Phosphorylation-dependent 14-3-3 binding to LRRK2 is impaired by common mutations of familial Parkinson's disease.

Abstract

Recent studies show that mutations in Leucine Rich Repeat Kinase 2 (LRRK2) are the cause of the most common inherited and some sporadic forms of Parkinson's disease (PD). The molecular mechanism underlying the pathogenic role of LRRK2 mutations in PD remains unknown.
Using affinity purification and mass spectrometric analysis, we investigated phosphorylation sites and binding proteins of LRRK2 purified from mouse brain. We identified multiple phosphorylation sites at N-terminus of LRRK2 including S910, S912, S935 and S973. Focusing on the high stoichiometry S935 phosphorylation site, we developed an anti-pS935 specific antibody and showed that LRRK2 is constitutively phosphorylated at S935 in various tissues (including brain) and at different ages in mice. We find that 14-3-3 proteins (especially isoforms γ and η) bind LRRK2 and this binding depends on phosphorylation of S935. The binding of 14-3-3, with little effect on dimer formation of LRRK2, confers protection of the phosphorylation status of S935. Furthermore, we show that protein kinase A (PKA), but not LRRK2 kinase itself, can cause the phosphorylation of LRRK2 at S935 in vitro and in cell culture, suggesting that PKA is a potential upstream kinase that regulates LRRK2 function. Finally, our study indicates that the common PD-related mutations of LRRK2, R1441G, Y1699C and G2019S, decrease homeostatic phosphorylation levels of S935 and impair 14-3-3 binding of LRRK2.
LRRK2 is extensively phosphorylated in vivo, and the phosphorylation of specific sites (e.g. S935) determines 14-3-3 binding of LRRK2. We propose that 14-3-3 is an important regulator of LRRK2-mediated cellular functions. Our study suggests that PKA, a cAMP-dependent kinase involved in regulating dopamine physiology, is a potential upstream kinase that phosphorylates LRRK2 at S935. Furthermore, the reduction of phosphorylation/14-3-3 binding of LRRK2 due to the common familial PD-related mutations provides novel insight into the pathogenic mechanism of LRRK2-linked PD.

Full text

Page 1

Phosphorylation-Dependent 14-3-3 Binding to LRRK2 Is
Impaired by Common Mutations of Familial Parkinson’s
Disease
Xianting Li1, Qing Jun Wang2¤, Nina Pan1, Sangkyu Lee3, Yingming Zhao3, Brian T. Chait2, Zhenyu Yue1*
1Department of Neurology and Neuroscience, Mount Sinai School of Medicine, New York, New York, United States of America, 2Laboratory of Mass Spectrometry and
Gaseous Ion Chemistry, The Rockefeller University, New York, New York, United States of America, 3Ben May Department for Cancer Research, The University of Chicago,
Chicago, Illinois, United States of America
Abstract
Background: Recent studies show that mutations in Leucine Rich Repeat Kinase 2 (LRRK2) are the cause of the most
common inherited and some sporadic forms of Parkinson’s disease (PD). The molecular mechanism underlying the
pathogenic role of LRRK2 mutations in PD remains unknown.
Methodology/Principal Findings: Using affinity purification and mass spectrometric analysis, we investigated
phosphorylation sites and binding proteins of LRRK2 purified from mouse brain. We identified multiple phosphorylation
sites at N-terminus of LRRK2 including S910, S912, S935 and S973. Focusing on the high stoichiometry S935 phosphorylation
site, we developed an anti-pS935 specific antibody and showed that LRRK2 is constitutively phosphorylated at S935 in
various tissues (including brain) and at different ages in mice. We find that 14-3-3 proteins (especially isoforms c and g) bind
LRRK2 and this binding depends on phosphorylation of S935. The binding of 14-3-3, with little effect on dimer formation of
LRRK2, confers protection of the phosphorylation status of S935. Furthermore, we show that protein kinase A (PKA), but not
LRRK2 kinase itself, can cause the phosphorylation of LRRK2 at S935 in vitro and in cell culture, suggesting that PKA is a
potential upstream kinase that regulates LRRK2 function. Finally, our study indicates that the common PD-related mutations
of LRRK2, R1441G, Y1699C and G2019S, decrease homeostatic phosphorylation levels of S935 and impair 14-3-3 binding of
LRRK2.
Conclusions/Significance: LRRK2 is extensively phosphorylated in vivo, and the phosphorylation of specific sites (e.g. S935)
determines 14-3-3 binding of LRRK2. We propose that 14-3-3 is an important regulator of LRRK2-mediated cellular functions.
Our study suggests that PKA, a cAMP-dependent kinase involved in regulating dopamine physiology, is a potential
upstream kinase that phosphorylates LRRK2 at S935. Furthermore, the reduction of phosphorylation/14-3-3 binding of
LRRK2 due to the common familial PD-related mutations provides novel insight into the pathogenic mechanism of LRRK2-
linked PD.
Citation: Li X, Wang QJ, Pan N, Lee S, Zhao Y, et al. (2011) Phosphorylation-Dependent 14-3-3 Binding to LRRK2 Is Impaired by Common Mutations of Familial
Parkinson’s Disease. PLoS ONE 6(3): e17153. doi:10.1371/journal.pone.0017153
Editor: Tsuneya Ikezu, Boston University School of Medicine, United States of America
Received December 10, 2010; Accepted January 20, 2011; Published March 1, 2011
Copyright: ? 2011 Li 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.
Funding: This work was supported by grants from the US National Institutes of Health/National Institute of Neurological Disorders and Stroke (NIH/NINDS)
NS061152 (Z.Y.), NS060809 (Z.Y.), RNS055683A (Z.Y.), Michael J. Fox Foundation (Z.Y.). 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.
* E-mail: zhenyu.yue@mssm.edu
¤ Current address: Department of Molecular and Cellular Biochemistry, College of Medicine, University of Kentucky, Lexington, Kentucky, United States of America
Introduction
Parkinson’s disease (PD) is a major human neurodegenerative
disease clinically characterized by the motor function deficits as
well as non-motor impairment. The most prominent neuropath-
ological features are the loss of midbrain dopaminergic neurons
and deposit of intracellular Lewy Body consisting mainly of a-
synuclein. The pathogenic mechanism of PD remains largely
undefined. However, the recent identification of genetic mutations
that are associated with familial PD provides an entry to uncover
cellular and molecular pathways that lead to the disease.
Mutations in Leucine-Rich Repeat Kinase 2 (LRRK2) have been
linked to the most common familial form and some sporadic forms
of PD [1,2]. Because many LRRK2 mutation carriers exhibit
typical PD symptoms indistinguishable from idiopathic PD cases,
it is hypothesized that an understanding of the biology and
pathophysiology of LRRK2 will provide new opportunities to
develop effective treatments of PD.
LRRK2 is a large complex protein of 280 kD containing two
important enzymatic domains, serine/threonine kinase and ROC
GTPase; it also carries several conserved protein motifs including
leucine-rich repeats (LRR), the C-terminal of ROC (COR)
domain and a WD40 repeat. Previous biochemical analysis
demonstrates the kinase and GTPase activity of LRRK2, and
these activities are apparently altered by several PD-linked
mutations [3,4,5,6], leading to the hypothesis of ‘‘gain-of-
PLoS ONE | www.plosone.org1 March 2011 | Volume 6 | Issue 3 | e17153

Page 2

function’’ in LRRK2 mutants with enhanced kinase activity
[5,7,8]. LRRK2 protein can be modified posttranslationally (e.g.
by phosphorylation) and forms a dimer in vitro and in vivo [9,10,11].
These alterations may play a critical role in regulating LRRK2
biochemical activities and cellular functions. However, until now
little has been known about the normal function of LRRK2 or the
pathogenic pathways mediated by PD-linked mutant LRRK2.
Several studies show that the mutants of LRRK2 impair the
outgrowth or maintenance of neurites in primary neuronal
cultures, whereas reduced or abolished expression of LRRK2
has the opposite effect. In addition, inhibition of kinase activity
seems to alleviate the toxic phenotype caused by LRRK2 mutants
[8,12,13,14].
To elucidate the cellular pathway and pathogenic role of
LRRK2 in PD, we investigated LRRK2 protein modifications and
interactors in the brain. We show that LRRK2 is phosphorylated
at multiple sites. Our study reveals that 14-3-3s bind LRRK2 and
the binding depends on the phosphorylation of S935. Further-
more, we show that protein kinase A (PKA) causes phosphoryla-
tion of LRRK2 at S935 in vitro and in cell culture, implicating PKA
pathway in regulating LRRK2 function. Finally, our study
suggests that common PD mutations of LRRK2 impair phos-
phorylation levels of S935 as well as14-3-3 binding. Our data,
therefore, provide molecular insight into the regulation of LRRK2
and suggests a potential mechanism for LRRK2-mediated PD
pathogenesis.
Results
Identification of phosphorylation sites in LRRK2 from
mouse brain
We previously reported the purification of FLAG-tagged LRRK2
protein from BAC transgenic mice [3]. For phosphorylation site
identification, the purified LRRK2 protein was digested in-gel using
various proteasesand the resultingproteolyticpeptideswere analyzed
by multiple mass spectrometer methods including MALDI-QqTOF,
MALDI-ion trap (LCQ DECA XP), and nano-HPLC/ velos LTQ
Orbitrap. The resulting MS/MS data were used to identify proteins
and protein modifications. The results reveal 3 serine phosphoryla-
tion sites (S910, 935 and 973) from tryptic peptides and 1 serine
phosphorylation site (S912) in chymotryptic peptides of LRRK2,
respectively (Figure 1A) (Figures S1, S2, S3 and S4). Interestingly,
stoichiometry of all 4 serine phosphorylation appears high, as the
ratios of MS/MS spectra for modified peptides versus unmodified
peptides areall more than 30%. This result indicates the relative high
probability of LRRK2 phosphorylation at these sites in the brain
(Figures S1, S2, S3 and S4).
In this initial study, we focused on the analysis of the high
stoichiometry S935 phosphorylation. We developed an antibody
raised against phosphorylated S935 (pS935) peptide. The anti-
pS935 antibody detects a strong signal in purified FLAG-LRRK2
protein from BAC transgenic brains, while the signal is completely
abolished upon the treatment with calf-intestinal alkaline phos-
phatase (CIAP) (Figure 1B). The loss of phosphorylation at S935
with alkaline phosphatase treatment was also confirmed by mass
spectrometric analysis (Figure S5). The antibody also detected
pS935 signal in FLAG-LRRK2 protein isolated from transfected
HEK-293T cells. In contrast, FLAG-LRRK2 mutant S935A,
where serine 935 was replaced with alanine, was not recognized by
this antibody, even though much more mutant protein (compared
to wild type) was loaded in the gel (Figure 1C). We also examined
the pS935 levels in purified FLAG-LRRK2 from different tissues
and at different ages in the brain. The results show that FLAG-
LRRK2 is phosphorylated at S935 in lung, spleen and kidney in
addition to brain, and the pS935 levels relative to the total FLAG-
LRRK2 protein amount are comparable among these tissues
(Figure 1D). Moreover, the relative levels of pS935 do not change
significantly at different ages in the brain (Figure 1E). The above
results suggest that pS935 are maintained at a constant level under
normal condition.
Identification of 14-3-3s in LRRK2 protein complex and
specific 14-3-3 isoforms as preferential LRRK2-binding
proteins
We sought to identify LRRK2-binding proteins in the brain by
analyzing proteins that were affinity-isolated with FLAG-LRRK2
from BAC transgenic brain. We isolated only the proteins unique
to the transgenic (compared to non-transgenic control) and
performed mass spectrometry analysis. We identified several
isoforms of 14-3-3 proteins, such as c, g, f and e (Figure 2A,
arrow) (Figures S6, S7 and S8) that are specifically isolated by
FLAG-LRRK2. Using commercial 14-3-3 isoform-specific anti-
bodies, we found 14-3-3c, g, f as well as b, h in the
immunoprecipitated products (Figure 2B). To further evaluate
various 14-3-3 isoform binding to LRRK2, we co-expressed
FLAG-LRRK2 with individual myc-tagged 14-3-3 isoforms in
HEK-293T cells and tested their binding by co-immunoprecipi-
tation (co-IP) analysis. The results indicated that, although all six
isoforms co-IP with LRRK2, the c and g forms pulled down much
more LRRK2 protein than did the other isoforms (Figure 2C). In a
parallel experiment, we incubated each purified GST-14-3-3
isoform protein with BAC-FLAG-LRRK2 transgenic mouse brain
lysate and, followed by a GST pull-down assay. The result again
showed enhanced binding of 14-3-3c and g with LRRK2,
compared to 14-3-3f, e and h (Figure 2D). The above data
collectively suggests that LRRK2 prefers 14-3-3c and g as binding
partners and that the binding is likely mediated through direct
protein-protein interactions.
Binding of 14-3-3 to LRRK2 depends on phosphorylation
of S935
14-3-3 proteins are a family of conserved regulatory molecules
that control various cellular functions by directly binding target
signaling proteins, including kinases, phosphatases and receptors
[15]. 14-3-3 binding was found frequently in a phospho-serine
dependent manner and occurs in a consensus motif (RSXpSXP
and RXY/FXpSXP) [16]. We next investigated whether pS935 of
LRRK2 is relevant to the binding of 14-3-3 (although S935
surrounding sequence does not fit well with the known consensus
motif). First, we transfected HEK-293T cells with plasmids
containing Myc-14-3-3c and FLAG-LRRK2 or FLAG-LRRK2-
S935A. We found that anti-Myc antibody co-immunoprecipitated
14-3-3c with FLAG-LRRK2 but not FLAG-LRRK2-S935A
(Figure 3A). Reciprocal Co-IP assay using anti-FLAG antibody
showed a similar result that 14-3-3 was pulled down only with
LRRK2 wild type but not S935A mutant (Figure 3B). To further
test the role of phophorylated S935 in binding, we incubated
purified GST-14-3-3c and BAC-FLAG-LRRK2 transgenic mouse
brain lysate in the presence of phospho-peptide QRHSNpSLGP
(containing pS935) and non-phospho-peptide QRHSNSLGP
(control) at various concentrations. GST-14-3-3c pulled down
FLAG-LRRK2 in the absence of the peptides; however, adding
the pS935-containing peptide effectively inhibited the co-precip-
itation of GST-14-3-3c with FLAG-LRRK2 in a dosage-
dependent manner. By contrast, the control peptide has no
inhibitory effect toward the pull-down of GST-14-3-3c and
FLAG-LRRK2 under the same conditions (Figure 3C). These
Phosphorylation and 14-3-3 Binding of LRRK2 and PD
PLoS ONE | www.plosone.org2 March 2011 | Volume 6 | Issue 3 | e17153

Page 3

results indicate that the phospho-peptide (but not the control
peptide) competed for 14-3-3 binding, therefore demonstrating the
critical role of pS935 in 14-3-3 binding with LRRK2.
Furthermore, we co-transfected FLAG-LRRK2 with Myc-14-3-
3c wild type, mutant K50E or V181D, which was reportedly
impaired in binding phospho-serine of target proteins [17,18]. In
Figure 1. Identification of LRRK2 phosphorylation sites in BAC transgenic brain. (A) Table of the identified phosphorylation sites within
tryptic peptides from purified brain FLAG-LRRK2 protein. The estimate of the stoichiometry of phosphorylation is based on the ratio of number of MS/
MS spectra observed. The exact location of each proteolytic peptide is shown in the position column. (B) Purified brain FLAG-LRRK2 protein was
treated with or without Calf-intestinal alkaline phosphatase (CIAP), followed by Western blot analysis with anti-pS935 antibody or anti-FLAG antibody.
(C) FLAG-LRRK2 WT or FLAG-LRRK2 S935A mutant plasmid was transfected into HEK-293T cells; FLAG-tagged proteins were then immunoprecipitated
and analyzed by anti-pS935 antibody or anti-FLAG antibody via Western blot analysis. (D) FLAG-LRRK2 was purified from different BAC-transgenic
tissues and western blot was performed to detect pS935 and total FLAG-LRRK2 levels. No significant difference in the ratio of pS935/total LRRK2 was
observed between different tissues. Data are presented as mean value (6 SEM) from three mice. (E) FLAG-LRRK2 was purified from 3 months and 18
months transgenic mouse brain and western blot was performed to detect pS935 and total FLAG-LRRK2 levels. No significant difference in the ratio of
pS935/total LRRK2 was observed between 3 months and 18 months. The Western blot signals were quantified by using LI-COR and Odyssey software.
Data are presented as mean value (6 SEM) from three mice.
doi:10.1371/journal.pone.0017153.g001
Phosphorylation and 14-3-3 Binding of LRRK2 and PD
PLoS ONE | www.plosone.org3 March 2011 | Volume 6 | Issue 3 | e17153

Page 4

contrast to wild type 14-3-3, mutant K50E or V181D completely
lost the binding to LRRK2 as shown in experiments that either use
anti-FLAG or anti-Myc antibody for the co-IP (Figure 3 D, E),
consistent with the notion that 14-3-3 binding depends on
phospho-serine in LRRK2.
To study the relationship of 14-3-3 binding and S935
phosphorylation, we transfected FLAG-LRRK2 in the absence or
presence of 14-3-3 wild type or mutant K50E/V181D, which was
previously shown as a dominant negative mutant (DN) that
abolishes wild type14-3-3 binding to the target proteins [19]. The
result showed that the relative pS935 levels, as measured with anti-
pS935 staining over anti-FLAG staining, were significantly higher
with co-transfection of Myc-14-3-3c and FLAG-LRRK2 than
FLAG-LRRK2 transfection alone. In contrast, co-transfection of
FLAG-LRRK2withMyc-14-3-3DN markedly reduced therelative
pS935 levels of LRRK2 (Figure 3F). These results suggest that 14-3-
3 binding prevents the dephosphorylation of LRRK2 at S935.
S935A mutation does not affect LRRK2 dimerization
One of the important roles of 14-3-3 binding is to provide
structural support for the target protein dimer formation via the
self-dimerization of 14-3-3 proteins [20]. Since LRRK2 is known
to form dimers in vitro and in vivo [9,10,11], we next investigated
whether 14-3-3 interaction of LRRK2 is important for the
dimerization of LRRK2. HEK-293T cells were co-transfected
with HA-tagged LRRK2 and FLAG-LRRK2 or FLAG-LRRK2
S935A mutant. The reciprocal co-IP experiments show that
FLAG-LRRK2 wild type and FLAG-LRRK2 S935A mutant did
Figure 2. Identification of 14-3-3s as LRRK2 interaction proteins. (A) Coomassie blue staining showing the affinity-purified FLAG-LRRK2 and
its binding proteins from FLAG-LRRK2 transgenic brain. The positions of the FLAG-LRRK2 and 14-3-3 bands are labeled with arrows. Wild-type mouse
brain was used as the control for background proteins during purification. (B) Western blot analysis was performed to detect the presence of various
14-3-3 isoforms in association with the affinity purified FLAG-LRRK2 protein. Antibodies against specific isoforms of 14-3-3 were used. Wild type
mouse tissue was used as the control. (C) FLAG-LRRK2 and different Myc-14-3-3 isoform plasmids were co-transfected into HEK-293T cells and
Immunoprecipitation was performed using anti-Myc antibody, followed by the analysis of protein levels using indicated antibodies. (D) FLAG-LRRK2
plasmid was transfected into HEK-293T cells and transfected cell lysate was incubated with different purified GST-14-3-3 isoforms as indicated. The
GST-pull down assay was assayed by Western blot analysis of the FLAG-LRRK2.
doi:10.1371/journal.pone.0017153.g002
Phosphorylation and 14-3-3 Binding of LRRK2 and PD
PLoS ONE | www.plosone.org4 March 2011 | Volume 6 | Issue 3 | e17153

Page 5

Figure 3. 14-3-3 binding to LRRK2 S935 is phosphorylation-dependent. (A, B) FLAG-LRRK2 Wt or S935A mutant plasmid was transfected
with or without Myc-14-3-3c into HEK-293T cells. Co-immunoprecipitation experiments were performed using either anti-myc (A) or anti-FLAG (B)
antibody, followed by Western blot analysis using the indicated antibodies. (C) Sepharose bead-conjugated GST-14-3-3c was incubated with FLAG-
LRRK2 transgenic brain lysate in the presence of various concentrations of phosphorylated peptide (pS935) or non-phosphorylated peptide. The
binding FLAG-LRRK2 protein was pulled down and analyzed by Western blot analysis using anti-FLAG antibody. (D, E) FLAG-LRRK2 was transfected
alone, with Myc-14-3-3c Wt, with Myc-14-3-3 g K50E mutant, or with Myc-14-3-3c V181D mutant into HEK-293T cells. The co-IP assay was performed
using either anti-Myc (D) or anti-FLAG (E) antibody, followed by Western blot analysis using indicated antibodies. (F) FLAG-LRRK2 was transfected
alone, with Myc-14-3-3c wt or Myc-14-3-3c dominant negative (DN) mutant (K50E/V181D) into HEK-293T cells. The total FLAG-LRRK2 was
immunoprecipitated and analyzed by Western blot using anti-pS935 and anti-FLAG antibodies. The Western blot shows the results from three
independent transfections. The quantification of the signals was done by using LI-COR imaging and Odyssey software. The data was analyzed by One-
way ANOVA (** P,0.01). Data are presented as mean value (6 SEM) from three independent experiments.
doi:10.1371/journal.pone.0017153.g003
Phosphorylation and 14-3-3 Binding of LRRK2 and PD
PLoS ONE | www.plosone.org5 March 2011 | Volume 6 | Issue 3 | e17153