Valproic Acid Inhibits RhoA-Mediated Vascular Smooth Muscle Cell Contraction

Abstract

Background

Vascular smooth muscle cells (VSMCs) play an important role in regulating vessel diameter and blood pressure. Dysregulation of VSMC contraction contributes to the development of coronary and post-subarachnoid hemorrhagic (SAH) vasospasms. We investigated the molecular mechanisms by which valproic acid (VPA) inhibits Ras homolog family member A (RhoA)-mediated VSMC contraction in rat VSMCs and isolated aortas.

Methods

In rat VSMCs, western blot analyses, quantitative real-time reverse transcription-polymerase chain reaction, ectopic expression of the constitutively active (CA)-RhoA gene or wild-type (WT)-histone deacetylase (HDAC) 5 gene, and inhibitor studies were performed. Active RhoA-GTP levels and Rho-associated protein kinase activity in VSMCs were also measured. We performed a phenylephrine (PE)-induced aortic contraction assay using isolated rat aortas, as well as post hoc analyses of an endothelium-dependent aortic relaxation assay using aortas from VPA-administered mice.

Results

VPA decreased the phosphorylation of the myosin light chain at Ser19 (p-MLC-Ser¹⁹) in a dose- and time-dependent manner. Interestingly, VPA significantly decreased RhoA mRNA and protein expression, as well as the active RhoA-GTP level. Furthermore, ectopic expression of CA-RhoA gene almost completely reversed VPA-inhibited p-MLC-Ser¹⁹. VPA markedly increased the acetylation levels of histone 3 (H3K9ac/K14ac). VPA and sodium butyrate but not valpromide significantly decreased the expression levels of both RhoA and p-MLC-Ser¹⁹ in VSMCs. However, this decrease was not reversed by overexpression of the WT-HDAC5 gene, indicating that HDAC5 was not responsible for this decrease. Consistent with the in vitro results, VPA attenuated PE-induced aortic contraction, decreased RhoA and p-MLC-Ser¹⁹ expression, and increased H3K9ac/K14ac levels in isolated rat aortas. The post hoc analysis revealed that the VPA-inhibited RhoA pathway accounted for 30% of the total aortic relaxation induced by VPA.

Conclusion

This study showed that VPA inhibits RhoA-mediated VSMC and vessel contraction by decreasing RhoA expression, which is mediated by the inhibitory action of VPA on HDACs. These results suggest that VPA may be useful in the treatment and prevention of spastic vascular diseases, including coronary and post-SAH vasospasms.

Graphical Abstract

INTRODUCTION

Valproic acid (VPA) has long been prescribed for the treatment of diverse neurological diseases, such as epilepsy, mood disorders, and intractable headaches.1 The exact mechanism of VPA action is unknown; however, one of the main mechanisms is known to inhibit the excessive activity of neurons by increasing the concentration of γ-aminobutyric acid (GABA), a neurotransmitter in the brain.2 In addition to potentiating GABA activity, VPA has been reported to inhibit several types of histone deacetylase (HDAC) isoforms, including class I HDACs (HDAC1, 2, 3, and 8) and class IIa HDACs (HDAC4, 5, 7, and 9).3,4 The inhibitory activity of VPA on HDACs has been shown to exert various biological effects, such as anti-cancer5 and vasoprotective effects.6,7

Contraction of vascular smooth muscle cells (VSMCs), which are typically located along the inner walls of blood vessels, plays a crucial role in controlling vessel diameter and modulating blood flow.8,9 VSMC contraction is triggered by various chemical and physiological signals within the vessel, primarily driven by an increase in Ca²⁺ concentration.10 The influx of Ca²⁺ into VSMCs induces the contraction of the actin-myosin complex.11 Myosin light chain kinase (MLCK) is essential for mediating VSMC contractions.12 MLCK is activated in a Ca²⁺/calmodulin-dependent manner and phosphorylates myosin light chain (MLC) at Ser19.12 Phosphorylated MLC induces VSMC contraction, which constricts the blood vessels and increases blood pressure.13,14 In contrast to MLCK, myosin light chain phosphatase (MLCP) plays an opposing role in VSMC contraction by decreasing phosphorylation of the myosin light chain at Ser19 (p-MLC-Ser¹⁹).12,15

Ras homolog family member A (RhoA), a member of the small GTPase family along with Rac1 and cdc42, functions as a molecular switch by cycling between an active GTP-bound state and an inactive GDP-bound state.16 RhoA/Rho-associated protein kinase (ROCK) also mediates VSMC contraction by regulating the activity of MLCK and MLCP in a Ca²⁺/calmodulin-independent pathway, and hyperactivation of RhoA/ROCK is linked to many cardiovascular diseases, including coronary artery spasm and hypertension.17,18 Excessive contractions and/or abnormalities in VSMCs contribute to the onset and exacerbation of coronary and post-subarachnoid hemorrhagic (SAH) vasospasms.19 Therefore, preventing or treating excessive VSMC contraction could be crucial for the management of these pathological conditions.

Although VPA has been shown to exhibit vasoprotective effects, including the enhancement of angiogenesis and nitric oxide (NO) production,6,7 the effects of VPA on VSMC contraction and the underlying molecular mechanisms remain unclear. Therefore, in the current study, we investigated the molecular mechanism by which VPA inhibits RhoA-mediated VSMC contraction in rat VSMCs and isolated aortas.

METHODS

Materials

VPA and phenylephrine (PE) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Valpromide (VM) was purchased from Selleckchem (Houston, TX, USA) and sodium butyrate (SB) from Gibco (Grand Island, NY, USA). Antibodies against RhoA, MLC, p-MLC-Ser¹⁹, myosin phosphatase target subunit 1 (MYPT1), p-MYPT1-Thr⁶⁹⁶, H3K9ac, H3K9/K14ac, H3K27ac, protein phosphatase 1β (PP1β), HDAC5, and actin were purchased form Cell Signaling Technology (Beverly, MA, USA). Antibodies against myc and ROCK1 were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA) and Novus Biologicals (Littleton, CO, USA), respectively. Dulbecco’s modified Eagle medium (DMEM) was obtained from Thermo Fisher Scientific (Waltham, MA, USA). Fetal bovine serum was purchased from Welgene (Gyeongsan, Korea). Penicillin and streptomycin were supplied by HyClone Laboratories Inc. (Logan, UT, USA). Trypsin-ethylenediaminetetraacetic acid solution was purchased from Cytiva (Marlborough, MA, USA). The cell culture plasticware was purchased from SPL Life Sciences (Pocheon, Korea). All other chemicals used were of the purest analytical grade.

Cell culture and drug treatments

Rat aortic VSMCs were isolated and cultured as previously described.20 Cells between passages three and seven were used for the experiments. VSMCs grown to 90% confluence in 60-mm culture dishes were incubated in the absence or presence of various concentrations of VPA for 24 hours, or in 2 mM VPA for the indicated duration in serum-free DMEM, followed by stimulation with 1 μM PE for 10 minutes. In some experiments, the cells were co-treated with the indicated drugs or chemicals for the indicated duration.

Western blot analyses

Total proteins were extracted from VSMCs or rat aortic tissues, and subjected to western blot analyses, as described previously.20 The primary antibody dilutions used for western blot analyses were as follows; MLCK (1:3,000), MLC (1:3,000), p-MLC-Ser¹⁹ (1:1,000), ROCK1 (1:3,000), MYPT1 (1:3,000), p-MYPT1-Thr⁶⁹⁶ (1:1,000), RhoA (1:3,000), PP1β (1:3,000), HDAC5 (1:1,000), H3K9ac (1:3,000), H3K9/K14ac (1:3,000), H3K27ac (1:1,000), myc (1:3,000), and actin (1:3,000).

Quantitative real-time reverse transcription-polymerase chain reaction (qRT-PCR)

After VSMCs were treated with various doses of VPA or vehicle in serum-free DMEM for 24 hours, total RNA was isolated using TRIzol Reagent (Invitrogen, Carlsbad, CA, USA), and then subjected to qRT-PCR, as described previously.21 The relative expression of RhoA mRNA to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA was quantitated using the ∆∆Ct method. The following PCR primer pairs were designed to detect rat genes: RhoA-F, 5′-GGCAGAGATATGGCAAACAGG-3′; RhoA-R, 5′-TCCGTCTTTGGTCTTTGCTGA-3′; GAPDH-F, 5′-ACAAGATGGTGAAGGTCGGTGTGA-3′; GAPDH-R, 5′-AGCTTCCCATTCTCAGCCTTGACT-3′.

RhoA activity assay

The RhoA activity assay was performed using a Rho activation assay kit (Merck Millipore, Burlington, MA, USA), according to the manufacturer’s instructions. The GTP-bound RhoA (active RhoA) levels were measured by adding 20 µL of Rho assay reagent slurry (Rhotekin RBD-agarose) to 400 µL of supernatants containing equal amounts of proteins (500 µg).

ROCK activity assay

The ROCK activity assay was performed using a ROCK activity assay kit (Cell Biolabs Inc., San Diego, CA, USA), according to the manufacturer’s instructions. The assay reaction was initiated by adding 90 µL of supernatants to a ROCK substrate-coated 96-well plate. ROCK activity was measured colorimetrically by reading absorbance at 450 nm, and the results were normalized to the protein concentration and expressed as a percentage of the control.

Transfection of constitutively active (CA)-RhoA gene or wild-type (WT)-HDAC5 gene

The N-terminal myc-tagged CA-RhoA cDNA construct and the C-terminal Flag-tagged WT-HDAC5 cDNA construct were gifts from Professor In-San Kim (Korea Institute of Science and Technology, Seoul, Korea) and Professor Han-Kwang Yang (Seoul National University College of Medicine, Seoul, Korea), respectively. The pcDNA3 vector containing human CA-RhoA cDNA with a point mutation of GGA→GTA at the 14th codon (G14V mutant) or the pcDNA3 vector containing human WT-HDAC5 cDNA was transfected into VSMCs using Solfect reagent from Biosolyx (Daegu, Korea), as described previously.20 Equal amounts of empty vectors were also transfected for use as controls.

Animals

Six-week-old male Sprague-Dawley rats were maintained for 1 week at the beginning of the experiment in a temperature- and humidity-controlled room (22 ± 1°C and 50 ± 10%, respectively) under a 12-hours alternate light/dark cycle. All rats were provided water and standard chow (Purina Mills, LLC, St. Louis, MO, USA) ad libitum throughout the experiments.

Measurement of aortic vessel contraction

Aortic vessel contractions were measured in the thoracic aortic rings, as previously described.21 The rings were stretched to a resting tension of 2 g and equilibrated for 30 minutes in an organ bath filled with the Krebs-Henseleit (KH) solution. The rings were sequentially exposed to 65 mM KCl and KH solutions at least twice, and PE (10⁻⁸–10⁻⁴ M) was added cumulatively to assess aortic vessel contraction.

Post hoc analyses of endothelium-dependent aortic vessel relaxation

The endothelium-dependent aortic vessel relaxation assay was previously performed in our laboratory using thoracic aortic rings from male institute of cancer research (ICR) mice administered orally with or without VPA (500 mg/kg daily for 2 weeks), as described in our already published paper.7 The percentage of the VPA-inhibited RhoA pathway in the total aortic vessel relaxation induced by VPA was evaluated through post hoc analyses of previously obtained raw data, comparing it with the degree of NO-mediated aortic vessel relaxation.

Statistical analyses

All results are expressed as mean ± standard deviation, with n indicating the number of experiments. The statistical significance of differences between the two points was evaluated using a Student’s t-test. All differences were considered significant at a P value < 0.05.

Ethics statement

All animal experiments were conducted in accordance with the approved institutional guidelines for animal care and use at Yeungnam University (approval No. YUMC-AEC2022-035).

RESULTS

VPA decreases p-MLC-Ser¹⁹ in rat VSMCs in dose- and time-dependent manners

So far, the effect of VPA on VSMC contraction and the underlying molecular mechanisms remain elusive. Therefore, we first investigated whether VPA inhibits rat VSMC contractility. To achieve this, we examined whether VPA decreases p-MLC-Ser¹⁹ levels in VSMCs because the primary function of p-MLC-Ser¹⁹ is to facilitate muscle contraction.15 As shown in Fig. 1A, VPA decreased the p-MLC-Ser¹⁹ levels in a dose-dependent manner. Treatment with 4 mM VPA reduced p-MLC-Ser¹⁹ to 40% compared to that in vehicle (water) control (Fig. 1A). In addition, 2 mM VPA decreased p-MLC-Ser¹⁹ levels in a time-dependent manner (Fig. 1B). The maximal inhibitory effect of VPA on p-MLC-Ser¹⁹ was observed at 24 hours, which was 50% of the control at baseline (Fig. 1B). These results suggest that VPA inhibits VSMC contractility by downregulating the intracellular signaling pathway that mediates MLC phosphorylation.

Fig. 1. VPA decreases p-MLC-Ser¹⁹ in rat VSMCs. (A) VSMCs were treated with various doses of VPA (0, 1, 2, or 4 mM) for 24 hours, followed by stimulation with 1 μM PE for 10 minutes, and then levels of p-MLC-Ser¹⁹, MLC, and actin expression were detected using western blotting. (B) VSMCs were treated with 2 mM VPA for various durations (0, 3, 8, 24 hours), followed by stimulation with 1 μM PE for 10 minutes, and then levels of p-MLC-Ser¹⁹, MLC, and actin expression were detected using western blotting. All experiments were performed independently at least four times, and the blots shown are representative of these experiments (n = 4). Bar graphs depict the mean below the control (± standard deviation).

VPA = valproic acid, p-MLC-Ser¹⁹ = phosphorylation of the myosin light chain at Ser19, VSMC = vascular smooth muscle cell, PE = phenylephrine, MLC = myosin light chain.

Differences were considered statistically significant at ^**P < 0.01 and ^***P < 0.001.

VPA mediates the inhibition of p-MLC-Ser¹⁹ by decreasing RhoA expression and active RhoA-GTP levels

Next, we examined whether VPA affects the expression levels of MLCK and MLCP, as these molecules directly regulate MLC phosphorylation.22 As shown in Supplementary Fig. 1, the levels of MLCK protein and MLCP subunits, including p-MYPT1-Thr⁶⁹⁶, MYPT1, and PP1β, did not change in VPA-treated VSMCs, suggesting that MLCK and MLCP are unlikely to directly mediate the VPA-induced decrease in p-MLC-Ser¹⁹. In addition to MLCK and MLCP, RhoA, a small GTP-binding protein, has also been reported to be involved in the regulation of VSMC contraction by activating ROCK.23 Therefore, we investigated whether VPA decreases RhoA expression. As shown in Fig. 2A, VPA decreased the RhoA protein expression in a dose-dependent manner. Treatment with 4 mM VPA reduced RhoA protein expression by 40% compared to that in vehicle control (Fig. 2A). Furthermore, treatment with 2 mM VPA decreased the RhoA protein expression in a time-dependent manner (Fig. 2B). To clarify whether VPA decreases RhoA expression at the transcriptional or post-translational step, RhoA mRNA expression was assessed in VPA-treated cells. As shown in Fig. 2C and D, VPA repressed RhoA mRNA expression in a dose- and time-dependent manner, suggesting that the VPA-induced decrease in RhoA expression was attributable to the inhibition of its mRNA expression. Next, to ensure that reduced RhoA expression indeed correlates with a decrease in active RhoA contents, cellular RhoA-GTP, an active RhoA, levels were examined. Similarly to the reduction of RhoA expression, VPA decreased RhoA-GTP levels in a dose-dependent manner (Fig. 2E). To clarify the role of RhoA in VPA-inhibited p-MLC-Ser¹⁹, myc-tagged CA-RhoA DNA constructs were introduced into VSMCs. Overexpression of CA-RhoA was successful, as evidenced by the restoration of the VPA-induced decrease in RhoA and myc expression in CA-RhoA-transfected cells (Fig. 2F). As expected, VPA-inhibited p-MLC-Ser¹⁹ was almost completely reversed in CA-RhoA-transfected VSMCs (Fig. 2G). These results clearly show that the VPA-induced decrease in RhoA expression and, consequently, RhoA-GTP levels, plays a pivotal role in the inhibition of p-MLC-Ser¹⁹ in VSMCs. Next, we tested whether VPA affects ROCK expression and its activity, as RhoA is well-established to be linked to the regulation of VSMC contraction through the activation of ROCK.23 However, unexpectedly, ROCK1 expression, as well as total cellular ROCK activity, was not altered in the VPA-treated VSMCs (Supplementary Fig. 2), which is consistent with the finding that p-MYPT1-Thr⁶⁹⁶ levels were not changed by VPA (Supplementary Fig. 1D). These results suggest that the VPA-induced decrease in RhoA expression may mediate the inhibition of p-MLC-Ser¹⁹ and VSMC contraction through an as-yet unidentified effector or signaling pathway, distinct from the well-established ROCK1/MYPT1/MLC signaling pathway.

Fig. 2. VPA inhibits p-MLC-Ser¹⁹ by reducing RhoA expression and the levels of active RhoA-GTP. (A) VSMCs were treated with various doses of VPA (0, 1, 2, or 4 mM) for 24 hours, followed by stimulation with 1 μM PE for 10 minutes, and then levels of RhoA and actin expression were detected using western blotting. (B) VSMCs were treated with 2 mM VPA for various duration (0, 3, 8, 24 hours), followed by stimulation with 1 μM PE for 10 minutes, and then levels of RhoA and actin expression were detected using western blotting. (C) VSMCs were treated with various doses of VPA (0, 1, 2, or 4 mM) for 24 hours, followed by stimulation with 1 μM PE for 10 minutes, and then levels of RhoA mRNA expression relative to GAPDH mRNA expression were detected using qRT-PCR. (D) VSMCs were treated with 2 mM VPA for various duration (0, 3, 8, 24 hours), followed by stimulation with 1 μM PE for 10 minutes, and then levels of RhoA mRNA expression relative to GAPDH mRNA expression were detected using qRT-PCR. (E) After VSMCs were treated with various doses of VPA (0, 1, 2, or 4 mM) for 24 hours, followed by stimulation with 1 μM PE for 10 minutes, RhoA activity assay was performed and levels of active RhoA-GTP, RhoA, and actin expression were determined using western blotting. (F) After VSMCs were transfected with the CA-RhoA mutant gene or empty vector and treated with 2 mM VPA for 24 hours, followed by stimulation with 1 μM PE for 10 minutes, levels of RhoA, myc, and actin expression were detected using western blotting. (G) After VSMCs were treated as described above, levels of p-MLC-Ser¹⁹, MLC, and actin expression were detected using western blotting. All experiments were performed independently at least four times, and the blots shown are representative of these experiments (n = 4). Bar graphs depict the mean above/below the control (± standard deviation).

VPA = valproic acid, p-MLC-Ser¹⁹ = phosphorylation of the myosin light chain at Ser19, RhoA = Ras homolog family member A, GTP = guanosine-5'-triphosphate, VSMC = vascular smooth muscle cell, PE = phenylephrine, GAPDH = glyceraldehyde-3-phosphate dehydrogenase, qRT-PCR = quantitative real-time reverse transcription-polymerase chain reaction, CA = constitutively active, MLC = myosin light chain.

Differences were considered statistically significant at ^*P < 0.05, ^**P < 0.01, and ^***P < 0.001.

VPA-inhibited HDAC signaling pathway mediates downregulations of RhoA and p-MLC-Ser¹⁹ expression

VPA inhibits class I and IIa HDACs, including HDAC1, 2, 3, 4, 5, 7, 8, and 9.3,4 HDACs regulate gene expression by controlling the acetylation levels of diverse proteins, including transcription factors,24 that affect various cellular functions.25 Based on these reports, we investigated whether VPA exerts its inhibitory effect on VSMC contractility by inhibiting HDACs. As shown in Fig. 3A and B, VPA markedly elevated the levels of H3K9/K14ac in a dose- and time-dependent manner. Interestingly, treatment with 2 mM VPA almost maximally increased the H3K9/K14ac at an earlier time point at 3 hours (Fig. 4B). Similar to the increased H3K9/K14ac, acetylation levels at other acetylation sites of histone 3, including H3K9 and H3K27, were markedly increased in VPA-treated VSMCs (Supplementary Fig. 3). Next, inhibitor studies were performed to confirm whether the VPA-mediated decrease in RhoA and p-MLC-Ser¹⁹ expression in VSMCs is due to VPA’s inhibitory action on HDACs. Elevated H3K9/K14ac levels were observed in VSMCs treated with VPA or SB, an HDAC inhibitor structurally related to VPA (Fig. 3C). In contrast, VM, a structural VPA analog that does not inhibit HDACs,26,27 did not affect H3K9/K14ac levels (Fig. 3C). As expected, treatment with 2 mM VPA or SB, but not VM, significantly decreased RhoA and p-MLC-Ser¹⁹ expression in VSMCs (Fig. 3D and E). These results show that VPA decreases RhoA and p-MLC-Ser¹⁹ expression in VSMCs via the HDAC signaling pathway. Next, to specify which HDAC isoform(s) or HDAC class(es) mediate VPA’s inhibitory effects on RhoA and p-MLC-Ser¹⁹ expression, we introduced the WT-HDAC5 gene into VSMCs and examined changes in the levels of H3K9ac, H3K9/K14ac, and H3K27ac. This is because knockdown of HDAC5 gene expression has been reported to decrease RhoA mRNA and protein levels in rat VSMCs.28 Furthermore, recent studies have shown that class I HDACs, such as HDAC1, 2, and 3, exert various physiological effects by deacetylating H3K9ac,29,30 while class IIa HDACs, such as HDAC5, mediate physiological impacts by deacetylating H3K14ac and H3K27ac.31,32 As shown in Supplementary Fig. 4A and B, the levels of H3K27ac and H3K9ac/K14ac, but not H3K9ac, were significantly attenuated in WT-HDAC5 gene-overexpressed VSMCs. However, to our surprise, RhoA and p-MLC-Ser¹⁹ expression were further decreased in an additive manner with VPA in these cells (Supplementary Fig. 4C and D). These results clearly demonstrate that HDAC5 is not responsible for the VPA-mediated inhibition of RhoA and p-MLC-Ser¹⁹ expression and suggest that class I HDACs may be involved in the observed inhibitory effects of VPA.

Fig. 3. VPA-inhibited HDAC signaling pathway is involved in the decrease of RhoA and p-MLC-Ser¹⁹ expression. (A) VSMCs were treated with various doses of VPA (0, 1, 2, or 4 mM) for 24 hours, followed by stimulation with 1 μM PE for 10 min, and then levels of H3K9/K14ac and actin expression were detected using western blotting. (B) VSMCs were treated with 2 mM VPA for various duration (0, 3, 8, 24 hours), followed by stimulation with 1 μM PE for 10 minutes, and then levels of H3K9/K14ac and actin expression were detected using western blotting. (C) VSMCs were treated with 2 mM of VPA, VM, or SB for 24 hours, followed by stimulation with 1 μM PE for 10 minutes, and then levels of H3K9/K14ac and actin expression were detected using western blotting. (D) After VSMCs were treated as described above, levels of RhoA and actin expression were detected using western blotting. (E) After VSMCs were treated as described above, levels of p-MLC-Ser¹⁹, MLC, and actin expression were detected using western blotting. All experiments were performed independently at least four times, and the blots shown are representative of these experiments (n = 4). Bar graphs depict the mean above/below the control (± standard deviation).

VPA = valproic acid, HDAC = histone deacetylase, RhoA = Ras homolog family member A, p-MLC-Ser¹⁹ = phosphorylation of the myosin light chain at Ser19, VSMCs = vascular smooth muscle cells, PE = phenylephrine, H3K9/K14ac = acetylated histone 3 at lysine residues 9 and 14, VM = valpromide, SB = sodium butyrate, MLC = myosin light chain, DMSO = dimethyl sulfoxide.

Differences were considered statistically significant at ^**P < 0.01 and ^***P < 0.001.

Fig. 4. VPA decreases PE-induced vessel contraction, p-MLC-Ser¹⁹, and RhoA expression in rat aortas. (A, B) Rat thoracic aortas were prepared and the vessel contraction assay performed. After endothelium-deprived aortic rings were treated with vehicle (water) or 2 mM VPA for 24 hours, PE-induced aortic contraction was measured by cumulative treatment with PE (10⁻⁸ M–10⁻⁴ M). The tension curves indicate PE-induced aortic contraction in response to vehicle (water) or 2 mM VPA, and the line graph represents the mean ± standard deviation at each point (n = 6). (C) In a separate experiment, after endothelium-deprived aortic tissues were treated with vehicle (water) or 2 mM VPA for 24 hours, followed by stimulation with 1 μM PE for 10 minutes, levels of p-MLC-Ser¹⁹, MLC, and actin expression were detected using western blotting. (D) After endothelium-deprived aortic tissues were treated as described above, levels of RhoA and actin expression were detected using western blotting. (E) After endothelium-deprived aortic tissues were treated as described above, levels of H3K9/K14ac and actin expression were detected using western blotting. All experiments were performed independently at least 4 times, and the blots shown are representative of these experiments (n = 4). Bar graphs depict the mean above/below the control (± standard deviation).

VPA = valproic acid, PE = phenylephrine, p-MLC-Ser¹⁹ = phosphorylation of the myosin light chain at Ser19, RhoA = Ras homolog family member A, MLC = myosin light chain, H3K9/K14ac = acetylated histone 3 at lysine residues 9 and 14.

Differences were considered statistically significant at ^*P < 0.05 and ^**P < 0.01.

VPA attenuates PE-induced vessel contraction, and decreases p-MLC-Ser¹⁹ and RhoA expression in rat aortas

To verify the inhibitory effects of VPA on RhoA-mediated vessel contraction, an ex vivo aortic vessel contraction assay was performed using endothelium-deprived rat aortas. As shown in Fig. 4A and B, the cumulative addition of PE at 10⁻⁸–10⁻⁴ M to the control organ bath containing the aortic ring markedly induced aortic vessel contraction. Maximal contraction force was observed to be 2 g in control aortas when 10⁻⁴ M PE was added, whereas aortic vessel contraction in VPA-treated aortas was remarkably attenuated by 65% compared to that in control aortas (Fig. 5A and B). In line with in vitro results and aortic vessel contraction results, VPA significantly decreased p-MLC-Ser¹⁹ and RhoA expression in isolated rat aortas (Fig. 4C and D). Additionally, H3K9/K14ac levels were increased in VPA-treated aortas (Fig. 4E). These results indicate that VPA attenuated PE-induced vessel contraction in rat aortas by inhibiting RhoA-mediated vessel contraction.

Fig. 5. RhoA-mediated vessel contraction inhibited by VPA accounts for 30% of the total aortic vessel relaxation induced by VPA. (A, B) Endothelium-intact thoracic aortas were obtained from ICR mice that were administered 500 mg/kg of VPA orally once daily for 2 weeks, or were not treated with VPA. ACh-induced aortic vessel relaxation assays were performed using these aortas. After aortic contraction was induced by adding 0.1 μM NE, ACh-induced aortic relaxation was measured by cumulative treatment with ACh (10⁻⁹–10⁻⁶ M). The tension curves show ACh-induced aortic relaxation in vehicle (water)-treated or VPA-treated mouse aortas, and the line graph represents the mean ± standard deviation at each point (n = 8). (C) Aortic relaxation in VPA-administered mouse aortas relative to control aortas was assessed by dividing the aortic tension of control mouse aortas by that of VPA-administered mouse aortas at each point. All experiments were performed independently 8 times (n = 8). Bar graphs depict the mean percentage of aortic relaxation in VPA-administered mouse aortas relative to control aortas (± standard deviation).

RhoA = Ras homolog family member A, VPA = valproic acid, ICR = institute of cancer research, ACh = acetylcholine, NE = norepinephrine.

Differences were considered statistically significant at ^*P < 0.05 and ^**P < 0.01.

VPA-inhibited RhoA-mediated vessel contraction contributes to 30% of the total aortic vessel relaxation induced by VPA

Finally, to assess the contribution of VPA-inhibited RhoA-mediated vessel contraction to the total vessel relaxation induced by VPA, we conducted post hoc analyses of our previous in vivo experiments using ICR mice.7 Similar to the findings of PE-induced vessel contraction inhibition in VPA-treated rat aortas (Fig. 4A and B), norepinephrine (NE)-induced aortic vessel contraction was also significantly attenuated in VPA-administered mouse aortas (Fig. 5A and B). Furthermore, the aortic vessel relaxation in VPA-administered mouse aortas was 130% of that in control aortas when 0.1 μM NE alone or a very low amount of acetylcholine (ACh) (10⁻⁹ M) was added (Fig. 5C). On the other hand, the aortic vessel relaxation in VPA-administered mouse aortas was 200% of that in control aortas when 0.1 μM NE and a sufficient amount of ACh (10⁻⁷ M) were added cumulatively (Fig. 5C). These results demonstrate that the NO-independent RhoA-mediated contraction pathway inhibited by VPA contributed to 30% of the total aortic vessel relaxation, while the ACh-induced NO production, enhanced by VPA, accounted for 70% of the total aortic vessel relaxation.

DISCUSSION

VPA, a short-branched fatty acid, has long been used to treat epilepsy and bipolar disorder because it prolongs the duration of action of GABA, an inhibitory neurotransmitter, in the central nervous system.1,2 In addition to its neuromodulatory effects, VPA exhibits beneficial effects in various human diseases, such as the inhibition of cancer cell proliferation.5,33 However, the effects of VPA on VSMC contraction and underlying molecular mechanisms remain unclear. The current results demonstrated for the first time that VPA inhibited RhoA-mediated VSMC contractility and aortic vessel contraction by decreasing RhoA expression through the inhibitory action of VPA on HDACs (Fig. 6). These results suggest that VPA can be used to treat and prevent vasospastic diseases such as coronary vasospasm and post-SAH vasospasm because aberrant and excessive contraction of VSMCs plays a critical role in the development and exacerbation of these diseases.34,35

Fig. 6. A schematic illustration of inhibitory action of VPA on RhoA-mediated vessel contraction. VPA inhibits HDAC, which reduces the expression of RhoA. The VPA-repressed RhoA expression accompanies a decrease in cellular RhoA-GTP, the active form of RhoA. Ultimately, decreased RhoA activity mediates the dephosphorylation of MLC, resulting in the inhibition of VSMC contraction.

VPA = valproic acid, RhoA = Ras homolog family member A, HDAC = histone deacetylase, GTP = guanosine-5'-triphosphate, MLC = myosin light chain, VSMC = vascular smooth muscle cell, PE = phenylephrine, GEF = guanine nucleotide-exchange factor, GAP = guanosine triphosphatase activating protein, GDP = guanosine-5'-diphosphate.

Notably, VPA decreased RhoA expression and consequently activated RhoA-GTP content in VSMCs (Fig. 2) through the inhibitory action of VPA on HDACs (Fig. 3), which led to the inhibition of VSMC contractility and aortic vessel contraction (Figs. 3E, 4 and 5). VPA inhibits class I and class IIa HDACs, including HDAC1, 2, 3, 4, 5, 7, 8, and 9,3,4 and elevates histone acetylation and induces relaxation of the chromatin structure, ultimately changing the transcriptional state of various genes.36,37,38 Recently, Bai et al.28 showed that the deletion of HDAC5 attenuated angiotensin II (Ang II)-induced hypertension and aortic vessel contraction in Hdac5 knockout mice.28 Furthermore, the authors revealed that knockdown of HDAC5 gene expression decreased the levels of RhoA mRNA and protein expression in rat VSMCs,28 although they did not show whether overexpression of the RhoA gene reversed the attenuation of Ang II-induced VSMC contractility in HDAC5-knockdown VSMCs. Furthermore, recent studies have shown that class I HDACs, such as HDAC1, 2, and 3, exert various physiological effects by deacetylating H3K9ac,29,30 and class IIa HDACs, such as HDAC5, mediate physiological impacts by deacetylating H3K14ac and H3K27ac,31,32 despite the fact that HDACs are known not to have high substrate selectivity for lysine residues in histone 3.39 However, under our experimental conditions, RhoA and p-MLC-Ser¹⁹ expression were further decreased in an additive manner with VPA in WT-HDAC5 gene-overexpressed VSMCs (Supplementary Fig. 4C and D), although the levels of H3K27ac and H3K9ac/K14ac, but not H3K9ac, were significantly attenuated in these cells (Supplementary Fig. 4A and B). Our current findings reveal that HDAC5 is not involved in the VPA-mediated inhibition of RhoA and p-MLC-Ser¹⁹ expression and narrow down class I HDACs, rather than class IIa HDACs including HDAC5, as the HDAC class that mediates the inhibitory effect of VPA on RhoA and p-MLC-Ser¹⁹ expression.

In general, VSMC contraction and relaxation are determined by the balance between phosphorylation and dephosphorylation of MLC at Ser19 by MLCK and MLCP, respectively.22 The current results showed that VPA did not change expressions of MLCK and MLCP subunits such as p-MYPT1-Thr⁶⁹⁶, MYPT1, and PP1β (Supplementary Fig. 1), suggesting that canonical MLCK and MLCP signaling pathways are not responsible for the VPA-inhibited p-MLC-Ser¹⁹. Instead of a VPA effect on MLCK and MLCP, the current results demonstrated that VPA decreased RhoA expression in rat VSMCs and aortas (Figs. 2 and 4D), ultimately resulting in a decrease in p-MLC-Ser¹⁹. Furthermore, our results showed that overexpression of the CA-RhoA gene markedly restored the VPA-induced decrease in p-MLC-Ser¹⁹ expression (Fig. 2G). In this respect, RhoA-GTP, the active form of RhoA, binds to and activates ROCK, which directly phosphorylates MLC at Ser19, inducing VSMC contraction.23 ROCK can also directly phosphorylate MYPT1 at Thr696, leading to the inhibition of MLCP and the subsequent VSMC contraction.23,40 ROCK activated by RhoA-GTP can also increase VSMC contraction through an indirect pathway involving CPI-17, in which ROCK phosphorylates CPI-17 at Thr38, and p-CPI-17-Thr38 binds to the MLCP subunit MYPT1, resulting in the inhibition of MLCP activity.23,41,42 However, under our experimental conditions, we observed that ROCK1 expression, as well as total ROCK enzymatic activity, was not altered in the VPA-treated VSMCs (Supplementary Fig. 2), which is consistent with the finding of no change in p-MYPT1-Thr⁶⁹⁶ levels upon VPA treatment (Supplementary Fig. 1D). Our current results suggest that the VPA-induced reduction in RhoA expression may mediate the inhibition of p-MLC-Ser¹⁹ and VSMC contraction through an as-yet unidentified effector or signaling pathway, separate from the well-established ROCK1/MYPT1/MLC signaling pathway. However, further multidisciplinary investigations are needed to clarify this issue, which is beyond the scope of the present study.

The dysregulation of VSMC contraction in arterial vessels contributes to the development and exacerbation of various vascular diseases, including hypertension and pathological vasospasms.12,34,35 Although Ca²⁺-dependent and -independent signaling pathways of VSMC contraction are indicated as the underlying mechanisms for the development of hypertension, the level of p-MLC-Ser¹⁹ in VSMCs is the final determinant of VSMC contraction and the development of hypertension.12 Furthermore, coronary artery spasm, characterized by severe vasoconstriction of the coronary arteries leading to complete or nearly complete vessel blockage, is a significant contributor to myocardial ischemic syndromes, such as stable and unstable angina, acute myocardial infarction, and sudden cardiac death.43 Post-SAH vasospasm is a significant complication that occurs after bleeding into the space surrounding the brain, the subarachnoid space.19 Similar to excessive VSMC contraction in the development of hypertension, tonic VSMC contraction and the subsequent development of pathological vasospasms are attributable to the elevation of p-MLC-Ser¹⁹ by the activation of the Ca²⁺-dependent MLCK pathway and/or Ca²⁺-independent RhoA signaling pathway.34,44 Based on these reports and our current results showing that VPA inhibited p-MLC-Ser¹⁹ and aortic vessel contraction by decreasing RhoA expression (Figs. 1, 2, and 4), VPA may be useful for treating pathological vasospasms, including coronary artery and post-SAH vasospasms.

VPA has been clinically used to treat many neurological diseases, such as epilepsy, bipolar disorder, and intractable headaches.1,45 Although the therapeutic plasma concentration of VPA is 50–100 mg/L (0.35–0.70 mM) in patients receiving VPA medication for treating these neurological diseases,46 higher concentrations (up to 5 mM) exhibit other beneficial effects, including anti-cancer and anti-hypertensive effects in several types of cells and mice.7,33,47 Similar to these results, the current study showed that VPA decreased p-MLC-Ser¹⁹ in a dose-dependent manner (Fig. 1A) and RhoA expression (Fig. 2A and C), with the greatest effects observed at 4 mM VPA. Furthermore, it has been reported the peak plasma concentration of VPA in patients can increase to 1,414 mg/L (9.9 mM).48 Based on these previous and current findings, VPA concentrations used in the current experiments (0–4 mM) seem compatible and comparable to those of various in vitro and clinical studies. In support this notion, ex vivo results using isolated aortic vessels clearly showed that treatment with 2 mM of VPA sufficiently inhibited PE-induced vessel contraction and p-MLC-Ser¹⁹ and RhoA expression in aortic tissues (Fig. 4). Furthermore, these results showed that VPA decreased VSMC contractility and vessel contraction at the same concentration (2 mM) in vitro and ex vivo by inhibiting RhoA expression, suggesting that the inhibitory action of VPA on VSMC contraction is possibly mediated by similar mechanisms in vitro, ex vivo, and perhaps in vivo, and its relevance to physiological function.

In conclusion, the current results demonstrate that VPA inhibits RhoA-mediated VSMC contractility and aortic vessel contraction by decreasing RhoA expression through the inhibitory action of VPA on HDACs. These results suggest that VPA may be useful in the treatment and prevention of vasospastic diseases, including coronary and post-SAH vasospasms.

SUPPLEMENTARY MATERIALS

Supplementary Fig. 1

VPA does not change the expressions of MLCK, p-MYPT-Thr⁶⁹⁶, MYPT, and PP1β in rat VSMCs.

Supplementary Fig. 2

VPA does not alter ROCK1 expression and total ROCK activity in rat VSMCs. (A) VSMCs were treated with various doses of VPA (0, 1, 2, or 4 mM) for 24 hours and followed by stimulation with 1 μM PE for 10 minutes, and then levels of ROCK1 and actin expression were detected using western blotting. (B) VSMCs were treated with 2 mM VPA for various duration (0, 3, 8, 24 hours), followed by stimulation with 1 μM PE for 10 minutes, and then levels of ROCK1 and actin expression were detected using western blotting. (C) VSMCs were treated with various doses of VPA (0, 1, 2, or 4 mM) for 24 hours, followed by stimulation with 1 μM PE for 10 minutes. Total ROCK activity was then measured colorimetrically by reading absorbance at 450 nm, and the results were normalized to the protein concentration and expressed as a percentage of the control. All experiments were performed independently at least 4 times, and the blots shown are representative of these experiments (n = 4). Bar graphs depict the mean above/below the control (± standard deviation).

Supplementary Fig. 3

VPA increases the levels of H3K9ac and H3K27ac, as well as H3K9ac/K14ac in rat VSMCs. VSMCs were treated with various doses of VPA (0, 1, 2, or 4 mM) for 24 hours and followed by stimulation with 1 μM PE for 10 minutes, and then levels of H3K9ac, H3K27ac, H3K9ac/K14ac, and actin expression were detected using western blotting. All experiments were performed independently at least four times, and the blots shown are representative of these experiments (n = 4). Bar graphs depict the mean above the control (± standard deviation).

Supplementary Fig. 4

HDAC5 is not involved in the VPA-induced inhibition of RhoA and p-MLC-Ser¹⁹ expression in rat VSMCs. (A) After VSMCs were transfected with the WT-HDAC5 gene or empty vector and treated with 2 mM VPA for 24 hours, followed by stimulation with 1 μM PE for 10 minutes, levels of H3K27ac, H3K9/K14ac, HDAC5, and actin expression were detected using western blotting. (B) After VSMCs were treated as described above, levels of H3K9ac, HDAC5, and actin expression were detected using western blotting. (C) After VSMCs were treated as described above, levels of RhoA, HDAC5, and actin expression were detected using western blotting. (D) After VSMCs were treated as described above, levels of p-MLC-Ser¹⁹, MLC, HDAC5, and actin expression were detected using western blotting. All experiments were performed independently at least four times, and the blots shown are representative of these experiments (n = 4). Bar graphs depict the mean above/below the control (± standard deviation).