Omega-3 and omega-6 DPA equally inhibit the sphingosylphosphorylcholine-induced Ca2+-sensitization of vascular smooth muscle contraction via inhibiting Rho-kinase activation and translocation

We previously reported that eicosapentaenoic acid (EPA), an omega-3 polyunsaturated fatty acid (n-3 PUFA), effectively inhibits sphingosylphosphorylcholine (SPC)-induced Ca2+-sensitization of vascular smooth muscle (VSM) contraction which is a major cause of cardiovascular and cerebrovascular vasospasm, and EPA is utilized clinically to prevent cerebrovascular vasospasm. In this study, we clearly demonstrate that docosapentaenoic acid (DPA), which exists in two forms as omega-3 (n-3) and omega-6 (n-6) PUFA, strongly inhibits SPC-induced contraction in VSM tissue and human coronary artery smooth muscle cells (CASMCs), with little effect on Ca2+-dependent contraction. Furthermore, n-3 and n-6 DPA inhibited the activation and translocation of Rho-kinase from cytosol to cell membrane. Additionally, SPC-induced phosphorylation of myosin light chain (MLC) was inhibited in n-3 and n-6 DPA pretreated smooth muscleVSM cells and tissues. In summary, we provide direct evidence that n-3 and n-6 DPA effectively equally inhibits SPC-induced contraction by inhibiting Rho-kinase activation and translocation to the cell membrane.


Effects of n-3 and n-6 DPA on SPC and 40 mM K + depolarization-induced contraction of VSM.
The chemical structures of n-3 and n-6 DPA are shown in Supplementary Fig. S1. First, we observed the direct effects of DPA on SPC-induced contraction in tissue strips. Both n-3 and n-6 DPA (60 μ M) strongly inhibited SPC-induced Ca 2+ -sensitization of VSM contraction by 74.03 ± 6.42% and 86.87 ± 3.74%, respectively, (Fig. 1a,c), compared with vehicle control (see Supplementary Fig. S2). The inhibitory effects of n-3 and n-6 DPA on SPCinduced contraction were dependent on concentration (Fig. 1e,f). In contrast, n-3 and n-6 DPA (60 μ M) showed little inhibitory effect (11.48 ± 8.41% and 14.57 ± 9.71%, respectively) in 40 mM K + depolarization-induced contraction of VSM (Fig. 1b,d) compared with vehicle control (see Supplementary Fig. S2). Next, we investigated the effects of DPA pretreatment on SPC-induced contraction in tissue strips. After pre-treating the tissue strips with DPA for 30 min, SPC was added to the media to stimulate contraction. SPC stimulation caused a minor contraction in tissue strips pretreated with 60 μ M n-3 and n-6 DPA compared with strong contraction in vehicle control (Fig. 2a-c), showing that pretreatment with n-3 and n-6 DPA significantly inhibited SPC-induced contraction by 73.00 ± 2.31% and 95.00 ± 2.89%, respectively (Fig. 2f). Pretreatment with DPA did not affect the 40 mM K + depolarization-induced contraction of VSM (Fig. 2d,e). We also observed that the inhibitory effect of DPA is irreversible when we applied SPC (without DPA) to the media after we observed a direct inhibitory effect on SPCinduced contraction by DPA after washout for 30 minutes (Fig. 1g,h).

Effects of n-3 and n-6 DPA on [Ca 2+
] i level during SPC and 40 mM K + depolarization-induced contraction of VSM. Because n-3 and n-6 DPA slightly inhibited the 40 mM K + depolarization-induced contraction of VSM, we wished to assess whether they could affect the levels of [Ca 2+ ]i. To investigate whether n-3 and n-6 DPA alter the levels of [Ca 2+ ]i, we simultaneously assessed changes in force and [Ca 2+ ]i. As shown in Fig. 3c,e, SPC induced the contraction of VSM with little or no change in the concentration of Ca 2+ compared with vehicle control (Fig. 3a), which was consistent with a previous report 6,8 . Additionally, n-3 and n-6 DPA inhibited the SPC-induced contraction of VSM without changing the concentration of Ca 2+ compared with that of vehicle control-treated VSM (Fig. 3a), suggesting that these inhibitory effects on SPC-induced contraction of VSM were Ca 2+ -independent. Conversely, n-3 and n-6 DPA slightly inhibited the 40 mM K + depolarization-induced contraction of VSM and the level of [Ca 2+ ]i (Fig. 3d,f) compared with vehicle control-treated VSM (Fig. 3b). Based on these results, the slight inhibition of the [Ca 2+ ]i level by n-3 or n-6 DPA could account for their inhibition of the 40 mM K + depolarization-induced contraction of VSM. However, the inhibitory effects of n-3 and n-6 DPA on SPC-induced contraction of VSM are Ca 2+ -independent.

Effects of direct pretreatment with n-3 and n-6 DPA on VSM cell contraction induced by SPC.
To further investigate the mechanism of n-3 and n-6 DPA on SPC-induced VSM contraction, we used cultured coronary artery smooth muscle cells (CASMCs). First, we examined the effects of direct pretreatment with n-3 and n-6 DPA on SPC-induced contraction of CASMCs by recording the time course of changes in CASMC cell morphology. Based on previous reports 32, 33 , after the cells were 80-90% confluent, the cultures were switched to serum-free and growth factor-free media. After 2 days, a spindle-shaped, contractile type of cells was observed (see Supplementary Fig. S3). Time-lapse observations recorded that a contraction in CASMCs was induced by SPC in the absence and presence of n-3 or n-6 DPA. As shown in Fig. 4, SPC induced a noticeable contraction in CASMCs, with the elongated cells becoming short or round (see Supplementary Video 1). In contrast, CASMCs pretreated with n-3 or n-6 DPA showed little change in morphology after being stimulated with SPC for 30 min (see Supplementary Videos 2 and 3).
We then used the ArrayScan V TI High Content Screening Reader to analyze and quantify changes in cellular area and length induced by SPC-stimulation in CASMCs pretreated with DPA or vehicle control. As shown in Fig. 5, after stimulation with SPC, vehicle control-treated CASMCs began to contract. The area and length of CASMCs decreased to 72.7 ± 7.56% and 85.1 ± 0.69% compared with those of no SPC-stimulated cells (100%) during the 45 min stimulation with SPC (Fig. 5b,c,f). However, the area and length of cells pretreated with n-3 or n-6 DPA did not change during the 45 min stimulation with SPC, indicating that n-3 or n-6 DPA inhibited SPC-induced contraction in CASMCs (Fig. 5b,d  contraction by suppressing Rho-kinase translocation remains unclear. Consistent with our previous report, SPC induced the translocation of Rho-kinase from cytosol to the periphery of the cell as measured by immunofluorescent staining. In contrast, Rho-kinase translocation was not observed in CASMCs pretreated with n-3 or n-6 DPA (Fig. 6a,b), suggesting that n-3 and n-6 DPA inhibited the SPC-induced translocation of Rho-kinase from cytosol to the cell membrane.

n-3 and n-6 DPA inhibited Rho-kinase activity induced by SPC in VSM tissues and cells.
Although n-3 and n-6 DPA inhibited SPC-induced Rho-kinase translocation, it was yet unclear whether they inhibited the SPC-induced activation of Rho-kinase. To examine if n-3 and n-6 DPA inhibited VSM contraction by preventing SPC from activating Rho-kinase, we measured threonine phosphorylation on site 853 of MYPT1 (MYPT1 Thr853) as a readout of Rho-kinase activity 34,35 . After stimulation with SPC, Rho-kinase activity in VSM tissues and cells increased 4.57 ± 0.82 and 1.99 ± 0.11 fold, respectively, compared with that in controls (Fig. 7). In contrast, n-3 DPA and n-6 DPA suppressed SPC-induced phosphorylation of MYPT1 Thr853 to 1.69 ± 0.42 and 1.06 ± 0.05, respectively (Fig. 7b), consistent with n-3 and n-6 DPA inhibition of SPC-induced activation of Rho-kinase. Similarly, MYPT1 Thr853 phosphorylation in VSM cells pretreated with n-3 and n-6 DPA decreased to 1.27 ± 0.11 and 1.12 ± 0.15, respectively (Fig. 7c,d).

n-3 and n-6 DPA inhibited MLC phosphorylation induced by SPC in VSM tissue and cells. It is
known that a contraction of smooth muscle is produced when the 20-kDa myosin light chain (MLC) is phosphorylated at Ser19 [36][37][38] . Activated Rho-kinase can also phosphorylate MLC at Ser19 to increase the activity of myosin ATPase in vitro 39  phosphorylation. Indeed, SPC-induced phosphorylation of MLC in smooth muscle tissue pretreated with n-3 or n-6 DPA was significantly lower (1.14 ± 0.22 and 0.68 ± 0.35, respectively) compared with that in vehicle control-treated tissue (2.68 ± 0.63) (Fig. 8a,b). Similarly, both n-3 and n-6 DPA significantly inhibited phosphorylation of MLC in smooth muscle cells (Fig. 8c,d).

Discussion
n-3 PUFAs are thought to have beneficial effects on prevention of cardiovascular diseases, while n-6 PUFAs are considered proinflammatory and thought to negatively affect the cardiovasculature. In this study, we found that both n-3 and n-6 DPA inhibited SPC-induced Ca 2+ -sensitization of VSM contraction by inhibiting the activity and translocation of Rho-kinase. To the best of our knowledge, this study is the first to show that DPA could specifically inhibit SPC-induced contraction without changing Ca 2+ concentration. Here, we suggest that including n-6 PUFAs in the diet may be beneficial for the prevention of cardiovascular diseases arising from Ca 2+ -sensitization-mediated contraction of VSM such as in coronary artery and cerebral vasospasm. However, the reported proinflammatory effects of n-6 PUFAs need to be further investigated.
Abnormal contraction induced by SPC mediated Ca 2+ -sensitization of VSM, has been proposed as a major cause of cardio and cerebrovascular diseases such as coronary artery and cerebral vasospasm 42,43 . In this study, n-3 and n-6 DPA showed a much stronger inhibition of SPC-induced contraction than did the 40 mM K +   a a a a a a a a a a a a a a a a a a   depolarization-induced contraction of VSM, suggesting that n-3 and n-6 DPA effectively inhibited SPC-induced abnormal contraction of VSM with little effect on Ca 2+ -dependent contraction. Furthermore, simultaneous measurement of force and [Ca 2+ ]i showed that inhibitory effects of n-3 and n-6 DPA on SPC-induced contraction were Ca 2+ -independent, indicating that n-3 and n-6 DPA specifically inhibited the SPC-induced abnormal contraction of VSM. n-3 and n-6 DPA slightly inhibited the 40 mM K + depolarization-induced contraction of VSM and [Ca 2+ ]i, indicating that their inhibition of [Ca 2+ ]i occurs by an unknown mechanism that is involved in inhibition of 40 mM K + depolarization-induced contraction of VSM. Previously we showed that EPA, an n-3 PUFA, inhibited SPC-induced Ca 2+ -sensitization 8,10 . Thus, it is not surprising that n-3 DPA effectively inhibited the SPC-induced Ca 2+ -sensitization of VSM contraction because n-3 DPA is structurally similar to n-3 EPA. Surprisingly, the n-6 DPA, an omega-6 form of DPA, inhibited the SPC-induced abnormal contraction of VSM, providing direct evidence that n-6 PUFAs are possibly beneficial for the prevention of cardiovascular diseases.
Next, we examined the mechanisms by which n-3 and n-6 DPA inhibit SPC-induced contraction. We previously demonstrated that activation and translocation of Rho-kinase to the cell membrane are involved in SPC-induced contraction 7 . In this study, our results showed that the SPC-induced translocation of Rho-kinase to the cell membrane and its subsequent activation were significantly inhibited in cells pretreated with n-3 or n-6 DPA. Although n-3 and n-6 DPA are not shown to directly inhibit Rho-kinase activity in an in vitro assay, they suppressed SPC-induced Rho-kinase translocation and MYPT1 phosphorylation on the site of Thr853. Additionally, our   previous study demonstrated that the SPC-induced activation of Rho-kinase is inhibited by the specific Rho-kinase inhibitor of Y27632 44 . These results suggest that SPC-induced Rho-kinase activation is inhibited by DPA through indirect suppression of Rho-kinase activity. Furthermore, SPC-induced MLC phosphorylation was significantly inhibited in both smooth muscle tissue and CASMCs pretreated with n-3 or n-6 DPA. We concluded that n-3 and n-6 DPA inhibit SPC-induced Ca 2+ -sensitization of VSM by inhibiting the activity and translocation of Rho-kinase.
Previously, we demonstrated that EPA inhibited the translocation of Fyn, a member of the Src family of tyrosine kinases (SrcPTK), which is involved in SPC-induced Ca 2+ -sensitization of VSM contraction 27 . Hence, we examined whether DPA could inhibit the translocation of Fyn, and found that DPA could not inhibit its translocation (see Supplementary Fig. S4). Although the structures of DPA and EPA are similar, the mechanisms by which they inhibit SPC-induced Ca 2+ -sensitization of VSM contraction are different. EPA is a carboxylic acid with a 20-carbon chain, while DPA has a 22-carbon chain (see Supplementary Fig. S1). The locations of the double bonds in these two compounds are also different. Further studies are needed to clarify the molecular basis of this difference.
n-3 PUFAs are metabolized in vivo by desaturation and elongation enzymes to form DPA and DHA, which are the major final products of EPA. n-3 DPA is formed by chain elongation of EPA performed by elongase-2 and -5, and DPA is converted to DHA by the activity of desaturase and β -oxidation 45 . n-6 DPA is the final product in the n-6 PUFA metabolism pathway. Morisaki et al. reported that smooth muscle cells have strong elongation activity and little or no desaturase activity 46 , suggesting that DPA may not be converted into DHA in smooth muscle. Another two studies also provide evidence that supplemented n-3 DPA shows no increase in DHA in endothelial cells 47 and in primary hepatocytes 48 within a period of 24 h. Additionally, Dyerberg et al. reported that administering linoleic acid (LLA) orally to humans for 7 days was not long enough to cause significant increases in the plasma levels of EPA and DHA 49 , suggesting that the metabolism of PUFAs in vivo is a very slow process. Therefore, we speculate that in the present study, there is little or no possibility of converting DPA to other vasoactive metabolites in smooth muscle tissue or cells, especially over a short period (treatment with DPA for 30 min).
In conclusion, the results of our present study demonstrated that two forms of DPA (n-3 and n-6) inhibit SPC-induced Ca 2+ -sensitization of VSM contraction. This inhibition occurs via suppression of SPC-induced Rho-kinase activation and translocation to the cell membrane, and subsequent inhibition of MLC phosphorylation (Fig. 9). Importantly, we also found that n-6 DPA effectively inhibits SPC-induced Ca 2+ -sensitization of VSM contraction. This study may provide new insights for the prevention of cardiovascular and cerebrovascular vasospasm by n-6 PUFAs.
Cell culture. Human coronary artery smooth muscle cells (CASMCs) were cultured in HuMedia SG2 (Kurabo, Japan) supplemented with 5% fetal bovine serum (FBS), 0.5 ng/ml human epidermal growth factor (hEGF), 2 ng/ml human fibroblast growth factor-B (hFGF-B), 5 μ g/ml insulin, 50 μ g/ml gentamycin, and 50 ng/ml amphotericin B. Cells were maintained at 37 °C in a humidified atmosphere of 5% CO 2 and 95% air. Human CASMCs were used for experiments at passage numbers < 10 splitting cycles. Preparation of VSM strips. All the procedures were subject to approval by the Institutional Animal Care and Use Committee of Yamaguchi University and were conducted in conformity with institutional guidelines. Porcine left anterior descending arteries were obtained from a public abattoir (Kitakyushu Municipal Meat Inspection and Control Center, Japan). Tissue specimensof coronary arteries were placed into ice-cold Krebs solution (123 mM NaCl, 4.7 mM KCl, 15.5 mM NaHCO 3 , 1.2 mM KH 2 PO 4 , 1.2 mM MgCl 2 , 1.25 mM CaCl 2 , and 11.5 mM D-glucose) for transport to our laboratory. An optical microscope was used to view fat tissue and adventitia, which were removed with minimum stretch to VSM. The arteries were cut into helical strips lacking endothelium. The complete removal of the endothelium from strips was confirmed by the lack of any relaxation response to 1 μ M BK.

Measurement of force in VSM.
Measurements of isometric force in VSM strips (1.0 × 4.0 mm) were carried out as described previously 32,33 . Briefly, after recording steady responses to repeated applications of 118 mM K + -depolarization, an injection of 30 μ M SPC or 40 mM K + was used to induce a contraction of VSM. Then, the effects of n-3 or n-6 DPA were examined at the plateau phase of the sustained contraction induced by 30 μ M SPC or 40 mM K + . The inhibitory extent of contraction induced by n-3 or n-6 DPA was described as a percentage of the response to the contraction induced by 30 μ M SPC or 40 mM K + (see Supplementary Fig. S2).

Simultaneous Measurement of [Ca 2+ ]i and contraction of VSM. We measured the changes in [Ca 2+ ]
i and force simultaneously using porcine coronary arterial strips as described previously 7 . Briefly, strips of porcine coronary arteries were loaded with 12.5 μ M Fura-2/AM for 4 h at 37 °C. Then, changes in [Ca 2+ ]i were continuously monitored using a spectrofluorometer (CAM230, JASCO) to assess fluorescence intensity at excitation of 340 nm (F340) and 380 nm (F380) as well as the fluorescence ratio (r = F340/F380).

Immunofluorescent staining.
For immunofluorescence studies, human CASMCs were cultured on 0.3% gelatin-coated glass coverslips. After stimulation with SPC, cells were fixed in 2% paraformaldehyde in phosphate buffered saline (PBS) for 10 min, permeabilized with 0.1% Triton X-100 in PBS for 2 min, and blocked in NanoBio blocker solution (Nano Bio Tech Co., Ltd) diluted in PBS for 60 min at room temperature. Cells were then incubated with the primary anti-Rho-kinase-2 antibody for 60 min. The antigen-antibody complex was visualized with Alexa Fluor 488-conjugated secondary antibody (Thermo Fisher Scientific). Coverslips were washed with PBS, rinsed with deionized water, and mounted with PermaFluor aqueous mounting medium (Thermo Fisher Scientific). Specimens were observed using fluorescent microscopy (KEYENCE, Japan).  (Funakoshi, Japan). Protein was extracted from freeze-dried tissues by the addition of 100 μ l RIPA buffer (Wako) and protease inhibitor cocktail (Sigma), and samples were kept at 95 °C for 5 min for western blot analysis.

Western blotting analysis.
Cell lysates were prepared as follows. Human CASMCs were serum starved for 24 h, then stimulated with SPC (30 μ M for 15 min) with pretreatment with n-3 or n-6 DPA (60 μ M for 30 min) or vehicle control. Cells were lysed in lysis buffer (50 mM Tris-HCl, pH 7.4; 150 mM NaCl; 0.5% NP-40; 0.1% Triton X-100; 1 mM DTT; 0.2 mM Na 3 VO 4 ; 10 mM NaF; 1 mg/ml aprotinin; and protease inhibitor cocktail (Sigma)). The lysates were centrifuged, and the supernatants were boiled in sodium dodecyl sulfate (SDS) loading buffer. Tissue or cell samples were separated using 10% SDS-polyacrylamide gel electrophoresis, transferred to Amersham Hybond-P PVDF membrane (GE Healthcare Life Sciences), and blocked in 5% non-fat milk in 0.05% Tween-20 phosphate-buffered saline (TBS-T) for 60 min at room temperature. The membranes were incubated with primary antibodies for 1 h at room temperature or overnight at 4 °C, followed by incubation with peroxidase-conjugated secondary antibodies for 1 h at room temperature. Immunoreactive bands were detected using a Supersignal West Pico Chemiluminescent Substrate kit (Thermo Fisher Scientific). Glyceraldehyde 3 phosphate dehydrogenase (GAPDH) was used as loading control.
Time-lapse recording of VSM cell contraction. Human CASMCs were grown in 35-mm glass-based dishes (Iwaki, Japan). When cell confluence reached 90-100%, FBS and growth factor-free HuMedia SB2 (Kurabo, Japan) medium was changed to obtain the hypercontractile type of CASMCs according to previously existing methods 34,35 . After treatment with HuMedia SB2 for 48 h, cells were pretreated with n-3 or n-6 DPA (60 μ M) for 30 min at 37 °C. Then, 30 μ M SPC was added to the medium and time-lapse recording of VSM cell contraction was performed under the fluorescent microscope (KEYENCE, Japan).
Automated image capture and analysis. Human CASMCs were seeded in 96-well plates. Before stimulation with SPC, media were removed and cells were stained with 5-chloromethylfluorescein diacetate (Cell-Tracker Green CMFDA; Molecular Probe) at a final concentration of 1 μ M in serum-free HuMedia SB2 (Kurabo, Japan). After 30 min incubation at 37 °C, the dye was removed, replaced with fresh, pre-warmed HuMedia SB2, and incubated for 30 min at 37 °C. After pretreatment with DPA (60 μ M for 30 min at 37 °C), cells were stimulated with SPC. Then, cells were fixed in 7.4% formaldehyde (Sigma Aldrich) in PBS containing Hoechst 33258 nuclear stain (Thermo Fisher Scientific) for 15 min at room temperature. The fixation buffer was removed and replaced with 100 μ l PBS. Plate analysis and image acquisition were performed immediately using ArrayScan V TI High Content Screening Reader (Thermo Fisher Scientific, USA). Images were acquired in two channels: channel 1 for the cytoplasmic chloromethyl fluorescein diacetate (CMFDA) stain and channel 2 for the Hoechst nuclear stain. The parameters set for nuclei recognition ensured that any artifacts or cell clusters were absent in the populations analyzed. The images were automatically stored and analyzed by the ArrayScan ™ Software; at least five fields were analyzed in each well with a 10X objective, corresponding to at least 1,000 cells counted.
Statistical Analysis. All data are expressed as mean ± SEM. Differences between groups were analyzed by Student's t-test. Values of P < 0.05 were considered to represent a statistically significant difference.