You are on page 1of 39

Author’s Accepted Manuscript

Unique Mechanistic Insights into the Beneficial


Effects of Angiotensin-(1–7) on the Prevention of
Cardiac Fibrosis: A Metabolomic Analysis of
Primary Cardiac Fibroblasts

Yun-lin Chen, Jinqi Fan, Li Cao, Ting-li Han,


Mengying Zeng, Yanping Xu, Zhiyu Ling, Yuehui
Yin www.elsevier.com/locate/yexcr

PII: S0014-4827(19)30094-1
DOI: https://doi.org/10.1016/j.yexcr.2019.03.006
Reference: YEXCR11342
To appear in: Experimental Cell Research
Received date: 16 October 2018
Revised date: 28 February 2019
Accepted date: 3 March 2019
Cite this article as: Yun-lin Chen, Jinqi Fan, Li Cao, Ting-li Han, Mengying
Zeng, Yanping Xu, Zhiyu Ling and Yuehui Yin, Unique Mechanistic Insights
into the Beneficial Effects of Angiotensin-(1–7) on the Prevention of Cardiac
Fibrosis: A Metabolomic Analysis of Primary Cardiac Fibroblasts, Experimental
Cell Research, https://doi.org/10.1016/j.yexcr.2019.03.006
This is a PDF file of an unedited manuscript that has been accepted for
publication. As a service to our customers we are providing this early version of
the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting galley proof before it is published in its final citable form.
Please note that during the production process errors may be discovered which
could affect the content, and all legal disclaimers that apply to the journal pertain.
Unique Mechanistic Insights into the Beneficial Effects of Angiotensin-(1-7) on the Prevention

of Cardiac Fibrosis: A Metabolomic Analysis of Primary Cardiac Fibroblasts

1,1 1, 5 ,1 1 2, 3, 4 1
Yun-lin Chen, PhD, Jinqi Fan, PhD, Li Cao, PhD, Ting-li Han, PhD, Mengying Zeng, MD,

1 1 1, *
Yanping Xu, PhD, Zhiyu Ling PhD, Yuehui Yin, MD

1
Department of Cardiology, the 2nd Affiliated Hospital of Chongqing Medical University

2
Department of Obstetrics and Gynecology, the 1st Affiliated Hospital of Chongqing Medical University

3
Liggins Institute, University of Auckland, New Zealand

4
Mass Spectrometry Centre, China-Canada-New Zealand Joint Laboratory of Maternal and Foetal

Medicine, Chongqing Medical University

5
Departments of Biomedical Engineering and Pediatrics, Emory University, Atlanta, GA 30322, USA

*Address correspondence to Prof. Yuehui Yin, MD, Department of Cardiology, the 2nd Affiliated

Hospital of Chongqing Medical University, E-mail: yinyh63@163.com, Tel. & Fax: +86-023-63693

1 These authors contributed equally to this work.

1
Abstract

Background:

Cell metabolic pathways are highly conserved among species and change rapidly in response to drug

stimulation. Therefore, we explore the effects of angiotensin-(1-7) in a primary cell model of cardiac

fibrosis established in angiotensin II-stimulated cardiac fibroblasts via metabolomics analysis and

further clarify the potential protective mechanism of angiotensin-(1-7).

Methods and Results:

After exposing cardiac fibroblasts to angiotensin II and/or angiotensin-(1-7), 172 metabolites in these

cells were quantified and identified by gas chromatography-mass spectrometry. The data were

subsequently analyzed by orthogonal partial least squares discriminant analysis to shortlist

biochemically significant metabolites associated with the antifibrotic action of angiotensin-(1-7). Seven

significant metabolites were identified: 10,13-dimethyltetradecanoic acid, arachidonic acid, aspartic

acid, docosahexaenoic acid (DHA), glutathione, palmitelaidic acid, and pyroglutamic acid. By

metabolic network analysis, we found that these metabolites were involved in six metabolic pathways,

including arachidonic acid metabolism, leukotriene metabolism, and the γ-glutamyl cycle. Since these

metabolic pathways are related to calcium balance and oxidative stress, we further verified that

angiotensin-(1-7) suppressed the abnormal extracellular calcium influx and excessive accumulation of

intracellular reactive oxygen species (ROS) in angiotensin II-stimulated cardiac fibroblasts.

Furthermore, we found that angiotensin-(1-7) suppressed the abnormal calcium- and ROS-dependent

activation of calcium/calmodulin-dependent protein kinase II delta (CaMKIIδ), the increased

expression of CaMKIIδ-related proteins (NADPH oxidase 4 (Nox4), cellular communication network

factor 2 (CTGF), and p-ERK1/2), and excessive collagen deposition in vitro and in vivo.

2
Conclusions: Angiotensin-(1-7) can ameliorate the angiotensin II-stimulated metabolic perturbations

associated with cardiac fibroblast activation. These metabolic changes indicate that modulation of

calcium- and ROS-dependent activation of CaMKIIδ mediates the activity of angiotensin-(1-7) against

cardiac fibrosis. Moreover, pyroglutamic acid and arachidonic acid may be potential biomarkers for

monitoring the antifibrotic action of angiotensin-(1-7).

Keywords

Angiotensin-(1-7), Angiotensin II, Primary cardiac fibroblasts, Cell culture metabolomics,

2+
Ca /calmodulin-dependent protein kinase II

Introduction

Cardiac fibrosis is an important global health issue associated with nearly all forms of cardiac disease

[1] and is characterized by the uncontrolled activation of cardiac fibroblasts (CFs) (excessive

proliferation and differentiation), which leads to arrhythmogenesis, disruption of myocardial

excitation-contraction coupling, and impaired cardiac architecture [2]. The renin-angiotensin system

(RAS), in which angiotensin II (Ang II) is the predominant effector, can promote the development of

cardiac fibrosis [3]. As a newly identified component of the RAS, angiotensin-(1-7) (Ang-(1-7))

counteracts Ang II. Recent studies, including ours, show that treatment with Ang-(1-7) can attenuate

the cardiac fibrosis induced by Ang II in association with CF activation inhibition [4-8]. Traditional

molecular biology studies have provided useful information regarding the mechanisms of Ang-(1-7)

action. However, the comprehensiveness of investigations into the underlying mechanisms of Ang-(1-7)

through a systems biology approach remains limited.

Metabolomics, an important technique in systems biology, can provide an instantaneous and

comprehensive snapshot of ongoing metabolic changes within a living system [9]. Metabolic changes

3
are the most proximal reporters of alterations in the body in response to a disease process or drug

therapy [10]. Integrating metabolic alterations with prior biological knowledge enables us to rapidly

uncover the mechanisms associated with the pathogenesis of disease or the action of a drug [11]. In

past decades, metabolomic studies have contributed substantially to cardiovascular pathophysiology

and biomarker discovery. Most of these studies have been performed in in vivo systems, including

human patients and animal disease models [12, 13]. In vitro metabolomic approaches such as cell

culture metabolomics can offer some advantages over in vivo metabolomics, including stricter control

of external variables and greater reproducibility; thus, such approaches are attractive for use in drug

discovery and development [14]. Historically, most drug screening studies have utilized

tumor-originated cell lines or other immortalized mammalian cells. The implementation of primary cells

is more suitable for monitoring drug toxicity and exploring drug action mechanisms because primary

cells retain functions that are more physiologically and clinically relevant to those observed in vivo [15].

However, the application of primary cells in cardiovascular drug research has rarely been reported.

In this study, we aim to systematically explore the mechanism underlying the protective activity of

Ang-(1-7) against cardiac fibrosis. After CFs were treated with Ang II and/or Ang-(1-7), a metabolomic

analysis of the treated CFs and normal CFs was performed to elucidate the mechanisms underlying

the protective activity of angiotensin-(1-7) against cardiac fibrosis. Multivariate analyses, including

orthogonal projections to latent structures discriminant analysis (OPLS-DA) and shared and unique

structures (SUS) plot analysis, were used to identify biochemically significant metabolites linked to the

mechanisms underlying the antifibrotic action of Ang-(1-7). Finally, the biological events associated

with these metabolites were verified by traditional molecular biology experiments in vivo and in vitro.

4
Materials and Methods

This research protocol was approved by the Institutional Animal Care and Use Committee of

Chongqing Medical University and was conducted in compliance with the Guide for the Care and Use

of Laboratory Animals [16].

Animal models

A total of 18 male Sprague-Dawley rats weighing approximately 240 g were purchased from

Chongqing Medical University Laboratory Animal Center and fed a chow diet in a 12-hour light/12-hour

dark environment maintained at 25°C. These Sprague-Dawley rats were treated with Ang II (500

ng/kg/min, n = 6), Ang II plus Ang-(1-7) (500 ng/kg/min, n = 6), and physiological saline solution (the

normal group, n = 6) for 28 days using osmotic minipumps (model 2004; Alzet, Cupertino, USA) as

previously described [17].

Isolation and culture of neonatal rat CFs

CFs were isolated from neonatal Sprague-Dawley rats (1–3 days old) as previously described [5].

Briefly, rat hearts were rapidly excised and digested with 0.08% collagenase type II (Invitrogen,

California, USA) and 0.125% trypsin (Beyotime, Shanghai, China). After digestion, the cells were

cultured in high-glucose Dulbecco’s modified Eagle’s medium (Invitrogen, California, USA)

supplemented with 10% fetal bovine serum (Invitrogen, California, USA) plus 1% penicillin–

streptomycin (Beyotime, Shanghai, China) and incubated in a humidified incubator containing 5% CO2

at 37°C. The CFs at passage one or two were used for subsequent experiments.

Quantitative reverse transcription polymerase chain reaction (RT-qPCR)

Total RNA was isolated from the treated CFs using TRIzol reagent (TaKaRa, Dalian, China).

Approximately 1 µg of RNA was used for reverse transcription reactions with PrimeScript RT Master

5
Mix (TaKaRa, Dalian, China) using the standard protocol described in the manufacture user manual,

with a 15-min RT step. Subsequently, qPCR reactions were run on an Applied Biosystems 7500

Real-Time PCR System using TB Green Premix Ex Taq II (TaKaRa) and the probes for

rat calcium/calmodulin-dependent protein kinase II delta (CaMKIIδ), NADPH oxidase 4 (Nox4), cellular

communication network factor 2 (CTGF) and glyceraldehyde 3-phosphate dehydrogenase (GAPDH)

(Supplemental Table 1). The relative expression levels of these genes were assessed by the

delta/delta Ct method.

Masson’s trichrome staining

Rat hearts were fixed in formaldehyde, embedded in paraffin, and sectioned at 5-μm thickness for

Masson’s trichrome staining (Solarbio Life Science, Beijing, China). Fibrotic areas were measured and

quantified using Image-Pro Plus (version 6.0, Rockville, USA).

Immunohistochemistry

Paraffin-embedded rat ventricular tissues were sliced into 5-μm sections. After deparaffinization,

hydration and antigen retrieval, the sections were washed in PBS. Peroxidase activity was blocked

with 3% H2O2 for 10 minutes, and the sections were then incubated with primary antibodies against

α-SMA (1:400, Sigma-Aldrich, Missouri, USA) overnight at 4°C. Next, the sections were washed and

incubated with HRP-conjugated secondary antibodies for 1 hour at room temperature. Finally, the

sections were processed with DAB working solution and mounted. Staining area quantification was

performed using Image-Pro Plus (version 6.0, Rockville, USA).

Western blot analysis

Snap-frozen heart tissues and cultured CFs were lysed in RIPA buffer (Beyotime, Shanghai, China)

containing a protease and phosphatase inhibitor cocktail (MedChemExpress, New Jersey, USA).

6
Protein lysates were boiled at 100°C for 5 minutes, and the protein concentration was assessed with a

BCA protein assay. A total of 30 μg of each sample was separated by 10% SDS-PAGE, transferred

onto PVDF membranes (Roche, Basel, Switzerland), and blocked in PBS containing 5% BSA

(Beyotime, Shanghai, China) and 0.1% Tween 20 (Beyotime, Shanghai, China). The membranes were

incubated with primary antibodies against collagen I, collagen III, α-SMA, CTGF, CaMKIIδ,

p-Thr(287)-CaMKIIδ, ox-Met(281/282)-CaMKIIδ, Nox4, ERK1/2, p-ERK1/2, and GAPDH (1:1000)

overnight at 4°C (Abcam, Cambridge, UK; Sigma-Aldrich, Missouri, USA; GeneTex, California, USA).

Then, the membranes were incubated with HRP-conjugated secondary antibodies (1:8000, Beyotime,

Shanghai, China) for 1 hour at room temperature. Protein bands were detected using the

WesternBright ECL HRP substrate (Advansta, CA, USA) and a Bio-Rad ChemiDoc MP system

(Bio-Rad, California, USA). The band intensities were quantified by Image Lab software version 5.1

(Bio-Rad, California, USA).

Immunofluorescence staining

Frozen left ventricular tissue samples for immunofluorescence staining were prepared according to the

reference [18]. To prepare cell samples for immunofluorescence staining, the treated CFs were rinsed

with PBS, fixed with 4% paraformaldehyde in PBS for 30 minutes at room temperature, permeabilized

with PBS containing 0.125% Triton X-100, and blocked with PBS containing 10% goat serum. Next,

these samples were incubated with primary antibodies against α-SMA (1:400, Sigma-Aldrich, Missouri,

USA) and ox-Met(281/282)-CaMKIIδ(1:100, Gene Tex, California, USA) overnight at 4°C, followed by

Alexa Fluor 594-conjugated secondary antibodies (1:1000, Gene Tex, California, USA) for 1 hour at

room temperature. Nuclei were stained with DAPI (Beyotime, Shanghai, China). Image acquisition and

analysis were performed using a Leica TCS-SP5 laser scanning confocal microscope and LAS AF Lite

7
software (Hessen, Germany).

Cell proliferation assay

CFs were cultured in a 96-well plate (5000 cells per well) and treated with Ang II (10 μM,

MedChemExpress, New Jersey, USA) and/or Ang-(1-7) (10 μM, MedChemExpress, New Jersey, USA)

for 48 hours. To assess cell proliferation, 10 μl of Cell Counting Kit-8 (Dojindo, Tokyo, Japan) solution

was added to each well; the absorbance at 540 nm was measured with a microplate reader

(Bio-Tek, Vermont, USA) after the plate was incubated for 2 hours under the conditions described

above.

Intracellular calcium measurement

2+ 2+
Measurements of the intracellular Ca concentration ([Ca ]i) were performed as described previously

[19]. CFs were plated onto 35-mm glass-bottom dishes (NEST, Jiangsu, China) and grown to a

confluence of approximately 70%. The cells were incubated with Fluo-4 AM (10 μM, Invitrogen,

California, USA) at 37°C for 30 minutes in Hanks’ balanced salt solution (HBSS, 137.93 mM NaCl, 5.33

mM KCl, 4.17 mM NaHCO3, 0.441 mM KH2PO4, 0.338 mM Na2HPO4, and 5.56 mM D-glucose)

2+ 2+
without Ca . The CFs were then stimulated with Ang II or Ang-(1-7) in HBSS with or without Ca (1.26

mM CaCl2). Fluorescence images of a group of cells (≥ 10 cells) were acquired and analyzed with the

Leica TCS-SP5 laser scanning confocal microscope (excitation at 494 nm and emission at 516 nm)

2+
and LAS AF Lite software. Changes in [Ca ]i were reported as the Fluo-4 ratio F/F0, where F indicates

the fluorescence intensity at different time points and F0 is the baseline fluorescence intensity recorded

at the beginning of the experiment.

Reactive oxygen species (ROS) detection

-
To measure the intracellular H2O2 and O2 levels, the fluorescent dyes dichloro-dihydro-fluorescein

8
diacetate (DCFH-DA, 10 μM, Beyotime, Shanghai, China) and dihydroethidium (DHE, 10 μM,

Sigma-Aldrich, Missouri, USA), respectively, were used as described previously [20, 21]. After

treatment with Ang II and/or Ang-(1-7) for 15 minutes, cells in 6-well culture dishes were washed 3

times with HBSS. Subsequently, fluorescence was detected by an Olympus IX53 fluorescence

microscope (Tokyo, Japan). The mean fluorescence intensity in different groups was quantified using

NIH ImageJ software (Maryland, USA).

Metabolite extraction and derivatization and gas chromatography-mass spectrometry (GC-MS)

analysis

The metabolite extraction and derivatization protocols were modified from those of Smart et al. [22, 23].

After 48 hours of treatment with Ang II and/or Ang-(1-7), CFs were quenched by liquid nitrogen.

Metabolites were extracted from these cells using cold methanol/chloroform (9:1 v/v, Sigma-Aldrich,

Missouri, USA), and the internal standard, namely, 2,3,3,3-d4-alanine (20 µl for every 1.5 ml of

extraction solvent), was concurrently introduced into every sample. Next, the cells were removed with

cell scrapers, and the solvent extracts were transferred to Eppendorf tubes. After centrifugation at

17,000 g for 15 minutes at 4°C, the supernatant was collected from each tube, divided into two tubes,

and dried in a SpeedVac (Labconco Corp., Missouri, USA) at room temperature for 4−5 hours. Then,

one tube of each sample was used to measure the protein concentration by a BCA assay for biomass

normalization, while the supernatant in the other tube was chemically derivatized by the methyl

chloroformate (MCF) approach prior to GC-MS analysis. The MCF derivatives were analyzed in an

Agilent GC7890B system coupled to an MSD5977A mass selective detector (EI) set at 70 eV. The GC

column installed for metabolite analysis was a ZB-1701 GC capillary column (30 m × 250 µm id × 0.15

µm with a 5-m guard column, Phenomenex). The GC analysis parameters were implemented

9
according to the protocol of Smart et al. [24]. Chromatogram deconvolution was performed by the

Automated Mass Spectral Deconvolution and Identification System (AMDIS), and compounds were

identified by an in-house MCF-derivatized mass spectral library and the NIST mass spectral library.

The relative abundance levels of the identified metabolites were extracted by an in-house

XCMS-based R script.

Statistical and bioinformatic analyses

For the GC-MS data, the metabolite abundance levels were normalized to the concentration of the

internal standard and the total protein detected in each sample. Batch variation was removed by

median centering, and the metabolite data were log transformed, mean centered, and Pareto scaled.

Then, the data were analyzed by principal component analysis (PCA) and OPLS-DA in SIMCA 14.1

(Umetrics, UMEÅ, Sweden) [25]. Seven-fold cross-validation was employed as a measure of the

robustness of the multivariate models. Loading plots with jackknife confidence intervals, variable

importance in projection (VIP) scores, and S-plots from OPLS-DA were used to identify biochemically

interesting metabolites. The significant metabolites were selected based on the following criteria: VIP

scores > 1, |p (corr)| > 0.5, |p|> 0.1, and jackknife confidence intervals excluding zero [26, 27]. The

terms p (corr) and p indicate the correlation and covariance loading profiles, respectively. An SUS plot

was used to identify the significant metabolites associated with the antifibrotic action of Ang-(1-7) that

were shared and unique between the OPLS-DA models (normal vs Ang II; Ang II + Ang-(1-7) vs Ang II).

Then, a specific metabolic network based on the Kyoto Encyclopedia of Genes and Genomes (KEGG)

database was built by MetScape 2 to unravel the biological implications of these shared metabolites

[28].

The molecular biological data are expressed as the means ± SEMs. The assumptions of normality and

10
variance homogeneity were tested by the Shapiro-Wilk test (p > 0.1) and Levene’s test (p > 0.1),

respectively. Statistical significance across groups was determined by one-way or two-way ANOVA

followed by Tukey’s honest significant difference (HSD) post hoc test using R software; p-values < 0.05

were considered statistically significant.

Results

Ang-(1-7) inhibits the proliferation and differentiation of CFs induced by Ang II

To determine whether Ang-(1-7) attenuates the differentiation of CFs induced by Ang II, we tested the

effects of Ang II and Ang-(1-7) on the expression of α-SMA, collagen I, and collagen III by Western

blotting and/or immunofluorescence staining. As shown in Figure 1a and 1b, Ang II significantly

increased the expression levels of α-SMA, collagen I, and collagen III compared with those of the

normal cells, but pretreatment with Ang-(1-7) prevented these increases. We then assessed the

proliferation of CFs by the CCK-8 assay. The exposure of CFs to Ang II led to a 1.76-fold increase in

cell proliferation; this increase was suppressed by pretreatment with Ang-(1-7) (Figure 1c).

Ang-(1-7) ameliorates the perturbations in arachidonic acid (AA) and glutathione metabolism in

CFs stimulated by Ang II

A total of 275 metabolite peaks were detected in the primary cell cultures, 172 of which were identified

by the in-house MCF GC-MS mass spectral library. PCA was performed to obtain an overview of the

metabolic profiles. The PCA score plot in Figure 2 shows a clear separation between the control group

(a cardiac fibrosis cell model established in Ang II-stimulated CFs) and the two treatment groups

(normal CFs and Ang II-stimulated CFs treated with Ang-(1-7)). The principal components t1 and t2

2
explained 64.9% and 12.1% of the total variance, respectively (Q values: t1 = 0.61, t2 = 0.7),

11
suggesting that the cellular metabolic phenotypes were significantly different between the control

group (Ang II group) and the treatment groups (normal and Ang II + Ang-(1-7) groups).

OPLS-DA was then used to screen statistically significant metabolites that contributed to the

classification of each of the treatment groups and the control group. As shown in Figure 3a and Figure

4a, both OPLS-DA models (model 1, normal vs Ang II; model 2, Ang II + Ang-(1-7) vs Ang II) yielded

2 2
good classification under different conditions (R = 0.87, Q = 0.73, 1 predictive component + 2

2 2
orthogonal components; R = 0.98, Q = 0.75, 1 predictive component + 5 orthogonal components).

According to the cross-validated score plots, no sample was misclassified, which further indicated the

good performance of these models for class separation.

In the OPLS-DA models, both loading plots with confidence intervals and VIP scores were used to

assess variable significance (VIP scores > 1 and jackknife confidence intervals excluding zero) [25].

The models identified 30 statistically significant metabolites (Figure 3b) for the normal group and 27 for

the Ang II + Ang-(1-7) group (Figure 4b). An S plot combining the covariance (p (corr), the

concentration of metabolites) and correlation (p, the contribution to class separation) from the

OPLS-DA model was generated to shortlist biochemically significant metabolites (statistically

significant metabolites with |p (corr)| > 0.5 and |p| > 0.1) [26, 27]. A total of 17 and 7 metabolites met

the selection criteria for model 1 and model 2, respectively (Figures 3c and 4c). Compared to the Ang II

group, the normal and Ang II + Ang-(1-7) groups exhibited decreased concentrations of all

biochemically interesting metabolites in both models.

Furthermore, an SUS plot was used to compare the outcomes of the two established OPLS-DA models

(normal vs Ang II and Ang II + Ang-(1-7) vs Ang II). Seven biochemically interesting metabolites near

the diagonal line ((-1, -1) to (1, 1)) were equally affected in both models (Figure 5). In other words, the

12
changes in these seven metabolites represent the rescue effect of Ang-(1-7) on Ang II-stimulated CFs.

Finally, the metabolomic data were linked to prior knowledge of metabolic pathways (KEGG analysis)

and visualized by MetScape 2, as shown in Figure 6. Among the identified metabolites, 56 were used

to reconstruct a metabolic network associated with the abnormal activation of CFs. The seven

biochemically interesting metabolites were covered by six major subnetworks, including AA

metabolism, the γ-glutamyl cycle, leukotriene metabolism, omega-6 fatty acid metabolism, and

prostaglandin biosynthesis.

Ang-(1-7) suppresses the calcium overload triggered by Ang II in CFs

Next, we explored whether the antifibrotic effect of Ang-(1-7) is associated with the reestablishment of

calcium hemostasis in CFs. Figures 7a and 7b show that Ang II triggered a lower and transient peak of

2+ 2+ 2+
[Ca ]i in the absence of extracellular Ca but a higher and sustained peak of [Ca ]i in the presence of

2+ 2+
extracellular Ca . Notably, the Ang II-induced increase in [Ca ]i was suppressed by Ang-(1-7)

2+
pretreatment in the presence of extracellular Ca but was unaffected in the absence of extracellular

2+
Ca (Figure 7c and 7d). These results suggested that the extracellular calcium influx contributes to the

2+
increase in [Ca ]i in Ang II-stimulated CFs. This phenomenon can be abolished in the presence of

Ang-(1-7) (Figure 7e).

Furthermore, we investigated the effect of Ang-(1-7) on CaMKIIδ and the associated downstream

signaling pathways by RT-qPCR and Western blot analyses. Supplemental Figure 1 shows that the

mRNA expression levels of CaMKIIδ and CTGF were significantly increased in the Ang II-stimulated

CFs. This effect can be significantly suppressed by Ang-(1-7) treatment. Correspondingly, the protein

expression levels of p-Thr(287)-CaMKIIδ and total CaMKIIδ were significantly higher in the Ang II

group than in the normal group. This increased expression was restored to normal levels by

13
pretreatment with Ang-(1-7) (Figure 5f). The ratio of p-Thr287 expression to total CaMKIIδ expression

was higher in the Ang II group than in the normal group, but pretreatment with Ang-(1-7) restored the

increased ratio to the normal level. In addition, treatment with Ang II significantly increased the protein

levels of p-ERK1/2 and CTGF over the levels observed in the normal group, but pretreatment with

Ang-(1-7) significantly inhibited the Ang II-induced phosphorylation of ERK1/2 and the increased

expression of CTGF (Figure 7f). Ang-(1-7) pretreatment decreased the ratio of p-ERK1/2 to ERK1/2

compared to that observed in the Ang II group (Figure 7f).

Ang-(1-7) ameliorates the oxidative stress increase induced by Ang II in CFs

To explore whether Ang-(1-7) suppresses the activation of CFs induced by Ang II via modulation of

intracellular oxidative stress, the effects of Ang-(1-7) on the level of ROS and the expression of Nox4

and ox-CaMKIIδ were studied in Ang II-stimulated CFs. CFs were stimulated with Ang II, which

resulted in increased production of intercellular ROS, including hydrogen peroxide (Figure 8a) and the

superoxide anion (Figure 8b). These increases were alleviated by pretreatment with Ang-(1-7) (Figure

8a and 8b). Furthermore, we observed a significant increase in the expression levels of Nox4 (mRNA

and protein levels) and ox-CaMKIIδ (protein level) in CFs treated with Ang II, but this increase was

abolished by pretreatment with Ang-(1-7) (Figure 8c and Supplemental Figure 1).

Ang-(1-7) infusion protects against Ang II-induced cardiac fibrosis by suppressing abnormal

CaMKIIδ activation of CFs in rat models

To further validate the effect of Ang-(1-7) on cardiac fibrosis induced by Ang II in vivo, Ang II-infused

rats were treated with Ang-(1-7). Masson’s trichrome staining of paraffin sections showed that Ang II

infusion promoted extensive fibrosis in the left ventricular sections of the rats (Normal: 3.28±0.81% vs

Ang II 14.75±5.17%, p < 0.05, Figure 9a). This extensive cardiac fibrosis can be reversed by treatment

14
with Ang-(1-7) (Ang II + Ang-(1-7): 7.48±0.59% vs Ang II: 14.75±5.17%, p < 0.05, Figure 9a).

Consistently, immunohistochemistry staining revealed a significant increase in the expression of

α-SMA after Ang II infusion, whereas the Ang-(1-7) treatment abolished the increase in the expression

of α-SMA (Figure 9b). In addition, Western blot analysis revealed elevated protein levels of CTGF and

p-ERK1/2 in Ang II-infused rats but not in Ang II-infused rats treated with Ang-(1-7) (Figure 9c). No

change was observed in the protein expression of ERK1/2 after the Ang II and Ang-(1-7) infusion

(Figure 9c).

Furthermore, we explored whether Ang-(1-7) ameliorates Ang II-induced cardiac fibrosis by inhibiting

CaMKIIδ overactivation of CFs in vivo. As shown in Figure 10a, immunofluorescence colocalization of

ox-CaMKIIδ (a marker of CaMKII oxidative activation) and α-SMA (a marker of CF activation) was

observed in ventricular sections of Ang II-infused rats, implying that CF activation is likely accompanied

by oxidative activation of CaMKIIδ. Regrettably, we were unable to show the colocalization of

p-Thr(287)-CaMKIIδ (a marker of CaMKIIδ calcium-induced activation) and α-SMA in Ang II-infused

hearts due to the incompatibility of the antibody with immunofluorescence staining. Western blot

analyses showed that Ang II infusion significantly increased the expression levels of CaMKIIδ,

p-CaMKIIδ, ox-CaMKIIδ and Nox4 in rat ventricles, whereas these protein expression levels were

significantly suppressed by Ang-(1-7) treatment (Figure 10b, 10c, 10d, and 10e).

Discussion

The results of our previous studies suggest that Ang-(1-7) attenuates cardiac fibrotic remodeling by

reducing the Ang II-induced proliferation and differentiation of CFs [4-6]. However, the antifibrotic

mechanism of Ang-(1-7) remains unclear. This study was the first to explore the mechanism underlying

Ang-(1-7) activity in Ang II-stimulated CFs via an untargeted metabolomic analysis. We successfully

15
determined that Ang-(1-7) suppressed the metabolic perturbations induced by Ang II in CFs, including

those affecting AA metabolism and the γ-glutamyl cycle. Considering that metabolic changes in AA

metabolism and the γ-glutamyl cycle are associated with calcium imbalance and oxidative stress,

respectively, we further verified that Ang-(1-7) inhibited Ang II-induced CF activation by suppressing

calcium overload, excessive oxidative stress, and the activity of CaMKIIδ-related signaling pathways in

cell and animal models (Figure 11).

The change in the AA level in CFs leads us to postulate that Ang-(1-7) exerts its antifibrotic effect by

suppressing the calcium overload induced by Ang II. The results of our metabolomic analysis revealed

that Ang II enhanced AA production in CFs, and subsequent metabolic network analysis showed that

AA was involved in multiple metabolic pathways associated with the intracellular calcium balance, such

as AA metabolism, leukotriene metabolism, and prostaglandin biosynthesis [29]. Previous studies

reveal that a calcium imbalance can result in abnormal activation of CFs [19, 30-32]. Extensive

2+
evidence demonstrates that AA can increase [Ca ]I via transient receptor potential channels, which

2+
are the main membrane channels mediating Ca influx into CFs [33-37]. Consistent with the results of

previous studies, our study also showed that Ang II increased the endogenous AA level and promoted

2+
the influx of extracellular Ca into CFs [38, 39]. More importantly, Ang-(1-7) not only reduced the

endogenous AA concentration but also profoundly decreased the excessive influx of calcium into Ang

II-stimulated CFs. These novel findings provide direct evidence that Ang-(1-7) can reverse the

deleterious effect of Ang II on CFs by maintaining calcium homeostasis and that AA has potential as a

biomarker for monitoring the antifibrotic effect of Ang-(1-7).

The alterations in the γ-glutamyl cycle reflect the antioxidant role of Ang-(1-7) against Ang II-triggered

ROS production in CFs. Our metabolic network analysis highlighted the upregulation of the γ-glutamyl

16
cycle in Ang II-stimulated CFs. Dysregulation of this cycle has been reported to be related to

uncontrolled oxidized stress during cardiac fibrotic remodeling [40, 41]. The γ-glutamyl cycle governs

the biosynthesis of glutathione; this process involves the ligation of glutamate with two amino acids

(cysteine and glycine) to form glutathione, the catabolism of glutathione to pyroglutamic acid (also

called 5-oxoproline) and the subsequent degradation of pyroglutamic acid into glutamate [42]. Here,

we reported that glutathione and pyroglutamic acid levels were positively correlated with the ROS level

in Ang II-stimulated CFs. Similarly, Atze et al. demonstrated that accumulation of pyroglutamic acid in

cardiomyocytes elevated the ROS level in mice with heart failure [43]. Intriguingly, Ang-(1-7) can

counteract the accumulation of pyroglutamic acid in CFs treated with Ang II. Furthermore, in our

experiments in vivo and in vitro, the expression of Nox4, which is a major source of ROS in CFs, was

increased after Ang II treatment. This result is consistent with the findings reported by other studies

showing that Nox4-mediated increases in H2O2 levels promote CF differentiation and proliferation and

that Nox4 upregulation in the myocardium causes cardiac fibrosis in rats [17, 44]. Interestingly, our

results showed that treatment with Ang-(1-7) inhibited Nox4 overexpression and excessive generation

of H2O2 in the CFs incubated with Ang II. Ang-(1-7) also markedly reduced excessive

collagen deposition and increased the protein expression levels of α-SMA and Nox4 in Ang II-infused

rats. These results are in line with those reported in a recent animal study published by Wang et al. [45].

Therefore, our findings suggest that Ang-(1-7) can ameliorate cardiac fibrosis by eliminating

Nox4-mediated ROS production in CFs and that this antioxidant outcome can be reflected by the level

of pyroglutamic acid.

CaMKIIδ acts as a central enzymatic node connecting the upstream calcium overload and ROS

perturbations with downstream profibrotic signaling pathways in the heart [46]. Calcium overload and

17
oxidative stress can promote posttranslational activation of CaMKIIδ through autophosphorylation of

threonine 287 (p-T287) and Nox4-mediated oxidation of methionine 281/282 (ox-M281/282),

respectively [47-51]. Indeed, our cell and animal results revealed that Ang II-induced excessive CF

activation was accompanied by overactivation of CaMKIIδ in response to the upstream calcium

overload and excessive ROS production. On the other hand, previous studies revealed that CTGF-

and ERK1/2-dependent signaling pathways are downstream cascades in the CaMKIIδ-induced

activation of CFs [52-54]. Consistently, we also showed that the downstream molecules of CaMKIIδ,

such as CTGF and p-ERK1/2, were upregulated after Ang II treatment. Most profoundly, pretreatment

with Ang-(1-7) can abolish both the abnormal activation of CaMKIIδ and the upregulation of its

downstream cascade observed in the Ang II-induced cardiac fibrosis. The effective inhibition of

CaMKIIδ by Ang-(1-7) may be the key step for protection against cardiac fibrosis triggered by Ang II.

This study has several limitations. Our study focused only on the downstream signaling pathways in

response to the metabolic changes in AA metabolism and the γ-glutamyl cycle in CFs treated with Ang

II and Ang-(1-7), and the upstream regulating mechanisms remain to be investigated. Furthermore,

multiple inbred rodent strains should be implicated to validate our hypothesis in vivo because a recent

study suggested that the heterogeneous nature of CFs is genetically diverse in rodent strains [55].

These limitations should be addressed in future studies.

Conclusions

In this study, the results of our mass spectrometry-based metabolomic analysis support the hypothesis

that Ang-(1-7) exhibits protective activity against cardiac fibrosis via the modulation of AA metabolism

and the γ-glutamyl cycle. These metabolic reprogramming observations guided us to unravel the

18
protective mechanism of Ang-(1-7), which acts by restoring calcium homeostasis, protecting against

ROS damage, and suppressing downstream CaMKIIδ-related signaling pathways in Ang II-stimulated

CFs. This enhanced mechanistic understanding of Ang-(1-7) activity could further promote the

development of Ang-(1-7) as a potential therapeutic intervention for cardiac fibrosis. Moreover,

pyroglutamic acid and AA are the biomarkers most likely to be useful for monitoring the protective

activity of Ang-(1-7) against cardiac fibrosis in the clinical setting.

Acknowledgments

Chen YL, Cao L, Ling ZY and Zeng MY performed the experiments. Chen YL and Han TL analyzed the

data. Chen YL, Fan JQ, and Yin YH designed the experiments. Chen YL, Fan JQ, and Han TL wrote

the manuscript, and all authors contributed to the preparation of the manuscript.

We acknowledge Yang Y for assisting with GC-MS sample preparation.

Funding

This study was supported by grants from the National Natural Science Foundation of China (grant nos.

81570302 and 81500250).

Disclosures

None.

References

[1] A. Leask, Getting to the heart of the matter: new insights into cardiac fibrosis, Circulation

research 116 (2015) 1269-1276.

19
[2] J.G. Travers, F.A. Kamal, J. Robbins, K.E. Yutzey, B.C. Blaxall, Cardiac Fibrosis: The

Fibroblast Awakens, Circ Res 118 (2016) 1021-1040.

[3] T.M. Hale, Persistent phenotypic shift in cardiac fibroblasts: impact of transient renin

angiotensin system inhibition, J Mol Cell Cardiol 93 (2016) 125-132.

[4] J. Fan, L. Zou, K. Cui, K. Woo, H. Du, S. Chen, Z. Ling, Q. Zhang, B. Zhang, X. Lan, L. Su, B.

Zrenner, Y. Yin, Atrial overexpression of angiotensin-converting enzyme 2 improves the

canine rapid atrial pacing-induced structural and electrical remodeling. Fan, ACE2 improves

atrial substrate remodeling, Basic research in cardiology 110 (2015) 45.

[5] X. Tao, J. Fan, G. Kao, X. Zhang, L. Su, Y. Yin, B. Zrenner, Angiotensin-(1-7) attenuates

angiotensin II-induced signalling associated with activation of a tyrosine phosphatase in

Sprague-Dawley rats cardiac fibroblasts, Biology of the cell 106 (2014) 182-192.

[6] L.P. Wang, S.J. Fan, S.M. Li, X.J. Wang, J.L. Gao, X.H. Yang, Protective role of

ACE2-Ang-(1-7)-Mas in myocardial fibrosis by downregulating KCa3.1 channel via ERK1/2

pathway, Pflugers Archiv : European journal of physiology 468 (2016) 2041-2051.

[7] J.L. Grobe, A.P. Mecca, M. Lingis, V. Shenoy, T.A. Bolton, J.M. Machado, R.C. Speth, M.K.

Raizada, M.J. Katovich, Prevention of angiotensin II-induced cardiac remodeling by

angiotensin-(1-7), Am J Physiol Heart Circ Physiol 292 (2007) H736-742.

[8] M. Iwata, R.T. Cowling, D. Gurantz, C. Moore, S. Zhang, J.X. Yuan, B.H. Greenberg,

Angiotensin-(1-7) binds to specific receptors on cardiac fibroblasts to initiate antifibrotic and

antitrophic effects, Am J Physiol Heart Circ Physiol 289 (2005) H2356-2363.

[9] R.W. McGarrah, S.B. Crown, G.F. Zhang, S.H. Shah, C.B. Newgard, Cardiovascular

Metabolomics, Circulation research 122 (2018) 1238-1258.

20
[10] G.D. Lewis, A. Asnani, R.E. Gerszten, Application of metabolomics to cardiovascular

biomarker and pathway discovery, J Am Coll Cardiol 52 (2008) 117-123.

[11] M. Cuperlovic-Culf, D.A. Barnett, A.S. Culf, I. Chute, Cell culture metabolomics: applications

and future directions, Drug Discov Today 15 (2010) 610-621.

[12] J.R. Ussher, S. Elmariah, R.E. Gerszten, J.R. Dyck, The Emerging Role of Metabolomics in

the Diagnosis and Prognosis of Cardiovascular Disease, J Am Coll Cardiol 68 (2016)

2850-2870.

[13] M. Ruiz-Canela, A. Hruby, C.B. Clish, L. Liang, M.A. Martinez-Gonzalez, F.B. Hu,

Comprehensive Metabolomic Profiling and Incident Cardiovascular Disease: A Systematic

Review, J Am Heart Assoc 6 (2017).

[14] S. Hayton, G.L. Maker, I. Mullaney, R.D. Trengove, Experimental design and reporting

standards for metabolomics studies of mammalian cell lines, Cell Mol Life Sci 74 (2017)

4421-4441.

[15] R. Eglen, T. Reisine, Primary cells and stem cells in drug discovery: emerging tools for

high-throughput screening, Assay Drug Dev Technol 9 (2011) 108-124.

[16] U. Albus, Guide for the Care and Use of Laboratory Animals (8th edn), Laboratory Animals 46

(2012) 267-268.

[17] Q.D. Zhao, S. Viswanadhapalli, P. Williams, Q. Shi, C. Tan, X. Yi, B. Bhandari, H.E. Abboud,

NADPH oxidase 4 induces cardiac fibrosis and hypertrophy through activating Akt/mTOR and

NFkappaB signaling pathways, Circulation 131 (2015) 643-655.

[18] X. Tian, T. Hu, H. Zhang, L. He, X. Huang, Q. Liu, W. Yu, L. He, Z. Yang, Y. Yan, X. Yang, T.P.

Zhong, W.T. Pu, B. Zhou, Vessel formation. De novo formation of a distinct coronary vascular

21
population in neonatal heart, Science 345 (2014) 90-94.

[19] Y. Yu, S. Chen, C. Xiao, Y. Jia, J. Guo, J. Jiang, P. Liu, TRPM7 is involved in angiotensin II

induced cardiac fibrosis development by mediating calcium and magnesium influx, Cell

calcium 55 (2014) 252-260.

[20] S.K. Kakarla, J.C. Fannin, S. Keshavarzian, A. Katta, S. Paturi, S.K. Nalabotu, M. Wu, K.M.

Rice, K. Manzoor, E.M. Walker, Jr., E.R. Blough, Chronic acetaminophen attenuates

age-associated increases in cardiac ROS and apoptosis in the Fischer Brown Norway rat,

Basic research in cardiology 105 (2010) 535-544.

[21] W.O. Carter, P.K. Narayanan, J.P. Robinson, Intracellular hydrogen peroxide and superoxide

anion detection in endothelial cells, J Leukoc Biol 55 (1994) 253-258.

[22] Z. Leon, J.C. Garcia-Canaveras, M.T. Donato, A. Lahoz, Mammalian cell metabolomics:

experimental design and sample preparation, Electrophoresis 34 (2013) 2762-2775.

[23] M.A. Lorenz, C.F. Burant, R.T. Kennedy, Reducing time and increasing sensitivity in sample

preparation for adherent mammalian cell metabolomics, Anal Chem 83 (2011) 3406-3414.

[24] K.F. Smart, R.B. Aggio, J.R. Van Houtte, S.G. Villas-Boas, Analytical platform for metabolome

analysis of microbial cells using methyl chloroformate derivatization followed by gas

chromatography-mass spectrometry, Nat Protoc 5 (2010) 1709-1729.

[25] A. Sussulini, Metabolomics: from fundamentals to clinical applications, Springer, 2017.

[26] S. Wiklund, E. Johansson, L. Sjöström, E.J. Mellerowicz, U. Edlund, J.P. Shockcor, J. Gottfries,

T. Moritz, J. Trygg, Visualization of GC/TOF-MS-based metabolomics data for identification of

biochemically interesting compounds using OPLS class models, Analytical chemistry 80 (2008)

115-122.

22
[27] Y. Ni, M. Su, J. Lin, X. Wang, Y. Qiu, A. Zhao, T. Chen, W. Jia, Metabolic profiling reveals

disorder of amino acid metabolism in four brain regions from a rat model of chronic

unpredictable mild stress, FEBS Lett 582 (2008) 2627-2636.

[28] A. Karnovsky, T. Weymouth, T. Hull, V.G. Tarcea, G. Scardoni, C. Laudanna, M.A. Sartor, K.A.

Stringer, H.V. Jagadish, C. Burant, B. Athey, G.S. Omenn, Metscape 2 bioinformatics tool for

the analysis and visualization of metabolomics and gene expression data, Bioinformatics 28

(2012) 373-380.

[29] C.M. Jenkins, A. Cedars, R.W. Gross, Eicosanoid signalling pathways in the heart, Cardiovasc

Res 82 (2009) 240-249.

[30] E.R. Olson, P.E. Shamhart, J.E. Naugle, J.G. Meszaros, Angiotensin II-induced extracellular

signal-regulated kinase 1/2 activation is mediated by protein kinase Cdelta and intracellular

calcium in adult rat cardiac fibroblasts, Hypertension 51 (2008) 704-711.

[31] K. Ikeda, T. Nakajima, Y. Yamamoto, N. Takano, T. Tanaka, H. Kikuchi, G. Oguri, T. Morita, F.

Nakamura, I. Komuro, Roles of transient receptor potential canonical (TRPC) channels and

reverse-mode Na+/Ca2+ exchanger on cell proliferation in human cardiac fibroblasts: effects

of transforming growth factor beta1, Cell calcium 54 (2013) 213-225.

[32] Y. Liu, H. Qi, M. E, P. Shi, Q. Zhang, S. Li, Y. Wang, Y. Cao, Y. Chen, L. Ba, J. Gao, W. Huang,

H. Sun, Transient receptor potential vanilloid-3 (TRPV3) activation plays a central role in

cardiac fibrosis induced by pressure overload in rats via TGF-beta1 pathway,

Naunyn-Schmiedeberg's archives of pharmacology 391 (2018) 131-143.

[33] H.Z. Hu, R. Xiao, C. Wang, N. Gao, C.K. Colton, J.D. Wood, M.X. Zhu, Potentiation of TRPV3

channel function by unsaturated fatty acids, Journal of cellular physiology 208 (2006) 201-212.

23
[34] J. Vriens, G. Owsianik, B. Fisslthaler, M. Suzuki, A. Janssens, T. Voets, C. Morisseau, B.D.

Hammock, I. Fleming, R. Busse, B. Nilius, Modulation of the Ca2 permeable cation channel

TRPV4 by cytochrome P450 epoxygenases in vascular endothelium, Circulation research 97

(2005) 908-915.

[35] H. Watanabe, J. Vriens, J. Prenen, G. Droogmans, T. Voets, B. Nilius, Anandamide and

arachidonic acid use epoxyeicosatrienoic acids to activate TRPV4 channels, Nature 424 (2003)

434.

[36] D.J. Beech, Y.M. Bahnasi, A.M. Dedman, E. Al-Shawaf, TRPC channel lipid specificity and

mechanisms of lipid regulation, Cell calcium 45 (2009) 583-588.

[37] L. Yue, J. Xie, S. Nattel, Molecular determinants of cardiac fibroblast electrical function and

therapeutic implications for atrial fibrillation, Cardiovasc Res 89 (2011) 744-753.

[38] A.M. Gallagher, T.D. Bahnson, H. Yu, N.N. Kim, M.P. Printz, Species variability in angiotensin

receptor expression by cultured cardiac fibroblasts and the infarcted heart, The American

journal of physiology 274 (1998) H801-809.

[39] A.M. Gallagher, H. Yu, M.P. Printz, Bradykinin-induced reductions in collagen gene expression

involve prostacyclin, Hypertension 32 (1998) 84-88.

[40] Y. Watanabe, K. Watanabe, T. Kobayashi, Y. Saito, D. Fujioka, T. Nakamura, J.E. Obata, K.

Kawabata, H. Mishina, K. Kugiyama, Chronic depletion of glutathione exacerbates ventricular

remodelling and dysfunction in the pressure-overloaded heart, Cardiovasc Res 97 (2013)

282-292.

[41] M.A. Forgione, A. Cap, R. Liao, N.I. Moldovan, R.T. Eberhardt, C.C. Lim, J. Jones, P.J.

Goldschmidt-Clermont, J. Loscalzo, Heterozygous cellular glutathione peroxidase deficiency

24
in the mouse: abnormalities in vascular and cardiac function and structure, Circulation 106

(2002) 1154-1158.

[42] A.K. Bachhawat, S. Yadav, The glutathione cycle: Glutathione metabolism beyond the

gamma-glutamyl cycle, IUBMB Life 70 (2018) 585-592.

[43] A. van der Pol, A. Gil, H.H.W. Sillje, J. Tromp, E.S. Ovchinnikova, I. Vreeswijk-Baudoin, M.

Hoes, I.J. Domian, B. van de Sluis, J.M. van Deursen, A.A. Voors, D.J. van Veldhuisen, W.H.

van Gilst, E. Berezikov, P. van der Harst, R.A. de Boer, R. Bischoff, P. van der Meer,

Accumulation of 5-oxoproline in myocardial dysfunction and the protective effects of OPLAH,

Sci Transl Med 9 (2017).

[44] J.T. Colston, S.D. de la Rosa, J.R. Strader, M.A. Anderson, G.L. Freeman, H2O2 activates

Nox4 through PLA2-dependent arachidonic acid production in adult cardiac fibroblasts, FEBS

Lett 579 (2005) 2533-2540.

[45] J. Wang, W. He, L. Guo, Y. Zhang, H. Li, S. Han, D. Shen, The ACE2-Ang (1-7)-Mas receptor

axis attenuates cardiac remodeling and fibrosis in post-myocardial infarction, Mol Med Rep 16

(2017) 1973-1981.

[46] M.M. Kreusser, L.H. Lehmann, S. Keranov, M.O. Hoting, U. Oehl, M. Kohlhaas, J.C. Reil, K.

Neumann, M.D. Schneider, J.A. Hill, D. Dobrev, C. Maack, L.S. Maier, H.J. Grone, H.A. Katus,

E.N. Olson, J. Backs, Cardiac CaM Kinase II genes delta and gamma contribute to adverse

remodeling but redundantly inhibit calcineurin-induced myocardial hypertrophy, Circulation

130 (2014) 1262-1273.

[47] P. De Koninck, H. Schulman, Sensitivity of CaM kinase II to the frequency of Ca2+ oscillations,

Science 279 (1998) 227-230.

25
[48] G. Dupont, G. Houart, P. De Koninck, Sensitivity of CaM kinase II to the frequency of Ca2+

oscillations: a simple model, Cell calcium 34 (2003) 485-497.

[49] J.R. Erickson, M.L. Joiner, X. Guan, W. Kutschke, J. Yang, C.V. Oddis, R.K. Bartlett, J.S.

Lowe, S.E. O'Donnell, N. Aykin-Burns, M.C. Zimmerman, K. Zimmerman, A.J. Ham, R.M.

Weiss, D.R. Spitz, M.A. Shea, R.J. Colbran, P.J. Mohler, M.E. Anderson, A dynamic pathway

for calcium-independent activation of CaMKII by methionine oxidation, Cell 133 (2008)

462-474.

[50] B.J. He, M.L. Joiner, M.V. Singh, E.D. Luczak, P.D. Swaminathan, O.M. Koval, W. Kutschke,

C. Allamargot, J. Yang, X. Guan, K. Zimmerman, I.M. Grumbach, R.M. Weiss, D.R. Spitz, C.D.

Sigmund, W.M. Blankesteijn, S. Heymans, P.J. Mohler, M.E. Anderson, Oxidation of CaMKII

determines the cardiotoxic effects of aldosterone, Nat Med 17 (2011) 1610-1618.

[51] J. Palomeque, O.V. Rueda, L. Sapia, C.A. Valverde, M. Salas, M.V. Petroff, A. Mattiazzi,

Angiotensin II-induced oxidative stress resets the Ca2+ dependence of Ca2+-calmodulin

protein kinase II and promotes a death pathway conserved across different species,

Circulation research 105 (2009) 1204-1212.

[52] A.G. Rokita, M.E. Anderson, New therapeutic targets in cardiology: arrhythmias and

Ca2+/calmodulin-dependent kinase II (CaMKII), Circulation 126 (2012) 2125-2139.

[53] W. Zhang, D.Q. Chen, F. Qi, J. Wang, W.Y. Xiao, W.Z. Zhu, Inhibition of

calcium-calmodulin-dependent kinase II suppresses cardiac fibroblast proliferation and

extracellular matrix secretion, Journal of cardiovascular pharmacology 55 (2010) 96-105.

[54] T.P. Martin, A. Lawan, E. Robinson, D.J. Grieve, R. Plevin, A. Paul, S. Currie, Adult cardiac

fibroblast proliferation is modulated by calcium/calmodulin-dependent protein kinase II in

26
normal and hypertrophied hearts, Pflugers Archiv : European journal of physiology 466 (2014)

319-330.

[55] S. Park, S. Ranjbarvaziri, F.D. Lay, P. Zhao, M.J. Miller, J.S. Dhaliwal, A. Huertas-Vazquez, X.

Wu, R. Qiao, J.M. Soffer, C. Rau, Y. Wang, H.K.A. Mikkola, A.J. Lusis, R. Ardehali, Genetic

Regulation of Fibroblast Activation and Proliferation in Cardiac Fibrosis, Circulation 138 (2018)

1224-1235.

27
Data

Figure 1. Ang-(1-7) inhibited the cardiac fibroblast differentiation and proliferation induced by Ang II. Cardiac
fibroblasts were cultured in microplates and incubated with Ang II (10 μM) and/or Ang-(1-7) (10 μM) for 48 hours. a)
Immunofluorescence images indicating that Ang-(1-7) inhibited the Ang II-induced differentiation of cardiac fibroblasts,
as evidenced by the reduction in α-SMA (red) expression. Nuclei were stained with DAPI (blue). The scale bars
represent 50 μm. b) Western blot analysis of α-SMA, collagen I, and collagen III and quantification of the protein bands.
c). Assessment of cardiac fibroblast proliferation with a CCK-8 assay. These experiments were repeated at least 3
times with reproducible results. The data are expressed as the means ± SEMs. * indicates p < 0.05 compared to the
Ang II group. Significant differences between conditions were determined by one-way ANOVA and Tukey’s multiple
comparisons test.

28
Figure 2. Principal component analysis (PCA). PCA score plot of the normal (red boxes), Ang II (green boxes), and
2 2
Ang II + Ang-(1-7) (blue boxes) groups (R = 0.77; Q = 0.70; 2 cumulative principal components).

26
Figure 3. Identification of biochemically significant metabolites between the normal and Ang II groups by
2 2
OPLS-DA. a) Cross-validated score plots for the OPLS-DA model (R = 0.87; Q = 0.73; 1 predictive component + 2
orthogonal components). The red and green labels indicate the normal and Ang II groups, respectively. The modeled
score (t [1]) and the cross-validated score (tcv [1]) have identical orthogonal scores (to [1]). b) Loading plot with jackknife
confidence intervals. Both the black and the red bars represent metabolites with a VIP score > 1, but the red bars also
indicate metabolites identified as biochemically significant using the S-plot. c) S-plot. Seventeen biochemically
significant metabolites (statistically significant metabolites with a relatively high concentration and reliability) are
indicated by the red circles. The black circles represent metabolites with a VIP > 1. The white circles indicate statistical
nonsignificance.

27
Figure 4. Identification of biochemically significant metabolites between the Ang II + Ang-(1-7) and Ang II
2 2
groups by OPLS-DA. a) Cross-validated score plots for the OPLS-DA model (R = 0.98; Q = 0.75; 1 predictive
component + 5 orthogonal components). The blue and green labels represent the Ang II + Ang-(1-7) and Ang II groups,
respectively. The modeled score (t [1]) and the cross-validated score (tcv [1]) have identical orthogonal scores (to [1]). b)
Loading plot with jackknife confidence intervals. Both the gray and the blue bars represent metabolites with a VIP
score > 1, but the blue bars also indicate metabolites identified as biochemically significant using the S-plot. c) S-plot.
Twenty-seven biochemically significant metabolites (statistically significant metabolites with a relatively high
concentration and reliability) are indicated by the blue circles. The gray circles represent metabolites with a VIP > 1.
The white circles indicate statistical nonsignificance.

28
Figure 5. Identification of biochemically significant metabolites common to the normal and Ang II + Ang-(1-7)
groups using an SUS plot. The red and blue circles indicate the biochemically significant metabolites common to the
two OPLS-DA models (normal vs Ang II and Ang II + Ang-(1-7) vs Ang II). To improve clarity, all nonsignificant
metabolites were eliminated from each model.

29
Figure 6. The global metabolic network of all identified metabolites based on the KEGG database. The larger
circles indicate the identified metabolites. Biochemically significant metabolites for the normal group are shown as red
circles. Biochemically interesting metabolites for the Ang II + Ang-(1-7) group are highlighted by blue halos. Red circles
with a blue halo indicate biochemically significant metabolites common to the normal and Ang II + Ang-(1-7) groups.
The shared metabolites can be explained by their presence in the six metabolic pathways illustrated on the right.

30
Figure 7. Ang-(1-7) suppressed calcium overload and excessive activation of the associated signaling pathway
2+
in Ang II-stimulated CFs. The concentration of intracellular calcium (Ca ) is indicated by the green fluorescence
2+
intensity. a-d) The effect of Ang II (10 µM) on the intracellular Ca concentration under the following conditions: a) in
2+ 2+
the absence of extracellular Ca ; b) in the presence of extracellular Ca (1.8 mM); c) under pretreatment with
2+
Ang-(1-7) (10 µM for 48 h) in the absence of extracellular Ca ; and d) under pretreatment with Ang-(1-7) (10 µM for 48

(1.8 mM). ○ concentration; ○


2+ 2+
h) after the addition of extracellular Ca B indicates the baseline intracellular Ca P

2+
indicates the peak intracellular Ca concentration. The scale bars represent 50 μm. e) Summary of the results
2+
presented in a-d. The mean peak intracellular Ca values are plotted against each condition (*p < 0.05 vs condition a;
#p < 0.05 vs condition b). Significant differences between different conditions were determined by two-way ANOVA and

Tukey’s multiple comparison test. f) Western blot analysis showing the expression of p-T287-CaMKIIδ, CaMKIIδ, CTGF,

p-ERK1/2, and ERK1/2 in normal CFs, Ang II-stimulated CFs, and Ang II-stimulated CFs treated with Ang-(1-7). These
experiments were repeated at least 3 times with reproducible results. The data are presented as the means ± SEMs. *p
< 0.05 vs the Ang II group. Significant differences between different conditions were determined by one-way ANOVA
with Tukey’s multiple comparison test.

31
Figure 8. Ang-(1-7) suppressed abnormal oxidative stress and excessive activation of the associated signaling
pathway in Ang II-stimulated CFs. a) H2O2 production in normal CFs, Ang II-stimulated CFs, and Ang II-stimulated
CFs treated with Ang-(1-7). The green fluorescence intensity reflects the level of H2O2. The scale bars represent 100
-
μm. b) O2 production in normal CFs, Ang II-stimulated CFs, and Ang II-stimulated CFs treated with Ang-(1-7). The red
-
fluorescence intensity reflects the level of intracellular O2 . The scale bars represent 100 μm. c) Western blot analysis

showing the expression of Nox4, ox-M281/282-CaMKIIδ, and CaMKIIδ in normal CFs, Ang II-stimulated CFs, and Ang

II-stimulated CFs treated with Ang-(1-7). These experiments were repeated at least 3 times with reproducible results.
The data are presented as the means ± SEMs. *p < 0.05 vs the Ang II group. Significant differences between different
conditions were determined by one-way ANOVA with Tukey’s multiple comparison test.

32
Figure 9. Ang-(1-7) treatment protected against Ang II-induced cardiac fibrosis in rats. a) Masson’s trichrome
staining for fibrotic areas in hearts from rats infused with Ang II (control), Ang II plus Ang-(1-7), and physiological saline

solution. The scale bars represent 100 μm. b) Immunohistochemical staining of α-SMA in the left ventricular sections of

the rats treated with Ang II (control), Ang II plus Ang-(1-7), and physiological saline solution. The scale bars represent
200 μm. c) Western blot analyses showing the protein levels of CTGF, p-ERK1/2, and ERK1/2 in hearts from rats
treated with Ang II (control), Ang II plus Ang-(1-7), and physiological saline solution. These experiments were repeated
at least 6 times with reproducible results. The data are presented as the means ± SEMs. *p < 0.05 vs the Ang II group.
Significant differences between different treatments were determined by one-way ANOVA with Tukey’s multiple
comparison test.

33
Figure 10. Ang-(1-7) suppressed the excessive CaMKIIδ activation of CFs in the rat model of Ang II-induced

cardiac fibrosis. a) The immunofluorescence colocalization of positive α-SMA and ox-CaMKIIδ cell markers in

DAPI-stained cells indicates that CaMKIIδ activation of CFs occurred in hearts from Ang II-infused rats. The scale bar
indicates 100 μm. Western blotting (b, c, d) and quantification analysis (e) showing the protein expression levels of

CaMKIIδ, p-T287-CaMKIIδ, ox-M281/282-CaMKIIδ, and Nox4 in the hearts of rats treated with Ang II (control), Ang II

plus Ang-(1-7), and physiological saline solution. These experiments were repeated at least 6 times with reproducible
results. The data are presented as the means ± SEMs. *p < 0.05 vs the Ang II group. Significant differences between
different treatments were determined by one-way ANOVA with Tukey’s multiple comparison test.

34
,

Figure 11. The hypothetical mechanism underlying the protective activity of Ang-(1-7) against cardiac fibrosis.
In Ang II-stimulated CFs, AA induces calcium overload through TRP channels. Nox4 is a membrane protein that
converts oxygen (blue circles) to ROS (red circles). The increased ROS level is reflected by the upregulation of
γ-glutamyl cycle activity, which increases the levels of glutamic acid, glycine, GSH (or GSSH), and pyroglutamic acid.
The accumulation of calcium and ROS activates CaMKIIδ by the posttranslational modifications of phosphorylation at
threonine 287 (p-T287) and oxidation at methionine 281/282 (ox-M281/282). Activated CaMKIIδ upregulates the
downstream CTGF- and ERK1/2-dependent signaling pathways, which leads to cardiac fibroblast proliferation and
differentiation into myofibroblasts. Pretreatment of Ang II-stimulated CFs with Ang-(1-7) can abolish calcium overload,
decrease Nox4-mediated ROS generation, downregulate CaMKIIδ-related pathways, and finally attenuate fibroblast
proliferation as well as reduce the expression of collagen I/II and α-SMA in myofibroblasts. Furthermore, Ang-(1-7)
lowers the levels of AA, GSH (or GSSG), and pyroglutamic acid.

35