Abstract
Hypertension, the leading cause of cardiovascular disease and premature
mortality, is characterized by increased vessel stretch and alterations
in vascular smooth muscle cells (VSMCs). In this study, a
vessel‐on‐a‐chip model is developed to simulate both physiological and
pathological stretch conditions alongside a mouse model of
hypertension. Proteomics analysis is applied to investigate changes in
VSMCs using the vessel‐on‐a‐chip system and compared these findings
with data from the mouse model. The results demonstrates that
physiological stretch enhances the expression of contractile markers in
VSMCs. Additionally, the chip effectively replicates cellular responses
to pathological stretch and stress, including the upregulation of ERK
signaling, calcium ion transport pathway, integrin signaling pathway,
endoplasmic reticulum stress, toll‐like receptor activation, oxidative
stress, and synthesis of sphingolipids and ceramides. These findings
indicate that the vessel chip successfully mimics in vivo biological
events associated with hypertension. The vessel‐on‐a‐chip system holds
promise for advancing the study of vessel‐related diseases and
facilitating the development of novel hypertension therapeutics.
Keywords: hypertension, microfluidics, proteomics, smooth muscle cells,
vessel‐on‐a‐chip
__________________________________________________________________
A vessel‐on‐a‐chip system is developed to replicate both physiological
and pathological vessel cyclic stretches. Proteomic analysis reveals
pressure‐overload‐induced vascular remodeling, including cellular
response to mechanical cues, cellular oxidative stresses, ER stress,
and ceramide biosynthesis activation in the vessel‐on‐a‐chip system,
well aligned with the hypertensive mouse model by Ang II infusion.
graphic file with name ADVS-12-2415024-g001.jpg
1. Introduction
Hypertension is the primary contributor to cardiovascular diseases
(CVDs) and a leading factor in premature mortality worldwide,
associated with vascular changes characterized by vascular
remodeling.^[ [46]^1 , [47]^2 , [48]^3 ^] Vascular smooth muscle cells
(VSMCs) comprise the medial layer of the vessel wall, playing a
critical role in maintaining vascular integrity and responding to
pathological changes.^[ [49]^4 ^] In their quiescent state within the
medial layer, VSMCs are largely inactive and maintain a highly
differentiated contractile phenotype. However, following vascular
injury, VSMCs undergo phenotypic transformation, acquiring
proliferative, secretory, and migratory properties. In hypertension,
VSMCs dedifferentiate, losing their contractile phenotype and
exhibiting increased proliferation, migration, and inflammation.^[
[50]^2 ^]
2D cell culture models are widely employed in vitro to study VSMCs
changes associated with CVDs. Wang et al.^[ [51]^5 ^] demonstrated that
in vitro knockdown of the orphan receptor GPRC5C in human aortic VSMCs
reduced angiotensin II (Ang II)‐induced inositol phosphate production
and myosin light chain phosphorylation, thereby decreasing Ang II
binding to AT1 receptors and alleviating established arterial
hypertension. These cell models offer a simplified system that allows
for controlled experimental conditions. However, such models diverge
significantly from living organism, lacking essential in vivo
environmental features such as mechanical stretch and fluid shear
forces.^[ [52]^6 ^]
Animal models are also widely utilized in CVDs research, with Ang
II‐treated mice being particularly common for investigating pressure
overload‐induced remodeling and related pathologies.^[ [53]^7 , [54]^8
^] Using an Ang II‐treated mouse model, Wu et al.^[ [55]^8 ^]
identified cytosolic citrate accumulation within VSMCs during Ang
II‐induced vascular remodeling in mice, promoting the transformation of
VSMCs toward a pro‐inflammatory phenotype. However, animal models are
constrained by species differences and ethical concerns, and often
require prolonged periods to observe disease progression. Furthermore,
it is estimated that only ≈20% of CVDs drugs effective in animal models
demonstrate therapeutic efficacy in humans.^[ [56]^9 ^] This limitation
underscores the challenges of relying on animal models for studying
human diseases.
Traditional mechanistic studies predominantly rely on in vitro cell
experiments and animal models, both of which insufficiently replicate
human physiological characteristics. Moreover, clinical sample analyses
lack precise control over individual pathogenic factors, highlighting
the urgent need for a novel technological platform to bridge in vitro
models, animal studies, and clinical samples. Such a platform would
facilitate the deconstruction and biomimicry of in vivo
microenvironments, enabling the stepwise isolation of potential
pathogenic factors critical to understanding vascular remodeling
mechanisms. Advanced platforms, such as organ‐on‐a‐chip models and
organoids, have been proposed to meet these needs in CVDs research.
Organ‐on‐a‐chip systems integrate microengineering and microfluidic
technologies to recreate cellular environments and biological events in
vivo, including the simulation of biophysical forces.^[ [57]^10 ^]
Meijer et al.^[ [58]^11 ^] constructed a vessel chip using gelatin
methacryloyl to replicate the 3D substrate of VSMCs in vivo,
highlighting the critical role of the microenvironment in regulating
VSMCs behavior. Their findings demonstrated that VSMCs cultured on
gelatin methacryloyl substrates with varied functionalization and
Young's modulus were inclined toward a contractile phenotype. Chen et
al.^[ [59]^12 ^] developed a vessel chip to investigate the role of
mechanical cues in VSMC phenotype switching. Cyclic stretch at 1 Hz
enhanced the contractile phenotype of VSMCs, whereas higher frequencies
induced inflammation, underscoring the impact of mechanical stretch.
However, these studies primarily elucidate phenotype switching in
VSMCs, and a more comprehensive understanding of VSMCs dynamics during
disease progression is needed. Integrating these models with advanced
analytical methodologies could help bridge this knowledge gap.
Proteins are essential functional molecules in living organisms, making
their direct characterization vital to understanding biological
functions. Liquid chromatography‐tandem mass spectrometry
(LC‐MS/MS)‐based proteomics enables comprehensive protein profiling
across cells, tissues, and organisms under specific conditions,
encompassing protein identification, quantification, and
post‐translational modifications.^[ [60]^13 ^] Recently, with the
advances of mass spectrometry techniques and artificial intelligence,^[
[61]^14 , [62]^15 ^] proteomic analysis from a low amount of sample is
possible, further advancing personalized risk stratification, early
diagnosis, and precision therapy.^[ [63]^16 , [64]^17 ^] Therefore,
integrating advanced organ‐on‐a‐chip systems with cutting‐edge
proteomics techniques is feasible and essential for gaining a
comprehensive understanding of disease mechanisms, such as vascular
remodeling in hypertension.
In this study, we developed a vessel‐on‐a‐chip system comprising a cell
culture layer and a pneumatic layer separated by a polydimethylsiloxane
(PDMS) membrane. By applying varying cyclic pressures to the pneumatic
layer, the chip can simulate both physiological and hypertensive
vascular conditions. Proteomic analysis was conducted to characterize
protein and pathway changes in VSMCs under these conditions.
Additionally, an in vivo mouse model of hypertension was constructed to
enable a comparative analysis with the in vitro vessel chip model. Our
results demonstrated that the vessel chip effectively replicates
several hypertension‐associated characteristics in VSMCs, including
their responses to mechanical cues and cellular stress. This approach
offers significant potential for advancing research into vessel‐related
diseases and facilitating the development of novel therapeutic
strategies.
2. Results and Discussion
2.1. The Vessel‐On‐A‐Chip System
To mimic the in vivo arterial environment, we constructed a
vessel‐on‐a‐chip system. The device consists of a cell culture layer
and a pneumatic layer separated by an elastic membrane (Figure
[65]1A–C). PDMS was selected for fabricating the microfluidic chip due
to its transparency and biocompatibility. Additionally, PDMS's
elasticity allows it to deform under pressure. The upper layer includes
two columns of trapezoidal pillars, dividing it into a central cell
culture channel and two flanking medium channels, each 1 mm in width.
The gap between adjacent pillars is 40 µm. Although the central channel
is not fully enclosed, the pillars effectively confine liquid within
this channel due to surface tension. A pre‐polymer mixture of rat tail
collagen and the mouse aortic smooth muscle cell line (MOVAS) was
introduced into the cell culture channel. After 15 min of
polymerization, the cells were encapsulated within the collagen
structure (Figure [66]1D).
Figure 1.
Figure 1
[67]Open in a new tab
Illustration of the vessel‐on‐a‐chip system. A,B) Structure of the
vessel chip. The chip consisted of a cell culture layer and a pneumatic
layer separated by a PDMS membrane. The upper cell culture layer
contained two columns of trapezoidal pillars to separate the layer into
cell culture channel and flanking medium channels. A mixture of rat
tail collagen type I and the mouse aortic smooth muscle cell line MOVAS
was injected into the middle channel and polymerized, followed by
introducing culture medium into the two flanking medium channels. C)
Image of the vessel chip viewed from above. D) Cells cultured in the
vessel chip. E) Deformation of the cell‐containing collagen during
cyclic stretch. F) Measurement of the deformation of the collagen using
polystyrene microspheres.
To induce cyclic stretch, the bottom layer of the device was connected
to a push‐pull system that provided periodic pressure, with the
frequency set at 1 Hz. During each cycle, pressure application deformed
the PDMS membrane, compressing the collagen‐encapsulated cells; upon
release, both the membrane and collagen returned to their original
shapes (Figure [68]1E; Video [69]S1, Supporting Information). To
quantify deformation, polystyrene microspheres were mixed with rat tail
collagen and injected into the channel, where their displacement
reflected collagen stretching (Figure [70]1F; Video [71]S2, Supporting
Information). Deformation was defined as the ratio of the distance
traveled by a microsphere in one cycle to its minimum distance from the
channel center. Significant displacement can be observed in the x‐axis.
There was little displacement in the y‐axis (only ≈0.9% to the total
displacement of the beads in x‐ and y‐ axis). There should also be
significant displacement in the z‐axis based on the design of the
microfluidic chip, which, however, cannot be measured using the
microscopic characterization system.
In normal aortic and other elastic arteries, vessels expand by ≈10%
with each heartbeat.^[ [72]^18 ^] By contrast, hypertensive vessels are
exposed to increased cyclic stretch.^[ [73]^19 ^] In vitro models
typically classify 5–10% cyclic stretch as physiological and cyclic
stretch exceeding 15% as pathological.^[ [74]^20 ^] Accordingly, we
applied cyclic stretch to simulate both physiological and pathological
conditions. When a pressure range of 0 to 30 mbar was applied, the
average x‐axis deformation measured 7.6 ± 4.1% based on a total of 450
microspheres from the left and right sides randomly selected from three
independent chips (150 microspheres per chip). The three chips showed
highly reproducible results, where the average ± standard deviation of
the three chips based on their respective average x‐axis deformation
was 7.6 ± 0.2%. Increasing the pressure range to 0–110 mbar resulted in
an average x‐axis deformation of 19.0 ± 9.3% and a reproducibility
evaluation of 19.0 ± 1.0% across three chips. These deformation values
closely correspond to physiological and pathological vessel expansion,
respectively. Thus, we selected 30 mbar and 110 mbar as cyclic
pressures to model physiological and pathological vessel stretching,
respectively.
Notably, the extracellular matrix (ECM) exhibits nonelastic properties,
yet mechanical forces can induce ECM remodeling.^[ [75]^21 , [76]^22 ^]
Under high strain, collagen may reorganize, potentially modifying
mechanical signaling and influencing cell function. To assess the
stability of the microfluidic system over time, we measured collagen
deformation at 0h (7.6 ± 4.1% for 30 mbar, 19.0 ± 9.3% for 110 mbar)
and 24 h (7.3 ± 3.5% for 30 mbar, 18.3 ± 13.2% for 110 mbar). No
significant changes were observed under either condition
(Figure [77]S1, Supporting Information), indicating that cells
experienced a consistent mechanical environment and pressure throughout
the 24‐h period.
2.2. Physiological Stretch Promotes the Contractile Phenotype of MOVAS
Untargeted proteomic analysis was performed on MOVAS cells in both
static conditions and under physiological cyclic stretch (30 mbar)
after 24 h. A total of 9083 proteins (Table [78]S1, Supporting
Information) were identified and quantified across the two groups.
Principal component analysis (PCA) clearly separated the static and
physiological stretch groups, indicating a significant difference
between them (Figure [79]2A). Proteins that exhibited significant
changes were identified based on criteria of p‐value < 0.05 and fold
change (FC), resulting in 669 upregulated (FC >1.5) and 161
downregulated (FC< 2/3) proteins (Figure [80]2B; Table [81]S2,
Supporting Information). Heatmap analysis of all identified proteins
further showed distinct clustering between the two groups
(Figure [82]2C).
Figure 2.
Figure 2
[83]Open in a new tab
Proteomic and immunofluorescence characterization of the MOVAS cells
under physiological stretch and static culture. A) PCA score plots for
proteomic data comparing physiologically stretched and static cultured
cells. B) Volcano plots displaying identified proteins in MOVAS with
log[2]FC as the horizontal axis and −log[10] p‐value as the vertical
axis. The thresholds were defined as 0.05 for the p‐value and 0.585 for
|log[2]FC|. FC, fold change in protein expression of the
physiologically stretched group compared to the static group. C)
Heatmap clustering of physiological stretched and static group cells
based on the z‐scores for protein contents. Pearson correlation was
used as the distance measurement for both the columns and rows. D,E)
KEGG and GO enrichment analysis of the significantly changed proteins
obtained from (C). BP, Biological Process; CC, Cellular Component. F)
Fluorescence images of the static and physiological stretched groups
stained for MYH11. The scale bar indicates 100 µm. G) Quantification of
MYH11 fluorescence intensity. Data presented as mean±standard
deviation, n = 3, p‐values were calculated using two‐tailed t‐test. ^**
p < 0.01.
Among the upregulation proteins, several associated with VSMC phenotype
regulation and inflammation were identified, including myosin
regulatory light polypeptide 9 (MYL9), four and a half LIM domains
protein 2 (FHL2), and tropomyosin alpha‐1 chain (TPM1) (Figure [84]S2,
Supporting Information). MYL9 and TPM1 are positively correlated with
the maintenance of the VSMC contractile phenotype,^[ [85]^23 , [86]^24
^] while FHL2 is linked to inflammation suppression.^[ [87]^25 ^] The
observed upregulation of these proteins suggested that physiological
stretch enhanced the contractile phenotype of the MOVAS cells. KEGG
pathway enrichment analysis and Gene Ontology (GO) enrichment analysis
revealed notable enrichment of pathways related to the cell
cytoskeleton (Figure [88]2D,E). The cytoskeleton, which includes
intermediate filaments, actin, and other components, is crucial for the
contractile function and phenotype of VSMCs.^[ [89]^26 ^] Intermediate
filaments play a significant role in the development of smooth muscle
force. Upon activation, they undergo spatial rearrangement that is
associated with cytoskeletal reorganization.^[ [90]^27 ^] Actin also
contributes to cytoskeletal dynamics, which is essential for the
contraction of smooth muscle tissues.^[ [91]^28 ^]
In healthy arteries, VSMCs typically adopt a differentiated,
contractile phenotype, characterized by high expression of smooth
muscle myosin heavy chain 11 (MYH11).^[ [92]^29 ^] We employed
immunofluorescence to quantify MYH11 expression in VSMCs within our
microfluidic chips. The physiological stretch group showed
significantly higher fluorescence intensity compared to the static
group (Figure [93]2F,G), indicating a healthier state of MOVAS cells
under physiological stretching. These results collectively demonstrate
the successful establishment of a vessel‐on‐a‐chip system that
recapitulates physiological conditions.
2.3. Proteomics Uncovers Dysregulation of MOVAS Under Pathological Stretch
We conducted untargeted proteomics on MOVAS cells subjected to both
physiological (30 mbar) and pathological (110 mbar) stretches. In
total, 9102 proteins (Table [94]S3, Supporting Information) were
identified and quantified across both groups. PCA revealed a clear
separation between the physiological and pathological stretch groups,
indicating significant changes in MOVAS cells under pathological
stretch (Figure [95]3A). Heatmap analysis of all identified proteins
further showed distinct clustering between the two groups (Figure
[96]S3, Supporting Information). Proteins that exhibited significant
changes were identified using the criteria of p‐value < 0.05 and FC,
resulting in 829 upregulated (FC > 1.5) and 952 downregulated (FC <
2/3) proteins (Figure [97]3B; Table [98]S4, Supporting Information).
KEGG pathway enrichment analysis indicated that pathological stretch
impacted multiple pathways (Figure [99]3C), notably those regulating
the actin cytoskeleton, reactive oxygen species, sphingolipid
metabolism, glycolysis and fatty acid metabolism. Mechanical stress
triggers the cytoskeleton remodeling through various signaling
pathways.^[ [100]^30 , [101]^31 ^] The enrichment of actin cytoskeleton
regulation reflected a cellular response to pathological stretch.
Additionally, mechanical cues can affect a range of cellular functions,
including morphology, differentiation, proliferation, adhesion, growth,
and migration.^[ [102]^32 ^] The enrichment of pathways related to
reactive oxygen species and sphingolipid metabolism suggested a
heightened cellular stress response.^[ [103]^33 , [104]^34 ^] Moreover,
the enrichment of glycolysis and fatty acid metabolism pathways implied
metabolic reprogramming in MOVAS, which is crucial for phenotypic
switching.^[ [105]^35 ^] In hypertension and other vascular disorders,
the contractile phenotype of VSMCs is compromised, typically reflected
by decreased expression of the contractile marker MYH11.^[ [106]^29 ,
[107]^36 ^] Immunofluorescence analysis confirmed this in MOVAS cells
(Figure [108]3D,E), showing reduced MYH11 expression under pathological
stretch and indicating loss of the contractile phenotype.
Figure 3.
Figure 3
[109]Open in a new tab
Proteomic and immunofluorescence characterization of the MOVAS cells
under physiological and pathological stretches. A) PCA score plots for
proteomic data comparing cells cultured under pathological and
physiological stretches. B) Volcano plots displaying identified
proteins in MOVAS with log[2]FC as the horizontal axis and −log[10]
p‐value as the vertical axis. The thresholds were set as 0.05 for the
p‐value and 0.585 for |log[2]FC|. FC, fold change in protein expression
of the pathological stretched group compared to the physiological
group. C) Pathways from KEGG enrichment analysis of the significantly
changed proteins obtained from (B). D) Fluorescence images of MYH11
staining in the physiological and pathological stretched groups after
two days of culture. The scale bar indicates 100 µm. E) Quantification
of MYH11 fluorescence intensity. Data presented as mean±standard
deviation, n = 3, p‐values were calculated using two‐tailed t‐test. ^**
p < 0.01.
To corroborate the vessel chip findings, we established a hypertensive
mouse model using Ang II (Figure [110]S4A, Supporting Information). Ang
II administration via mini‐osmotic pumps significantly elevated both
systolic and diastolic blood pressure relative to saline‐injected
controls (Figure [111]S4B, Supporting Information). Vessels harvested
following Ang II infusion displayed marked collagen deposition (Figure
[112]S4C,D, Supporting Information). Quantitative analysis of trichrome
staining images showed that the collagen area ratio significantly
increased in the Ang II infusion group, indicating vascular remodeling
(Figure [113]S4E, Supporting Information). After 14 days of Ang II
treatment, proteomic analysis of the mouse aortas identified 8234
proteins, with 751 upregulated and 612 downregulated in the Ang
II‐treated group compared to the controls (Figure [114]S5A and Tables
[115]S5 and [116]S6, Supporting Information). PCA clearly distinguished
the Ang II‐treated group from saline‐injected controls (Figure
[117]S5B, Supporting Information), which was further confirmed by
heatmap clustering (Figure [118]S5C, Supporting Information). KEGG
enrichment analysis of the dysregulated proteins indicated that
hypertensive mouse vessels also responded to pathological mechanical
signals and cellular stress, along with undergoing metabolic
reprogramming (Figure [119]S5D, Supporting Information).
2.4. Proteomics Reveals MOVAS Response to Mechanical Cues
KEGG enrichment analysis suggested that cells respond to mechanical
cues during pathological stretch. To further validate these findings,
we performed GO enrichment on the significantly changed proteins. Terms
related to integrin, calcium, and extracellular signal‐regulated kinase
(ERK)—previously known to be linked to mechanical stimuli^[ [120]^37 ,
[121]^38 , [122]^39 ^]—were detected. Both our in vitro vessel chip
model and the in vivo Ang II‐treated mice model showed enrichment of
these terms (Figure [123]4A; Figure [124]S6, Supporting Information).
Multiple proteins involved in ERK signaling were dysregulated
(Figure [125]4B). ERK signaling regulates fundamental cellular
processes in response to extracellular cues, including protein
synthesis, differentiation, cell‐cycle entry, cell survival, and cell
motility.^[ [126]^37 , [127]^40 , [128]^41 ^] It mediates
phosphorylation of ERK1/2, promoting transcription of genes associated
with growth and mitogenic signals. Key mediators, such as Protein
kinase C (PKC) and ERK1/2, play essential roles in responding to
mechanical stress through intricate signaling networks. For instance,
PKCε forms subcellular‐targeted signaling complexes with ERKs, wherein
the mitochondrial PKCε‐ERK module mediates cardioprotective effects by
phosphorylating and inactivating Bad.^[ [129]^42 ^] Shear stress also
activates ERK1/2 in endothelial cells, influencing VSMC proliferation
and migration via endothelial cell‐dependent mechanisms.^[ [130]^43 ^]
IPA upstream regulator analysis predicted ERK activation in our in
vitro model (Figure [131]4C), while gene set enrichment analysis (GSEA)
indicated positive regulation of the ERK1 and ERK2 cascades under
pathological stretch (Figure [132]4D). Proteins involved in the ERK
signaling pathway exhibited an upregulation trend in the pathological
stretch group (Figure [133]S7, Supporting Information). Consistent with
these findings, both GSEA and IPA confirmed ERK pathway activation in
the in vivo mice model (Figure [134]S8A,B, Supporting Information).
Collectively, these results demonstrate that ERK signaling was
activated under pathological stretch and in Ang II‐treated mice,
aligning with established knowledge of hypertension pathogenesis.^[
[135]^44 , [136]^45 ^]
Figure 4.
Figure 4
[137]Open in a new tab
Proteomic analysis of MOVAS responses to mechanical cues under the
pathological stretch. A) Pathways related to cellular responses to
mechanical stress from GO enrichment analysis of the significantly
changed proteins between pathological and physiological stretched
groups. B) The PKC/ERK signaling pathway. C) IPA upstream regulator
analysis predicted ERK activation. Dashed lines indicate an indirect
relationship. D,E) GSEA predicted the activation of ERK1 and ERK2
cascade, as well as calcium ion transport. F) Upregulation of TRPM7 in
the pathological stretched group. Data presented as mean±standard
deviation, n = 4, p‐values were calculated using two‐tailed t‐test. ^*
p < 0.05.
The calcium ion transport pathway was also activated under pathological
stretch and Ang II‐induced hypertension in mice, as indicated by GSEA
enrichment (Figure [138]4E; Figure [139]S8C, Supporting Information).
Calcium ions serve as crucial second messengers that initiate and
regulate VSMCs contraction.^[ [140]^46 ^] Excessive mechanical stretch
can disrupt intracellular calcium homeostasis, leading to contractile
dysfunction in VSMCs.^[ [141]^47 , [142]^48 ^] Inositol
1,4,5‐trisphosphate‐gated calcium channels (ITPRs), primarily located
on the endoplasmic reticulum (ER) membrane, are pivotal for calcium
signaling.^[ [143]^38 ^] Specifically, ITPRs bind inositol
1,4,5‐trisphosphate (IP3) and, upon activation, mediate the release of
calcium ions into the cytoplasm. In our in vitro vessel chip model,
ITPR1 and ITPR3 displayed an upregulation trend under pathological
stretch, further suggesting intracellular calcium dysregulation (Figure
[144]S9, Supporting Information).
Transient receptor potential cation channel subfamily M member 7
(TRPM7), a mechanosensitive ion channel,^[ [145]^49 ^] is essential for
maintaining and regulating intracellular calcium levels^[ [146]^50 ^]
as well as mediating mechanical force sensing.^[ [147]^51 ^] Previous
studies associate TRPM7 upregulation with loss of the contractile
phenotype in VSMCs.^[ [148]^50 ^] Consistently, we observed elevated
TRPM7 expression under pathological stretch (Figure [149]4F),
indicating calcium dysregulation and a cellular response to mechanical
cues. Integrins, which are involved in cell differentiation,
proliferation, and survival, also contribute to mechano‐sensing.^[
[150]^39 , [151]^52 ^] In hypertension, integrin αV (ITGAV) expression
is increased, driving vascular eutrophic inward remodeling.^[ [152]^52
^] The enrichment of integrin signaling (Figure [153]4A) and the
elevated trend of ITGAV (Figure [154]S10, Supporting Information)
likewise support a cellular response to mechanical stress in
pathologically stretched MOVAS.
2.5. Proteomics Discloses Cellular Stresses Induced by Pathological Stretch
KEGG enrichment analysis also suggested cellular stress induced by
pathological stretch. To further validate these findings, GO enrichment
analysis was performed to investigate cellular stress responses in
MOVAS cells and mouse vessels (Figure [155] 5A; Figure [156]S11,
Supporting Information). Under pathological stretch of cells or Ang II
treatment of mice, multiple pathways were activated, particularly those
related to ER stress, oxidative stress, pattern recognition receptors,
and sphingolipid metabolism.
Figure 5.
Figure 5
[157]Open in a new tab
Proteomic characterization of MOVAS cellular stress responses under
pathological stretch. A) Pathways related to cellular stresses
identified through GO enrichment analysis of significantly altered
proteins between the pathological and physiological stretched groups.
B) ATF6 activation predicted by IPA upstream regulator analysis. Dashed
lines indicate an indirect relationship, while solid lines indicate a
direct relationship. C–E) Downregulation of SIRT1 (D) and SOD1 (E) in
the pathological stretch group contributed to oxidative stress (data
presented as mean±standard deviation, n = 4, p‐values calculated using
two‐tailed t‐test. ^* p < 0.05, ^** p < 0.01). F) NF‐κB activation
predicted by IPA upstream regulator analysis. G) TLR4 activation
predicted by IPA upstream regulator analysis. H) Schematic
representation of the ceramide biosynthesis pathway.
As the primary intracellular calcium reservoir, ER plays a fundamental
role in protein maturation and folding. Impairment of its
protein‐folding capacity leads to the accumulation of unfolded or
misfolded proteins, triggering ER stress.^[ [158]^53 ^] This stress
response is implicated in numerous CVDs, including hypertension. To
mitigate ER stress, cells initiate the unfolded protein response (UPR),
an adaptive mechanism aimed at restoring protein‐folding homeostasis
and preserving cellular function. ATF6 is a crucial regulatory factor
of the UPR. Under conditions of ER stress, ATF6 translocates from the
ER to the Golgi apparatus, where proteolytic cleavage releases its
active form. Once activated, ATF6 facilitates proper protein folding
and drives the degradation of misfolded proteins, ultimately restoring
ER function.^[ [159]^54 , [160]^55 ^] Our GSEA further revealed that ER
stress and UPR pathways were activated in both the pathological stretch
cell model and the Ang II‐treated mice model (Figure [161]S12A,B and
[162]S13A,B, Supporting Information). Additionally, IPA upstream
regulator analysis predicted the activation of ATF6 (Figure [163]5B;
Figure [164]S14A, Supporting Information), indicating that both MOVAS
cells under pathological stretch and vessels in hypertensive mice
respond to ER stress.
Oxidative stress, defined by an imbalance between the generation and
elimination of reactive oxygen species (ROS), mediates cellular
dysfunction, inflammatory responses, and apoptosis through mechanisms
such as lipid peroxidation, protein oxidation, and DNA damage.^[
[165]^56 ^] It is a critical pathogenic factor in various diseases,
including hypertension and other CVDs.^[ [166]^33 ^] Pathways related
to oxidative stress were significantly enriched in both Ang II‐treated
mice and MOVAS cells under pathological stretch (Figure [167]5A; Figure
[168]S11, Supporting Information). GSEA indicated hyperactivation of
the cellular response to oxidative stress, suggesting a disruption of
ROS homeostasis (Figures [169]S12C and [170]S13C, Supporting
Information). Numerous proteins associated with oxidative stress were
dysregulated in both models (Figure [171]5C). NAD‐dependent protein
deacetylase sirtuin‐1 (SIRT1) is recognized for its protective role in
various diseases, including vascular remodeling during hypertension and
ROS regulation.^[ [172]^57 , [173]^58 ^] Our data showed that SIRT1 was
downregulated under pathological stretch (Figure [174]5D), potentially
exacerbating oxidative stress and contributing to the pathological
phenotype. In the pathological stretch group, superoxide dismutase
[Cu‐Zn] (SOD1) was downregulated (Figure [175]5E). SOD1 typically
protects against oxidative stress by catalyzing the dismutation of
superoxide anion (O[2] ^•−) into hydrogen peroxide (H[2]O[2]), which is
subsequently broken down into water by catalase (CAT) and the
glutathione pathway. Therefore, reduced SOD1 expression may lead to
elevated ROS levels, exacerbating cellular oxidative stress.^[ [176]^59
^] Additionally, CAT and glutathione pathway components (GLRX2, GPX4,
GPX7) exhibited upregulating trends (Figure [177]S15, Supporting
Information), indicating the activation of compensatory antioxidant
defense mechanisms.^[ [178]^60 ^]
Our data also showed activation of the NF‐κB complex under pathological
stretch (Figure [179]5F), a ROS‐sensitive transcription factor that
plays a complex and pivotal role in oxidative stress and inflammatory
responses.^[ [180]^61 ^] Excessive ROS production serves as a major
trigger for NF‐κB activation. SIRT1 can inhibit the NF‐κB signaling
pathway through multiple mechanisms,^[ [181]^62 ^] one of which
involves enhancing antioxidant defenses—such as increasing the
expression of superoxide dismutase (SOD)—to reduce ROS levels, and
thereby indirectly suppress NF‐κB activation. Accordingly, the observed
downregulation of SIRT1 likely contributed to NF‐κB activation in
concert with increased ROS. Furthermore, endoplasmic reticulum
oxidoreductase 1 (ERO1), which generates hydrogen peroxide in the ER,^[
[182]^63 ^] was upregulated in the pathologically stretched group,
potentially exacerbating ROS burden and oxidative stress (Figure
[183]S16, Supporting Information). Consistently, analyses of vessels
from hypertensive mice confirmed reduced SOD1 expression and SIRT1
inhibition, along with NF‐κB activation and ERO1 upregulation, aligning
with our vessel chip model findings (Figure [184]S14B–E, Supporting
Information).
Toll‐like receptors 4 (TLR4), a key mediator of inflammatory responses,
is associated with vascular remodeling and hypertension.^[ [185]^64 ^]
When TLR4 is elevated, it can drive cell proliferation and
inflammation, thereby promoting vascular pathology. Our results
revealed the activation of TLR4 in both the hypertensive mouse model
and the pathological stretch cell model (Figure [186]5G; Figures
[187]S12D, [188]S13D, and [189]S14F, Supporting Information),
indicating that pattern recognition receptors mediate cellular stress
responses under these conditions.
Sphingolipids are a class of complex lipids widely present in cell
membranes, characterized by a sphingosine backbone that is amide‐linked
to fatty acids to form ceramide, which in turn gives rise to a diverse
array of sphingolipid species. In addition to serving as integral
structural components of cell membranes, sphingolipids act as signaling
molecules in various physiological and pathological processes,
including apoptosis, inflammation, and vascular function.^[ [190]^65 ^]
Previous studies have linked VSMCs injury and vascular remodeling to
the synthesis or accumulation of ceramide and other sphingolipids.^[
[191]^66 , [192]^67 ^] Our results demonstrated that sphingolipid and
ceramide biosynthetic pathways were significantly enriched in both the
hypertensive mouse model and the pathological stretch cell model, as
revealed by GSEA (Figure [193]5H; Figures [194]S12E,F and [195]S13E,F,
Supporting Information). Moreover, proteins involved in ceramide
synthesis displayed mainly upward trends in both pathological stretched
cells and vessels from hypertensive mice (Figures [196]S17 and
[197]S18, Supporting Information), suggesting that ceramide
biosynthesis is activated under cellular stress, aligning with previous
research.^[ [198]^68 , [199]^69 ^]
2.6. Limitation Discussion of the Vessel‐On‐A‐Chip Platform
Our vessel chip model successfully recapitulated several cellular
processes relevant to hypertension observed in vivo. Compared with
traditional 2D cultures, a notable advantage of this vessel‐on‐a‐chip
platform is its ability to closely mimic the native vascular
microenvironment while enabling precise hypertension modeling through
the modulation of cyclic mechanical strain. Nonetheless, several
limitations remain. First, the vasculature comprises multiple cell
types—including VSMCs, endothelial cells, monocytes/macrophages, and
other immune cells—all of which function synergistically under
physiological conditions. A single‐cell vessel chip model, lacking
co‐culture with endothelial and immune cells, cannot fully capture the
vascular dysfunction associated with hypertension. Second, in vivo
hypertension triggers remodeling of ECM components (e.g., collagen,
elastin), whereas microfluidic chips typically rely on a simplified ECM
that fails to adequately reflect these changes. Moreover, mechanical
cues can induce ECM reorganization in vessel chip model,^[ [200]^21 ,
[201]^22 ^] underscoring the need to investigate collagen remodeling
for understanding hypertension pathophysiology. Advancements in
biomaterial development—such as artificial hydrogels^[ [202]^70 ^] and
inorganic biomaterials^[ [203]^71 ^]—are therefore critical for more
accurately reproducing the native ECM environment. Third, hypertension
is a multifaceted pathological state involving intricate crosstalk
among various organs and cell types, with immune regulation by the
kidney and brain playing a particularly important role.^[ [204]^72 ^]
Proteomic analysis of hypertensive mouse vessels revealed significant
enrichment of pathways related to macrophage immune responses,
highlighting the interplay between hypertension and immunology (Figure
[205]S19, Supporting Information). Nevertheless, our current vessel
chip model lacks multi‐organ integration, precluding a comprehensive
representation of these immune signals. Fourth, hypertension is a
chronic condition, and short‐term microfluidic studies cannot
adequately model the long‐term vascular remodeling it entails. Future
developments should thus focus on developing more advanced models that
better capture the in vivo immune interactions, the native ECM
environment, the multi‐organ crosstalk, and the protracted remodeling
processes associated with hypertension.
3. Conclusion
In summary, we developed a vessel‐on‐a‐chip model to simulate both
healthy and hypertensive vascular conditions. A pressure of 30 mbar was
applied to mimic physiological stretch, whereas 110 mbar simulated
pathological stretch. Under physiological stretch, contractile marker
expression was upregulated, reflecting a contractile phenotype and a
healthier state in MOVAS cells. Conversely, pathological stretch
resulted in the downregulation of these contractile markers, signifying
a loss of contractile phenotype. Proteomic analysis demonstrated that
MOVAS cells responded to mechanical cues and cellular stress under
pathological stretch. To validate these findings, we established a
mouse hypertension model using Ang II infusion. Observations from the
vessel chip were consistent with the in vivo model, demonstrating
comparable pressure‐overload‐induced remodeling of VSMCs. Notably, key
pathways—including ERK signaling, calcium ion transport,
integrin‐mediated signaling, ER stress response, oxidative stress
response, TLR4 signaling, and ceramide synthesis—exhibited similar
changes in the vessel‐on‐a‐chip system and the mouse hypertension
model, signifying successful replication of in vivo conditions. This
vessel‐on‐a‐chip system holds promise for advancing our understanding
of vascular disease and facilitating the development of novel
antihypertensive therapies. Future research could integrate multiple
organs‐on‐a‐chip platforms to more accurately reflect the intricate
cellular interactions underlying hypertensive pathology.
4. Experimental Section
Microfluidic Device Design, Fabrication, and Assembly
The microfluidic device was adapted from a previously reported
heart‐on‐a‐chip design.^[ [206]^73 ^] It consisted of an upper cell
culture layer and a lower pneumatic layer, separated by a PDMS
membrane. The upper channel was 3 mm in width and 170 µm in height and
was divided into three channels by two columns of trapezoidal pillars.
Each pillar measures 160 µm, with a 40 µm gap between pillars within a
column and a 1 mm distance between the two columns. The master molds
were fabricated using standard photolithography with SU‐8 2050 and SU‐8
2075 photoresist (MicroChem, USA) for the upper and lower channels,
respectively, on a 3 inch silicon wafer substrate (Figure [207]S20,
Supporting Information). A direct write optical lithography system
(Durham Magneto Optics Ltd., UK) was used to generate patterns. The
PDMS pre‐polymer was mixed with a curing agent at a 10:1 weight ratio,
degassed, and then poured into the master mold. Polymerization was
carried out at 95 °C for 2 h.
The PDMS stamps were peeled from the molds after curing. Inlets and
outlets for cells and culture medium were created using a steel needle
with an outer diameter (OD) of 1.2 mm and inner diameter (ID) of 0.8
mm. The upper layer and the membrane were activated with oxygen plasma
before assembly, followed by incubation at 95 °C for at least 2 h to
ensure irreversible bonding. The pneumatic inlet and outlet were
created with a needle of OD 0.8 mm and ID 0.6 mm and then the block was
bonded with the pneumatic layer using oxygen plasma. Cells and culture
medium were introduced into the microfluidic device using an 18‐gauge
syringe with a steel needle (OD 1.2 mm, ID 0.8 mm) and
polytetrafluoroethylene (PTFE) tubing (OD 2 mm, ID 1 mm). For the
pneumatic layer, a 22‐gauge syringe with a steel needle (OD 0.8 mm, ID
0.6 mm) and PTFE tubing (OD 1.6 mm, ID 0.6 mm) were used. All syringes,
needles, and tubing were washed with 75% ethanol and distilled water
before being connected to the microfluidic chip with PDMS as a sealant.
Cell Culture in the Microfluidic Device
The mouse aortic smooth muscle cell line MOVAS was obtained from Meisen
CTCC (China). MOVAS cells were cultured in a DMEM medium (Solarbio,
China) supplemented with 10% fetal bovine serum (FBS, ExCell, China),
penicillin, and streptomycin. The cells were maintained and expanded in
T25 cell culture flasks (Nest, China) until they were ready for use in
the microfluidic device.
The microfluidic device was sterilized with 75% ethanol for 2 min and
then washed twice with distilled water. After drying, the device was
exposed to UV light for an additional 2 h for further sterilization.
Before the cell culture, the cell culture channel was treated with
dopamine hydrochloride (Sigma, USA) for 30 min and subsequently washed
with distilled water. For cell seeding, MOVAS cells were digested and
suspended in rat tail collagen type I (Corning, USA, 9.45 mg mL^−1).
The volume ratio of cell suspension to rat tail collagen to NaOH
solution (0.13 mol L^−1) was set at 2:1:0.15 to achieve a final
concentration of 6 × 10^6 cells mL^−1. After thorough mixing, the cell
suspension was injected into the chip. The device was then placed in a
cell incubator for 15 min to facilitate the polymerization of the rat
tail collagen. Subsequently, the two medium channels were connected to
a syringe pump (LongerPump, China), and DMEM medium was injected at a
flow rate of 1 µL min^−1 to supply nutrients to the cells. After one
day's static culture, the pneumatic channel was connected to a
push‐pull pump (Fluigent, France) to provide cyclic pressure to the
chip, applying pressures ranging from 0 to 30 mbar or 0 to 110 mbar at
1 Hz for 24 h, adapting from the previously described method.^[
[208]^74 ^] Statically cultured vessel chips were used as controls.
Cyclic Strain Characterization
Cyclic strain characterization was conducted as previously described by
tracking the displacement of polystyrene microspheres with diameters
ranging from 5 to 5.9 µm (2.5% w/v, Aladdin, China) during
stretching.^[ [209]^75 ^] The microspheres were first diluted in PBS
(Solarbio, China) at a 1:15 ratio and then combined with rat tail
collagen (9.45 mg mL^−1, Corning, USA) in a 2:1 ratio, yielding a final
concentration of 0.1% w/v. A concentrated NaOH solution (0.13 mol L^−1)
was added to adjust the pH to 7, facilitating collagen polymerization.
This mixture was then injected into the middle channel of the upper
layer and allowed to polymerize for 15 min. After polymerization, the
side channels were filled with PBS.
Mechanical characterization was performed to deform the upper layer at
a frequency of 1 Hz, with pressures varying between 0 and 30 mbar or 0
and 110 mbar. During cyclic stretching, microsphere movement was
recorded using an inverted microscope (Cewei, China). The recorded
frames were analyzed with ImageJ software,^[ [210]^75 ^] and a total of
150 microspheres from the left and right sides were randomly selected
to calculate the deformation for each group.
Deformation along the x‐axis was calculated as follows:
[MATH: Displacementoftheleftmicrospheres=Δxx0
−xmax=xmax−xminx0−xmax<
/mrow> :MATH]
(1)
[MATH:
Displacementoftherightmicrospheres=Δxxmin−x
0=xmax−xminx<
mrow>min−x<
/mi>0 :MATH]
(2)
where x[0] represents the center position, and x[max] and x[min] denote
the maximum and minimum x‐coordinates of the microspheres during
stretching. The ratio of displacement along the y‐axis contributing to
the total displacement was calculated as:
[MATH:
Displacementcontributionalongy−axis=1−xmax−xminymax−ymin2+xmax−xmin2 :MATH]
(3)
where y[max] and y[min] represent the maximum and minimum y‐coordinates
of the microspheres during stretching. The mean and standard deviation
of deformation along the x‐axis were also calculated.
To evaluate the repeatability of the device, the average deformation of
the collagen was measured under pressures of 0–30 mbar and 0–110 mbar
for three chips separately. The mean and standard deviation of the
deformation were calculated. Additionally, the deformation of the three
chips under both pressure conditions was measured at 0 and 24 h after
cyclic stretch to characterize the device's consistency.
Immunofluorescence Staining
Immunofluorescence analysis was performed on the microfluidic chip for
cell samples statically cultured and physiologically and pathologically
stretched. Cell samples were sequentially incubated with 4%
paraformaldehyde (Solarbio, China) for 10 min for fixation, 0.1%
TritonX‐100 (Biosharp, China) for 8 min to permeabilize the cells, and
10% goat serum (Solarbio, China) for 90 min to block nonspecific
binding. Following blocking, samples were incubated overnight with a
rabbit anti‐MYH 11 antibody (1:100 dilution in PBST, Abcam, UK) in the
presence of 2% goat serum. After washing with PBST, samples were
further incubated with Alexa Fluor 488‐conjugated goat anti‐rabbit
secondary antibody (1:250 dilution in PBST, ThermoFisher Scientific,
USA) for 1 h. Fluorescence imaging was recorded using a Zeiss LSM 980
Inverted Confocal Microscope (Zeiss, Germany) with a 10x objective
lens. The excitation wavelength was set at 488 nm, while the receiving
wavelength was set as 490–551 nm. The excitation and max emission
wavelengths of the secondary antibody were 499 and 520 nm. All the
images were analyzed and quantified using ImageJ software and GraphPad
Prism 10.
In Vivo Assays for Hypertension Model
C57BL/6 mice (8–10 weeks old) were obtained from SiPeiFu Biotechnology
Company (Beijing, China) and housed in groups of five in individually
ventilated cages under controlled conditions (19–23 °C, 12 h light‐dark
cycle). Mice were randomly divided into a treatment group and control
group and anesthetized by intraperitoneal injection of a 4% chloral
hydrate solution at a dose of 10 µL g^−1 body weight. Following
anesthesia, mini‐osmotic pumps (Model 1002, Alzet, USA) were implanted
subcutaneously, with the catheter directed toward the mouse's head.
Pumps were loaded with Ang II (Sigma–Aldrich, USA) diluted in saline
according to the manufacturer's instructions. Each mouse received a
continuous subcutaneous infusion of AngII (700 ng kg^−1 per minute) or
saline over a 14‐day period to induce hypertension.^[ [211]^76 ^] Mice
were monitored daily, and blood pressure was measured pre‐surgery and
on days 7 and 14 post‐surgery using the non‐invasive CODA blood
pressure monitoring system (ADInstruments, New Zealand). After 14 days
of Ang II infusion, mice were euthanized by CO₂ inhalation, and aortic
tissue samples were collected for subsequent histological and
proteomics analysis.
Tissue Preparation and Paraffin Embedding
The vessels were fixed in 4% paraformaldehyde for 48 h, followed by
running water rinsing for 6 h and three washes with distilled water,
each lasting 3 min. Then the samples were dehydrated through a graded
ethanol (Sinopharm, China) series and cleared using xylene (Sinopharm,
China) as follows: 70% ethanol for 1 h; 80% ethanol for 1 h; 90%
ethanol for 1 h; 95% ethanol for 2 h; anhydrous ethanol for 30 min
(repeated twice); xylene for 8 min (repeated twice). For paraffin
infiltration, the tissues were immersed in molten paraffin at 60 °C for
1 h, followed by two consecutive changes of fresh paraffin, each for 1
h. Finally, the paraffin‐infiltrated tissues were embedded in fresh
paraffin using embedding molds. The paraffin blocks were allowed to
cool and solidify, completing the preparation for subsequent sectioning
and analysis.
Masson's Trichrome Staining
Paraffin blocks were trimmed and sectioned at 5 µm thickness using a
rotary microtome, ensuring complete vascular cross‐sections in each
slide. Sections were floated on a 40 °C water bath, mounted onto clean
glass slides, and dried in a 60 °C oven for 2 h to ensure adherence.
Masson trichrome staining was performed using a commercial kit (Masson
Staining Kit, Yifan Biotechnology, China). Slides were deparaffinized
and rehydrated through the following sequence: xylene for 10 min
(repeated twice); anhydrous ethanol for 5 min; 95% ethanol for 5 min;
80% ethanol for 5 min; 70% ethanol for 5 min, followed by rinsed with
distilled water. For staining, sections were treated with Weigert's
iron hematoxylin for 5 min to stain nuclei, rinsed with distilled water
for 2 min, differentiated in 1% hydrochloric acid ethanol for a few
seconds, and blued in distilled water for 5 min. Cytoplasm was stained
using Ponceau‐acid fuchsin for 10 min, followed by a minute rinse in
distilled water. Collagen fibers were differentiated with
phosphomolybdic acid solution for 5 min, then stained with aniline blue
for 5 min. Finally, sections were immersed in 1% acetic acid for 1 min
to fix the staining. Dehydration and clearing were conducted through
the following sequence: 70% ethanol for 2 min; 80% ethanol for 2 min,
95% ethanol for 2 min; anhydrous ethanol for 2 min; xylene for 5 min
(repeated twice). Finally, the sections were mounted with neutral
balsam. The sections were imaged using a digital slide scanner (JFBIO,
China). All the images were analyzed and quantified using ImageJ
software and GraphPad Prism 10.
Proteins Extraction
The proteome samples were prepared using the EasyPept Micro Proteins
Pre‐Processing and Preparation Kit (Omicsolution, China) as previously
described.^[ [212]^77 ^] For the vessel chip model, after 48 h of
incubation, the rat tail collagen was enzymolyzed by collagenases
(Sigma, USA) diluted in TESCA buffer (Solarbio, China) at a
concentration of 0.5 mg mL^−1 until the cells were completely released
from the collagen. Cells were collected and washed with PBS to remove
the culture medium and then treated with the lysis buffer (reagent 0).
The samples were lysed on ice for 10 min and then heated at 95 °C for
10 min followed by centrifuging at 12 000 rpm for 20 min at 4 °C to
obtain the supernatants.
For vessel protein extraction, a lysis buffer composed of 1% sodium
dodecyl sulfate (Sinopharm, China), 8M urea (Sigma–Aldrich, USA), and
Halt Protease Inhibitor Cocktail EDTA‐free (ThermoFisher Scientific,
USA) was added, followed by homogenization using a mechanical grinder
(Jingxin, China). After grinding twice with 3 steel balls (−40 °C,
70Hz, on 120 s, off 180 s), the mixture was centrifuged at 12 000 rpm
for 20 min at 4 °C to collect the supernatant. For each sample, 30 µg
of protein was retrieved according to the quantification results using
the Pierce BCA protein assay kit (ThermoFisher Scientific, USA).
Acetone (Sinopharm, China) was added at a volume ratio of
acetone:supernatant = 5:1 to precipitate protein at −20 °C overnight.
Then, precipitation was washed with cold acetone 3 times and dried to
remove acetone completely. 20 µL of reagent 0 of the EasyPept kit was
added to each sample to dissolve protein, followed by centrifugation of
the solution at 12 000 rpm for 20 min at 4 °C to obtain the
supernatants.
Reagent A was added and heated at 95 °C for 5 min for protein
denaturation and cysteine reduction/alkylation. After the solution was
cooled, regent B containing trypsin was added to digest protein at
37 °C for 2 h and stopped by adding reagent C. The solution was
centrifugated at 20 000 g for 1 min at 4 °C to obtain the supernatants.
The supernatants were then desalinated through loading, washing, and
elution. Pierce quantitative colorimetric peptide assay (ThermoFisher
Scientific, USA) was used to quantify the peptides from mouse vessels,
while a NanoDrop Microvolume Spectrophotometer (ThermoFisher
Scientific, USA) was used to quantify the peptides from vessel chips.
Peptides were collected and dried under a vacuum.
LC‐MS/MS Analysis
Dried peptides were dissolved in water containing 0.1% formic acid to
reach a final concentration of 200 ng µL^−1. For analysis, 1 µL peptide
solution was analyzed by a nanoElute liquid chromatography system
(Bruker, Germany) coupled with a trapped ion‐mobility spectrometry
quadrupole time‐of‐flight mass spectrometer (timsTOF Pro2, Bruker,
Germany) using an AUR3‐15075 C18 column (150 mm × 75 µm, 1.7 µm, 120 Å
pore size, IonOpticks, Fitzroy, Australia) at a column temperature of
60 °C. The mobile phases consisted of water (with 0.1% formic acid) as
solvent A and acetonitrile (with 0.1% formic acid) as solvent B, at a
flow rate of 300 nL min^−1. The gradient elution profile was programmed
as follows: 0 min at 2% B; 0–4 min, 2% to 3% B; 4–5 min, holding at 3%
B; 5–46 min, 3% to 22% B; 46–52 min, 22% to 35% B; 52–57 min, 35% to
72% B; 57–60 min, holding at 72% B. The mass spectrometer operated in
diaPASEF mode for data‐independent acquisition (DIA) to acquire mass
spectrometry signal. Each analytical run was conducted over a total
duration of 60 min, with precursor isolation windows set to 32 × 26 Th
and the m/z scanning range from 400 to 1201 Da. During the MS/MS
scanning, collision energy (CE) increased linearly from 20 eV (1/K[0] =
0.6 Vs cm^−2) to 59 eV (1/K[0] = 1.6 Vs cm^−2).
Proteomics Data Analysis
Spectronaut 19 (Biognosys, Switzerland) was used to process and analyze
the raw data through directDIA method with default settings. All the
statistical analysis of the proteomic data was processed by R project
versions 4.3, MetaboAnalyst 6.0 ([213]https://www.metaboanalyst.ca/),
ggplot2, and GraphPad Prism 10. For each protein, if the number of
missing values exceeded 50% of the sample size, the protein was not
included in the statistical analysis. Missing values were imputed with
one‐fifth of the minimum identified value for the protein.
Significantly changed proteins were defined as those with a FC greater
than 1.5 or less than 2/3 as well as p < 0.05 (two‐tailed t‐test). The
interaction and relation networks of significantly changed proteins
were referred to the Kyoto Encyclopedia of Genes and Genomes (KEGG)
pathway database ([214]https://www.kegg.jp/) and Gene Ontology (GO)
knowledgebase ([215]https://geneontology.org/), with significance
determined using hypergeometric testing. The prediction of pathway
activation was performed using Gene Set Enrichment Analysis (GSEA)
([216]https://www.gsea‐msigdb.org/gsea/index.jsp). Ingenuity Pathway
Analysis (IPA, Qiagen, USA,
[217]https://digitalinsights.qiagen.com/IPA) was used to analyze and
generate figures through upstream regulator analysis as well as
diseases and functions analysis.
Statistical Analysis
All in vitro experiments were performed at least in triplicate.
Numerical data were presented as mean±standard deviation with a group
number n described in the figure caption. Statistically significant
differences between groups were assessed using a two‐tailed unpaired
Student's t‐test, with significance defined as p < 0.05. All
statistical analyses were carried out using GraphPad Prism 10.
Ethics Statement
All experimental procedures were approved by the Animal Care and Use
Committee of Shanghai Medical College, Fudan University
(#DSF‐2020‐014).
Data and Materials Availability
All proteome data have been deposited to the ProteomeXchange via the
iProX partner^[ [218]^78 , [219]^79 ^] repository with the dataset
identifiers PXD057724 and IPX0010186000.
Conflict of Interest
The authors declare no conflict of interest.
Author Contributions
Y.L. performed the experiments, analyzed the data, and wrote the first
draft of the manuscript. J.Z. assisted with data processing. L.Z. and
Z.W. assisted with protein extraction and immunofluorescence staining.
D.Z. assisted with LC‐MS/MS analysis. H.L., X.Z., K.M., and X.Y.
conducted the hypertensive mouse model and assisted with the related
analysis. D.Z. assisted with photolithography. L.Q. and L.L. designed
and supervised the study, provided funding, and finalized the
manuscript.
Supporting information
Supporting Information
[220]ADVS-12-2415024-s004.pdf^ (3.2MB, pdf)
Supporting Table 1‐6
[221]ADVS-12-2415024-s003.xlsx^ (3.5MB, xlsx)
Supplemental Video 1
[222]Download video file^ (4.5MB, avi)
Supplemental Video 2
[223]Download video file^ (4.5MB, avi)
Acknowledgements