Abstract
Alterations of serine/threonine phosphorylation of the cardiac proteome
are a hallmark of heart failure. However, the contribution of tyrosine
phosphorylation (pTyr) to the pathogenesis of cardiac hypertrophy
remains unclear. We use global mapping to discover and quantify
site-specific pTyr in two cardiac hypertrophic mouse models, i.e.,
cardiac overexpression of ErbB2 (TgErbB2) and α myosin heavy chain
R403Q (R403Q-αMyHC Tg), compared to control hearts. From this, there
are significant phosphoproteomic alterations in TgErbB2 mice in right
ventricular cardiomyopathy, hypertrophic cardiomyopathy (HCM), and
dilated cardiomyopathy (DCM) pathways. On the other hand, R403Q-αMyHC
Tg mice indicated that the EGFR1 pathway is central for cardiac
hypertrophy, along with angiopoietin, ErbB, growth hormone, and
chemokine signaling pathways activation. Surprisingly, most myofilament
proteins have downregulation of pTyr rather than upregulation.
Kinase-substrate enrichment analysis (KSEA) shows a marked
downregulation of MAPK pathway activity downstream of k-Ras in TgErbB2
mice and activation of EGFR, focal adhesion, PDGFR, and actin
cytoskeleton pathways. In vivo ErbB2 inhibition by AG-825 decreases
cardiomyocyte disarray. Serine/threonine and tyrosine phosphoproteome
confirm the above-described pathways and the effectiveness of AG-825
Treatment. Thus, altered pTyr may play a regulatory role in cardiac
hypertrophic models.
Subject terms: Proteomics, Cardiac hypertrophy
__________________________________________________________________
Quantitative global phosphotyrosine proteomics of two mouse models of
hypertrophic cardiomyopathy reveals that tyrosine kinase inhibitors may
be a future therapeutic approach for treating hypertrophic
cardiomyopathy.
Introduction
Familial hypertrophic cardiomyopathy (HCM) increases the left ventricle
(LV) wall thickness; abnormal loading conditions cannot explain this
alteration. Mutations in genes encoding sarcomere proteins are common
in HCM patients (40–60%)^[54]1. Studying post-translational
modifications (PTMs) of the sarcomere proteins (or Ca^2+ handling
proteins) can offer a unique opportunity to understand better how
genetic disorders lead to cardiac dysfunction and discover potential
targets for therapy^[55]2–[56]4. For instance, PTMs of cardiac Troponin
I (cTnI), a sarcomere protein centrally involved in myocardial
contractility regulation, have been extensively studied, particularly
for the functional role of protein kinase A-dependent
phosphorylation^[57]5. Notably, phosphorylation of cTnI-S22/23—one of
the most relevant regulatory sites of cTnI—is downregulated in human
heart failure (HF)^[58]6 and leads to contractile dysfunction^[59]7.
Tyrosine phosphorylation (pTyr) is essential for cardiac structural
development and myofibril organization during
embryogenesis^[60]8,[61]9. For example, several tyrosine phosphatases
have been linked to heart disease and even proposed as a therapeutic
target for some conditions. Moreover, mutations of the tyrosine-protein
phosphatase non-receptor type 11 (PTPN11) can lead to HCM or dilated
cardiomyopathy (DCM)^[62]10,[63]11. In the heart, PTPN11 likely plays a
role in systolic dysfunction produced by pressure-overload^[64]12. Acid
phosphatase 1 (ACP1) is another tyrosine phosphatase associated with
cardiac pathophysiology. Accordingly, its deletion protects against
stress-induced cardiomyopathy^[65]13. Our group first showed that
cTnI-Y26 phosphorylation is readily detected in healthy human hearts
and downregulated in human HF and DCM^[66]14. However, how these
alterations contribute to the onset and progression of cardiac disease
remains poorly understood. A better grasp of additional site-specific
changes of individual tyrosine-phosphorylated sites would help in this
direction substantially.
The present study applied a label-free and tandem mass tagging (TMT)
quantitative global phosphotyrosine proteomics approach to determine
which sarcomere sites have altered amounts of Tyr phosphorylation at
specific sites in two unrelated models of HCM. The first model is
secondary to the overexpression of the tyrosine kinase receptor
ErbB2^[67]15,[68]16. The second recapitulates features of the human
disease, more specifically, an R403Q mutation of the myofilament
protein myosin heavy chain, distinctive of familial HCM^[69]17. TgErbB2
mice initially develop a striking, concentric cardiac hypertrophy,
which evolves to diffuse fibrosis and myocyte disarray^[70]16 with HCM.
This line of mice also has abnormal calcium handling, is prone to
arrhythmias, and to developing hypertrophic obstructive
cardiomyopathy^[71]2. Similarly, R403Q-αMyHC mice reproduce human
familial HCM by progressing from mild hypertrophy and fibrosis to overt
myocyte disarray, HF, and arrhythmias^[72]17.
The main impetus behind this study was to determine whether triggering
HCM through different mechanisms elicits similar pTyr-related
pathways/regulatory sites within the heart. In doing so, the
manipulation of these pTyr sites would, in turn, offer new
opportunities for therapeutic targeting in different forms of human HCM
Results
Immunoblotting reveals decreased cardiac pTyr in HCM models
First, we performed immunoblot analysis of the myocardium of TgErbB2,
R403Q-αMyHC, and non-transgenic (Ntg) mice to estimate pTyr. There was
a global decline in pTyr in whole heart homogenates of TgErbB2
(p = 0.0298) and R403Q-αMyHC (p = 0.003) mice compared to Ntg
(Fig. [73]1a, b). The pTyr signal was normalized to troponin I
expression levels (Fig. [74]1a). Although proto-oncogene
tyrosine-protein kinase Src (c-Src) is downstream of ErbB2 signaling,
we did not anticipate its increased expression in R403Q-αMyHC mice
(Fig. [75]1c, d, p = 0.0414) when normalized to GAPDH western blot
signals. The c-Src activity (c-Src Tyr416 phosphorylation) appeared to
be enhanced in R403Q-αMyHC mice compared to Ntg; however, it did not
reach statistical significance (Fig. [76]1e, f, one-way ANOVA was used
to compare groups). Altogether these data suggest that pTyr is a
broadly distributed PTMs in the myocardium. Hence, its upregulation or
downregulation may play a regulatory role in the disease progression of
non-ischemic cardiomyopathies, as in the case of familial
cardiomyopathy (HCM) and DCM, as part of the pathophysiologic response
to the underlying disease.
Fig. 1. Global pTyr is reduced in cardiac hypertrophy associated with ErbB2
overexpression or myosin heavy chain mutation.
[77]Fig. 1
[78]Open in a new tab
a Western Blots show a decrease in pTyr on TgErbB2 (p = 0.0298) mice
and R403Q-α-MyHC (p = 0.003) heart homogenates. The Troponin I signal
shows corresponding total protein loading. b Densitometry analysis
shows that after the pTyrosine signal was normalized to total cTnI, the
TgErbB2 and R403Q-α-MyHC mice significantly decreased in pTyr. c
Western Blots show expression levels of total Src and p-Src (Tyr 416)
and GAPDH as a loading control. d Densitometry analysis revealed that
Src levels are significantly elevated in R403Q-α-MyHC (p = 0.0282),
whereas e and f Show that p-Src phosphorylation levels are elevated in
R403Q-α-MyHC but did not reach statistical significance. Signals in d
and e were normalized to total GAPDH, whereas in f pSRC was normalized
to total SRC. Ntg (n = 3), TgErbB2 (n = 3), and R403Q-α-MyHC (n = 3),
error bars represent S.E.M. One-way ANOVA was used to compare groups.
pTyr in myofilaments and cross-talk with serine/threonine
kinases/phosphatases
Next, we adopted a global TMT phospho-tyrosine proteomic approach to
compare the whole heart pTyr profile of Ntg, TgErbB2, and R403Q-αMyHC
mice (Fig. [79]2). We hypothesized that unbiased global whole heart
pTyr profiling would provide a complete landscape about specific
tyrosine sites in essential cardiac-specific proteins and tyrosine
kinases, thus providing clues about which tyrosine kinases mediate
cardiac pTyr. A total of 1,800 peptide spectra matched (PSM) were
collected from the heart ventricle whole proteome, and 50% were
identified with high confidence resulting in 213 (R403Q-αMyHC), 214
(Ntg), 217 (TgErbB2) unique tyrosine-phosphorylated peptides
corresponding to 499 proteins (Fig. [80]2g). Tables with all identified
phosphoproteins and phosphopeptides are provided in Supplementary
Data [81]3. The mass spectrometry data quality, intensity, and
distribution post-median sweep normalization are included in
Supplementary Fig. [82]1a, b. Confidence of phosphorylated site
localization was evaluated for annotation, and a score of more than 49%
was required. This dataset indicates a comparable yield of peptide
identification was achieved, reproducibility on the enrichment of pTyr
peptides, and high-quality MS/MS data using similar approaches to
others^[83]18–[84]20.
Fig. 2. Workflow for global whole heart TMT phospho-tyrosine proteomics
profiling.
[85]Fig. 2
[86]Open in a new tab
a Ventricular tissues were rinsed with PBS and subjected to Heat
Stabilization using a T1 Tissue Stabilizor to maximize phosphorylation
preservation, then snap-frozen in liquid nitrogen and stored at −80 °C
until further use. b Ntg, TgErbB2, and R403Q-αMyHC mice hearts (n = 5
per group) were processed in parallel to obtain whole heart lysates;
30 mg of protein lysates were used from each mouse heart for
in-solution trypsin digestion, a total of 150 mg of trypsinized protein
per genotype were C18 desalted and lyophilized at −20 °C for three
days. c A total of 150 mg of tryptic peptides (pooled material from 5
hearts per group) were enriched for pTyrosine using immunoaffinity
precipitation (Protein Agarose Beads + 300 micrograms of mAb Anti-pTyr)
and desalted on C18 stage tips. d RP-HPLC ESI (electrospray ionization)
MS/MS was performed on an LTQ Orbitrap Elite (Thermo Scientific) for a
120 min on a linear acetonitrile gradient (4−40%). Raw data were
searched with Mascot 2.3, and label-free quantification with MS1
extracted ion chromatograms were performed using MaxQuant. This
approach was repeated three times with samples from all three groups
running parallel; 15 hearts per genotype in three technical replicates.
e A fraction of the flow-through of pTyr enrichment was used for a
9-plexTMT; three replicates of each genotype (Ntg, TgErbB2, and
R403Q-αMyHC) were pooled, fractionated in RPLC for full proteome
quantification, f Data from pTyrosine and Full Proteome were processed
in a similar way, pTyrosine signals were normalized to matched total
protein expression to normalize data. Analyses are detailed in the text
and [87]supplemental data. g Shows the overlap of phosphorylated
proteins found in the three groups, whereas h shows the overlapping
tyrosine-phosphorylated peptides among the three groups.
The global TMT phospho-tyrosine proteomics approach identified 499
different tyrosine-phosphorylated proteins, 294 tyrosine-phosphorylated
proteins that overlap among the Ntg, and the myocardium from two
cardiac hypertrophic models (TgErbB2 and R403Q-αMyHC). At the peptide
level, 217 pTyr sites were detected with an overlap of 178
tyrosine-phosphorylated residues (Fig. [88]2g, h).
Next, the focus was on the cardiac-specific proteins from the
myofilament apparatus. Nine major myofilament proteins were
tyrosine-phosphorylated. They harbor multiple pTyr amino acid sites,
many of which are novel (See Supplementary Fig. [89]2); the best
example is Titin, with 36 pTyr amino acid sites (See Supplementary
Fig. [90]3). Also, seventeen tyrosine kinases had detectable
tyrosine-phosphorylated peptides; thus, they could potentially be
involved in regulating cardiac myofilament pTyr.
In contrast, only two tyrosine phosphatases (Ptpn11 and Ptpra) had
detectable tyrosine-phosphorylated peptides; therefore, they are likely
to play a role in regulating cardiac pTyr levels in the sarcomere. In
addition to the number of tyrosine kinases found with phosphorylated
peptides, fifteen serine/threonine kinases had detectable
tyrosine-phosphorylated peptides; however, only one serine/threonine
phosphatase (Ppp1r12b) demonstrated tyrosine-phosphorylated peptides.
These data show that pTyr is broadly distributed in the heart proteome
(cardiac sarcomere and other sub-proteomes) and that a cross-talk
between tyrosine kinases/phosphatases with serine/threonine
kinases/phosphatases might be regulating cardiac sarcomere pTyr.
ErbB2 cardiac overexpression and R403Q-αMyHC point mutation remodeled the
cardiac pTyr proteome
Next, we undertook bioinformatic analyses to define the impact of ErbB2
cardiac overexpression and R403Q mutation of the cardiac myosin heavy
chain on the specific pTyr changes. We determined the signaling
transduction pathways associated with two unrelated types of HCM.
Spectral signal normalization of pTyr was performed over the whole
global heart TMT isobaric probes that quantified full proteome
(Fig. [91]2e, f), and statistical methods were performed, as previously
described^[92]21,[93]22. Then, principal component analysis (PCA) and
hierarchical clustering were applied to determine if the pTyr profiles
were similar within the groups. This approach revealed that the three
groups segregate by principal component 1 (PC1 = 46%) in full proteome
quantification (Fig. [94]3a). This finding was also evident after pTyr
normalization to full proteome; the three groups segregate along the
principal component (PC1 = 52.2%) (Fig. [95]3b). Of note, one
technical/biological replicate of R403Q-αMyHC was removed because it
had more than 50% absent pTyr peptides compared to the other two
replicates. Hence, it was considered a technical failure. Heatmaps of
hierarchically clustered up- or downregulated normalized pTyr peptides
helped visualize the technical reproducibility and the specificity of
how the genotype of cardiac hypertrophy largely influenced the pTyr
proteome remodeling (Fig. [96]3c, p < 0.05 by ANOVA).
Tyrosine-phosphorylated peptides are color-coded according to their
extracted chromatogram MS1 signal intensities normalized to TMT
isobaric label peptide MS2 intensity. Yellow is upregulated, and blue
is downregulated. This approach provided us with the global heart
proteome tyrosine peptide abundance. These results demonstrate that
Ntg, TgErbB2, and R403Q-αMyHC mice have a pTyr proteome with a distinct
signature, evidenced by PCA and heatmaps clustering in unsupervised
unbiased statistical methods.
Fig. 3. PTMSigDB reveals common pTyr signatures of cardiac proteome in HCM.
[97]Fig. 3
[98]Open in a new tab
a Unsupervised PCA of the full cardiac Proteome of Non-transgenic,
TgErbB2, and R403Q-αMyHC mice show distinct segregation of the
experimental groups. The groups segregate along with the first
principal component (PC #1), which accounts for 46% of the total
correlation in the expression of Full Proteome (normalized, mean = 0,
variance = 1). b Unsupervised PCA of the pTyrosine peptides signal
normalized to full proteome expression. The groups also segregate along
with the first principal component (PC #1), accounting for 52.2% of the
total correlation. c A heatmap was used to visualize the unsupervised
hierarchical clustering of pTyrosine peptides normalized to full
proteome expression with a p-value < 0.05 by ANOVA. Overrepresented
(yellow) and underrepresented (Blue). d Volcano plot showing the Log2
fold-change and –Log10 p-value of each pTyr peptide in the pairwise
comparison of Ntg and TgErbB2. The data points highlighted in red show
a significantly lower intensity, whereas the data in green show
significantly higher intensity. The significant sites were magnified in
the figure (dotted squares) to label some peptide sites. Magenta data
points match genes significantly associated with MsigDB disease
pathways. Cyan data point did not reach statistical significance
although point to EGFR1 pathway. e Volcano plot showing the Log2
fold-change and –Log10 p-value of each pTyr peptide in the pairwise
comparison of Ntg and R403Q-αMyHC. Red and green data points were
color-coded as above. With a similar magnification as above, cyan data
points show a significant association to EGFR1 Pathway by PTMsigDB. f
Molecular Signature DB (MSigDB) and Molecular PTMs Signature DB
(PTMsigDB) of Global pTyr data identified pathways that were
significantly altered.
Statistically significant changes (Log[2] fold change >1, -Log p-value
< 1.3) were detected in 23 phosphosites corresponding to 18 proteins in
TgErbB2 mice (Fig. [99]3d) and 45 phosphosites in 35 proteins in the
R403Q-αMyHC Tg mice (Fig. [100]3f) using ANOVA (p < 0.05). It is
noteworthy that pTyr of several peptides of the alpha-myosin heavy
chain (Myh7-Y386, Y410, Y1375), one peptide of the beta-myosin heavy
chain (Myh6-Y1349), and two peptides of Titin (Ttn-Y2118, Y21190) are
downregulated while cardiac Ttn Y-31324 pTyr is upregulated on TgErbB2
mice hearts (Fig. [101]3d). Interestingly, Myh7-Y410 phosphorylation is
downregulated in TgErbB2 and upregulated in R403Q-αMyHC Tg mice hearts
(Fig. [102]3e). In contrast, cTnI Y113 (Tnni3-Y113) pTyr is
downregulated in both HCM models but only reached statistical
significance in R403Q-αMyHC Tg mice hearts. Titin phosphorylation
(TtnY1881, Y1901, and Y33864) was significantly downregulated in the
latter model.
We used MATLAB, a custom-made PhoshoEnrichment software (M. Ayati),
which integrates the Molecular Signatures Database (MsigDB)^[103]23 and
a novel database of PTMs site-specific phosphorylation signatures of
kinases, perturbations, and signaling pathways (PTMsigDB)^[104]24. The
Broad Institute created the databases, which comprehend a well-curated
collection of annotated gene data sets and PTMs. PTMsigDB accounts for
phospho site-specific changes and their impact on activating or
inhibiting a given pathway, together with the most common experimental
systems biology perturbations. The advantage of these bioinformatics
tools is to assess the statistical significance of each pathway and
perturbation (gene-level and site level) using a hypergeometric test.
Here, the number of identified phosphosites and genes is used as the
population for the hypergeometric model parameter instead of all known
genes.
A pathway enrichment analysis was performed on significant genes/sites
using a gene-protein level with MSigDB. TgErbB2 mice’s significant
pathways are arrhythmogenic right ventricular cardiomyopathy (ARVC),
HCM, DCM, and Integrin AlphaIIB Beta3 Signaling (Fig. [105]3f).
R403Q-αMyHC Tg most significant pathways are the angiopoietin receptor
pathway, ErbB signaling pathway, growth hormone receptor signaling, and
chemokine signaling pathway (Fig. [106]3f).
Both Tg animal models showed a remarkable alteration in plakophilin-2
(PKP2) pTyr at the site Y119, although in TgErbB2 hearts, it was
down-regulated, and in R403Q-αMyHC Tg hearts, it was upregulated
(Fig. [107]3e). MsigDB analysis of global whole heart TMT flow through
the entire proteome is displayed in Supplementary Data [108]3. PKP2 is
a critical component of the myocardium’s desmosomes. Mutations in this
gene are associated with arrhythmogenic cardiomyopathy^[109]25–[110]27.
Therefore, disturbed phosphorylation of Plakophilin-2 could have a
functional impact.
In comparison, the PTMsigDB revealed that pathway analysis at the
phosphorylated site-specific level did not reach statistical
significance for the EGFR1 Pathway in TgErbB2 mice; only two
site-specific phosphosites matched the EGFR1 pathway; they are
highlighted in cyan color at the Volcano plot (Fig. [111]3d) and
PTMsigDB table (Fig. [112]3f). Intriguingly, in R403Q-αMyHC mice, the
same pathway was statistically significant; see the ten site-specific
changes that matched the EGFR1 pathway (p = 0.001) represented in the
volcano plot as cyan data points (Fig. [113]3e) and PTMsigDB table
(Fig. [114]3f). These data show the utility of phosphorylation
site-specific databases to narrow the search for a biologically
relevant pathway on phosphoproteomics data sets, particularly in
understudied phosphorylation events such as pTyr with small data sets
to date.
TMT-labeled quantitative proteomics of TgErbB2 confirms cardiac sarcomere
tyrosine phosphoproteome dysregulation
We used a TMT quantitative labeling proteomics to gain more specific
insight into the myofilament pTyr changes. More specifically, we
hypothesized that myofilament enrichment would enhance the number of
site-identification in myofilament proteins, especially those with low
abundance phospho-tyrosine modifications that may have been missed
using the global approach. To this end, we utilized myofilaments
freshly isolated from Ntg and TgErbB2 mice, as described in
“Methods”^[115]28, with minor modifications. We used three hearts per
genotype to characterize the myofilament pTyr proteome (Fig. [116]4a,
b), as described for the heart failure phosphoproteome^[117]20.
Briefly, we performed a large-scale 6-plex TMT experiment where most of
TMT-labeled peptides were enriched for pTyr (Fig. [118]4c). We adopted
a workflow similar to whole heart TMT phospho-tyrosine proteomics for
further statistical data analysis (Fig. [119]4d). The peptides
identified with single spectra were removed, which led to 24,727 PSM.
After median sweep normalization, 1116 peptides corresponded to 1,092
proteins. See Supplementary Fig. [120]4a, b for spectral intensity
distribution before and after normalization (Supplementary
Fig. [121]4c, d). For the pTyr proteome, 4391 PSM were collected and
followed the same quality control curation of the entire proteome to
remove missing data and unique spectra, which led to 3064 PSM. Raw data
of pTyr proteome spectral intensity distribution are shown before
normalization (Supplementary Fig. [122]5a, b) and after normalization
(Supplementary Fig. [123]5c, d) and PCA (Fig. [124]5b). After median
sweep normalization, 832 peptides corresponded to 184 different
proteins. Specific pTyr for 146 peptides was quantified because they
had corresponding peptides from the full proteome’s expression data.
Fig. 4. Workflow for cardiac sarcomere pTyr identification of normal and
TgErbB2 hypertrophic myofilaments by TMT quantitative proteomics.
[125]Fig. 4
[126]Open in a new tab
a Myofilament from Ntg (n = 3) and ErbB2 (n = 3) transgenic mice hearts
were freshly isolated on ice-cold buffers containing Proteinase and
Phosphatase Inhibitors Cocktails (Roche). b All material was
resuspended in TEAB, then reduced and alkylated. Tryptic peptides were
desalted and labeled with 6-plex isobaric tandem mass tags (TMT). c The
digested and labeled peptides were pooled and desalted with C[18]
SEP-PAK. The enrichment for phosphotyrosine was performed with PTMScan
Phospho-Tyrosine Rabbit mAb (P-Tyr-1000) kit (Cell Signaling
Technology). The eluted peptide samples were desalted using C18 STAGE
tips d Easy-nanoLC 1200 nanoflow liquid chromatography system coupled
to Orbitrap Fusion Lumos Tribrid.
Fig. 5. TgErbB2 cardiac sarcomere pTyr proteome remodeling.
[127]Fig. 5
[128]Open in a new tab
a Full Proteome was labeled with TMT, and the logged and summarized
protein signal intensities indicate that experimental groups segregate
along with the first principal component (PC #1), which accounts for
53.9% of the total correlation in the expression proteome (normalized,
mean = 0, variance = 1). b PCA of the phospho-Tyr Proteome reveals a
similar trend, and the groups segregate along the PC#1, which accounts
for 47.2% of the correlation. c Signal intensities of the phospho-Tyr
proteome were normalized to Full Proteome protein expression and
analyzed by principal component analysis. The distinctive signature of
pTyr is still apparent by PC#1 and accounts for 57% of the correlation.
d A heatmap was used to visualize the unsupervised hierarchical
correlation clustering of protein expression and e pTyr Proteome
expression with p < 0.05 by LIMMA moderated 2-samples t-test comparison
of TgErbB2 and Ntg group. f Volcano plot showing the Log2 fold-change
and –Log10 p-value of each pTyr peptide in the pairwise comparison of
Ntg and TgErbB2. The data points in red showed a significantly lower
intensity, whereas the data points in green showed a significantly
higher intensity. For clarity of presentation, only some peptides were
highlighted. g Molecular Signature DataBase (MSigDB) confirmed targets
and the involvement of Dilated and HCM KEEG Pathways also found on
TgErbB2 mice in Fig. [129]3.
The first step of the analysis uses unsupervised PCA and hierarchical
clustering. PCA analysis showed the correlation between members in the
same group (Fig. [130]5a–c), full proteome, pTyr proteome, and
normalized pTyr proteome segregate well along PC1, showing a
correlation of expression of 57.6%, 82.4%, and 62.1%, respectively.
Heatmaps of hierarchically clustered expression helped visualize how
TgErbB2 cardiac overexpression largely influenced the genotype
clustering. Both patterns, proteins from the whole proteome
(Fig. [131]5d) and the normalized intensity of pTyr proteome
(Fig. [132]5e), suggested a mirrored remodeling (377 proteins for the
entire proteome and 51 pTyr peptides had a p < 0.05 by LIMMA moderated
2-sample t-test comparison).
The full proteome detected 1116 peptides which corresponded to 1092
proteins. A comparison of the Log[2] fold-change (FC) of Ntg/TgErbB2
showed 377 proteins with statistically significant differences (Log[2]
FC > 1 and p-value < 0.05). Like whole global heart TMT
phospho-tyrosine normalized to full proteome data, the statistically
significant protein expression changes were subjected to MSigDB for
pathway analysis. The results are shown in Supplementary Data [133]4.
We used 146 phospho-sites corresponding to 50 proteins to analyze the
pTyr proteome phosphorylation or normalized pTyr proteome. The TgErbB2
mice showed significant changes in 21 pTyr sites (Fig. [134]5g)
corresponding to 15 proteins. A substantial increase of pTyr was
detected on MLC1-Y82, Y139, Myh6-Y1310, α-Tm-Y261, Actin-Y168,
MyBP-C-Y544, and Actinin2-Y200, among other proteins. The MsigDB
pathway analysis yielded results similar to the global TMT
phospho-tyrosine approach on whole hearts (Fig. [135]3). TgErbB2
sarcomere pTyr data also show that the most significant pathways are
DCM, and HCM (Fig. [136]5g, h). Interestingly, when normalized to total
ErbB2, the ErbB2-Y1006 phosphorylation levels decrease. On the other
hand, the PTMsigDB pathways analysis did not match any pathways because
many sites are new and not reported in the PTMsigDB database.
Our data show that the enrichment of sarcomere proteins detected more
myofilament pTyr peptides that were missed in the global approach of
whole heart lysates. However, the total number of
tyrosine-phosphorylated peptides and corresponding proteins was lower.
KSEA implicated downregulation of MAPK in TgErbB2 and upregulation of EGFR in
R403Q-αMyHCTg
KSEA was used to characterize genotype-induced signaling changes by
estimating the kinase’s relative activity in TgErbB2 and R403Q-αMyHC
mice using the global whole heart pTyr TMT and TMT quantitative
sarcomere pTyrosine proteomics data, using their respective Ntg groups
as a reference. KSEA^[137]29 is a method that infers the kinases’
differential activity based on the differential phosphorylation of its
substrates and computes scores that reflect the directional change in
each kinase’s activity. This method assumes that the differential
activity of kinases is correlated with phosphorylation changes in its
substrates. A positive score corresponds to a kinase with
phosphorylated substrates in Tg mice relative to Ntg control. Likewise,
a negative score is a hypophosphorylation in Tg relative to their Ntg
control. The kinase-substrate interaction data was downloaded from the
PhosphoSitePlus^[138]30 website. Next, the KSEA method was applied to
all the pTyr sites identified in these experiments. We identified 34
phosphosites that have associated kinases reported on PhosphoSitePlus.
However, 49 unique kinases were scored in the combined data sets. A
KSEA heatmap inferred clusters of kinases downregulated (blue color
scale) or upregulated (red color scale) is shown in Fig. [139]6a.
TgErbB2 mice displayed significant downregulation of MAP2K4, MAP3K6,
MAP3K5, MAP2K3, and MAP2K6, and marked upregulation of EPHA4. On the
other hand, R403Q-αMyHC Tg showcased a substantial downregulation of
GSK3B (see highlighted blue rectangle in Fig. [140]6a). R403Q-αMyHC Tg
displayed a larger cluster of significantly upregulated kinases (see
highlighted red rectangle in Fig. [141]6a).
Fig. 6. KSEA & MoBaS analysis.
[142]Fig. 6
[143]Open in a new tab
a KSEA on whole hearts pTyr enrichment and TMT pTyr Myofilament
enrichment data. Heat maps report KSEA results according to normalized
scores (i.e., TgErbB2/Ntg, R403Q-αMyHC/Ntg, etc.). Only kinases shared
between data sets are included, therefore present in Tg and Ntg.
Asterisks indicate the statistical significance of p < 0.05. Red
represents a positive score, and blue is a negative score. Red
rectangle highlights groups of activated kinases clustered together and
are predominant in R403Q-αMyHC, whereas the blue rectangle highlights
suppressed kinases that clustered and are predominant on TgErbB2 whole
hearts. b MoBaS identified a tightly interconnected subnetwork involved
in growth factors signaling (ERBB, PDGFRA, and EGFR1). Represented by
the top-scoring PPI subnetwork called Module 1 was identified on
TgErbB2 whole heart pTyr enrichment and compared to R403Q-αMyHC. c KSEA
on Module 1 group of substrates, MsigDB and PTMsigDB. d Second
top-scoring PPI subnetwork called Module 2 was identified on
R403Q-αMyHC whole heart pTyr enrichment and compared to TgErbB2. e KSEA
on Module 2 group of substrates show activated kinases in R403Q-αMyHC.
Results from PTMsigDB and MSigDB analysis identified EGFR1, Actin
cytoskeleton, ERBB and Focal Adhesion signaling. Each node is a
phosphoprotein represented by its most significant peptide on PPI
modules. Node color is based on Log2 fold-change relative to NTG. The
scale is noted by FC and a bar colored from blue to red.
Conventional pathway analysis can miss protein groups because pathway
algorithms are predefined and rigid. A protein–protein interaction
(PPI) network approach might better capture signaling responses in
these models that are not typically detected by pathway analysis. To do
so, MoBaS Analysis^[144]22,[145]31 was employed to identify densely
connected subnetworks that are related and might exhibit differential
phosphorylation in Tg models. Several modules were identified, and we
focused on the top two statistically significant PPI modules from the
global whole heart TMT flow through dataset; for interpretation
purposes, they were designated as Module 1 and Module 2. The substrates
of modules 1 and 2 are illustrated in Fig. [146]6b, [147]d,
respectively.
Interestingly, PTMsigDB site-level molecular signature analysis showed
that module 1 contained significant site-specific modifications for the
EGFR1 pathway. In addition, using MSigDB Gene-Protein Level analysis,
several pathways, such as the ErbB signaling, PDGFRA, and focal
adhesion pathways, showed statistical significance. KSEA of module 1
protein-protein network highlights significant activation of AURKB
(aurora kinase B) and CSNK2A1 (Fig. [148]6c).
Similarly, PTMSigDB analysis for module 2 confirmed the involvement of
EGFR1 Pathway while MsigDB detected actin cytoskeleton, ErbB signaling,
and focal adhesion pathways. KSEA of module 2 protein-protein network
highlights significant activation of SRC, MAP2K1, HCK, LCK, SYK, JAK3,
JAK2, FYN, PTK6, and ABL1 in R403Q-αMyHC mice (Fig. [149]6e). These
data confirm that MoBAS identified functionally relevant modules of PPI
among identified pTyr peptides, and PTMsigDB pointed to EFGR1 Pathway
as a common pathway for HCM from two unrelated mouse models.
Tyrphostin AG-825 administration decreased cardiomyocyte disarray in the
TgErbB2 mouse model
TgErbB2 mouse hearts develop a concentric type of HCM. This alteration
rapidly progresses to pathologic HCM, showing features such as
fibrosis, cardiomyocyte disarray, perturbed Ca^2+ handling,
arrhythmias, and sudden death^[150]16. On these grounds, we
hypothesized that counteracting the activity of cardiac ErbB2
overexpression pharmacologically, a tyrosine kinase receptor, using
AG-825 (an ErbB2 inhibitor), would halt the progression of the
histopathological phenotype. To that end, TgErbB2 mice (ages 6–9
months) with established HCM and Ntg controls were treated with AG-825
or DMSO vehicle for two weeks, subcutaneously twice daily at a dose of
1 mg/Kg^[151]16. Next, myocardial fibrosis was evaluated by Masson’s
trichrome staining. Untreated TgErbB2 mice developed variable amounts
of fibrosis of the left ventricle (Fig. [152]7a, b); however, TgErbB2
treated mice displayed less fibrosis (Fig. [153]7b). In contrast,
cardiac histology was not affected by AG-825 Treatment in Ntg mice. We
used CytoSpectre^[154]32, a web graphic user interphase, to
characterize the effects of AG-825 on cardiomyocyte disarray; this
interphase determines the local orientation of structures, including
cardiac myocytes, across histological sections, similar to the analysis
described previously by the Seidman’s laboratory^[155]33. Randomly
chosen regions of interest are depicted in Fig. [156]7a, and increase
the magnification of microphotographs from 5X to 40X (Fig. [157]7b).
The myocyte orientations angle variance was evaluated, referred to as
Circular Variance and means orientation angles variance. Ntg treated
with the vehicle (Fig. [158]7c) compared to Ntg AG-825 treated
(Fig. [159]7c) was not significantly different (Fig. [160]7d), as
already described^[161]33. In contrast, TgErbB2 mice cardiomyocytes
treated with the vehicle displayed an orientation angle variance that
was substantially different from Ntg-Vehicle control; see the
heterogeneous and broader shape of the orientation plot in
Fig. [162]7c. Surprisingly, cardiomyocytes from TgErbB2 mice AG-825
treated displayed significantly less disarray, as demonstrated by the
significant return of circular variance and shape of the plot to Ntg-
Vehicle or AG-825 treated controls (Fig. [163]7e, p = 0.0014 by One-way
ANOVA). This analysis suggests that cardiomyocytes from TgErbB2 mice
AG-825 treated displayed significantly less myocyte disarray, as
indicated by reduced circular variance.
Fig. 7. Tyrosine kinase receptor inhibitor AG-825 reduces cardiac myocyte
disarray without effects on contractile function in TgErbB2 mice.
[164]Fig. 7
[165]Open in a new tab
a Masson’s trichrome stained sections (5X) from vehicle-treated Ntg and
AG-825–treated Ntg and vehicle-treated TgErbB2 and AG-825-treated
TgErbB2 mice, yellow arrowheads point to fibrotic areas. Scale bar
1 mm. b Masson’s trichrome stained sections (40X) from representative
regions of interest show increased Fibrosis in TgErbB2 mice myocardium,
that fibrosis improved after 2 weeks of AG-825 Treatment. The same
areas were used to determine the level of cardiac myocyte disarray,
which was achieved by analyzing cell orientation using CytoSpectre
Software from Vehicle and AG-825 treated Ntg and TgErbB2 mouse heart
sections (scale bar 50 μm). c Plots of the distribution of cardiac
cells orientation angles. Note how Ntg angles are homogeneous and did
not change in the group treated with AG-825. In contrast, Tg treated
with vehicle showed less homogenous and different orientation angles,
confirming cardiomyocyte disarray. AG-825 Treatment reduces
cardiomyocyte disarray, and the orientation plot is similar to the
AG-825 treated NTG group. d Mean orientation angle of cardiac myocytes,
and e Circular variance, another measure of isotropy, in Ntg and
TgErbB2 mice with and without AG-825 treatment (n = 5 animals per
condition). Graph bars are expressed as mean ± s.e.m.; a one-way ANOVA
test was performed for statistical comparisons showing a significant
improvement in TgErbB2 mice treated with AG-825 (p-value < 0.0014). f
Cardiac echocardiography from the four groups shows that AG-825
Treatment for 14 days did not modify the contractile function,
estimated as % Fraction Shortening.
Next, we evaluated the direct in vivo impact of ErbB2 inhibition by
AG-825 on contractility, echocardiograms, and tissue-doppler imaging to
assess diastolic function. These studies were performed on conscious
(non-anesthetized) mice, at baseline and after two weeks of Treatment.
We used a Vevo 2100 High-Resolution Imaging System with a 40 MHz
transducer (VisualSonics®, Toronto). The data were subsequently
analyzed with an Advanced Cardiovascular Package Software. The
parameters studied were chamber dimensions, fractional shortening,
ejection fraction, and tissue Doppler velocity dynamics. Left ventricle
functional data obtained by parasternal short-axis echocardiographic
imaging determined no significant change in the mean fractional
shortening over time in TgErbB2 mice treated with vehicle solution
compared with the Ntg (Fig. [166]7f). Similarly, TgErbB2 mice treated
with AG-825 did not display significant changes in contractile function
over time compared to Ntg AG-825 treated (Fig. [167]7f). Tissue Doppler
imaging detected no significant differences in diastolic function; see
complete Echocardiography data in Supplementary Data [168]6. These data
suggest that ErbB2 pharmacological inhibition halted or reversed
pathological remodeling by reducing the fibrotic response and restoring
cardiomyocyte alignment. Yet, this response was not paralleled by in
vivo preservation of LV contractile function in TgErbB2 mice.
Extensive phosphoproteomics validate EGFR1 pathway activation on cardiac
hypertrophy of TgErbB2 that is inhibited by tyrosine kinase inhibitor
(AG-825)
We revisited the phosphoproteomics of TgErbB2 and NTg to validate
bioinformatic findings of KSEA and to study the impact of the tyrosine
kinase inhibitor AG-825 on the proteome. We used a TMT 13-plex labeling
approach to compare the whole heart phosphorylation profile of Ntg and
TgErbB2 mice treated with vehicle or AG-825 (ErbB2 inhibitor). Briefly,
a TMT 13-plex was performed as follows; Ntg-Vehicle (n = 3), Ntg-AG-825
treatment (n = 3), TgErbB2-Vehicle (n = 3) and TgErbB2-AG-825 treatment
(n = 4). Whole heart protein lysates were isolated from thermally
stabilized hearts and processed as described in Fig. [169]2 and
“Material and methods”. Most of the material was used to enrich
pS/pT/pY peptides, and a small portion is labeled to quantify full
proteome with TMT. As described in the previous experiment in
Fig. [170]2, the intensity of pS/pT/pY peptides is normalized to the
intensity of full proteome TMT signals. A total of 6614 proteins, 4732
phosphorylated serine and threonine peptides, and 346 pTyr peptides
were quantified by TMT reporter ions. The full dataset is available in
Supplementary Data [171]7.
Heatmaps of hierarchically clustered up- or down-regulated normalized
serine/threonine phosphorylated, or tyrosine-phosphorylated peptides
intensity confirm the influence of the genotype on all phosphoproteome
remodeling, serine and threonine phosphorylation (Fig. [172]8a) and
pTyr (Fig. [173]8b). We directly interrogated data with MSigDB and
PTMSigDB to compare Ntg with TgErbB2 mice. As anticipated, we confirmed
the involvement of EGFR1 Pathway by PTMSigDB in TgErbB2 mice cardiac
hypertrophy treated with vehicle (Fig. [174]8c, d) or AG-825
(Fig. [175]8e, f). Also, we confirmed the involvement of focal adhesion
and extracellular matrix pathways by MSigDB in vehicle-treated mice and
dilated cardiomyopathy, HCM, and arrhythmogenic right ventricle
cardiomyopathy (ARCV) in AG-825 treated mice. However, AG-825 Treatment
for 14 days had a subtle effect in TgErbB2 mice that was not evident by
MsigDB or PTMsigDB. Since we included all phosphoproteome, we estimated
kinase activity in serine and threonine phosphorylation peptide. The
top upregulated and downregulated kinases were identified using KSEA,
and the result is presented in Fig. [176]8g. The complete KSEA and
pathway analysis are available in Supplementary Data [177]5 and
Supplementary Data [178]7. We compared the activity of kinases in
AG-825 treated animals relative to their control (vehicle treatment).
We identified a small cluster of kinases significantly inhibited or
down-regulated by AG-825 Treatment and not by vehicle treatment (See
highlighted red rectangle in Fig. [179]8g). To get more insights on the
functional impact from a system biology point of view, we explored the
relation of these kinases using The Signaling Network Open Resource
(Signor 2.0) visualization software^[180]34. Figure [181]8h shows that
most of these kinases are downstream of EGFR, highlighted in blue
dotted circles. Altogether, our data indicate that EGFR1 is central to
hypertrophy of TgErbB2 and that AG-825 Treatment inhibits ErbB2 and
EGFR; therefore, a significant subset of targets downstream of EGFR
signaling cascade are affected.
Fig. 8. Comprehensive serine/threonine/tyrosine phosphoproteomics confirm
EGFR1 pathway involvement in TgErbB2 hypertrophy and tyrosine kinase receptor
inhibition by AG-825.
[182]Fig. 8
[183]Open in a new tab
a A heatmap was used to visualize the unsupervised hierarchical
clustering of Ser/Thr peptides, and b pTyrosine peptides normalized to
full proteome expression with p-value < 0.05 by ANOVA. Overrepresented
(yellow) and underrepresented (Blue). c Volcano plots showing the Log2
fold-change and –Log10 p-value of each pSer/pThr peptide and d pTyr
peptides in the pairwise comparison of Ntg and TgErbB2 after two weeks
of Vehicle Treatment. e Volcano plots showing the Log2 fold-change and
–Log10 p-value of each pSer/pThr peptide and f pTyr peptides in the
pairwise comparison of Ntg and TgErbB2 after two weeks of AG-825
Treatment, (g) KSEA shows in a red rectangle the kinases significantly
inhibited by AG-825. KSEA and Pathway Enrichment Analysis of combined
Ser/Thr/Tyr phosphorylation data analysis using PTMSigDB confirm that
EGFR1 Pathway is significantly involved. h The Signaling Network Open
Resource (Signor 2.0) analysis shows that nine kinases are involved in
the EGFR pathway. Kinases found in the network and also in KSEA are
circled in dotted blue circles. The symbol code of the network is
displayed next to the pathway.
Discussion
The present study demonstrates that the pTyr proteome of the heart is
altered in proteins related to cardiomyopathies, serine/threonine, and
tyrosine kinases in two pathogenetically-unrelated models of
hypertrophy cardiomyopathy (HCM), namely cardiac-specific
overexpression of ErbB2 and allelic expression of R403Q-αMyHC. In
addition, the serine/threonine phosphoproteome of the heart is also
altered in the TgErbB2 mouse model cardiac hypertrophy. Altogether our
data point that EGFR pathway activation is centrally involved in
cardiac hypertrophy that rises from two very different mouse models.
Heart failure contractile dysfunction is, in part, due to myofilament
Ca^2+ desensitization. Altered serine/threonine phosphorylation of the
cardiac sarcomere and reduced myofilament Ca^2+ sensitivity are
features of human and experimental models of heart failure and play a
role in the pathophysiology of HCM and DCM^[184]7,[185]35,[186]36. We
have previously identified pTyr in human cardiac Troponin I Y26 that
was downregulated in human heart failure and DCM^[187]14. However, no
studies have yet addressed the impact of global dysregulated cardiac
pTyr in the pathogenesis of cardiac disease.
Here, we show evidence that in TgErbB2 mice’s heart, tyrosine
phosphoproteome is altered in a differential fashion, upregulation, and
downregulation of tyrosine phosphopeptides. MsigDB pathway analysis
demonstrated the involvement of ARVC, HCM, DCM, and integrin alphaIIB
beta3 signaling in TgErbB2 mice hearts. Most of the sarcomere proteins
had significant downregulation of pTyr, such as Myh7-Y386, Y410, Y1375,
Myh6-Y1349, and Ttn-Y2118, Y21190, except for Ttn-Y31324 that was
upregulated. The directional change of pTyr in these sarcomere proteins
is downregulation, just like it was reported for human cTnI Y26 in
heart failure and DCM. Very little is known about the role of pTyr of
sarcomere proteins. In fact, only one group^[188]37 has examined the
potential tyrosine kinases that phosphorylate cTnI. They demonstrated
in vitro that c-Src and another Src kinases family member, Lyn,
phosphorylates the cTnIY26, reporting phosphorylation and
pseudo-phosphorylation mimicking at this site reduced myofilament Ca^2+
sensitivity, as evidenced by the altered force-calcium response of the
myofilaments. Noteworthy, cTnI Y113 phosphorylation, a novel discovered
site, was downregulated in R403Q-αMyHC mice. We also demonstrated
significant downregulation of pTyr in Ttn-Y1881, Y1902, and Y33864.
Contrary to TgErbB2 mice, Myh7-Y410 was upregulated in R403Q-αMyHC
mice. pTyr at Myh7-410 is near R403Q αMyHC and could have a functional
impact since it is in the motor domain head of αMyHC. The site-specific
implications of these changes are worthy of further study; however, it
is beyond the scope of the present work. Interestingly, in the
R403Q-αMyHC hearts, pTyr proteome is also altered differently. However,
in R403Q-αMyHC mice, more phosphotyrosine peptides reached statistical
significance. Unlike TgErbB2, the MsigDB pathway analysis demonstrated
that R403Q-αMyHC mice had alterations in tyrosine kinase receptor
pathways, angiopoietin receptor, ErbB, and growth hormone receptor
signaling. Whereas PTMsigDB revealed the alteration of the EGFR1
Pathway and squamous cell carcinoma. We reasoned that overexpression of
a receptor tyrosine kinase in TgErbB2 mice hearts would result in a
more significant alteration of pTyr proteome. However, the data suggest
that a single sarcomere point mutation, as in the case of R403Q-αMyHC
mice, is incisive enough to alter the cardiac pTyr proteome
significantly. Top-down proteomics^[189]38 demonstrated that consistent
alterations of cTnI, ENH2, and other Z-disk protein phosphorylation in
HCM myocardium correlated better with the phenotype, regardless of the
single sarcomere point disease-causing mutation. The mechanisms that a
pathogenic mutation in the myosin heavy chain (R403Q-αMyHC) may affect
pTyr regulation are unknown and warrant further investigation.
Global pTyr dysregulation in the TgErbB2 mouse model is essential
because it connects pronounced cardiac hypertrophy, sarcomere
dysfunction, and abnormal calcium handling^[190]16 with alterations in
tyrosine kinase pathways. On the other hand, pharmacological inhibition
of ErbB2 (HER2/neu) with the inhibitor lapatinib in breast cancer
patients treated with doxorubicin increases the risk of developing
heart failure compared to patients treated with doxorubicin
alone^[191]39. It suggests that maintaining pTyr homeostasis may be
necessary for regulating cardiomyocyte function and homeostasis.
A wide variety of other functionally relevant targets were identified
in this study with pTyr, from sarcomere proteins to z-disk,
intermediate filaments (such as desmin), desmosome components, focal
adhesion, and adherence junction proteins, membrane receptors, kinases,
and phosphatases. For instance, pTyr in several z-disk-associated
proteins was noted (for a complete list, see Supplementary Data [192]2,
[193]3, and [194]6). Z-disk proteins are crucial for muscle contraction
and mechanical stress, growth, and metabolic signaling^[195]40. Also,
the alteration in the phosphorylation levels of the desmosome key
component Plakophilin-2 protein (Pkp2) peptide Pkp2-Y119 was noted in
both mouse models, upregulation in R403Q-αMyHC Tg and downregulation in
TgErbB2. The functional effect of up- or downregulation of Pkp2-Y119
phosphorylation is not known. However, Pkp2 homozygous deletion
disrupts heart architecture and is lethal in the embryo^[196]41. In the
heart, Pkp2 is required to assemble the desmosome and the PKC
activity^[197]42. Autosomal dominant mutations in this gene are
responsible for 25 to 50% of ARVC. Interestingly, TgErbB2 mice have
increased susceptibility to arrhythmias and myofibrillar
disarray^[198]16, similar to patients with myosin mutations and HCM.
Pkp-Y119 phosphorylation changes could impact the phenotype.
The EGFR1 (ErbB1) pathway plays a central role in cardiac hypertrophy
in both Tg models. EGFR pharmacological inhibition, using AG-1478,
protects against Angiotensin II-induced cardiac hypertrophy in vitro
and in vivo^[199]43. The concentric hypertrophy associated with ErbB2
cardiac-specific overexpression can be reversed by early administration
of Lapatinib, which inhibits EGFR receptor tyrosine kinase^[200]12 in
addition to ErbB2. In this study, the ErbB2 receptor’s pharmacological
blockage by AG-825 in TgErbB2 adult mice reduced myocyte disarray but
did not preserve cardiac function or alter cardiac hypertrophy. We
think that the lack of response is related to the age of mice at the
time of administration: in the current study, we used adult mice,
whereas pups have been used in previous studies^[201]15. We also
evaluated the cardiac serine/threonine and pTyr proteome in TgErbB2
after vehicle or AG-825 Treatment. We found global alterations in the
phosphoproteome (Fig. [202]8) that confirmed the involvement of the
EGFR pathway. In addition, we confirmed the alterations in focal
adhesion, extracellular matrix, DCM, HCM, and ARCV pathways. We used
KSEA to pair the findings with a subset of activated kinases and narrow
it down to the top 40 most representative serine/threonine and tyrosine
kinases detected in the cardiac proteome of Ntg and TgErbB2 mice.
AG-825 Treatment could inhibit or significantly reduce the activation
of 14 kinases corresponding to the EGFR pathway (Fig. [203]8g, h),
suggesting that AG-825 Treatment was effective and specific. Altogether
these findings indicate that alterations of cardiac proteome span to
serine/threonine and pTyr, confirming the involvement of the
above-described pathways. In addition, these data suggest that AG-825
treatment-induced changes in the phosphoproteome correlated with an
improvement in cardiac myocyte disarray.
The KSEA implicated a marked downregulation of members of the canonical
MAPK pathway downstream of k-Ras in TgErbB2 mice. Conversely, in
R403Q-αMyHC mice, we observed an upregulation of focal adhesion and
PDGFR-beta signaling pathways (SRC, LYN, LCK, JAK2, INSR, MAPK1, and
HCK). More importantly, modularity-base scoring (MoBas) analysis
identified protein-protein interaction subnetworks enriched in ErbB,
PDGFRA, and focal adhesion pathways and activation of aurora kinase B
and CSNKA1. These data indicate that many targets overlap in both
transgenic models of cardiac hypertrophy. Although there are marked
differences in the upstream regulators of ErbB2 and myosin heavy chain,
many downstream effector molecules are common and can be targeted by
small-molecule inhibitors (designed initially to treat several types of
cancer) to block cardiac hypertrophy.
The MAPK pathway is involved in the adaptive and maladaptive response
that could lead to heart hypertrophy (for review^[204]44). The
RAS-RAF-MEK-ERK signaling pathway is an attractive target for oncology
therapeutic intervention. Several selective RAF and MEK small-molecule
inhibitors have been tested in clinical trials^[205]45. The present
study found increased phosphorylation at MAPK1-Y185, which activates
the kinase in R403Q-αMyHC, and MAPK1 activity directly relates to heart
hypertrophy^[206]46. The c-Src, the central kinase from its family,
could be one of the potential regulators for the pTyr proteome changes
observed in R403Q-αMyHC because it phosphorylates the EGFR receptor
upstream and phosphorylates Stat, among other targets, downstream.
c-Src affects the response to mechanical cardiomyocyte stretching by
triggering a cascade of intracellular signaling in cardiomyocytes
towards a hypertrophic response^[207]47,[208]48. Also, c-Src
phosphorylates PXN-Y118^[209]49, and this site had a 25-fold increase
in phosphorylation in R403Q-αMyHC mice. Interestingly, c-Src mediates
the activation of MAPK1 and MAPK3 in response to
pressure-overload^[210]47. Notably, a previous study has shown that
pressure-overload-induced cardiac hypertrophy is exacerbated in
R403-αMyHC Tg mice^[211]50, suggesting that the mechanism by which
R403-αMyHC mutation produces heart hypertrophy is by sensitizing cells
to pressure-overload-induced signaling via c-Src.
R403Q-αMyHC mice displayed enhanced JAK2, STAT5A, and STAT5B
phosphorylation in the activation sites (Y570, Y694, and Y699,
respectively), which indicates not only activation of MAPK signaling
but also activation of JAK-STAT signaling. Jak-Stat Signaling: IL-6
pathway is activated in response to the IL-6 family of cytokines (IL-6,
cardiotrophin 1, and leukemia inhibitor factor), cross-talks with the
EGFR pathway, and is involved in cardiac hypertrophy. This pathway has
cardioprotective effects, but chronic activation may lead to heart
hypertrophy (reviewed in ref. [212]44). Vakrou and colleagues performed
a pathway analysis on miRNA profiles of R403Q-αMyHC. These authors
found similarities with the findings described in this work, such as
overactivation of chemokine signaling (CXCR4), actin cytoskeleton, and
cardiac hypertrophy signaling pathways^[213]51. These results suggest
an essential involvement of pTyr regulation in myofilaments and other
cardiac proteins. The insight gained from these studies could inform
new therapeutic approaches in sarcomere mutation-related HCM and
potentially other conditions associated with HCM. The studies were
limited to two transgenic mouse models that develop cardiac
hypertrophy; other models of sarcomere mutations or pressure-overload
such as aortic banding could be explored and compared. Some relevant
tyrosine-phosphorylated sites might be missed due to their low
stoichiometry and technical challenges. This problem is particularly
true for membrane-bound proteins, which are difficult to evaluate in
phosphoproteomic studies.
This study reveals that altered pTyr patterns are striking in two
separate models of HCM. Moreover, despite some shared sites,
etiologically different forms of HCM may harbor specific pTyr
signatures that point EGFR pathway as central in both transgenic
models. This evidence, combined with the inhibition of a specific
receptor tyrosine kinase using tyrphostin AG-825, can reverse
cardiomyocyte disarray and rationalize approaches to manipulate the
pTyr proteome as a therapeutic approach for HCM. Furthermore, these
studies indicate that tyrosine kinase inhibitors, now used broadly in
cancer therapies, may change cardiac function by directly modifying the
heart’s tyrosine kinase and serine/threonine kinase profiles.
Methods
Western blot
Whole heart lysates from Heat Stabilized Tissue were resuspended in 1%
SDS buffer, 10–15 μg were separated by SDS-PAGE, and transferred to
nitrocellulose membranes. Membranes were blocked with 5% BSA in TBS-T
buffer (20 mmol/L Tris pH 7.4, 150 mmol/L NaCl, and 0.1% Tween 20) for
1 h at room temperature, then were incubated with primary antibody
dilution 1:1000 in 1% BSA-TBS-T; phospho-Tyrosine mouse mAb (pTyr-100,
cat. No. 9411SCST), TnI Rabbit Ab (cat. No. 4002SCTS), Src Rabbit Ab
(cat. No.2108SCST), Phopsho-Src Family (Tyr416) Rabbit mAb (D49G4, cat.
No.6943T CST), GAPDH Rabbit Ab (14C10, cat. No. 2118S CST) at 4 °C
overnight. After washing them five times, the secondary antibodies were
diluted 1:10,000 in 1% BSA-TBS-T (anti-Rabbit IgG-HRP linked (cat. No.
7074S CTS), Anti-Mouse m-IgGk BP-HRP (cat. No. sc-516102, Santa Cruz)
and incubated for 1 h at room temperature. Then, the membranes were
developed with super signal West Pico Chemiluminescent Substrate
(Thermo), and the immunoreactive bands were detected by
chemiluminescence with an iBright 1500 (Invitrogen). The images were
obtained and analyzed with Image J software. One-way ANOVA was used to
compare groups and determine statistical significance.
Global whole heart TMT phospho-tyrosine proteomics
Sample preparation
All protocols were performed following the “Guide for the Use and Care
of Laboratory Animals” published by the National Institutes of Health
and the Institutional Animal Care and Use Committee’s approval. The
TgErbB2 (B6SJLF1/J stain) and R403Q-αMyHC (C57/Bl6) mice were obtained
from Dr. K. Gabrielson and Dr. L. A. Leinwand,
respectively^[214]16,[215]17, to establish breeding colonies. Male or
female mice (6-9 months) were anesthetized with sodium pentobarbital IP
(75 mg/kg) or isoflurane (5%) overdose; the hearts were quickly
dissected, followed by thermal stabilization (Denator T1 Heat
Stabilizor, Sweden) and stored at −80 °C until analyzed. TgErbB2 and
R403Q-αMyHC mice hearts (n = 5 per group) were processed in parallel.
To obtain whole heart lysates, cardiac ventricles (~200 mg) were
homogenized on ice-cold buffer: 20 mM HEPES, pH 7.6, 1 mM, 1.5 mM
sodium pyrophosphate, PhosStop, and 9 M urea at 10 μl/mg (wet weight of
tissue). The homogenate was cooled on ice, followed by brief micro-tip
sonication on ice, centrifugation at 10,000 × g for 15 min at 4 °C. The
supernatant was retrieved, and protein concentration was determined by
the method of Lowry (Bio-Rad).
Trypsin digestion
Protein from heart lysates was reduced in 5 mM of dithiothreitol (DTT)
at 60 °C for 20 min and alkylated in 10 mM iodoacetamide (IDA) at room
temperature for 15 min in the dark. Each sample (30 mg per mouse, n = 5
mice per genotype) was digested with Proteomics grade Trypsin (Promega)
at a ratio of 1:200 in 2 M urea, 20 mM HEPES buffer, pH 8.0 at room
temperature overnight. The digestion was terminated with
trifluoroacetic acid (1% TFA). Samples were centrifuged (5 min at
1800g), and supernatants were desalted by solid-phase extraction
(SepPak C18 10cc cartridge, Waters). Elutes were lyophilized for three
days at −20 °C.
Enrichment of tyrosine phosphopeptides
A total of 150 mg of trypsinized peptides per genotype were pooled and
enriched for pTyrosine, as previously described^[216]52. Lyophilized
peptides were mixed in 1.4 ml immunoprecipitation buffer (IAP buffer
50 mM MOPS, pH 7.2, 10 mM sodium phosphate, 50 mM NaCl, pH 7.0). A
stock solution of protein agarose G beads (Santa Cruz Biotechnology)
80 μL of slurry were conjugated with 300 μg of Phospho-Tyrosine Mouse
monoclonal antibody (p-Tyr-100, Cell Signaling Technology). The
beads-Anti-p-Tyr antibody conjugate was transferred to the peptide’s
tube and incubated with gentle rotation for 2 h at 4 °C. The beads were
washed and eluted with 55 μl and 45 μl of 0.15% trifluoroacetic acid
(TFA), respectively. The two elution yields were pooled. The resulting
peptide mixtures were purified by solid-phase extraction (stage tips
C18, Thermo Scientific). The samples were dried by vacuum
centrifugation. This approach was repeated three times with samples
from all three groups run in parallel, 15 hearts per genotype in three
technical replicates. The immunoprecipitation buffer flow-through was
stored at −80 °C for subsequent TMT 9-plex. According to the
manufacturer’s instructions, tryptic peptides were desalted and labeled
with 9-plex isobaric TMT (Thermo Scientific). The labeling reaction was
carried out for 1 h at room temperature, followed by quenching with
100 mM Tris.HCl (pH 8.0).
LC-MS/MS analysis
Phosphopeptides were dissolved in 10 μl of 0.1% TFA, 2%ACN (v/v)
followed by RF-HPLC-ESI-MS/MS analysis. Phosphopeptides were separated
on a C18 reversed-phase column with a linear gradient of acetonitrile
(4−40%) for 120 min and then analyzed on an LTQ-Orbitrap Elite MS
(ThermoFisher Scientific) with neutro loss triggered HCD.
Peptide identification and quantification
Raw MS data were searched with Mascot 2.3, and label-free
quantification with MS1 extracted ion chromatograms were performed
using MaxQuant software.
TMT from flow-through for total proteomics
LC-MS/MS analysis and database search
Peptides were analyzed on Orbitrap Fusion Lumos Tribrid (Thermo
Scientific) coupled with Easy-nanoLC 1200 nanoflow liquid
chromatography system (Thermo Scientific). The peptides from each
fraction were reconstituted in 10% formic acid and loaded on Acclaim
PepMap100 Nano Trap Column (100 μm × 2 cm) (Thermo Scientific) packed
with 5 μm C[18] particles at a flow rate of 5 μl per minute. Peptides
were resolved at 250 nl/min flow rate using a linear gradient of 10 to
35% solvent B (0.1% formic acid in 95% acetonitrile) over 120 min on
the EASY-Spray column (50 cm × 75 µm ID, PepMap RSLC C[18,] and 2 µm
C[18] particles) (Thermo Scientific). It was fitted on an EASY-Spray
ion source that was operated at 2.0 kV voltage. Mass spectrometry
analysis was carried out in a data-dependent manner with full scans of
m/z 350 to 1500. Both MS and MS/MS were acquired and measured using
Orbitrap mass analyzer. Full MS scans were measured at a resolution of
120,000 at m/z 200. Precursor ions were fragmented using a
higher-energy collisional dissociation method (35) and detected at a
mass resolution of 30,000 at m/z 200. Proteome Discoverer 2.4 was used
for identification and quantification against
uniport_mouse_120119_UP000000589. The search parameters used are as
follows: (a) trypsin (up to two missed cleavages), (b) minimum amino
acid length: 6, (c) minimum peptides for protein: 1, (d) Fixed
modifications TMT6plex on any N-terminus and Lysis residue,
Carbamidomethylation of cysteine, (e) Dynamic modifications: Oxydation
on Methionine, acetyl on N-terminus, Met-loss on M and Met loss +acetyl
on M and (f) 1% false discovery rate for peptide and protein levels.
TMT cardiac sarcomere phospho tyrosine proteomics
Sample preparation and TMT pTyr enrichment
Myofilament enriched preparations were obtained by rinsing freshly
isolated hearts in ice-cold PBS (Proteinase Inhibitors Roche 1X and
PhosStop 1X, Roche), auricular tissue removed and ventricular tissue
minced in ice-cold standard buffer (60 mM KCl, 30 mM Imidazole pH 7.0,
2 mM MgCl[2], Proteinase Inhibitors Roche 1X and PhosStop 1X from
Roche). The ventricular tissues were homogenized with a homogenizer
(Omni tissue homogenizer TH) at maximum speed for 2–3 pulses of
10 seconds with the solution on ice, centrifuged at 12,000 g for
15 min. The myofilament pellets were resuspended in a skinning solution
(10 mK EGTA 8.2 mM MgCl[2], 14.4 mM KCl, 60 mM imidazole pH 7.0, 5.5 mM
ATP, 12 mM creatine phosphate, 10 U/mL creatine phosphokinase, 1%
Triton 100X, Proteinase Inhibitors Roche 1X and PhosStop 1X from Roche)
and incubated on ice for 30 min^[217]28. Myofilament pellets were
centrifuged at 1100 g for 15 min and washed in the standard buffer.
Freshly isolated myofilament pellets were diluted in standard buffer
and quantified with Lowery assay. Myofilament proteins (~7 mg) from
each mouse heart were resuspended in 8 M urea and 50 mM
triethylammonium bicarbonate (TEAB) (Sigma) followed by reduction with
10 mM dithiothreitol (Sigma) at room temperature for 1 h and alkylation
with 30 mM iodoacetamide (Sigma) for 20 min in the dark, see
Fig. [218]4a–d. The protein samples were then digested overnight at
37 °C using sequencing grade trypsin (1:50) (Promega). Tryptic peptides
were desalted and labeled with 6-plex isobaric tandem mass tags (TMT)
(Thermo Scientific) according to the manufacturer’s instructions. The
labeling reaction was carried out for 1 h at room temperature, followed
by quenching with 100 mM Tris.HCl (pH 8.0). The digested and labeled
peptides were pooled and desalted with C[18] SEP-PAK (Waters), followed
by pTyrsoine enrichment using PTMScan® Phospho-Tyrosine Rabbit mAb
(P-Tyr-1000) kit (Cell Signaling Technology), see Fig. [219]4c.
Briefly, the desalted peptides were reconstituted in 1.4 ml of
immunoaffinity purification (IP) buffer containing 50 mM MOPS pH 7.2,
10 mM sodium phosphate, 50 mM NaCl. Anti-phosphotyrosine antibody
(p-Tyr-1000, Cell Signaling Technology) was mixed with peptide solution
and incubated on a rotator at 4 °C for 2 h. After incubation, the beads
were washed with IP buffer and water two and three times, respectively.
The phosphotyrosine peptides were eluted using 0.1% TFA. The eluted
peptide samples were desalted using C18 STAGE tips, vacuum dried, and
kept at −80 °C before LC-MS analysis. From flow-through, a
quantification of the full proteome was made to serve as a reference
for pTyr proteome findings.
LC-MS/MS analysis
The phosphotyrosine peptides and non-phosphorylated peptides were
analyzed on Orbitrap Fusion Lumos Tribrid (Thermo Scientific, San Jose,
CA, USA) coupled with Easy-nanoLC 1200 nanoflow liquid chromatography
system (Thermo Scientific). The peptides from each fraction were
reconstituted in 10% formic acid and loaded on Acclaim PepMap100 Nano
Trap Column (100 μm × 2 cm) (Thermo Scientific) packed with 5 μm C[18]
particles at a flow rate of 5 μl per minute. Peptides were resolved at
250 nl/min flow rate using a linear gradient of 10–35% solvent B (0.1%
formic acid in 95% acetonitrile) over 95 min on the EASY-Spray column
(50 cm × 75 µm ID, PepMap RSLC C[18,] and 2 µm C[18] particles) (Thermo
Scientific) and it was fitted on EASY-Spray ion source that was
operated at 2.0 kV voltage. Mass spectrometry analysis was carried out
in a data-dependent manner with full scans in the range of m/z 350 to
1500. Both MS and MS/MS were acquired and measured using Orbitrap mass
analyzer. Full MS scans were measured at a resolution of 120,000 at m/z
200. Precursor ions were fragmented using a higher-energy collisional
dissociation method and detected at a mass resolution of 30,000 at m/z
200.
Peptide search and Identification
MaxQuant 1.5 was used for quantitation and identification against a
mouse RefSeq database (version 78) supplemented with frequently
observed contaminants. The search parameters used are as follows: (a)
trypsin as a proteolytic enzyme (with up to two missed cleavages); (b)
peptide mass error tolerance of 10 ppm; (c) fragment mass error
tolerance of 0.02 Da; (d) Carbamidomethylation of cysteine (+57.02 Da)
and TMT tags (+229.16 Da) on lysine residue and peptide N-termini as
fixed modification and oxidation of methionine (+15.99 Da) and
phosphorylation of serine/threonine/tyrosine (+79.96 Da) as variable
modifications. Two missed cleavages were allowed, and ‘match between
runs’ was enabled. Peptides and proteins were filtered at a 1% false
discovery rate. Proteins with a q-value lower than 0.05 will be
considered as differentially expressed with statistical significance.
Echocardiogram
Cardiac morphology and function were assessed by transthoracic
echocardiography using a high-resolution high-frequency system (Vevo
2100, VisualSonics, Canada) equipped with 40 MHz ultrasound probe in
conscious mice^[220]22,[221]53. Mechanical and chemical chest hair
removal was performed by shaving and using a commercially available
depilatory agent. Core temperature, monitored via a rectal probe, was
maintained at 36.5–37 °C using a heating lamp. Parasternal long and
short-axis views of the heart were captured in both B and M-modes at
optimized frame rates. LV end-diastolic (LVIDd, mm) and end-systolic
(LVIDs, mm) internal diameters, the end-diastolic wall thickness of
both interventricular septum (IVS, mm) and LV posterior wall (LVPW,
mm), LV ejection fractions (EF, %) and fractional shortening (FS, %)
were measured using M-Mode in long-axis view. Relative wall thickness
(RWT), a measure of hypertrophy, was calculated as IVSd+LVPWd/LVID.
Pulsed-Doppler recordings at the mitral inflow, left ventricular
outflow and basal septum was recorded. Digital images were stored and
analyzed offline using commercially available Vevo LAB software
(VisualSonics, Canada). An experienced echocardiography technician
blinded to experimental groups performed all the measurements.
Statistical analysis
Western Blot densitometry
The density of the bands was measured using ImageJ software and the
area under the curve corresponding to the bands of pSRC, SRC, GAPDH,
TnI, and total pTyr was calculated. Then, pTyr/TnI, SRC/GAPDH,
pSRC/GAPDH, and pSRC/SRC ratios were calculated and the ratios were
compared between groups using one-way ANOVA.
Mean orientation and circular variance from histological samples comparison
To characterize the effects of AG-825 on cardiomyocyte disarray, we
used the web graphic user interphase CytoSpectre^[222]32 to determine
the local orientation of cardiomyocytes across histological sections.
In randomly chosen regions, the myocyte mean orientation angle circular
variance was calculated. Statistical comparison between groups
(Ntg+Vehicle, Ntg+AG825, TgErbB2+Veh, TgErbB2 + AG825) was made with
one-way ANOVA.
Systems biology analysis
Statistical analysis of the label-free dataset was performed using
Perseus software^[223]54. Peptides with less than 50% of detection were
filtered, and the missing values of the intensities of the reminding
peptides were filled with k-nearest neighbors. Then, the log2
intensities were normalized by median subtraction, ANOVA, and false
discovery rate calculations. A heatmap was used to visualize the
unsupervised hierarchical clustering of proteins with a p-value <0.05
by ANOVA test. For statistical analysis of the TMT-labeled proteomics
data collected, we followed the workflow of Foster et al.^[224]21, with
few modifications. First, Partek software was used to calculate each
peptide intensity’s p-values, q-values, fold change, and ratio.
Phosphosites with less than 50% of missing density values were
subjected to statistical analysis, and the reminding missing values
were filled with the k-nearest neighbors. Then, a normalization of the
Log[2] densities by median subtraction and determined the fold-change,
ratio, and statistical significance (q and p values), from each of the
selected phosphosites of the dataset. The volcano plots were made with
GraphPad Prism 7®. Unsupervised PCA was used in both data sets
separately with Partek® software. The dataset was segregated along with
each component concerning the pTyr abundance (normalized, mean = 0,
variance = 1).
Normalization to the full proteome performed by log2 intensity
subtraction was done, followed by a two-sided t test and the moderated
p-values and q-values calculation using an algorithm developed by
Herbrich et al.^[225]55; R software was used for this analysis.
Pathway enrichment analysis
The hypergeometric p-value was used to identify the significantly
enriched processes in the proteins and phosphosites that are
differentially phosphorylated between case and control samples. For
this purpose, MsigDB^[226]23 was used to analyze the protein level
pathways and PTMsigDB^[227]24 to analyze the processes at the site
level. The population/background for which the enrichment is calculated
is restricted to all the proteins in which their sites are quantified
in the phosphoproteomics experiment, rather than all universally known
proteins/phosphosites.
KSEA
The KSEA seeks to identify kinases whose targets exhibit significantly
altered phosphorylation levels in a given condition. KSEA scores each
kinase k with a set of substrates S as follows:
[MATH: score(k<
mo>)=(P¯S−P¯)*∣S∣σ :MATH]
where
[MATH: P¯S :MATH]
Den otes the average log[2] of the fold change of all the substrates of
kinase k, and
[MATH: P¯
:MATH]
and σ represent the average and standard deviation of log[2] of the
fold change of all the identified phosphosites in the dataset. KSEA was
performed on the identified modules by restricting S to the substrates
in the module instead of all substrates in the dataset. The data
provided by PhosphoSitePLUS^[228]30 was used as the reference for
kinase-substrate associations. This tool is available from Dr. M. Ayati
website: [229]https://faculty.utrgv.edu/marzieh.ayati/software.html
Module Identification
First, networks were created in which nodes represent proteins and
edges represent the PPI obtained from BioGRID^[230]22,[231]56. The
proteins were assigned a score in the networks by computing the average
fold change of phosphosites residing on each protein obtained from
experiments individually (i.e., Ntg, TgErbB2, and R403Q-αMyHC Tg). Then
MoBaS^[232]22 was applied to identify subnetworks of highly connected
and differentially phosphorylated proteins. For visualization of
subnetworks, the proteins are colored based on the average fold change
of that protein in different conditions. If a protein in one dataset is
not identified in another dataset, the node is represented in gray
color.
Echocardiography comparison
One-way ANOVA was used to compare differences between groups to analyze
the echocardiography data. For the effect of AG-825 or vehicle
treatments, the pre- post- Treatment, we used a t-test.
Statistics and reproducibility
Experiments were conducted in replicates (see “Methods” section).
Mascot (version2.3) was used for peptide identification from MS raw
data, and MaxQuant (version 1.5) was used for quantification.
Statistical analysis was performed using R (version 4.0), Partek
Genomics Suit (version 7.0), GraphPad Prism 7.0, and Perseus (version
1.5). Additional details of data processing are described in the
“Methods” section.
Reporting summary
Further information on research design is available in the [233]Nature
Research Reporting Summary linked to this article.
Supplementary information
[234]Peer Review File^ (999.4KB, pdf)
[235]Supplementary Information^ (1.9MB, pdf)
[236]42003_2022_4021_MOESM3_ESM.pdf^ (495.1KB, pdf)
Description of Additional Supplementary Files
[237]Supplementary Data 1^ (35.8KB, xlsx)
[238]Supplementary Data 2^ (4MB, xlsx)
[239]Supplementary Data 3^ (476KB, xlsx)
[240]Supplementary Data 4^ (25.5KB, xlsx)
[241]Supplementary Data 5^ (21KB, docx)
[242]Supplementary Data 6^ (97.6MB, xlsx)
[243]Supplementary Data 7^ (38.5KB, xlsx)
[244]Reporting Summary^ (1.7MB, pdf)
Acknowledgements