Nucleotide transmitters ATP and ADP mediate intercellular calcium wave communication via P2Y12/13 receptors among BV-2 microglia

Nerve injury is accompanied by a liberation of diverse nucleotides, some of which act as ‘find/eat-me’ signals in mediating neuron-glial interplay. Intercellular Ca2+ wave (ICW) communication is the main approach by which glial cells interact and coordinate with each other to execute immune defense. However, the detailed mechanisms on how these nucleotides participate in ICW communication remain largely unclear. In the present work, we employed a mechanical stimulus to an individual BV-2 microglia to simulate localized injury. Remarkable ICW propagation was observed no matter whether calcium was in the environment or not. Apyrase (ATP/ADP-hydrolyzing enzyme), suramin (broad-spectrum P2 receptor antagonist), 2-APB (IP3 receptor blocker) and thapsigargin (endoplasmic reticulum calcium pump inhibitor) potently inhibited these ICWs, respectively, indicating the dependence of nucleotide signals and P2Y receptors. Then, we detected the involvement of five naturally occurring nucleotides (ATP, ADP, UTP, UDP and UDP-glucose) by desensitizing receptors. Results showed that desensitization with ATP and ADP could block ICW propagation in a dose-dependent manner, whereas other nucleotides had little effect. Meanwhile, the expression of P2Y receptors in BV-2 microglia was identified and their contributions were analyzed, from which we suggested P2Y12/13 receptors activation mostly contributed to ICWs. Besides, we estimated that extracellular ATP and ADP concentration sensed by BV-2 microglia was about 0.3 μM during ICWs by analyzing calcium dynamic characteristics. Taken together, these results demonstrated that the nucleotides ATP and ADP were predominant signal transmitters in mechanical stimulation-induced ICW communication through acting on P2Y12/13 receptors in BV-2 microglia.


Introduction
The intracellular roles of nucleotides (ATP, ADP, UTP, etc.) have long been recognized as energy currency or building blocks for DNA and RNA. However, in the extracellular compartment, mounting evidences reveal that they function as key signal molecules to participate in various processes including inflammation, infection and tissue damage through activation of membrane P2 nucleotide receptors [1,2]. For instance, during nerve injury, amount of ATP, ADP and UTP liberated from apoptotic neurons and acted as 'find-me' signal via activating P2Y receptors, resulting in oriented migration of microglia [3,4]. The 'eat-me' signal such as UDP will promote phagocytic clearance of dead cells and tissue debris by microglia through activating P2Y 6 receptor [5,6]. Moreover, extracellular nucleotide signaling by P1 and P2 receptors was found to regulate numerous inflammatory responses, which contributed to maintain the pro/anti-inflammatory balance of the central nervous system (CNS) [1,7].
In the CNS, microglia acted as the first and main immune defense against infectious and pathological events [8]. They could promptly survey changes of the surroundings and sense extracellular messengers via cell surface receptors [9], subsequently congregate together rapidly to build a protective barrier between normal and damaged tissues [10]. Intercellular Ca 2+ wave (ICW) communication was the principal approach by which microglia [11,12], astrocytes [13], and neurons [14] coordinate and synchronize with each other to execute immune defense and maintain homeostasis of CNS. ICWs could prefigure and guide the migration of microglia in response to neuronal damage [15].
Many reports showed that extracellular nucleotide ATP played as an important signal transmitter in ICWs [16][17][18]. However, it remains unclear whether and how other types of extracellular nucleotides contribute to ICW communication in microglia. In the present study, we evoked ICWs in an immortalized murine microglial cell line (BV-2) by localized micro-stimulation on single cell using glass microelectrodes. We investigated the involvement of potential intercellular messengers and corresponding receptors. Results provided compelling evidences that both ATP and ADP played pivotal roles in ICW communication via P2Y 12/13 receptors in BV-2 microglia, rather than other types of nucleotides.

Materials and methods
Cell culture BV-2 cell is an immortalized mouse microglial cell line that exhibits the morphological and functional characteristics of microglia [19]. The BV-2 cells, a generous gift from Dr. Linhua Jiang (Institute of Membrane and System Biology, Faculty of Biological Sciences, University of Leeds, Leeds, U.K.), were routinely cultured in Dulbecco's modified Eagle's medium (DMEM, Gibco, USA) supplemented with 10% (v/v) fetal bovine serum (FBS, Biological Industries, USA), 100 U/mL penicillin and 100 μg/mL streptomycin (Gibco, USA) at 37˚C under 5% CO 2 . Before test, cells were isolated with 0.25% trypsin + 0.04% EDTA (Gibco, USA) and plated on glass coverslips at a density of 2×10 4 cells/cm 2 to be adherent for 18 h. acquisition, the tip of the microelectrode was controlled to provoke an acute, short-lasting mechanical stimulation on target cells.

Ca 2+ imaging
Measurements of cytosolic calcium concentration ([Ca 2+ ] c ) in microglia was performed as described previously [11]. Briefly, BV-2 microglial cells were incubated with 3 μM calciumsensitive Fluo-4 AM (Invitrogen, USA) for 40 min at 37˚C in Hanks balanced salt solution (HBSS, 140 mM NaCl, 5 mM KCl, 2 mM CaCl 2 , 1 mM MgCl 2 Á6H 2 O, 10 mM glucose, 10 mM HEPES, pH 7.4) in the presence of 10% pluronic F-127 (Biotium, USA), subsequently washed for 10 min with HBSS for de-esterification. Ca 2+ dynamics was observed by the Axio observer D1 inverted fluorescent microscope. Fluo-4 was excited by a mercury with a 485/20 nm excitation filter, and fluorescence was collected by a fluar 40×/1.30 oil objective with 540/50 nm emission filter. Images were acquired by an electron multiplying charge coupled device (EMCCD) (DU-897D-CS0-BV, Andor, U.K.), which was controlled by MetaMorph software (Universal Imaging Corp., USA). Each fluorescence image was acquired for 50 ms with a 1 s interval between frames. The obtained data were quantitatively analyzed for changes of fluorescence intensities within the region of interest (ROI). The [Ca 2+ ] c level over time was presented as relative fluorescence intensity (F/F 0 , intensity after stimulation/basal intensity before stimulation).
Immunofluorescence BV-2 microglial cells were prepared as described above and seeded on a glass cover slip in a 6-well plate (Corning, USA) for 24 h. For living cell immunostaining, cells were incubated with anti-P2Y 12 (extracellular) rabbit antibody (1:500; Alomone, Israel) resolved in DMEM at 37˚C for 1h, followed by washing with HBSS for three times. Then, cells were stained with Alexa-488 goat anti-rabbit IgG (1:500; Abcam, UK) at 37˚C for 1h. After washing with HBSS for three times, images were obtained by 535 nm emission filter based on an Olympus IX51 inverted fluorescence microscope with 40× oil objective.

Statistical analysis
A cell was defined as responsive to mechanical stimulation if it showed a calcium transient with a peak magnitude not less than 1.1 times of baseline. The response rate was defined as the number of responsive neighboring cells (except the stimulated cell) divided by the total number of neighboring cells in each group. All data were analyzed using the IBM software SPSS (version 22). For multiple comparisons, the data were analyzed using unpaired Student's t-test and analysis of variance (ANOVA) followed by the Scheffé post hoc test. P < 0.05 was considered statistically significant.

Mechanical stimulation on a single BV-2 microglia initiates intercellular Ca 2+ wave propagations
To simulate local injury in the CNS, we mechanically stimulated BV-2 microglia using glass microelectrodes ( Fig 1A). Before stimulation treatment, all cells should be in resting states indicated by stable basal [Ca 2+ ] c (F/F 0 = 1 for Fig 1B). In addition, we paid attention to keeping the integrity of plasma membrane after treatment with mechanical stress tested by propidium iodide staining (S1 Fig). Fig 1B displayed the Ca 2+ traces of the stimulated cell and five representative neighboring cells. Fig 1C and S1 Movie showed a typical process of ICW propagations among BV-2 microglia in response to mechanical stimulus. It can be found that mechanical stimulation triggered a rapid rise of [Ca 2+ ] c that originated from the point of the stimulation and then spread throughout the whole cell body (Fig 1C, 1s and 2s). Upon reaching the cell boundaries, the Ca 2+ transient propagated to the neighboring cells in a wave-like manner, namely, intercellular calcium waves (ICWs). The velocity of these ICWs was 10.43 ± 2.66 μm/s.

ICW propagation is mediated by ATP-dependent P2Y purinergic signaling
Our data showed that the 1 st treatment with ATP/ADP could not only strongly suppress Ca 2+ mobilization strength in response to the 2 nd challenge, but also significantly delay response time (S2 Fig), indicating that purinoceptors could be desensitized in BV-2 cells with applications of nucleotides. Thus, to investigate whether ICWs were linked to ATP release, we applied different doses of exogenous ATP (0.3, 1 and 10 μM) prior to mechanical stimuli to desensitize corresponding purinoceptors. Results showed that ICWs were largely blocked by 0.3 μM ATP (Fig 3A), and completely suppressed by 1 and 10 μM ATP (Fig 3B and 3C). Besides, preincubation with apyrase (10 U/mL for 30 min), an ATP-hydrolyzing enzyme, evidently blocked the ICWs (Fig 3D), further demonstrating the importance of ATP during ICWs.
Because of the relevance between ATP and P2 purinergic receptors, we determined the involvement of P2 receptors by pretreatment of suramin (a broad-spectrum P2 receptor antagonist). Results showed that ICWs were almost abolished by 100 μM suramin ( Fig 3E). As summarized in Fig 3F, the response rate of ICWs (0-75 μm) was 55.4 ± 11.6% for control, 5.4 ± 11.2% for 0.3 μM ATP group, 0 ± 0% for 1 and 10 μM ATP, 10.2 ± 11.8% for apyrase and 2.9 ± 6.4% for suramin, respectively. It was known that P2 receptors could be classified into two families: ionotropic P2X ion channels, which mediate calcium influx, and metabotropic G protein-coupled P2Y receptors that initiate intracellular calcium release via IP 3 -sensitive calcium store [20]. Thus, considering that the ICWs were due to intracellular calcium release (Fig  2), the calcium response was mostly associated with P2Y receptors. These results together demonstrated that the mechanical stimulation-induced ICWs were mediated by extracellular nucleotides ATP via activation of P2Y subtypes in BV-2 microglia.
The key roles of ADP and P2Y 12/13

Quantitative estimation of extracellular ATP/ADP concentrations sensed by neighboring BV-2 microglia during Ca 2+ wave propagation
As a critical mechanism of [Ca 2+ ] c regulation, store operated calcium entry (SOCE) can be triggered by a depletion of intracellular calcium stores. Mechanical stimulus-induced ICWs were found to depend on IP 3 receptor-sensitive stores release in BV-2 microglia (Fig 2). However, readdition of 2 mM external calcium did not lead to a second rising of [Ca 2+ ] c in nonmechanically stimulated BV-2 microglia (cell 2-4 in Fig 5A), indicating the independence of SOCE. Thus, we used different concentration of exogenous ATP and ADP to mimic the pattern of calcium mobilization during ICWs. Data showed that application of 0.1 μM ATP induced a sustained and regular calcium oscillation in Ca 2+ -free HBSS (Fig 5B). 0.1 μM ADP only resulted in weak Ca 2+ responses (data not shown). Interestingly, similar to that of mechanical stimulation, application of 0.3 μM ATP ( Fig 5C) and 0.3 μM ADP (Fig 5E) induced Ca 2+ mobilization without SOCE. In contrast, 1 μM ATP ( Fig 5D) and 1 μM ADP (Fig 5F) could evoke evident SOCE. Taken together, we suggested that the concentration of ATP and ADP sensed by neighboring BV-2 microglia was about 0.3 μM during mechanical stimulationinduced ICWs.

Discussion
ICWs appear to be a widespread phenomenon by which various cell types communicate with each other to coordinate and synchronize their activities. For instance, ICWs prefigure and guide the migration of microglia in response to neuronal damage [15], as well as regulate the functions of glial cells in inflammation and immunity [23]. In the present study, we employed mechanical stimulation to an individual BV-2 microglia, and observed remarkable ICWs among adjacent cells. Subsequently, we determined the involvement of several naturally occurring nucleotides in these ICWs. Data showed that ATP and ADP, rather than UTP, UDP and UDP-glucose, were the predominant signal messengers for ICW communication. Furthermore, we indicated that calcium waves were mostly mediated by P2Y 12/13 receptors.
As significant paracrine messengers in the CNS, nucleotides were known to be released in response to biochemical or physical stimuli, such as oxygen radicals, mechanical forces and virus infection [24,25]. Nucleotides could be released from the stimulated cells through exocytosis pathway or conductive mechanisms (including pannexin channels, P2X 7 receptors etc.) [26][27][28]. The released nucleotides gradually diffused to adjacent cells, which subsequently evoked ICW propagation by activating membrane-bound P2 receptors [29]. In this study, we found that mechanical stimulation triggered remarkable ICWs among BV-2 microglia (Fig 1). Subsequently, we investigated the effects of five naturally occurring nucleotides (ATP, ADP, UTP, UDP and UDP-glucose) on the ICWs. It has been well established that prior challenge with agonists could result in desensitization of P2 receptors, especially the G protein-coupled P2Y receptors to a second challenge with the same agonists [30,31]. Our results showed that desensitization with ATP (Fig 3A to 3C) and ADP (Fig 4A) significantly blocked mechanical stimulation-induced Ca 2+ waves. In contrast, application of UTP (Fig 4B), UDP ( Fig 4C) and UDP-glucose (Fig 4E) did not abolish ICW propagation. Therefore, these results indicated that ATP and ADP were predominant transmitters during mechanical stimulation-induced ICW communication in BV-2 microglia.
This mechanism we proposed was differ from previous studies, such as P2Y 1,2 receptors mediated ICWs in rat astrocytes [37], P2Y 11 receptors contributed to ICWs in rat cardiac myofibroblasts [38] and N-methyl-D-aspartic acid (NMDA) receptor was involved in zebrafish microglial ICWs [15]. Importantly, in addition to participate in ICWs, extracellular nucleotides were associated with a wide array of physiological effects of microglia through activation of specific P2 receptors [39]. For instance, P2X 7 receptors are relevant to membrane pore  formation, microvesicle shedding and cell apoptosis [40,41]. P2X 4 as well as P2Y 12 receptors mediated microglial chemotaxis toward injured sites [42][43][44], and P2Y 6 receptor was critical player in promoting microglial phagocytosis resulted from Ca 2+ mobilization [34].
Subsequently, we estimated the concentrations of paracrine ATP and ADP sensed by BV-2 microglia during the ICW. The traditional method detecting concentration of ATP was luciferin-luciferase assay. However, this method was mostly used for the measurements of the overall level of ATP concentration and not applicable to ADP. In this work, we found that mechanical stimulation-evoked ICWs was due to calcium release rather than SOCE (Fig 5A). Therefore, we observed whether the [Ca 2+ ] c pattern in response to exogenous ATP and ADP at different concentration were in agreement with mechanical stimulation-induced ICW. Data showed that different concentration of nucleotides could induce different patterns of Ca 2+ transient, which was similar to that of Visentin's work [50]. Only application of 0.3 μM ATP or ADP exhibited a similar characteristic behavior to that of mechanical stimulation-evoked ICWs (Fig 5C and 5E), indicating that the concentration of ATP and ADP sensed by purinergic receptors of BV-2 microglia was about 0.3 μM. These findings might be significant in the field of drug analysis and development and could provide the basis for pharmacological experiments.
One of the main limitation on this study was the application of BV-2 microglia. Although BV-2 has been known as an immortalized mouse microglial cell line that exhibits the morphological and functional characteristics of primary microglia [19,51], there is no denying that it cannot fully represent the real physiological features of primary microglia in situ or in vivo [52]. Therefore, we plan to study and understand the procedure of primary microglia isolation. Much work will be performed on primary microglia that makes our research more interesting and significant.
In summary, our study demonstrated that ATP and ADP were predominant signal transmitters in mechanical stimulation-induced ICWs through P2Y 12/13 receptors activation in BV-2 microglia. The concentration of ATP/ADP sensed by BV-2 microglia was about 0.3 μM. Our research may bring new insights in the mechanism of cell-to-cell communication of microglia to investigation of its immune defense in the CNS.