Kim JE, White FM (2006) Quantitative analysis of phosphotyrosine signaling
networks triggered by CD3 and CD28 costimulation in Jurkat cells. Journal of
Immunology 176: 2833–2843.
Iwai LK, Benoist C, Mathis D, White FM (2010) Quantitative phosphoproteomic analysis of T cell receptor signaling in diabetes prone and resistant mice.
Journal of Proteome Research 9: 3135–3145.
Matsumoto M, Oyamada K, Takahashi H, Sato T, Hatakeyama S, et al. (2009)
Large-scale proteomic analysis of tyrosine-phosphorylation induced by T-cell
receptor or B-cell receptor activation reveals new signaling pathways.
Proteomics 9: 3549–3563.
Ruperez P, Gago-Martinez A, Burlingame AL, Oses-Prieto JA (2012)
Quantitative phosphoproteomic analysis reveals a role for serine and threonine
kinases in the cytoskeletal reorganization in early T cell receptor activation in
human primary T cells. Molecular & Cellular Proteomics 11: 171–186.
Faeder JR, Hlavacek WS, Reischl I, Blinov ML, Metzger H, et al. (2003)
Investigation of early events in FceRI-mediated signaling using a detailed
mathematical model. Journal of Immunology 170: 3769–3781.
Wofsy C, Torigoe C, Kent UM, Metzger H, Goldstein B (1997) Exploiting the
difference between intrinsic and extrinsic kinases: implications for regulation of
signaling by immunoreceptors. Journal of Immunology 159: 5984–5992.
Altan-Bonnet G, Germain RN (2005) Modeling T cell antigen discrimination
based on feedback control of digital ERK responses. PLoS Biology 3: e356.
PLOS ONE | www.plosone.org
13
August 2014 | Volume 9 | Issue 8 | e104240
Phosphorylation Site Dynamics of Early TCR Signaling
112.
113.
114.
115.
116.
117.
118.
119.
120.
121.
122.
123.
124.
125.
126.
127.
128.
129.
130.
131.
132.
133.
134.
135.
136.
137. June CH, Fletcher MC, Ledbetter JA, Samelson LE (1990) Increases in tyrosine
phosphorylation are detectable before phospholipase C activation after T cell
receptor stimulation. Journal of Immunology 144: 1591–1599.
138. Zhang W, Sloan-Lancaster J, Kitchen J, Trible RP, Samelson LE (1998) LAT:
the ZAP-70 tyrosine kinase substrate that links T cell receptor to cellular
activation. Cell 92: 83–92.
139. Weber JR, Ørstavik S, Torgersen KM, Danbolt NC, Berg SF, et al. (1998)
Molecular cloning of the cDNA encoding pp36, a tyrosine-phosphorylated
adaptor protein selectively expressed by T cells and natural killer cells. Journal
of Experimental Medicine 187: 1157–1161.
140. Malissen B, Aguado E, Malissen M (2005) Role of the LAT adaptor in T-cell
development and Th2 differentiation. Advances in Immunology 87: 1–25.
141. Zhu M, Janssen E, Zhang W (2003) Minimal requirement of tyrosine residues
of linker for activation of T cells in TCR signaling and thymocyte development.
Journal of Immunology 170: 325–333.
142. Gilliland LK, Schieven GL, Norris NA, Kanner SB, Aruffo A, et al. (1992)
Lymphocyte lineage-restricted tyrosine-phosphorylated proteins that bind
PLCc1 SH2 domains. Journal of Biological Chemistry 267: 13610–13616.
143. Zhang W, Trible RP, Zhu M, Liu SK, McGlade CJ, et al. (2000) Association of
Grb2, Gads, and phospholipase C-c1 with phosphorylated LAT tyrosine
residues. Effect of LAT tyrosine mutations on T cell antigen receptor-mediated
signaling. Journal of Biological Chemistry 275: 23355–23361.
144. Zhang W, Irvin BJ, Trible RP, Abraham RT, Samelson LE (1999) Functional
analysis of LAT in TCR-mediated signaling pathways using a LAT-deficient
Jurkat cell line. International Immunology 11: 943–950.
145. Dennis EA, Rhee SG, Billah MM, Hannun YA (1991) Role of phospholipase in
generating lipid second messengers in signal transduction. FASEB Journal 5:
2068–2077.
146. Fu G, Chen Y, Yu M, Podd A, Schuman J, et al. (2010) Phospholipase Cc1 is
essential for T cell development, activation, and tolerance. Journal of
Experimental Medicine 207: 309–318.
147. Oh-hora M, Rao A (2008) Calcium signaling in lymphocytes. Current Opinion
in Immunology 20: 250–258.
148. Schaeffer EM, Debnath J, Yap G, McVicar D, Liao XC, et al. (1999)
Requirement for Tec kinases Rlk and Itk in T cell receptor signaling and
immunity. Science 284: 638–641.
149. Bogin Y, Ainey C, Beach D, Yablonski D (2007) SLP-76 mediates and
maintains activation of the Tec family kinase ITK via the T cell antigen
receptor-induced association between SLP-76 and ITK. Proceedings of the
National Academy of Sciences of the United States of America 104: 6638–
6643.
150. Braiman A, Barda-Saad M, Sommers CL, Samelson LE (2006) Recruitment
and activation of PLCc1 in T cells: a new insight into old domains. EMBO
Journal 25: 774–784.
151. Clements JL, Yang B, Ross-Barta SE, Eliason SL, Hrstka RF, et al. (1998)
Requirement for the leukocyte-specific adapter protein SLP-76 for normal T
cell development. Science 281: 416–419.
152. Koretzky GA (2009) The role of SH2 domain-containing leukocyte
phosphoprotein of 76 kDa in the regulation of immune cell development
and function. Immune Network 9: 75–83.
153. Bubeck Wardenburg J, Pappu R, Bu JY, Mayer B, Chernoff J, et al. (1998)
Regulation of PAK activation and the T cell cytoskeleton by the linker protein
SLP-76. Immunity 9: 607–616.
154. Wunderlich L, Farago A, Downward J, Buday L (1999) Association of Nck with
tyrosine-phosphorylated SLP-76 in activated T lymphocytes. European Journal
of Immunology 29: 1068–1075.
155. Berry DM, Nash P, Liu SK, Pawson T, McGlade CJ (2002) A high-affinity
Arg-X-X-Lys SH3 binding motif confers specificity for the interaction between
Gads and SLP-76 in T cell signaling. Current Biology 12: 1336–1341.
156. Moran O, Roessle MW, Mariuzza RA, Dimasi N (2008) Structural features of
the full-length adaptor protein GADS in solution determined using small-angle
X-ray scattering. Biophysical Journal 94: 1766–1772.
157. Liu SK, Fang N, Koretzky GA, McGlade CJ (1999) The hematopoietic-specific
adaptor protein gads functions in T-cell signaling via interactions with the SLP76 and LAT adaptors. Current Biology 9: 67–75.
158. Liu Q, Berry D, Nash P, Pawson T, McGlade CJ, et al. (2003) Structural basis
for specific binding of the Gads SH3 domain to an RxxK motif-containing
SLP-76 peptide: a novel mode of peptide recognition. Molecular Cell 11: 471–
481.
159. Seet BT, Berry DM, Maltzman JS, Shabason J, Raina M, et al. (2007) Efficient
T-cell receptor signaling requires a high-affinity interaction between the Gads
C-SH3 domain and the SLP-76 RxxK motif. EMBO Journal 26: 678–689.
160. Lettau M, Pieper J, Janssen O (2009) Nck adapter proteins: functional
versatility in T cells. Cell Communication and Signaling 7: 1.
161. Lettau M, Pieper J, Gerneth A, Lengl-Janssen B, Voss M, et al. (2010) The
adapter protein Nck: role of individual SH3 and SH2 binding modules for
protein interactions in T lymphocytes. Protein Science 19: 658–669.
162. Zeng R, Cannon JL, Abraham RT, Way M, Billadeau DD, et al. (2003) SLP76 coordinates Nck-dependent Wiskott-Aldrich syndrome protein recruitment
with Vav-1/Cdc42-dependent Wiskott-Aldrich syndrome protein activation at
the T cell-APC contact site. Journal of Immunology 171: 1360–1368.
163. Kenney D, Cairns L, Remold-O’Donnell E, Peterson J, Rosen FS, et al. (1986)
Morphological abnormalities in the lymphocytes of patients with the WiskottAldrich syndrome. Blood 68: 1329–1332.
dynamic molecule with multiple regulatory interactions. Journal of Molecular
Biology 274: 757–775.
Resh MD (1994) Myristylation and palmitylation of Src family members: the
fats of the matter. Cell 76: 411–413.
Timson Gauen LK, Kong A-NT, Samelson LE, Shaw AS (1992) p59fyn
tyrosine kinase associates with multiple T-cell receptor subunits through its
unique amino-terminal domain. Molecular and Cellular Biology 12: 5438–
5446.
Badour K, Zhang J, Shi F, Leng Y, Collins M, et al. (2004) Fyn and PTPPEST-mediated regulation of Wiskott-Aldrich syndrome protein (WASp)
tyrosine phosphorylation is required for coupling T cell antigen receptor
engagement to WASp effector function and T cell activation. Journal of
Experimental Medicine 199: 99–112.
Dombroski D, Houghtling RA, Labno CM, Precht P, Takesono A, et al. (2005)
Kinase-independent functions for Itk in TCR-induced regulation of Vav and
the actin cytoskeleton. Journal of Immunology 174: 1385–1392.
Liao XC, Littman DR (1995) Altered T cell receptor signaling and disrupted T
cell development in mice lacking Itk. Immunity 3: 757–769.
Liu KQ, Bunnell SC, Gurniak CB, Berg LJ (1998) T cell receptor-initiated
calcium release is uncoupled from capacitative calcium entry in Itk-deficient T
cells. Journal of Experimental Medicine 187: 1721–1727.
August A, Gibson S, Kawakami Y, Kawakami T, Mills GB, et al. (1994) CD28
is associated with and induces the immediate tyrosine phosphorylation and
activation of the Tec family kinase ITK/EMT in the human Jurkat leukemic
T-cell line. Proceedings of the National Academy of Sciences of the United
States of America 91: 9347–9351.
Gibson S, August A, Branch D, Dupont B, Mills GM (1996) Functional LCK Is
required for optimal CD28-mediated activation of the TEC family tyrosine
kinase EMT/ITK. Journal of Biological Chemistry 271: 7079–7083.
Marengere LE, Okkenhaug K, Clavreul A, Couez D, Gibson S, et al. (1997)
The SH3 domain of Itk/Emt binds to proline-rich sequences in the cytoplasmic
domain of the T cell costimulatory receptor CD28. Journal of Immunology
159: 3220–3229.
Torgersen KM, Vang T, Abrahamsen H, Yaqub S, Horejsi V, et al. (2001)
Release from tonic inhibition of T cell activation through transient
displacement of C-terminal Src kinase (Csk) from lipid rafts. Journal of
Biological Chemistry 276: 29313–29318.
Palacios EH, Weiss A (2004) Function of the Src-family kinases, Lck and Fyn,
in T-cell development and activation. Oncogene 23: 7990–8000.
Sabe H, Okada M, Nakagawa H, Hanafusa H (1992) Activation of c-Src in
cells bearing v-Crk and its suppression by Csk. Molecular and Cellular Biology
12: 4706–4713.
Chow LM, Fournel M, Davidson D, Veillette A (1993) Negative regulation of
T-cell receptor signalling by tyrosine protein kinase p50csk. Nature 365: 156–
160.
Takeuchi S, Takayama Y, Ogawa A, Tamura K, Okada M (2000)
Transmembrane phosphoprotein Cbp positively regulates the activity of the
carboxyl-terminal Src kinase, Csk. Journal of Biological Chemistry 275:
29183–29186.
Hatada MH, Lu X, Laird ER, Green J, Morgenstern JP, et al. (1995)
Molecular basis for interaction of the protein tyrosine kinase ZAP-70 with the
T-cell receptor. Nature 377: 32–38.
Isakov N, Wange RL, Burgess WH, Watts JD, Aebersold R, et al. (1995) ZAP70 binding specificity to T cell receptor tyrosine-based activation motifs: the
tandem SH2 domains of ZAP-70 bind distinct tyrosine-based activation motifs
with varying affinity. Journal of Experimental Medicine 181: 375–380.
van Oers NSC, Killeen N, Weiss A (1994) ZAP-70 is constitutively associated
with tyrosine-phosphorylated TCR f in murine thymocytes and lymph node T
cells. Immunity 1: 675–685.
Ottinger EA, Botfield MC, Shoelson SE (1998) Tandem SH2 domains confer
high specificity in tyrosine kinase signaling. Journal of Biological Chemistry
273: 729–735.
Au-Yeung BB, Deindl S, Hsu LY, Palacios EH, Levin SE, et al. (2009) The
structure, regulation, and function of ZAP-70. Immunological Reviews 228:
41–57.
Matthews RJ, Bowne DB, Flores E, Thomas ML (1992) Characterization of
hematopoietic intracellular protein tyrosine phosphatases: description of a
phosphatase containing an SH2 domain and another enriched in proline-,
glutamic acid-, serine-, and threonine-rich sequences. Molecular and Cellular
Biology 12: 2396–2405.
Lorenz U, Ravichandran KS, Pei D, Walsh CT, Burakoff SJ, et al. (1994) Lckdependent tyrosyl phosphorylation of the phosphotyrosine phosphatase SHPTP1 in murine T cells. Molecular and Cellular Biology 14: 1824–1834.
Yasuda T, Bundo K, Hino A, Honda K, Inoue A, et al. (2007) Dok-1 and Dok2 are negative regulators of T cell receptor signaling. International
Immunology 19: 487–495.
Shah K, Shokat KM (2002) A chemical genetic screen for direct v-Src
substrates reveals ordered assembly of a retrograde signaling pathway.
Chemistry & Biology 9: 35–47.
Yamanashi Y, Baltimore D (1997) Identification of the Abl- and rasGAPassociated 62 kDa protein as a docking protein, Dok. Cell 88: 205–211.
Lock P, Casagranda F, Dunn AR (1999) Independent SH2-binding sites
mediate interaction of Dok-related protein with RasGTPase-activating protein
and Nck. Journal of Biological Chemistry 274: 22775–22784.
PLOS ONE | www.plosone.org
14
August 2014 | Volume 9 | Issue 8 | e104240
Phosphorylation Site Dynamics of Early TCR Signaling
164. Molina IJ, Kenney DM, Rosen FS, Remold-O’Donnell E (1992) T cell lines
characterize events in the pathogenesis of the Wiskott-Aldrich syndrome.
Journal of Experimental Medicine 176: 867–874.
165. Gallego MD, Santamaria M, Pena J, Molina IJ (1997) Defective actin
reorganization and polymerization of Wiskott-Aldrich T cells in response to
CD3-mediated stimulation. Blood 90: 3089–3097.
166. Facchetti F, Blanzuoli L, Vermi W, Notarangelo LD, Giliani S, et al. (1998)
Defective actin polymerization in EBV-transformed B-cell lines from patients
with the Wiskott-Aldrich syndrome. Journal of Pathology 185: 99–107.
167. Snapper SB, Rosen FS, Mizoguchi E, Cohen P, Khan W, et al. (1998) WiskottAldrich syndrome protein-deficient mice reveal a role for WASP in T but not B
cell activation. Immunity 9: 81–91.
168. Zhang J, Shehabeldin A, da Cruz LA, Butler J, Somani AK, et al. (1999)
Antigen receptor-induced activation and cytoskeletal rearrangement are
impaired in Wiskott-Aldrich syndrome protein-deficient lymphocytes. Journal
of Experimental Medicine 190: 1329–1342.
169. Thrasher AJ, Burns SO (2010) WASP: a key immunological multitasker.
Nature Reviews Immunology 10: 182–192.
170. Blundell MP, Bouma G, Metelo J, Worth A, Calle Y, et al. (2009)
Phosphorylation of WASp is a key regulator of activity and stability in vivo.
Proceedings of the National Academy of Sciences of the United States of
America 106: 15738–15743.
171. Rivero-Lezcano OM, Marcilla A, Sameshima JH, Robbins KC (1995) WiskottAldrich syndrome protein physically associates with Nck through Src homology
3 domains. Molecular and Cellular Biology 15: 5725–5731.
172. Rohatgi R, Nollau P, Ho HY, Kirschner MW, Mayer BJ (2001) Nck and
phosphatidylinositol 4,5-bisphosphate synergistically activate actin polymerization through the N-WASP-Arp2/3 pathway. Journal of Biological Chemistry
276: 26448–26452.
173. Holdorf AD, Green JM, Levin SD, Denny MF, Straus DB, et al. (1999) Proline
residues in CD28 and the Src homology (SH)3 domain of Lck are required for
T cell costimulation. Journal of Experimental Medicine 190: 375–384.
174. August A, Dupont B (1994) Activation of src family kinase lck following CD28
crosslinking in the Jurkat leukemic cell line. Biochemical and Biophysical
Research Communications 199: 1466–1473.
175. Yang WC, Ghiotto M, Barbarat B, Olive D (1999) The role of Tec proteintyrosine kinase in T cell signaling. Journal of Biological Chemistry 274: 607–
617.
176. Garcon F, Ghiotto M, Gerard A, Yang WC, Olive D, et al. (2004) The SH3
domain of Tec kinase is essential for its targeting to activated CD28
costimulatory molecule. European Journal of Immunology 34: 1972–1980.
177. Takeuchi K, Yang H, Ng E, Park SY, Sun ZY, et al. (2008) Structural and
functional evidence that Nck interaction with CD3e regulates T-cell receptor
activity. Journal of Molecular Biology 380: 704–716.
178. de la Cruz J, Kruger T, Parks CA, Silge RL, van Oers NSC, et al. (2011) Basal
and antigen-induced exposure of the proline-rich sequence in CD3e. Journal of
Immunology 186: 2282–2290.
179. Shi X, Bi Y, Yang W, Guo X, Jiang Y, et al. (2013) Ca2+ regulates T-cell
receptor activation by modulating the charge property of lipids. Nature 493:
111–115.
180. Blank U, Launay P, Benhamou M, Monteiro RC (2009) Inhibitory ITAMs as
novel regulators of immunity. Immunological Reviews 232: 59–71.
181. Yang J, Liu L, He D, Song X, Liang X, et al. (2003) Crystal structure of human
protein-tyrosine phosphatase SHP-1. Journal of Biological Chemistry 278:
6516–6520.
182. Evren S, Ma XZ, Sakac D, Branch DR (2012) SHP-1 protein tyrosine
phosphatase associates with the adaptor protein CrkL. Experimental
Hematology 40: 1055–1059.
183. Barber EK, Dasgupta JD, Schlossman SF, Trevillyan JM, Rudd CE (1989) The
CD4 and CD8 antigens are coupled to a protein-tyrosine kinase (p56lck) that
phosphorylates the CD3 complex. Proceedings of the National Academy of
Sciences of the United States of America 86: 3277–3281.
184. Rudd CE, Barber EK, Burgess KE, Hahn JY, Odysseos AD, et al. (1991)
Molecular analysis of the interaction of p56lck with the CD4 and CD8 antigens.
Advances in Experimental Medicine and Biology 292: 85–96.
185. van Oers NSC, Killeen N, Weiss A (1996) Lck regulates the tyrosine
phosphorylation of the T cell receptor subunits and ZAP-70 in murine
thymocytes. Journal of Experimental Medicine 183: 1053–1062.
186. Straus DB, Weiss A (1993) The CD3 chains of the T cell antigen receptor
associate with the ZAP-70 tyrosine kinase and are tyrosine phosphorylated after
receptor stimulation. Journal of Experimental Medicine 178: 1523–1530.
187. Guirado M, de Aós I, Orta T, Rivas L, Terhorst C, et al. (2002)
Phosphorylation of the N-terminal and C-terminal CD3-e-ITAM tyrosines is
differentially regulated in T cells. Biochemical and Biophysical Research
Communications 291: 574–581.
188. Love PE, Hayes SM (2010) ITAM-mediated signaling by the T-cell antigen
receptor. Cold Spring Harbor Perspectives in Biology 2: a002485.
189. Straus DB, Weiss A (1992) Genetic evidence for the involvement of the lck
tyrosine kinase in signal transduction through the T cell antigen receptor. Cell
70: 585–593.
190. Watts JD, Wilson GM, Ettenhadieh E, Clark-Lewis I, Kubanek C-A, et al.
(1992) Purification and initial characterization of the lymphocyte-specific
protein-tyrosyl kinase p56lck from a baculovirus expression system. Journal of
Biological Chemistry 267: 901–907.
PLOS ONE | www.plosone.org
191. Dianzani U, Shaw A, Al-Ramadi BK, Kubo RT, Janeway CA Jr (1992)
Physical association of CD4 with the T cell receptor. Journal of Immunology
148: 678–688.
192. Aivazian D, Stern LJ (2000) Phosphorylation of T cell receptor f is regulated by
a lipid dependent folding transition. Nature Structural Biology 7: 1023–1026.
193. Veillette A, Fournel M (1990) The CD4 associated tyrosine protein kinase
p56lck is positively regulated through its site of autophosphorylation. Oncogene
5: 1455–1462.
194. Hardwick JS, Sefton BM (1995) Activation of the Lck tyrosine protein kinase by
hydrogen peroxide requires the phosphorylation of Tyr-394. Proceedings of the
National Academy of Sciences of the United States of America 92: 4527–4531.
195. Yamaguchi H, Hendrickson WA (1996) Structural basis for activation of
human lymphocyte kinase Lck upon tyrosine phosphorylation. Nature 384:
484–489.
196. Holdorf AD, Lee KH, Burack WR, Allen PM, Shaw AS (2002) Regulation of
Lck activity by CD4 and CD28 in the immunological synapse. Nature
Immunology 3: 259–264.
197. D’Oro U, Sakaguchi K, Appella E, Ashwell JD (1996) Mutational analysis of
Lck in CD45-negative T cells: dominant role of tyrosine 394 phosphorylation
in kinase activity. Molecular and Cellular Biology 16: 4996–5003.
198. Xu H, Littman DR (1995) The kinase-dependent function of Lck in T-cell
activation requires an intact site for tyrosine autophosphorylation. Annals of the
New York Academy of Sciences 766: 99–116.
199. Abraham N, Veillette A (1990) Activation of p56lck through mutation of a
regulatory carboxy-terminal tyrosine residue requires intact sites of autophosphorylation and myristylation. Molecular and Cellular Biology 10: 5197–5206.
200. Paster W, Paar C, Eckerstorfer P, Jakober A, Drbal K, et al. (2009) Genetically
encoded Förster resonance energy transfer sensors for the conformation of the
Src family kinase Lck. Journal of Immunology 182: 2160–2167.
201. Brown K, Long JM, Vial SC, Dedi N, Dunster NJ, et al. (2004) Crystal
structures of interleukin-2 tyrosine kinase and their implications for the design
of selective inhibitors. Journal of Biological Chemistry 279: 18727–18732.
202. Heyeck SD, Wilcox HM, Bunnell SC, Berg LJ (1997) Lck phosphorylates the
activation loop tyrosine of the Itk kinase domain and activates Itk kinase
activity. Journal of Biological Chemistry 272: 25401–25408.
203. Gibson S, August A, Kawakami Y, Kawakami T, Dupont B, et al. (1996) The
EMT/ITK/TSK (EMT) tyrosine kinase is activated during TCR signaling:
LCK is required for optimal activation of EMT. Journal of Immunology 156:
2716–2722.
204. Mahajan S, Fargnoli J, Burkhardt AL, Kut SA, Saouaf SJ, et al. (1995) Src
family protein tyrosine kinases induce autoactivation of Bruton’s tyrosine
kinase. Molecular and Cellular Biology 15: 5304–5311.
205. Wang H, Kadlecek TA, Au-Yeung BB, Goodfellow HE, Hsu LY, et al. (2010)
ZAP-70: an essential kinase in T-cell signaling. Cold Spring Harbor
Perspectives in Biology 2: a002279.
206. Chan AC, Dalton M, Johnson R, Kong GH, Wang T, et al. (1995) Activation
of ZAP-70 kinase activity by phosphorylation of tyrosine 493 is required for
lymphocyte antigen receptor function. EMBO Journal 14: 2499–2508.
207. Mege D, Di Bartolo V, Germain V, Tuosto L, Michel F, et al. (1996) Mutation
of tyrosines 492/493 in the kinase domain of ZAP-70 affects multiple T-cell
receptor signaling pathways. Journal of Biological Chemistry 271: 32644–
32652.
208. Wange RL, Guitian R, Isakov N, Watts JD, Aebersold R, et al. (1995)
Activating and inhibitory mutations in adjacent tyrosines in the kinase domain
of ZAP-70. Journal of Biological Chemistry 270: 18730–18733.
209. Brdicka T, Pavlistova D, Leo A, Bruyns E, Korinek V, et al. (2000)
Phosphoprotein associated with glycosphingolipid-enriched microdomains
(PAG), a novel ubiquitously expressed transmembrane adaptor protein, binds
the protein tyrosine kinase Csk and is involved in regulation of T cell activation.
Journal of Experimental Medicine 191: 1591–1604.
210. Yasuda K, Nagafuku M, Shima T, Okada M, Yagi T, et al. (2002) Cutting
edge: Fyn is essential for tyrosine phosphorylation of Csk-binding protein/
phosphoprotein associated with glycolipid-enriched microdomains in lipid rafts
in resting T cells. Journal of Immunology 169: 2813–2817.
211. Zhang SQ, Yang W, Kontaridis MI, Bivona TG, Wen G, et al. (2004) Shp2
regulates SRC family kinase activity and Ras/Erk activation by controlling Csk
recruitment. Molecular Cell 13: 341–355.
212. Berg KL, Siminovitch KA, Stanley ER (1999) SHP-1 regulation of p62DOK
tyrosine phosphorylation in macrophages. Journal of Biological Chemistry 274:
35855–35865.
213. Torres E, Rosen MK (2006) Protein-tyrosine kinase and GTPase signals
cooperate to phosphorylate and activate Wiskott-Aldrich syndrome protein
(WASP)/neuronal WASP. Journal of Biological Chemistry 281: 3513–3520.
214. Banin S, Gout I, Brickell P (1999) Interaction between Wiskott-Aldrich
Syndrome protein (WASP) and the Fyn protein-tyrosine kinase. Molecular
Biology Reports 26: 173–177.
215. Paz PE, Wang S, Clarke H, Lu X, Stokoe D, et al. (2001) Mapping the Zap-70
phosphorylation sites on LAT (linker for activation of T cells) required for
recruitment and activation of signalling proteins in T cells. Biochemical Journal
356: 461–471.
216. Houtman JCD, Higashimoto Y, Dimasi N, Cho S, Yamaguchi H, et al. (2004)
Binding specificity of multiprotein signaling complexes is determined by both
cooperative interactions and affinity preferences. Biochemistry 43: 4170–4178.
15
August 2014 | Volume 9 | Issue 8 | e104240
Phosphorylation Site Dynamics of Early TCR Signaling
217. DeBell K, Graham L, Reischl I, Serrano C, Bonvini E, et al. (2007)
Intramolecular regulation of phospholipase C-c1 by its C-terminal Src
homology 2 domain. Molecular and Cellular Biology 27: 854–863.
218. Jordan MS, Sadler J, Austin JE, Finkelstein LD, Singer AL, et al. (2006)
Functional hierarchy of the N-terminal tyrosines of SLP-76. Journal of
Immunology 176: 2430–2438.
219. Bubeck Wardenburg J, Fu C, Jackman JK, Flotow H, Wilkinson SE, et al.
(1996) Phosphorylation of SLP-76 by the ZAP-70 protein-tyrosine kinase is
required for T-cell receptor function. Journal of Biological Chemistry 271:
19641–19644.
PLOS ONE | www.plosone.org
220. Yablonski D, Kadlecek T, Weiss A (2001) Identification of a phospholipase Cc1 (PLC-c1) SH3 domain-binding site in SLP-76 required for T-cell receptormediated activation of PLC-c1 and NFAT. Molecular and Cellular Biology 21:
4208–4218.
221. Bunnell SC, Diehn M, Yaffe MB, Findell PR, Cantley LC, et al. (2000)
Biochemical interactions integrating Itk with the T cell receptor-initiated
signaling cascade. Journal of Biological Chemistry 275: 2219–2230.
222. Qi Q, August A (2007) Keeping the (kinase) party going: SLP-76 and ITK
dance to the beat. Science’s STKE 2007: pe39.
16
August 2014 | Volume 9 | Issue 8 | e104240