Enhancement by Hydrogen Peroxide of Calcium Signals in Endothelial Cells Induced by 5-HT1B and 5-HT2B Receptor Agonists

Hydrogen peroxide, formed in the endothelium, acts as a factor contributing to the relaxation of blood vessels. The reason for this vasodilatory effect could be modulation by H2O2 of calcium metabolism, since mobilization of calcium ions in endothelial cells is a trigger of endothelium-dependent relaxation. The aim of this work was to investigate the influence of H2O2 on the effects of Ca2+-mobilizing agonists in human umbilical vein endothelial cells (HUVEC). We have found that H2O2 in concentration range 10-100 μM increases the rise of [Ca2+]i induced by 5-hydroxytryptamine (5-HT) and carbachol and does not affect the calcium signals of ATP, agonist of type 1 protease-activated receptor SFLLRN, histamine and bradykinin. Using specific agonists of 5-HT1B and 5-HT2B receptors CGS12066B and BW723C86, we have demonstrated that H2O2 potentiates the effects mediated by these types of 5-HT receptors. Potentiation of the effect of BW723C86 can be produced by the induction of endogenous oxidative stress in HUVEC. We have shown that the activation of 5-HT2B receptor by BW723C86 causes production of reactive oxygen species (ROS). Inhibitor of NADPH oxidases VAS2870 suppressed formation of ROS and partially inhibited [Ca2+]i rise induced by BW723C86. Thus, it can be assumed that vasorelaxation induced by endogenous H2O2 in endothelial cells partially occurs due to the potentiation of the agonist-induced calcium signaling.


Introduction
In the early 1930s, the phenomenon of oxidative (respiratory) burst at phagocytosis was described [1]. This work initiated enormous number of studies of the role of ROS in biological processes and systems. High concentrations of ROS normally are characteristic function of the so-called professional phagocytes-cells of innate immunity; in other cells, high concentration of ROS is a sign of oxidative stress and a cause of cell death [2]. Low concentrations of ROS are permanently formed in almost all cells of the body and perform the functions of second messengers in redox-sensitive signaling pathways [3,4]. The vascular endothelium plays a crucial role in maintaining homeostasis, and it is often both the main target and one of the sources of ROS [5,6]. In physiological conditions, the synthesis of ROS by NADPH oxidases (NOX) carries out a signaling function [7]; however, the disturbance of its metabolism and associated signaling pathways is the cause of many vascular pathologies [8]. In terms of signaling, H 2 O 2 is considered as the most stable and important kind of ROS. According to some data, the concentration of both endogenous and exogenous H 2 O 2 can reach hundred micromoles [9][10][11]. In vascular endothelial cells H 2 O 2 is produced by NOX4, and it was demonstrated that its targeted overexpression in endothelial cells results in a decrease in blood pressure and potentiation of relaxation under the action of acetylcholine and histamine [12]. It was assumed that the cause of vasodilation is hyperpolarization of the plasma membrane of smooth muscle cells induced by H 2 [13]. Recently, we demonstrated that H 2 O 2 in concentration range of 10 to 100 μM stimulated mobilization of Ca 2+ in HUVEC partially due to the activation of two-pore calcium channels [18]. Calcium ions released from endolysosomal vesicles via two-pore channels are operating as a trigger promoting Ca 2+ mobilization from more capacious intracellular calcium depots. The aim of this study was to estimate how H 2 O 2 influences upon the action of Ca 2+mobilizing agonists in HUVEC.

Cell
Culture. The endothelial cells were isolated from the human umbilical vein as described previously [19] with modifications [20]. The umbilical veins were washed with Hanks balanced salt solution containing antibiotics, filled with a M199 medium supplied with 0.1% collagenase (Sigma-Aldrich), and incubated at room temperature for 40 min. The cells were washed and grown in plastic vessels covered with 0.2% gelatin in M199 medium. The medium contained Earle's salts, 20% fetal bovine serum (Invitrogen, United States); 300 μg/mL endothelial growth supplement from the rabbit brain obtained according to [21]; 100 μg/mL heparin; and 100 μg/mL gentamycin. 2+ ] i were registered in HUVEC of 2-4 passages with the use of the fluorescent probe CalciumGreen (Thermo Fisher Scientific, USA) and a microplate spectrofluorometer Synergy 4 (BioTech, USA). The cells grown in 96-well plates were loaded with the probe at 37°C for 1 h in M199 medium, containing 1 μM CalciumGreen/AM and 100 μg/mL Pluronic F127. After that, the physiological salt solution (PSS, pH 7.4) containing 145 mM NaCl, 5 mM KCl, 10 mM Hepes, 1 mM MgCl 2 , 1 mM СaCl 2 , and 10 mM glucose was added to the cells. The cells were preliminarily washed with the PSS solution to remove the M199 medium. Fluorescence was registered at 485 nm (excitation) and 530 nm (emission) at room temperature. All data are represented as ratios of two values (ΔF/Fo): the increase in the fluorescence in response to an agonist (ΔF) and the basal fluorescence of the unstimulated cells (Fo). The plots show the averages of three or more measurements.

Detection of ROS in HUVEC. ROS generation in HUVEC
was assessed using 2 ′ ,7 ′ -dichlorodihydrofluorescein diacetate (DCFH-DA) as a probe. The cells grown in 96-well plates were incubated with 2 μM of DCFH-DA during 1 h at room temperature. The formation of ROS in the cells was determined by measuring the rate of DCFH oxidation into 2 ′ ,7 ′ -dichlorofluorescein (DCF). DCF fluorescence was measured at excitation wavelength of 485 ± 20 nm and with emitter bandpass of 527 ± 15 nm. The rate of fluorescence rise was taken as an index of ROS generation.
2.5. Statistics. Data are presented as mean ± SEM of 6-12 measurements. Statistical significance was calculated using Excel and MedCalc statistical software according the Student-Newman-Keuls test.

Results
In HUVEC, [Ca 2+ ] i is strongly increased in response to ADP, histamine, and the agonist of type 1 proteaseactivated receptor SFLLRN and to a lesser degree to bradykinin ( Figure 1(a)). The responses to these agonists are not affected by H 2 O 2 . 5-hydroxytryptamine and carbachol induced substantially less elevation of [Ca 2+ ] i in HUVEC, and H 2 O 2 caused a significant potentiation of their effects (Figure 1(a)). In some preparations of HUVEC, the rise of [Ca 2+ ] i in response to 5-HT was negligible, but in the presence of H 2 O 2 it increased severalfold ( Figure 1(b)). The reason for such strong potentiation of the [Ca 2+ ] i rise induced by 5-HT is not clear. We suppose that it might depend on the quality of the contacts between the cells in monolayer and the degree of synchronization of Ca 2+ responses of single cells. The synchronization of calcium signaling of endothelial cells in monolayer has been demonstrated earlier [22]. The increase in [Ca 2+ ] i in response to 5-HT and carbachol occurred when the H 2 O 2 concentration changed from 10 to 100 μM (Figure 2(a)). At 200 μM H 2 O 2 did not cause additional potentiation, while its own effect on the level of [Ca 2+ ] i was growing (Figure 2(b)). Catalase prevents potentiation by H 2 O 2 of 5-HT calcium signaling ( Figure S1).
The vascular endothelium expresses serotonin 5-HT1B and 5-HT2B receptors. It is known that serotonin or its selective agonists activate NO synthase and cause vascular relaxation [23,24]. In general, serotonin receptors of endothelial cells are very poorly investigated. There is no information in the literature on the participation of 5-HT1B receptors in the regulation of [Ca 2+ ] i in endothelial cells. Regarding 5-HT2B receptors, there is an evidence that their activation in human pulmonary artery endothelial cells causes elevation of [Ca 2+ ] i [25]. We investigated effects of 5-HT1B and 5-HT2B receptor agonists CGS12066B and BW723C86 in order to determine which of these receptors could be responsible for the enhancement of the Ca 2+ -mobilizing activity by H 2 O 2 ( Figure 3) Figure S3) as in the case of 5-HT (Figure 2(a)). The potentiating effect of H 2 O 2 is suppressed by the antioxidant N-acetylcysteine (Figures 4  and 5). In some HUVEC preparations, the calcium signals in response to CGS12066B and BW723C86 were quite strong. We suggested that this may be due to the formation of endogenous ROS, which might increase the calcium-mobilizing activity of 5-HT1B and 5-HT2B receptors in a manner similar to exogenous H 2 O 2 . In favor of this point of view, there is a data on the inhibition by N-acetylcysteine of [Ca 2+ ] i rise in response to CGS12066B and BW723C86 ( Figure 6). On the other hand, there was no suppression of the effect caused by the agonist of type 1 protease-activated receptor SFLLRN. These data assume the existence of a specific mechanism through which an increase in [Ca 2+ ] i occurs when 5-HT1B and 5-HT2B receptors are activated in HUVEC.
In order to test the hypothesis of an increase in the functional activity of 5-HT1B and 5-HT2B receptors under the action of endogenous ROS, we investigated whether their formation in HUVEC occurred in conditions of [Ca 2+ ] i measurement. For registration of the ROS, the fluorescent probe DCFH-DA penetrating into the cells was used. The increase in the fluorescence due to oxidation DCFH into 2 ′ ,7 ′ -dichlorofluorescein (DCF) reflects the accumulation of ROS in the cytoplasm. As can be seen in Figure 6(a), spontaneous formation of ROS occurs in HUVEC. In the presence of CGS12066B, the oxidation rate of the H2DCF does not change. We found that BW723C86 at a concentration of 100 μM increased the formation of ROS severalfold in HUVEC ( Figure 6). VAS2870, a selective inhibitor of NOX, almost completely blocked the formation of ROS induced by BW723C86 (Figure 6(b)). Thus, the data suggest the coupling of 5-HT2B receptors with NOX. The next task was to find out whether endogenous ROS generated by NOX can affect the calcium signaling in HUVEC. As shown in Figure 7(a), the NOX inhibitor VAS2870 does not change the response to 30 μM BW7323C86 and decreases by 15 4 + 3 8% (n = 12; p < 0 01 ) the [Ca 2+ ] i elevation caused by BW723C86 at a concentration of 100 μM. This indicates the contribution of NOX to the regulation of calcium signaling from 5-HT2B receptors. To evaluate the role of ROS endogenously formed in HUVEC in [Ca 2+ ] i regulation via 5-HT2B receptors, we used another approach: inducers of ROS formation were applied. Such properties are possessed by orthovanadate (Na 3 VO 4 ) [26] and substance BVT948 [27]. Na 3 VO 4 and BVT948 potentiated the rise of [Ca 2+ ] i in HUVEC caused by 30 μM BW723C86 by 1.7 and 2.5 times (Figure 7), respectively.

Discussion
In this paper, we have demonstrated that hydrogen peroxide dramatically increases the rise of [Ca 2+ ] i in endothelial cells in response to serotonin. It was found that the reason for this could be an increase in the functional activity of 5-HT1B and 5-HT2B receptors. To the best of our knowledge, this is the first demonstration of Ca-mobilizing activity of 5-HT1B receptors in endothelial cells. There is only one experimental work demonstrating that 5-HT2B receptors stimulate calcium mobilization in endothelial cells, which was conducted on the cells isolated from the human pulmonary artery [25]. The reason for the lack of published data on this topic seems to be the low functional activity of 5-HT1B and 5-HT2B receptors. Earlier we showed that in smooth muscle cells of blood vessels 5-HT2B receptors are functionally inactive and their transition to the active state occurs under the influence of ROS [28]. The results presented in this paper show that the same pattern of 5-HT receptor regulation exists in HUVEC. The increase in calcium signaling upon exposure to H 2 O 2 is selective with respect to the action of serotonin. There is an increase in the level of [Ca 2+ ] i in response to carbachol, though it is extremely low even in the presence of H 2 O 2 . The difference in the effect of ROS on the functional properties of 5-HT receptors, on the one hand, and histamine, ADP, and PAR1 receptors, on the other hand, is indicative of differences in the mechanisms of signal transduction. In the paper of Ullmer et al. [25], data were presented according to which 5-HT2B receptors induce the activation of ryanodine-sensitive channels in human pulmonary endothelial cells. We have shown that the main source of increasing [Ca 2+ ] i in HUVEC in response to the activation of 5-HT1B and 5-HT2B receptors could certainly be intracellular stores, but their nature has not yet been established.
In many cases, H 2 O 2 initially causes a transient rise of [Ca 2+ ] i , but its potentiating effect can be manifested without an apparent increase in the level of [Ca 2+ ] i in HUVEC. In a recent paper, we showed that H 2 O 2 activated two-pore calcium channels in HUVEC [18]. It was assumed that calcium ions released from endolysosomal vesicles through the two-pore channels serve as a trigger opening of the endo/sarcoplasmic reticulum channels. Perhaps this mechanism is implemented in this case, too.
Presently, it is a well-known fact that H 2 O 2 at low nontoxic concentrations can serve as a second messenger. The enhancement of [Ca 2+ ] i is one of the terminal links in serotonergic signaling. The primary targets of H 2 O 2 are known to be SH groups of PTEN and tyrosine phosphatases [29]. Moreover, tyrosine phosphatases are the target for different inducers of ROS formation, such as Na 3 VO 4 and BVT946 [27,30]. These substances have the same effect as H 2 O 2 , so there is a reason to think that the tyrosine phosphorylation system is involved in the serotonergic regulation of [Ca 2+ ] i in HUVEC. The involvement of tyrosine protein phosphatases and protein kinases of the Src family in calcium signaling of 5-HT2B receptors was shown in vascular smooth muscle cells [28].
In this work, we have shown that the activation of 5-HT2B receptor by BW723C86 induces ROS accumulation in the cytoplasm of HUVEC. Earlier, we showed the same effect of BW723C86 in rat aorta smooth muscle cells [28]. The formation of ROS in both cell types is suppressed specifically by the NADPH oxidase inhibitor VAS2870. This suggests the coupling of 5-HT2B receptors with this enzyme. The question about the type of NOX interacting with the 5-HT2B receptors requires further investigation.
Based on the data obtained, it might be suggested that the excessive production of H 2 O 2 by leukocytes as well as the formation of endogenous ROS in endothelial cells could influence the serotoninergic regulation of vascular tone and secretion from endothelial cells. Serotonin stimulates von Willebrand factor secretion via 5-HT1B receptors and the mechanism of this process is not completely clear [31]. Oxidative stress and simultaneous release of 5-HT from platelets could be a pathogenic factor of microvascular thrombosis. The stimulation of 5-HT2B receptors in endothelial progenitors by nordexfenfluramine and some other pharmacological agents causes valvular heart disease [32]. The potentiation of calcium response via 5-HT2B receptors in progenitors by H 2 O 2 can probably accelerate its development. On the other hand, the moderate formation of H 2 O 2 in endothelial cells can cause a physiologically relevant increase of [Ca 2+ ] i and promote vascular relaxation. Future work will be required to uncover whether H 2 O 2 potentiates vasorelaxation via 5-HT receptors.

Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
The authors confirm that there are no conflicts of interest.

Acknowledgments
These studies were supported by the Russian Science Foundation (grant no. 18-15-00417).  presence of 100 μM EGTA. The average of 3 parallel measurements ± SEM is presented. Figure S3: