Introduction

Vascular endothelial growth factor (VEGF) plays a pivotal role in cardiovascular disease. Tissue or serum levels of VEGF have been described to represent tissue hypoxia or angiogenesis status in patients with atrial fibrillation (AF) [1], acute myocardial infarction [2], and heart failure [3]. VEGF is highly expressed in atrial tissue and intra-atrium serum of patients with atrial arrhythmia [1, 4, 5]. In addition, VEGF in peri-atrial adipose tissue is highly correlated with atrial fibrosis levels in patients with AF [6]. A recent epidemiological study suggested that VEGF is an important risk factor for AF and ischemic stroke [7]. In hypertensive rats, VEGF is greatly coexpressed with myofibroblasts in the pulmonary veins-atrial junction [8]. These findings support the hypothesis that VEGF may play an important role in the pathophysiology of fibrotic atrial myopathy. However, the association between VEGF and fibrogenesis has not been fully elucidated. Previous studies have revealed that VEGF overexpressing mice have higher degrees of ventricular fibrosis than knockout mice [9]. VEGF can increase migratory and collagen production capabilities of human skin fibroblasts [10], and rat ventricular fibroblasts through ERK signaling pathway [11]. VEGF exerts its profibrotic effect by inducing myofibroblast transformation in conjunctival epithelial cells [12]. VEGF can activate higher collagen production in chondrocytes via the P38 signaling pathway [13]. Similarly, our recent study showed that VEGF can significantly increase collagen production in rat left atrial fibroblasts [14]. However, the cellular mechanisms underlying the effects of VEGF on atrial fibroblasts were not fully elucidated.

The calcium (Ca2+) signaling pathway has been demonstrated to be the downstream signaling of multiple profibrotic cytokines [15, 16]. Increasing Ca2+ influx augments myofibroblast differentiation [17], and proliferation [18], collagen production [19], and migratory capabilities of fibroblasts [20]. Moreover, VEGF can promote the proliferative capability of endothelial cells by increasing Ca2+ influx [21]. Nevertheless, the profibrotic effect of VEGF induced by Ca2+ signaling in human atrial fibroblasts is unclear. The purpose of this study was to examine whether VEGF regulates atrial fibrogenesis through modulation of fibroblast activity and Ca2+ homeostasis.

Materials and methods

Cell cultures

Human atrial fibroblasts were obtained from Lonza Research Laboratory (Walkersville, MD, USA). Cells were seeded on uncoated culture dishes as monolayers in Dulbecco’s modified Eagle’s medium (Thermo Fisher Scientific, Loughborough, UK) containing 10% fetal bovine serum (HyClone Laboratories, Logan, UT, USA) and 100 U/mL penicillin–streptomycin (Thermo Fisher Scientific, Loughborough, UK) at 37 °C with 5% CO2. Cells from passages 4–6 were used in this study to avoid the possible variations in cellular function.

Cell migration assay

Migration of atrial fibroblasts, treated with VEGF-A (300 pg/mL, 1000 pg/mL, Sigma-Aldrich, St. Louis, MO, USA), with or without ethylene glycol tetra-acetic acid (EGTA, 1 mmol/L, Sigma-Aldrich) or a Ca2+/calmodulin-dependent protein kinase II (CaMKII) inhibitor (KN93, 10 μmol/L, Sigma-Aldrich), was analyzed using a wound-healing assay 6 h after a cell monolayer in a six-well plate was scraped using a P200 pipette tip. Each gap length was assessed using SPOT software (SPOT imaging, Sterling Heights, MI, USA) and calculated from the average of 12 regions. The net migration distance after 6 h was subtracted from that at the time of the initial scratch [14].

Cell proliferation assay

Atrial fibroblast proliferation was measured using a commercial MTS kit (Promega, Madison, WI, USA) as previously described [22]. Briefly, atrial fibroblasts were seeded onto a 96-well culture dish at a density of 3000 cells/well. After growing to 50% confluence, the cells were incubated in serum-free medium with VEGF-A (300 pg/mL, 1000 pg/mL) for 24 h. Cell growth was analyzed using the MTS reagent, which was added 4 h before spectrophotometric analysis was performed.

Western blotting

Western blotting was performed as described previously [14]. Atrial fibroblasts treated with VEGF-A (300 pg/mL, 1000 pg/mL), with or without EGTA (1 mmol/L), or KN93 (10 μmol/L) were homogenized and lysed in radioimmunoprecipitation assay buffer containing 50 mmol/L Tris pH 7.4, 150 mmol/L NaCl, 1% Nonidet P P40, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate (SDS) and protease inhibitor cocktails (Sigma-Aldrich). The protein concentration was determined using a Bio-Rad protein assay reagent (Bio-Rad Laboratories, Hercules, CA, USA). Proteins were separated on a 10% SDS-polyacrylamide gel electrophoresis gel under reducing conditions and electrophoretically transferred onto an equilibrated polyvinylidene difluoride membrane (Amersham Biosciences, Buckinghamshire, UK). Blots were probed with primary antibodies against phosphorylated VEGF receptor 1 (pVEGFR1, 1:500, polyclonal, GeneTex Inc., Irvine, CA, USA), total VEGF receptor 1 (tVEGFR1, 1:1000, monoclonal, clone number: Y103, Abcam, Cambridge, UK), α-smooth muscle actin (SMA) (1:1000, monoclonal, clone number: 1A4, Abcam), pro-collagen type IA1(1:500, monoclonal, clone number: 3G3, Santa Cruz Biotechnology, Santa Cruz, CA, USA), pro-collagen type III(1:1000, monoclonal, clone number: FH7A, Abcam), phosphorylated ERK 1/2 (pERK, 1:1000, monoclonal, and clone number: D13.14.4E, Cell Signaling Technology, Beverly, MA, USA), phosphorylated p38 (pp38, 1:1000, Polyclonal, Cell Signaling Technology), phosphorylated CaMKII (1:2000, polyclonal, Abcam), and total CaMKII (1:2000, polyclonal, GeneTex Inc.). The blots were then incubated with secondary antibodies conjugated with horseradish peroxidase. Bound antibodies were detected using an enhanced chemiluminescence detection system (Millipore, Darmstadt, Germany) and analyzed with AlphaEaseFC software (Alpha Innotech, San Leandro, CA, USA). Targeted bands were normalized to the glyceraldehyde 3-phosphate dehydrogenase protein (Sigma-Aldrich) to confirm equal protein loading and then normalized to the value of control cells.

Patch-clamp experiments

A whole-cell patch clamp was used on detached single fibroblast using an Axopatch 1D amplifier (Axon Instruments, Foster City, CA, USA). Borosilicate glass electrodes (o.d., 1.8 mm) with tip resistances of 3–5 MΩ were used. The area under the capacitive current was initiated by a small hyperpolarizing step from a holding potential of −50 mV to a test potential of −55 mV for 80 ms. When measuring nonselective cation currents (INSC) through transient receptor potential (TRP) channels, we superfused the detached fibroblasts with Tyrode solution containing the following (in mmol/L): NaCl 140, tetraethylammonium chloride 5.4, MgCl 1.0, CaCl2 2.0, glucose 10, and 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) 10 with pH 7.4 adjusted with CsOH. The pipette solution contained the following (in mmol/L): CsCl2 135, CaCl2 0.1, EGTA 10, magnesium adenosine triphosphate (Mg-ATP) 4.0, MgCl2 1.0, HEPES 10, sodium guanosine triphosphate 0.3, and Na2-phosphocreatine 6.6 with pH 7.4 adjusted with CsOH. The currents were the differences before and after gadolinium (100 μmol/L; sigma) recorded by a voltage ramps for 3 s, ranging from −110 to +100 mV (0.07 mV/ms, 0.1 Hz) at 37 °C. Nifedipine (5 μmol/L) was used in external solution to block any L-type Ca2+ current.

When measuring potassium currents through intermediate-conductance Ca2+-activated K+ channels (KCa3.1), we superfused the detached fibroblasts with Tyrode solution containing the following (in mmol/L): NaCl 136, KCl 5.4, MgCl 1.0, CaCl2 1.8, glucose 10, and HEPES 10 with pH 7.4 adjusted with KOH. The pipette solution contained the following (in mmol/L): KCl 20, K-aspartate 120, MgCl2 1.0, EGTA 5.0, HEPES 10, GTP 0.1, Na2-phosphocreatine 5.0, Mg-ATP 5.0, and free Ca2+ 500 nmol/L, with pH 7.2 adjusted with KOH. The currents were measured as the differences before and after TRAM-34 (1 μmol/L, Sigma-Aldrich) recorded through a voltage STEP for 0.3 s, ranging from −100 to +60 mV with holding potential at −70 mV at 22 °C

Intracellular Ca2+ imaging

Ca2+ imaging was conducted as described previously [23]. Atrial fibroblasts on 3 cm glass bottom chamber slides were loaded with fura-2-acetoxymethyl ester (5 μmol/L, Life Technologies, Carlsbad, CA, USA) and Pluronic F-127 (20% solution in dimethyl sulfoxide; 2.5 μg/mL) in a Ca2+-free solution containing the following (in mmol/L) NaCl 120, KCl 5.4, KH2PO4 1.2, MgSO4 1.2, glucose 10, HEPES 6 and taurine 8 (pH 7.40) for 30 min at 36 °C in a humidified incubator with 5% CO2. Fura-2 fluorescence images were captured using a Polychrome V monochromator (Till Photonics, Munich, Germany) mounted on an upright Leica DMI 3000B microscope (Leica Microsystems, Buffalo Grove, IL, USA) with dual-excitation wavelengths of 340 and 380 nm, and an emission wavelength of 510 nm. Fura-2 images were analyzed using MetaFluor software version 7.7.6.0 (Molecular Devices, Sunnyvale, CA, USA). The ratio of fluorescence due to excitation at 340 nm (F340) to F380 was used as a marker of the relative level of intracellular Ca2+. To measure Ca2+ entry, cells were first exposed to the Ca2+-free solution for 8 min. The extracellular Ca2+ concentration was then increased to 2 mmol/L to measure Ca2+ entry via store-operated channels activated by Ca2+-store depletion. The change in intracellular Ca2+ from Ca2+-free solution to 2 mmol/L Ca2+ solution (∆F340/F380) was used to represent Ca2+ entry.

Statistical analysis

All quantitative data were expressed as the mean ± standard error of the mean. Paired t-test, Mann–Whitney rank-sum test, and one-way repeated measures ANOVA test with a post hoc Fisher’s least significant difference test were used to compare atrial fibroblasts under different conditions. A P value of <0.05 was considered statistically significant.

Results

VEGF effects on the proliferation, migration, and collagen production of atrial fibroblasts

Compared with control cells, after treatment with VEGF for 48 h, atrial fibroblasts exhibited greater migratory capability, and higher expression of pVEGFR1, pro-collagen type I protein, pro-collagen type III protein, and α-SMA (a myofibroblast differentiation marker) protein expression (Fig. 1). However, control and VEGF-treated atrial fibroblasts had similar tVEGFR1 protein expression, and proliferation rate (Fig. 1).

Fig. 1
figure 1

Cell migration, collagen production, myofibroblast differentiation, signal transduction, and proliferation capabilities in atrial fibroblasts treated with vascular endothelial growth factor (VEGF). a Photographs and averaged data reveal the migration assay results of atrial fibroblasts treated with VEGF. Left upper panels show the initial scratch (baseline) in different groups. Left lower panels show the images 6 h after the scratch was created (after migration) (n = 6 independent experiments). b Photographs and averaged data reveal the pro-collagen type I production (n = 7 independent experiments), pro-collagen type III production (n = 6 independent experiments), α-smooth muscle actin (SMA, n = 4 independent experiments), phosphorylated VEGF receptor type I (pVEGFR1; n = 7 independent experiments), and total VEGFR1 (tVEGFR1; n = 7 independent experiments) in atrial fibroblasts treated with VEGF for 48 h. c VEGF treatment for 24 h had no significant effect on the proliferation rate of atrial fibroblasts (n = 5 independent experiments). GAPDH was used as a loading control. *p < 0.05, ***p < 0.005

Full size image

Ca2+ entry and downstream signaling in VEGF-treated atrial fibroblasts

As shown in Fig. 2a, VEGF-treated (1000 pg/mL) atrial fibroblasts exhibited a greater Ca2+ entry. VEGF-treated (1000 pg/mL) atrial fibroblasts had upregulated currents through TRP and KCa3.1 channels (Fig. 2b). The enhanced Ca2+ entry in VEGF-treated (1000 pg/mL) atrial fibroblasts was attenuated with the TRP channel inhibitor (gadolinium, 100 μmol/L) or KCa3.1 inhibitor (TRAM-34, 1 μmol/L, Fig. 3a). Atrial fibroblasts treated with a combination of VEGF (1000 pg/mL) and the KCa3.1 inhibitor (TRAM-34, 1 μmol/L) exhibited lower INSC, than those treated with VEGF alone (Fig. 3b). As shown in Fig. 4a, VEGF-treated (1000 pg/mL for 5 min) atrial fibroblasts had a greater phosphorylated ERK than control atrial fibroblasts. However, control and VEGF-treated (1000 pg/mL for 5 min) atrial fibroblasts had similar phosphorylated p38 expression. Atrial fibroblasts treated with the combination of VEGF and the ERK inhibitor (PD98059, 50 μmol/L) had lower INSC than those treated with VEGF alone (Fig. 4b). The enhanced migratory, pro-collagen type I, pro-collagen type III, and α-SMA expression capabilities of VEGF-treated atrial fibroblasts were attenuated by EGTA at 1 mmol/L (Figs. 5 and 6). VEGF-treated atrial fibroblasts had a larger phosphorylated CaMKII than control atrial fibroblasts (Fig. 6). However, control and VEGF-treated atrial fibroblasts had similar total CaMKII (Fig. 6). The upregulated phosphorylated CaMKII of VEGF-treated atrial fibroblasts could also be attenuated with EGTA at 1 mmol/L (Fig. 6). Moreover, the increased migratory capability and the expressions of pro-collagen type I, pro-collagen type III, and α-SMA in VEGF-treated atrial fibroblasts were blocked by KN93 (Fig. 7).

Fig. 2
figure 2

Calcium entry, membrane currents of transient receptor potential (TRP) channels, and intermediate-conductance calcium-activated K+ channels (KCa3.1) in vascular endothelial growth factor (VEGF)-treated atrial fibroblasts. a Upper panels reveal representative intracellular calcium (Ca2+) tracing from control (upper tracing) and VEGF-treated (1000 pg/mL; lower tracing) atrial fibroblasts. Cells that had already been treated for 48 h, were first incubated with a Ca2+ free extracellular solution to deplete Ca2+ stores. Ca2+ entry was found after increasing extracellular Ca2+ to 2 mmol/L. lower panel shows the averaged change in intracellular Ca2+ from Ca2+ free solution to 2 mmol/L Ca2+ solution (change in the ratio of fluorescence due to excitation at 340 nm to that at 380 nm [∆F340/F380]). VEGF significantly increased Ca2+ influx in atrial fibroblasts (n = 32 cells from three independent experiments). b Upper panels reveal tracings and I/V relationship of the Gadolinium (100 μmol/L) sensitive nonselective cation current (INSC) from control (n = 15) and VEGF (1000 pg/mL)-treated (n = 18) atrial fibroblasts. Lower panels reveal tracings and I/V relationship of the TRAM-34 (1 μmol/L) sensitive potassium current from control (n = 11) and VEGF (1000 pg/mL)-treated (n = 12) atrial fibroblasts. *p < 0.05, **p < 0.01, ***p < 0.005. The insets in the current traces show the various clamp protocols

Full size image
Fig. 3
figure 3

Interaction between transient receptor potential (TRP) channels, intermediate-conductance calcium (Ca2+)-activated K+ channels (KCa3.1) and Ca2+ entry in vascular endothelial growth factor (VEGF)-treated atrial fibroblasts. a Representative Ca2+ influx tracing and average change in intracellular Ca2+ from Ca2+ free solution to 2 mmol/L Ca2+ solution (change in ratio of fluorescence due to excitation at 340 nm to that at 380 nm [∆F340/F380]) from atrial fibroblasts treated with VEGF (1000 pg/mL, n = 22), VEGF plus Gadolinium (Gd, 100 μmol/L, n = 24), or VEGF plus TRAM-34 (1 μmol/L, n = 24). b Tracings and I/V relationship reveal the gadolinium-sensitive (100 μmol/L) nonselective cation current (INSC) from atrial fibroblasts treated with VEGF (1000 pg/mL, n = 14), and those treated with VEGF plus TRAM-34 (1 μmol/L, n = 14). The insets in the current traces show the various clamp protocols. *p < 0.05, ***p < 0.005

Full size image
Fig. 4
figure 4

Downstream signaling of vascular endothelial growth factor (VEGF)-treated atrial fibroblasts. a Photographs and averaged data show the phosphorylated ERK (pERK, n = 6 independent experiments), and phosphorylated p38 (pp38, n = 6 independent experiments) in atrial fibroblasts treated with VEGF (1000 pg/mL) for 5 min. GAPDH was used as a loading control. b Tracings and I/V relationship reveal the gadolinium-sensitive (100 μmol/L) nonselective cation current (INSC) from atrial fibroblasts treated with VEGF (1000 pg/mL, n = 11), and those treated with VEGF plus ERK inhibitor (PD98059 50 μmol/L, n = 12). The insets in the current traces show the various clamp protocols. *p < 0.05

Full size image
Fig. 5
figure 5

The effects of ethylene glycol tetra-acetic acid (EGTA) on vascular endothelial growth factor (VEGF) increased migration in atrial fibroblasts. Photographs and averaged data show the migration assay in VEGF-treated (1000 pg/ mL) atrial fibroblasts with or without EGTA (1 mmol/L). Upper panels show the initial scratch (baseline) in different groups. Lower panels show the images 6 h after the scratch was created (after migration) (n = 8 independent experiments). ***p < 0.005

Full size image
Fig. 6
figure 6

The effects of ethylene glycol tetra-acetic acid (EGTA) on vascular endothelial growth factor (VEGF) increased profibrotic signaling in atrial fibroblasts. Photographs and averaged data show the pro-collagen type I production (n = 5 independent experiments), pro-collagen type III production (n = 7 independent experiments), α-smooth muscle actin (SMA, n = 7 independent experiments), phosphorylated Ca2+/calmodulin-dependent protein kinase II (pCaMKII, n = 6 independent experiments), and total CaMKII (tCaMKII, n = 6 independent experiments) in VEGF-treated (1000 pg/mL) atrial fibroblasts with or without EGTA (1 mmol/L) for 48 h. GAPDH was used as a loading control. *p < 0.05, ***p < 0.005

Full size image
Fig. 7
figure 7

The effects of KN93, the Ca2+/calmodulin-dependent protein kinase II (CaMKII) inhibitor, on vascular endothelial growth factor (VEGF)-treated atrial fibroblasts. a Photographs and averaged data show the migration assay in VEGF-treated (1000 pg/mL) atrial fibroblasts with or without KN93 (10 μmol/L). Upper panels show the initial scratch (baseline) in different groups. Lower panels show the images 6 h after the scratch was created (after migration) (n = 6 independent experiments). b Photographs and averaged data show the pro-collagen type I production (n = 6 independent experiments), pro-collagen type III production (n = 5 independent experiments), and α-smooth muscle actin (SMA, n = 6 independent experiments) in VEGF (1000 pg/mL)-treated atrial fibroblasts with or without KN93 (10 μmol/L) for 48 h. GAPDH was used as a loading control. *p < 0.05, **p < 0.01, ***p < 0.005

Full size image

Discussion

VEGF is associated with endothelial dysfunction and tissue hypoxia in cardiovascular diseases. Our previous study has shown that VEGF enhanced collagen production capability of left atrial fibroblasts [14], but the mechanisms were not fully elucidated. To our knowledge, in this study in human atrial fibroblasts, for the first time, we found that VEGF (1000 pg/mL) enhanced collagen production, and migration, and myofibroblast differentiation in atrial fibroblasts. Moreover, VEGF significantly increased Ca2+ entry, which is recognized as the fundamental signaling pathway in fibrogenesis. In addition, lower extracellular Ca2+ by EGTA blocked the profibrotic effects of VEGF on the fibroblasts migratory capability, collagen production ability, and myofibroblast differentiation capability, which suggests that augmented Ca2+ influx may contribute to the profibrotic effects of VEGF. The Ca2+ signaling pathway critically regulates the activities of multiple profibrotic growth factors. Activation of ryanodine channel-induced Ca2+ entry by the transforming growth factor was shown to increase collagen type I gene expression in fibroblasts [15]. The increase in intracellular Ca2+ by evoking inositol trisphosphate signaling may induce the genesis of collagen type I in fibroblasts [16]. Our findings provide benchside evidence leading to a potential novel strategy targeting atrial myopathy and arrhythmofibrosis. The concentration of VEGF in the present study was similar to the known higher plasma concentration in patients with AF (120–1400 pg/mL) [24]. Accordingly, the biological effects of VEGF in human atrial fibroblasts are expected to be clinically relevant.

TRP channels, which are Ca2+ entry channels activated by various pathological stimuli and cell stretch [25], are highly associated with cardiac fibrosis in a plethora of studies. TRP channels are upregulated in patients with atrial fibrosis [26]. Overexpressing TRP channels exhibited higher myocardial fibrosis in mice [27]. Blocking TRP channels suppresses angiotensin II-induced Ca2+ influx, collagen production, and myofibroblast differentiation in atrial fibroblasts [28]. Activated KCa3.1 channels induce membrane hyperpolarization and consequently trigger more persistent Ca2+ entry through TRP channels or other voltage-gated channels [29]. Blockade of KCa3.1 by TRAM-34 decreases Ca2+ entry [30], attenuates myocardial fibrosis in post-myocardial infarction rats [31] and decreases collagen production and myofibroblast differentiation capability of renal fibroblasts [32]. In line with previous studies, we found that VEGF significantly increased the currents of TRP and KCa3.1 channels, which contribute to the high Ca2+ entry in Ca2+ imaging assays. In addition, this study found that TRAM-34 attenuated the effects of VEGF on INSC in atrial fibroblasts, further confirming the interaction of TRP channels and KCa3.1 channels upon Ca2+ entry.

VEGF-A induces VEGFR1 auto-phosphorylation thereby transmitting various downstream signaling [33,34,35]. Autophosphorylated VEGFR1 activate profibrotic ERK [36] and p38 [37] signaling pathways [38]. In this study, VEGF was found to upregulate the expressions of phosphorylated VEGFR1 and phosphorylated ERK, but not p38, in atrial fibroblasts. Inhibition of the ERK signaling pathway decreased INSC through TRP channels in VEGF-treated atrial fibroblasts, which may indicate that VEGF activates TRP channels through ERK signaling. Similarly, a previous study showed that ERK inhibition attenuated TRP channel-dependent Ca2+ entry [39] and that TRP channels mediated inward currents can be activated through the ERK signaling pathway [40].

CaMKII is the downstream messenger of Ca2+ signaling. Binding of Ca2+/calmodulin triggers CaMKII auto-phosphorylation [41], thereby activating the profibrotic signaling, such as ERK [42], or NF-κB [43] signaling pathway. Pharmacological blockade of CaMKII by KN93 or genetically knocking down CaMKII decreases cardiac fibrosis in pathological left ventricular remodeling [44, 45]. In addition, the activation of CaMKII promotes collagen production and the migratory capability of fibroblasts [46, 47]. In this study, we found that VEGF upregulated phosphorylated CaMKII, but not total CaMKII, suggesting that VEGF modulates Ca2+ homeostasis in atrial fibroblasts through phosphorylated CaMKII, but not nonphosphorylated CaMKII. In addition, the profibrotic cellular activities and upregulated phosphorylated CaMKII of VEGF-treated atrial fibroblasts could be decreased with EGTA. Moreover, inhibiting CaMKII activity by KN93 decreased the VEGF-augmented migration, collagen production, and myofibroblast differentiation capabilities of atrial fibroblasts. These findings suggest that VEGF upregulated CaMKII signaling through Ca2+ influx and consequently triggered the profibrotic activities of atrial fibroblasts.

In conclusion, as summarized in Fig. 8, VEGF increases atrial fibroblast activity through CaMKII signaling by enhancing Ca2+ entry.

Fig. 8
figure 8

Illustration showing the proposed molecular mechanism of the profibrotic effects of vascular endothelial growth factor (VEGF) on atrial fibroblasts. VEGF increases Ca2+ entry via the ERK signaling and the interaction of transient receptor potential (TRP) and intermediate-conductance calcium-activated K+ (KCa3.1) channels thereby activating the collagen production, migratory capability and myofibroblast differentiation of atrial fibroblasts through the Ca2+/calmodulin-dependent protein kinase II (CaMKII)/ERK, NF-κB signal pathway. EGTA ethylene glycol tetra-acetic acid, KN93 CaMKII inhibitor

Full size image