Crosstalk between glial Cx43 and neuronal Panx1 and P2X7 receptors modulate presynaptic homeostatic plasticity

doi:10.21203/rs.3.rs-3146678/v1

Download PDF

Research Article

Crosstalk between glial Cx43 and neuronal Panx1 and P2X7 receptors modulate presynaptic homeostatic plasticity

https://doi.org/10.21203/rs.3.rs-3146678/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

The emerging role of glial cells in the modulation of neuronal excitability and synaptic strength is a growing field in neuroscience. An increased number of studies indicate that gliotransmission is mediated by both the classical Ca²⁺-dependent release and the non-vesicular route mediated by connexin and pannexin hemichannels. Recent reports highlighted the role of gliotransmission in the homeostatic adjustment of synaptic strength. In particular, glial-derived ATP arises as a key mediator in this process. However, very little is known about the glial non-vesicular ATP-release pathway and how this ATP could participate in the modulation of synaptic strength. Here we investigated the functional changes occurring in the presynaptic terminal and the role of the purinergic signaling, connexin43 and pannexin1 hemichannels in this process. By using hippocampal dissociated cultures, we show that glial connexin43-dependent ATP release triggers the homeostatic adjustment of presynaptic remodeling and function upon chronic activity silencing. Notably, activated neuronal P2X₇Rs play an essential role in this process by mediating the entrance of Ca²⁺ to the presynaptic cell; our data also suggest that the functional crosstalk between neuronal P2X₇Rs and pannexin1 hemichannels is required for the compensatory adjustment of presynaptic function. Interestingly, we show that neuronal permeability is regulated during chronic inactivity and, in particular, we found that neuronal pannexin1 hemichannels are gradually opened, increasing their activity upon 12h of neuronal inactivity. Hence, our results show for the first time the glial-dependent changes occurring in the presynaptic terminal upon chronic activity silencing and highlight the role of the interaction between presynaptic P2X₇Rs and pannexin1 in the homeostatic adjustment of presynaptic strength.

Glial ATP

P2X7 receptors

Cx43 hemichannels

Panx1 hemichannels

presynaptic plasticity

The nervous system (NS) is composed of both neurons and glia. Neurons have a well-known role in generating and propagating chemical signals and glial cells have an emerging role in shaping and modulating neuronal excitability and synaptic transmission through the release of gliotransmitters [1, 2]. Accumulating evidence indicate that gliotransmission is not only mediated by the classical Ca²⁺-dependent release but also involves a non-vesicular route mediated by connexin and pannexin hemichannels (CxHCs and PanxHCs, respectively) [1, 3–5].

In recent years, the role of gliotransmission in homeostatic synaptic plasticity (HSP), a form of long-term plasticity that refers to the ability of neurons to modulate their firing in response to network activity, has been addressed. For instance, studies performed in hippocampal dissociated and entorhino-hippocampal slice cultures revealed that the tumor necrosis factor alpha (TNFα) is required for synaptic scaling, a postsynaptic form of HSP [6, 7]. On the other hand, we have shown recently that glial ATP and Cx43HCs play a pivotal role in the homeostatic adjustment of presynaptic strength through the activation of neuronal purinergic P2X₇ receptors (P2X₇Rs) and Panx1HCs in hippocampal dissociated cultures [8]. However, the precise relationship between these molecular components is still unknown.

Here we analyze the functional changes that occur in the presynaptic terminal in response to chronic inactivity and the role of gliotransmission in this process. Notably, using hippocampal dissociated cultures, we show that glial Cx43HC-dependent ATP release modulates the compensatory adjustment of cytosolic Ca²⁺ abundance in neurons and thus presynaptic strength. Our data demonstrate that this increase in cytosolic Ca²⁺ is mediated by activated neuronal P2X₇Rs, which are able to permeate Ca²⁺ in non-depolarized terminals [9, 10]. Moreover, the compensatory enhancement in presynaptic strength is also supported by a P2X₇Rs-dependent increase in the size of the total recycling pool of vesicles. Our data also suggest that the functional crosstalk between neuronal P2X₇Rs and Panx1HCs is required for the homeostatic adjustment of presynaptic function, possible by mediating both the release of ATP and the entrance of Ca²⁺ to the presynaptic terminal by a positive feed-forward loop. Interestingly, we show for the first time that neuronal permeability is modulated during chronic inactivity and, in particular, we found that neuronal Panx1HCs are gradually opened, increasing their activity upon 12h of neuronal inactivity. Together, our results reveal the glial-dependent changes occurring in the presynaptic terminal upon chronic inactivity and highlight the role of the interaction between presynaptic P2X₇Rs and Panx1HCs in the homeostatic adjustment of synaptic transmission.

P2X₇Rs control the size of the recycling pool of synaptic vesicles upon chronic inactivity

In a previous work we have shown that glial-derived ATP plays a pivotal role in the activity-dependent modulation of synaptic strength by activating presynaptic P2X₇Rs [8]. Here, we explore the functional changes that occur in the presynaptic terminal that could lead to the P2X₇R-dependent compensatory upregulation of presynaptic strength. We first analyzed synaptic vesicle dynamics by using an antibody against the lumenal domain of the synaptic vesicle protein synaptotagmin 1 (Stg1). To label the total recycling pool of vesicles, we applied field stimulation (600 action potentials, at 10Hz) to dissociated hippocampal cocultures at 14 days in vitro (DIV) and quantified the Stg1-associated fluorescence intensity by immunocytochemistry [11] (Fig. 1a-l). Compatible with the structural changes described previously [8, 12, 13], chronically silencing synaptic activity with the voltage-gated sodium (Na⁺) channel blocker tetrodotoxin (TTX) (1µM, 36hs), increases the size of the total recycling pool of synaptic vesicles compared to untreated neurons (ut) (fold respectively: from 1.00 ± 0.09 to 1.89 ± 0.16; P < 0.0001) (Fig. 1m). Notably, treating cocultures with the competitive P2X₇R antagonist A-804598 (100nM), 15 min prior and during TTX treatment, prevents the compensatory increase in Stg1-associated fluorescence intensity, whereas applying the selective P2X₇R agonist BzATP (10µM, 36h) increases the synaptic amount of Stg1, relative to untreated cocultures (fold respectively: A-804598 1.07 ± 0.10, BzATP 1.51 ± 0.09; P < 0.0001) (Fig. 1m). Together, these results suggest that P2X₇Rs control the size of the total recycling pool of vesicles upon synaptic activity silencing.

P2X₇Rs mediate the chronic inactivity-dependent increase in neuronal Ca²⁺ levels

Previous reports indicate that adaptation to chronic inactivity involves an increase in presynaptic Ca²⁺, accompanied by an extensive remodeling of the presynapse that could lead to changes in the abundance of presynaptic Ca²⁺ channels [14–16]. Our previous work, revealing a pivotal role for P2X₇Rs in HSP, supports the possibility that the increase in intracellular Ca²⁺ could be, at least in part, mediated by P2X₇Rs activation upon chronic inactivity [8]. We test this possibility by performing Ca²⁺ imaging with the small-molecule fluorescent reagent (Fluo-4/AM) in hippocampal dissociated cocultures (14DIV). We first analyzed the sensitivity of neuronal Ca²⁺ levels in two experimental conditions that are reported to increase cytosolic Ca²⁺ concentration and thus presynaptic strength [8, 14, 15]. Hence, treating cultures with TTX (1µM, 36hs) increases basal Ca²⁺-associated fluorescence intensity relative to untreated controls (fold respectively: from ut 1.00 ± 0.02 to TTX 1.50 ± 0.05; P < 0.0001) (Supplementary Fig. 1). Alike, increasing the extracellular concentration of Ca²⁺ ([Ca²⁺]_o) from basal 2mM (ut) to 5mM triggered an increase in neuronal cytosolic Ca²⁺-associated fluorescence intensity (fold respectively: 5mM 1.43 ± 0.03; P < 0.0001) (Supplementary Fig. 1). These results confirmed that the fluorescent reagent Fluo-4/AM is sensitive to estimate changes in cytoplasmic Ca²⁺ levels and thus variations in presynaptic function in our experimental model. Next, we treated hippocampal dissociated cultures with TTX (1µM, 36hs), P2X₇R agonist BzATP (10µM, 36h) and the P2X₇R antagonist A-804598 (100nM), 15 min prior and during TTX treatment (Fig. 2). Again, chronic inactivity promotes an increase in basal neuronal Ca²⁺-associated fluorescence intensity compared to untreated controls (fold respectively: from 1.00 ± 0.03 to 1.39 ± 0.04; P < 0.0001) (Fig. 2k). Interestingly, blocking the activity of P2X₇Rs prevents the increase in cytoplasmic Ca²⁺ levels, whereas neurons treated with BzATP increased it relative to untreated cultures (fold respectively: A-804598 0.96 ± 0.02, BzATP 1.90 ± 0.05; P < 0.0001) (Fig. 2k). As expected, increasing the extracellular concentration of ATP, by adding ATP-disodium salt (ATP-Na₂) (500µM, 36h), promotes an increase in Ca²⁺-associated fluorescence intensity compared to controls (fold respectively: 1.76 ± 0.06, P < 0.0001) (Fig. 2k), mimicking the effect of BzATP. We have used ATP-Na₂, and not the widely used ATP-Mg, as P2X₇Rs agonist because different studies show that Mg²⁺ blocks P2XRs [17–19]. Taken together these results indicate that P2X₇Rs increase their permeability to Ca²⁺ upon activity blockade. As changes in cytoplasmic Ca²⁺ are important for synaptic potentiation [20, 21], our results could suggest that chronic inactivity activates P2X₇Rs, increasing their permeability to Ca²⁺, entering more Ca²⁺ to the presynaptic neuron and thus enhancing presynaptic function.

Chronic inactivity triggers Cx43HC-dependent ATP release

Next, considering that P2X₇Rs have low affinity to ATP, requiring high levels of extracellular ATP for activation [22], we speculate that chronic activity blockade may lead to an increase in extracellular ATP. In our previous work, we showed that the compensatory adjustment of presynaptic strength requires both glial-derived ATP and Cx43HCs [8]. Hence, are these ATP-permeable channels responsible for releasing ATP in response to chronic inactivity? To answer this question, we first analyzed if ATP is accumulated in the extracellular media upon activity blockade by using a bioluminescence assay for quantitative determination of ATP. Notably, media collected from TTX-treated cultures (1µM, 36hs) contains a significantly increased amount of extracellular ATP compared to media collected from untreated cultures (luminescence units, fold respectively: ut 1.09 ± 1.38, TTX 4.70 ± 1.05; P < 0.001) (Fig. 3j), suggesting that an increased amount of ATP is released to the extracellular media in response to chronic inactivity. Next, we investigated whether Cx43HCs are responsible for this increase in extracellular ATP by selectively blocking Cx43HCs activity with the mimetic peptide Gap26 (150µM), 15 min prior and during TTX treatment, and then measuring the bioluminescence associated to extracellular ATP. We found that the increase in extracellular ATP upon chronic inactivity is prevented when Cx43HCs activity is blocked (luminescence units, fold respectively: 1.42 ± 0.42) (Fig. 3j). Taking these results together, we suggest that in response to chronic activity silencing, Cx43HCs enhance the release of glial ATP to the extracellular media and thus activating presynaptic P2X₇Rs and Ca²⁺ entrance into neurons. To further corroborate this notion, we block Cx43HCs activity by treating hippocampal dissociated cultures (14DIV) with the peptide Gap26 (150µM), 15 min prior and during TTX treatment (1µM, 24hs), and measured the cytoplasmic Ca²⁺ levels in neurons. Notably, we found that blocking Cx43 activity prevents the compensatory increase in neuronal cytoplasmic Ca²⁺ -associated fluorescence intensity (fold respectively: ut 1.00 ± 0.01, TTX 1.26 ± 0.02, Gap26 0.92 ± 0.13; P < 0.001) (Fig. 3a-h, i). Furthermore, adding the P2X₇R agonist BzATP (10µM, 24h) to Gap26/TTX-pretreated cocultures, did not rescue the inactivity-dependent increase in cytoplasmic Ca²⁺ (rescue BzATP 0.99 ± 0.02) (Fig. 3i). Collectively, these data strongly suggest that chronic inactivity triggers Cx43HC-dependent ATP release from glia, increasing the extracellular concentration of ATP, activating presynaptic P2X₇Rs and thus enhancing their permeability to Ca²⁺. Moreover, consistent with our previous work indicating that glia-derived ATP is necessary but no sufficient to increase presynaptic strength upon inactivity [8], the rescue experiments could suggest that Cx43HCs could be releasing another gliotransmitter that cooperates with glial ATP in the modulation of presynaptic strength. On the other hand, as we explore below, another possibility is that glia-derived ATP initiates neuronal Panx1HCs opening and thus ATP release from neuronal source.

Neuronal Panx1HCs cooperate with P2X₇Rs to control the activity-dependent increase in Ca²⁺ levels

Panx1HCs are described to be functionally associated with P2X₇Rs under prolonged activation [23]. Moreover, presynaptic P2X₇Rs and neuronal Panx1HCs are thought to cooperate in the modulation of presynaptic strength in HSP [8]. As Panx1HCs are typically known as paths for ATP release, one possibility is that chronic inactivity promotes the release of ATP by neuronal Panx1HCs, potentiating P2X₇Rs activation and the entrance of Ca²⁺ to the presynaptic cell by a positive feed-back loop. We explore this possibility by analyzing the importance of Panx1HC in increasing the extracellular concentration of ATP. By using the bioluminescence assay we found that media collected from cultures treated with probenecid (PBD, 500µM), a Panx1HC blocker, 15 min prior and during TTX treatment (1µM, 36hs), shows no significantly differences in the amount of extracellular ATP compared to untreated media (luminescence units, fold respectively: ut 1.09 ± 0.38, TTX 4.70 ± 1.05, PBD 1.04 ± 0.24; P < 0.0001) (Fig. 4j). This suggests that ATP is also released by Panx1HCs in response to chronic inactivity. We next analyzed the relevance of Panx1HC in the P2X₇Rs-mediated Ca²⁺ entrance by quantifying cytoplasmic Ca²⁺-associated fluorescence intensity in hippocampal dissociated neurons treated with PBD (500µM), 15 min prior and during TTX treatment (1µM, 36hs) (Fig. 4a-h). We found that blocking Panx1HCs activity prevents the compensatory increase in cytosolic Ca²⁺ availability (fold respectively: ut 1.00 ± 0.01, TTX 1.36 ± 0.02, PBD 1.06 ± 0.03; P < 0.0001) (Fig. 4i). Notably, adding the P2X₇R agonist BzATP (10µM, 24h) to PBD/TTX-pretreated cocultures (24h), rescues the inactivity-dependent increase in cytoplasmic Ca²⁺ (fold respectively: rescue BzATP 1.59 ± 0.03; P < 0.01) (Fig. 4i), suggesting that P2X₇Rs functionally cooperates with Panx1HCs to increase the amount of cytosolic Ca²⁺ and potentiate presynaptic strength upon chronic inactivity. Thus, we next wonder whether Panx1HCs could be also controlling the size of the total recycling pool of synaptic vesicles upon activity blockade. We analyzed Stg1-associated fluorescence intensity after applying field stimulation (600 action potentials, at 10Hz) to dissociated hippocampal cocultures (14DIV), finding that PBD (500µM) exposure, 15 min prior and during TTX treatment (1µM, 36hs), prevents the compensatory increase in Stg1-associated fluorescence intensity compared to untreated cultures (fold respectively: ut 1.00 ± 0.09, TTX 1.89 ± 0.16; PBD 0.83 ± 0.05) (Fig. 4k). These results suggest that Panx1HCs mediate both the increase in neuronal Ca²⁺ and the size of the recycling pool of vesicles in, two parameters that modulate presynaptic efficacy.

In this sense, we have previously shown that neuronal Panx1HCs, but not the glial ones, play an essential role in HSP. Hence, to confirm that neuronal Panx1HCs are responsible for modulating the entrance of Ca²⁺ into neurons, we used hippocampal dissociated cultures where the glia monolayer come from the hippocampus of wild type (wt) mice pups and the neurons growing on top derived from panx1-deficient (panx1ko) hippocampus (wt glia/panx1ko neu). Notably, in absence of neuronal Panx1HCs the increase in neuronal cytoplasmic Ca²⁺-associated fluorescence intensity upon TTX treatment is prevented, relative to untreated cultures (fold respectively: ut wt glia/panx1ko neu 1.00 ± 0.01, TTX wt glia/panx1ko neu 0.97 ± 0.01) (Fig. 5b). As expected, control wt mice cultures, where both glia and neurons come from wt hippocampus, significantly increase the cytoplasmic Ca²⁺ levels upon TTX treatment in neurons, compared to untreated cultures (ut wt 1.00 ± 0.01, TTX wt 2.25 ± 0.01, P < 0.0001) (Fig. 5a). All together, our findings support the notion that neuronal Panx1HCs mediate the increase in cytoplasmic Ca²⁺ in response to chronic activity silencing by a P2X₇R-dependent signaling pathway.

The activity of neuronal Panx1HC is regulated during chronic inactivity

We have previously described no significant differences in the distribution nor abundance of Panx1 upon chronic inactivity [8]. Hence, the role of neuronal Panx1HCs in the activity-dependent upregulation of presynaptic strength could be due to an enhanced Panx1HCs activity rather than to an increase in hemichannel number. This raises the question whether the activity of neuronal Panx1HCs is regulated during HSP. To explore this possibility we first determined the temporal pattern of HSP induction by analyzing the time course of synaptic vGlut1 during chronic TTX treatment. The synaptic abundance of vGlut1 is modulated in an activity-dependent fashion and correlates with presynaptic strength [8, 24–26]. Thus, we evaluate changes in presynaptic strength by immunolabeling against vGlut1 hippocampal dissociated cocultures (14DIV) treated with TTX for 1h, 2h, 5h, 12h or 24h (Fig. 5c and Supplementary Fig. 2). Interestingly, only neurons treated for 12 or 24h with TTX significantly increase the vGlut1-associated fluorescence intensity relative to untreated cultures; suggesting that chronic inactivity-dependent upregulation of vGlut1 emerges 12h upon the onset of inactivity (ut 1.0 ± 0.03, TTX 1h 0.91 ± 0.02, TTX 2h 0.95 ± 0.03, TTX 5h 1.03 ± 0.02, TTX 12h 1.19 ± 0.04, TTX 24h 1.89 ± 0.05) (Fig. 5c). We also analyzed the density of vGlut1-positive synaptic contacts, another parameter that is homeostatically regulated [8]. Consistent with the changes observed in presynaptic strength, the density of synaptic contacts was also increased after 12h of TTX treatment (ut 4.76 ± 0.14, TTX 1h 4.69 ± 0.15, TTX 2h 4.89 ± 0.09, TTX 5h 5.13 ± 0.16, TTX 12h 5.50 ± 0.08, TTX 24h 8.14 ± 0.10) (Supplementary Fig. 2). Together these data could suggest that the activity-dependent increase in synaptic strength is set around 12h of inactivity and well established upon 24h.

Next, we ought to analyze whether the activity of neuronal Panx1HCs is regulated during chronic activity blockade by performing ethidium bromide (EtBr) uptake experiments, as an indicator of channel activity [23]. The neuronal EtBr-associated fluorescence intensity was measured in regions of interest selected around single neuronal somas in both wt cultures and wt glia/panx1ko neu cultures, treated with TTX for 1h, 2h, 5h, 12h or 24h (Fig. 5d-o). Strikingly, during the first hours of HSP induction (from 1 to 5h of TTX treatment), wt neurons showed a reduced EtBr permeability relative to untreated neurons, however, after 12h of TTX treatment the neuronal EtBr-associated fluorescence intensity is significantly increased (fold respectively: ut 1.00 ± 0.03, TTX 1h 0.78 ± 0.04, TTX 2h 0.84 ± XX, TTX 5h 0.73 ± 0.02, TTX 12h 1.17 ± 0.03, TTX 24h 1.31 ± 0.04; P < 0.0001) (Fig. 5p). These results are consistent with the temporal pattern of HSP induction revealed by the synaptic vGlut1 abundance, and suggest that during the first hours of synaptic inactivity neuronal channels could be less activated compared to basal conditions, and that upon 12h of inactivity the neuronal permeability increase. Hence, are the neuronal Panx1HCs mediating these changes in neuronal permeability during TTX treatment? Notably, the loss of neuronal Panx1 was sufficient to prevent the inactivity-dependent changes in EtBr-associated fluorescence intensity. On one hand, neurons from wt glia/panx1ko neu cultures show no downregulation of EtBr-associated fluorescence intensity during the first hours of HSP induction (from 1 to 5h of TTX treatment), compared to untreated neurons. On the other hand, no upregulation of neuronal permeability was detected in neurons at 12h or 24h of TTX treatment, in fact, a reduction of EtBr-associated fluorescence intensity was detected at these times compared to untreated controls (ut 1.01 ± 0.02, TTX 1h 0.95 ± 0.02, TTX 2h 1.01 ± 0.02, TTX 5h 0.98 ± 0.02, TTX 12h 0.83 ± 0.03, TTX 24h 0.82 ± 0.02 P < 0.0001) (Fig. 5q). Taking these results all together, we suggest that when HSP is settled, neuronal Panx1HCs are gradually opened, increasing their activity and the interchange between the extracellular and intracellular compartment. This could lead to an increase in Panx1HC-dependent ATP release and P2X₇R activation, potentiating the entrance of Ca²⁺ to the presynaptic neuron, and thus presynaptic strength, upon chronic activity silencing.

In this work we have systematically studied the functional changes that occur in the presynaptic neuron leading to the P2X₇R-dependent compensatory upregulation of presynaptic strength and the role of the neuron-glia crosstalk in this process. Monitoring synaptic vesicle dynamics by using an antibody against the lumenal domain of Stg1 we found that activated P2X₇Rs control the size of the total recycling pool of vesicles upon chronic inactivity, which is critical for maintaining normal synaptic function. As we mention above, different reports indicate that adaptation to chronic inactivity involves an increase in presynaptic Ca²⁺, accompanied by an extensive remodeling of the presynapse leading, for instance, to changes in the number of presynaptic Ca²⁺ channels [14–16]. However, our previous work supports the possibility that the increase in presynaptic Ca²⁺ could be, at least in part, mediated by P2X₇Rs activation upon chronic activity silencing [8]. This is also supported by the intrinsic properties of the channels, for instance, P2X₇R permeability to Ca²⁺ is described to be higher at negative potentials compared to depolarized membrane potentials. Thus, neurotransmitter release could be initiated upon P2X₇R activation, without presynaptic depolarization [27, 28]. Hence, neuronal P2X₇Rs appear as good candidates to mediate Ca²⁺ entrance in our experimental model where presynaptic HSP is induced by preventing presynaptic depolarization with chronic TTX treatment. Together, these properties turn P2X₇Rs to key actors in the modulation of presynaptic strength upon chronic inactivity. Consistent with this notion, we found that blocking P2X₇Rs prevents the compensatory increase in neuronal cytosolic Ca²⁺ and an opposite effect was observed when P2X₇Rs were long-term activated. Moreover this role of P2X₇Rs, as a route to modulate the abundance of free cytosolic Ca²⁺, is accompanied by an increase in the extracellular amount of ATP upon chronic TTX treatment. This is consistent with results published by others showing that P2X₇Rs have low affinity to ATP, requiring high levels of extracellular ATP for activation [22]. In our experimental model, P2X₇Rs are expressed presynaptically and their abundance and distribution are not altered upon chronic inactivity [8]; thus, the increased neuronal Ca²⁺ permeability seems to be due to an enhanced activation and prolonged P2X₇ channel opening rather to an increase in receptor number. Altogether our findings suggest that chronic activity blockade increases the extracellular concentration of ATP, enhancing the activated state of presynaptic P2X₇Rs, increasing the entrance of Ca²⁺ to the presynaptic cell and promoting the remodeling of the presynaptic terminal to support the compensatory adjustment of presynaptic function.

In our previous work we have described a pivotal role of the neuron-glia crosstalk in presynaptic homeostatic plasticity [8]. In recent years, growing evidence indicates that Cx43HCs are alternative non-vesicular routes for gliotransmitters release -such as ATP and glutamate- to the extracellular milieu [1]. Moreover, recent findings performed in Cx43 conditional KO mice (Cx43-/-), showed that Cx43HCs are essential for synaptic vesicles dynamics and distribution and, as a consequence, for the modulation of synaptic efficacy in hippocampal synapses under basal conditions [29]. In accordance, our results reveal that astrocytes modulate presynaptic function in trans by increasing the release of ATP through Cx43HCs and concomitantly the presynaptic P2X₇Rs-dependent Ca²⁺ entrance. In this sense, one possibility is that prolonged inactivity reduces the amount of glutamate released by neurons, which could be sensed by astrocytes to activate the purinergic pathway to promote the compensatory upregulation of presynaptic strength.

To further understand this notion, we performed rescue experiments where we blocked Cx43HCs activity during TTX treatment and followed by chronic P2X₇Rs activation. Interestingly under these conditions, the activation of P2X₇Rs is not sufficient to recover presynaptic function upon chronic activity silencing, suggesting that Cx43HCs could be releasing another gliotransmitter that cooperates with ATP in the modulation of presynaptic strength. Another possibility is that a purinergic channel other than P2X₇ is involved, but this hypothesis is ruled out by the fact that our previous work shows that glial ATP is not sufficient to elicit presynaptic homeostatic plasticity [8]. Together these data add more intricacy to the Cx43HCs-dependent molecular pathway as other factors such as glutamate or TNFα, both released by astrocytes, could be cooperating with this pathway to modulate presynaptic function upon inactivity [1, 6, 7].

Consistent with previous reports, we found that neuronal Panx1HCs are essential to elicit the homeostatic changes occurring in the presynaptic terminal to ultimately adjust presynaptic strength upon chronic inactivity [8]. In this sense, here we described that neuronal Panx1 controls the size of the total recycling pool of vesicles and the abundance of cytosolic Ca²⁺ in neurons. Moreover, Panx1HCs provide another pathway for ATP release as blocking these channels reduces the amount of extracellular ATP in hippocampal dissociated cultures. But how might Panx1 activation control presynaptic strength? Different lines of evidence suggest that Panx1HCs are functionally associated with P2X₇Rs. In this sense, reports indicate that an increase in cytosolic Ca²⁺ activates Panx1HCs and ATP release suggesting that the activation of Panx1HCs would be elicited by the increase in intracellular Ca²⁺ downstream P2X₇Rs activation [30, 31]. Notably, our rescue experiments where Panx1HCs were blocked during chronic TTX treatment followed by P2X₇Rs activation, showed that the compensatory adjustment of cytosolic Ca²⁺ was recovered, suggesting that Panx1 acts downstream P2X₇Rs-dependent Ca²⁺ entrance. Previously, we found that the abundance and distribution of neuronal Panx1HCs are not affected upon chronic inactivity [8]. Hence, we analyzed how chronic inactivity affects the activity of neuronal channels. Interestingly, we found that during the first hours of activity blockade the neuronal permeability is reduced compared to basal conditions, however, after 12hs of inactivity the neuronal permeability is significantly increased. Notably, this increase is mediated, at least in part, for Panx1HCs. Thus, taking all our results together it is tempting to postulate that chronic inactivity triggers the release of ATP by astrocytes via Cx43HCs, activating presynaptic P2X₇Rs, increasing the entrance of Ca²⁺ to the presynaptic terminal and activating presynaptic Panx1HCs. These channels may cooperate with P2X₇Rs by increasing both the release of ATP and the entrance of Ca²⁺ to the presynaptic terminal by a positive feed-forward loop (Fig. 6). Interactions between Cx43HCs, P2X₇Rs and Panx1HCs have been reported to modulate presynaptic function in the peripheral tripartite synapse, specifically in sensory ganglia, where they contribute to nociceptive processing [32, 33]. In this regard, opening of glial Cx43HCs in murine nodose-petrosal jugular complexes, enhances neuronal activity resulting in increased vagal discharge, an effect dependent on P2X₇Rs and neuronal Panx1HCs [32]. In the dorsal root ganglia, activation of glial P2X₇Rs leads to the release of glial ATP via Panx1HCs which diffuses towards surrounding sensory neurons to activate neuronal P2X₃ receptors and Panx1HCs resulting in both increased excitability and further ATP release. This paracrine communication between satellite glial cells and sensory neurons mediated by Cx43HCs, P2X₇Rs and Panx1HCs sensitizes dorsal root sensory neurons to subsequent peripheral stimulation yielding enhanced nociceptive afferent transmission in vivo [33].

Hence, our work provides for the first time functional data highlighting the role of Cx43HCs in the compensatory adjustment of presynaptic strength upon chronic inactivity and the functional interaction between presynaptic P2X₇Rs and Panx1HCs in this process.

Cell cultures

For neuronal culture, hippocampi were extracted from neonatal (P0-P1) Sprague Dawley rats, C57BL/6 mice or Panx1-deficient mice (provided by Genentech, US). Tissue was chemical disaggregated in a solution containing papain (15 units, Sigma), EDTA (0.5mM), CaCl2 (1.5mM), L-cysteine (0.2mg/ml), DNAsa (0.1mg/µl) for 20 minutes at 37°C.

Neurons were plated onto 12mm coverslips containing a feeder glial monolayer. Neuroglial cocultures were maintained in supplemented NB medium (Neurobasal, Gibco), containing glucose (34mM), B27 (2.5%, Gibco), glutamax (2mM, Gibco), penicillin/streptomycin (100U/ml; 100µg/ml, Gibco). Cells were grown at 37°C with 95% O2 and 5% CO₂. To prevent glial overgrowing, cells were treated with cytosine arabinoside (AraC 1µM) 24h after neurons were plated. HSP was induced with chronic TTX treatment (1µM; Tocris, 24-36h); non-treated cocultures were used as controls of basal synaptic function. Cells were used for experiments at 14–15 DIV.

Pharmacology

Pharmacological agents used were: A804598 (100nM), Benzoyl ATP (BzATP, 10 µM), Probenecid (PBD, 500 µM), Gap26 peptide (0.25mg/ml) and Na₂ATP (500µM). Reagents were provided by Sigma-Aldrich. Where noted, antagonists were incubated 15min prior and during TTX treatment.

Immunocytochemistry and live labeling

Immunocytochemistry was performed as previously described [8]. Briefly, neuroglial cocultures were fixed in 4% paraformaldehyde (PFA) at room temperature (RT) for 15 minutes, then permeabilized and blocked for 30 minutes in PBS containing 0.2M glycine, 10% FBS and 0.1% Triton X100. Primary antibodies were added in PBS containing 5% FBS and incubated for 2 hours at RT. The following primary antibodies were used: rabbit anti-vGlut1 (Synaptic Systems); mouse anti-Homer1 (Synaptic Systems); chicken anti-MAP2 (Abcam). After 3 washes in 0.1% triton X100, cells were incubated for 1h at RT with secondary fluorescently conjugated antibodies. The following secondary antibodies were used: Alexa Fluor 488 (Invitrogen); DyLight 649 (Jackson); Alexa Fluor 568 (Invitrogen); and mounted in Fluoromount™ (Sigma-Aldrich).

Synaptotagmin 1 live labeling: 600 action potentials were delivered to hippocampal cocultures by field stimulation at 10 Hz in the presence of Stg1 primary antibody (Synaptic Systems), diluted in external bath solution (EBS) containing (in mM): 137 NaCl, 2 CaCl₂, 2 MgCl₂, 10 D-glucose, 5 HEPES, 0.001 glycine, at RT. After two washes with EBS, cells were fixed in 4% PFA and the immunocytochemistry performed using the chicken anti-MAP2 primary antibody (Abcam) and secondary antibodies, as described above.

Bromide ethidium influx essay

hippocampal dissociated cocultures were exposed to a bromide ethidium solution (2.5 µM) for 15 minutes at 37°C with 5% CO₂, 95% O₂. After two washes with NB media, cells were fixed with 4% PFA.

Calcium imaging

To assess changes in neuronal cytosolic Ca²⁺ levels, the small-molecule fluorescence reagent Fluo-4/AM (Invitrogen; Cat: F14201) was used [34]. Each vial of Fluo-4/AM was reconstituted in 50µl DMSO, resulting in a 1 µg/µl stock solution. Cocultures were incubated in EBS solution containing 0.01% pluronic acid, 1x10^− 3µg/ml of Fluo-4/AM and specific agonist or antagonist for 30 minutes at 37°C with 5% CO₂, 95% O₂. After two washes with EBS, cells were mounted on a homemade chamber.

Image analysis

Images were acquired using a confocal microscope (model Zeiss Axio Observer ZM 800 or Leica TCS SP5II) or an epifluorescence microscope Nikon Eclipse E 600. Images were analyzed using FIJI software.

To quantify the fluorescent signal associated to synaptic vGlut1, individual synapses were identified for partial or completely overlapping signals between the presynaptic marker vGlut1 and postsynaptic marker Homer1, localized onto MAP2-positive cells. Integrated fluorescence values were measured and averaged per neuron and background signal (non-punctate regions) subtracted, as previously described [8]. The density of synaptic boutons was quantified within the first 30um from the soma and averaged per neuron [8].

To quantify Stg1 abundance, integrated fluorescence intensity from single Stg1-positive puncta was measured and averaged per neuron. Our analysis was limited to puncta localized within MAP2-positive cells. The background signal (non-punctate regions) was subtracted.

To quantify the BrEt-associated fluorescence signal in individual neurons, the neuronal soma (from MAP2-positive cells) was defined as the region of interest, and the integrated fluorescence intensity measured within those regions.

To quantify Fluo-4/AM-associated fluorescence signal, neuronal somas were delimited as regions of interest, and the integrated fluorescence intensity was measured within those regions.

For all quantifications we performed at least three independent experiments for each condition. The total number of neurons used is indicated in figures. Where noted, the ratio between averaged values in treated neurons with those in control neurons from the same experiment was determined.

ATP measurement

Extracellular ATP levels were quantified using the luciferin-luciferase bioluminescence assay (Invitrogen™; cat. A22066). Company’s protocol was followed for standard luciferin-luciferase solution preparation.

Three media samples of each condition were taken and added to 96 multiwell plate with white upper structure and clear bottom, containing the standard luciferin-luciferase solution. Luminiscence values were obtained using a luminometer Lumsistar Galaxy of BMG Lab Tech and values were averaged. Background signal was measured as the average luminescence of two wells containing standard solution and two wells containing supplemented NB medium. Finally, this signal was subtracted from luminescence values of treated and control samples.

Statistics

Statistical significance was analyzed by the two-tailed Student’s t test or one-way ANOVA with Tukey’s Multiple Comparison post-hoc test. All data are shown as the mean ± s.e.m. Statistical significance was taken as p < 0.05. In figures, *p < 0.05, **p < 0.01 and ***p < 0.001.

ACKNOWLEDGMENTS

The present study was supported by grants to N.V., V.A. and A.R. by the Comisión Sectorial de Investigación Científica (CSIC-UdelaR) and by the Program for the Development of Basic Sciences (PEDECIBA). A.R. and M.T. received a PhD and Master Fellowship from the Agencia Nacional de Investigación e Innovación (ANII) and CSIC- UdelaR, respectively.

Ethics Approvals

Animal care and protocols were approved by the Committee of Ethics in Animal Research (CHEA-UdelaR), Uruguay.

Consent to participate

Not applicable

Consent for publication

Not applicable

Availability of data and material

Not applicable

Competing interests

The authors have no relevant financial or non-financial interests to disclosure.

Funding

This work was supported by grants to N.V., V.A. and AR by the Comisión Sectorial de Investigación Científica (CSIC-UdelaR, Uruguay, grant number: 22520220100095UD, 22520220100243UD and 22320200200227UD, respectively) and by the Programa de Desarrollo de las Ciencias Básicas (PEDECIBA, Uruguay).

Authors' contributions

N.V. and V.A contributed to the manuscript conception and design. The manuscript was written by N.V., A.R. and M.T. and all the authors commented on previous versions of the manuscript. Figures were prepared by N.V., V.A. and A.C. A.R. performed P2X₇Rs- and Cx43HCs- based experiments, rescue experiments and ATP measurements. C.J. collaborated in setting Ca²⁺imaging and quantification. M.T. and A.C. performed panx1-based experiments and HSP time course.

Abudara V, Retamal MA, Del Rio R, Orellana JA (2018) Synaptic Functions of Hemichannels and Pannexons: A Double-Edged Sword. Front Mol Neurosci 11:1–24. https://doi.org/10.3389/fnmol.2018.00435
Perea G, Navarrete M, Araque A (2009) Tripartite synapses: astrocytes process and control synaptic information. Trends Neurosci 32:421–431. https://doi.org/10.1016/j.tins.2009.05.001
Cheung G, Chever O, Rouach N (2014) Connexons and pannexons: Newcomers in neurophysiology. Front Cell Neurosci 8:1–19. https://doi.org/10.3389/fncel.2014.00348
Ardiles AO, Flores-Muñoz C, Toro-Ayala G, et al (2014) Pannexin 1 regulates bidirectional hippocampal synaptic plasticity in adult mice. Front Cell Neurosci 8:1–11. https://doi.org/10.3389/fncel.2014.00326
Chever O, Lee CY, Rouach N (2014) Astroglial connexin43 hemichannels tune basal excitatory synaptic transmission. J Neurosci 34:11228–11232. https://doi.org/10.1523/JNEUROSCI.0015-14.2014
Stellwagen D, Malenka RC (2006) Synaptic scaling mediated by glial TNF-α. Nature 440:1054–1059. https://doi.org/10.1038/nature04671
Beattie EC, Stellwagen D, Morishita W, et al (2002) Control of Synaptic Strength by Glial TNF alpha. Science 295:2282–2285. https://doi.org/doi: 10.1126/science.1067859.
Rafael A, Cairus A, Tizzoni M, et al (2020) Glial ATP and Large Pore Channels Modulate Synaptic Strength in Response to Chronic Inactivity. Mol Neurobiol 57:2856–2869. https://doi.org/10.1007/s12035-020-01919-0
Pankratov Y, Lalo U, Krishtal O a., Verkhratsky a. (2009) P2X receptors and synaptic plasticity. Neuroscience 158:137–148. https://doi.org/10.1016/j.neuroscience.2008.03.076
Khakh BS (2001) Molecular physiology of P2X receptors and ATP signalling at synapses. Nat Rev Neurosci 2:165–174. https://doi.org/10.1038/35058521
Vitureira N, Letellier M, White IJ, Goda Y (2012) Differential control of presynaptic efficacy by postsynaptic N-cadherin and β-catenin. Nat Neurosci 15:81–89. https://doi.org/10.1038/nn.2995
Thiagarajan TC, Lindskog M, Tsien RW (2005) Adaptation to synaptic inactivity in hippocampal neurons. Neuron 47:725–737. https://doi.org/10.1016/j.neuron.2005.06.037
Murthy VN, Schikorski T, Stevens CF, Zhu Y (2001) Inactivity produces increases in neurotransmitter release and synapse size. Neuron 32:673–682. https://doi.org/10.1016/S0896-6273(01)00500-1
Zhao CJ, Dreosti E, Lagnado L (2011) Homeostatic synaptic plasticity through changes in presynaptic calcium influx. J Neurosci 31:7492–7496. https://doi.org/10.1523/JNEUROSCI.6636-10.2011
Lazarevic V, Schöne C, Heine M, et al (2011) Extensive remodeling of the presynaptic cytomatrix upon homeostatic adaptation to network activity silencing. J Neurosci 31:10189–10200. https://doi.org/10.1523/JNEUROSCI.2088-11.2011
Rátkai A, Tárnok K, Aouad H El, et al (2021) Homeostatic plasticity and burst activity are mediated by hyperpolarization-activated cation currents and T-type calcium channels in neuronal cultures. Sci Rep 11:1–17. https://doi.org/10.1038/s41598-021-82775-3
North R a (2002) Molecular physiology of P2X receptors. Physiol Rev 82:1013–67. https://doi.org/10.1152/physrev.00015.2002
Li M, Silberberg SD, Swartz KJ (2013) Subtype-specific control of P2X receptor channel signaling by ATP and Mg2+. Proc Natl Acad Sci U S A 110:E3455–E3463. https://doi.org/10.1073/pnas.1308088110
Acuña-Castillo C, Coddou C, Bull P, et al (2007) Differential role of extracellular histidines in copper, zinc, magnesium and proton modulation of the P2X7 purinergic receptor. J Neurochem 101:17–26. https://doi.org/10.1111/j.1471-4159.2006.04343.x
Zucker RS (1999) Calcium- and activity-dependent synaptic plasticity. Curr Opin Neurobiol 9:305–313. https://doi.org/10.1016/S0959-4388(99)80045-2
Winder DG, Sweatt JD (2001) Roles of serine/threonine phosphatases in hippocampel synaptic plasticity. Nat Rev Neurosci 2:461–474. https://doi.org/10.1038/35081514
Surprenant A, Kawashima E, Rassendren F, et al (1996) The Cytolytic P2z Receptor for Extracellular ATP identified as a P2x Receptor ( P2X7 ). Science 272:735–738. https://doi.org/doi: 10.1126/science.272.5262.735.
Pelegrin P, Surprenant A (2006) Pannexin-1 mediates large pore formation and interleukin-1 release by the ATP-gated P2X7 receptor. EMBO J 25:5071–5082. https://doi.org/10.1038/sj.emboj.7601378
Wilson NR (2005) Presynaptic Regulation of Quantal Size by the Vesicular Glutamate Transporter VGLUT1. J Neurosci 25:6221–6234. https://doi.org/10.1523/JNEUROSCI.3003-04.2005
de Jong APH, Schmitz SK, Toonen RFG, Verhage M (2012) Dendritic position is a major determinant of presynaptic strength. J Cell Biol 197:327–337. https://doi.org/10.1083/jcb.201112135
Herman M a., Ackermann F, Trimbuch T, Rosenmund C (2014) Vesicular glutamate transporter expression level affects synaptic vesicle release probability at hippocampal synapses in culture. J Neurosci 34:11781–11791. https://doi.org/10.1523/JNEUROSCI.1444-14.2014
Sperlagh B, Kofalvi A, Deuchars J, et al (2002) Involvement of P2X7 receptors in the regulation of neurotransmitter release in the rat hippocampus. J Neurochem 81:1196–1211. https://doi.org/10.1046/j.1471-4159.2002.00920.x.
Egan TM, Samways DSK, Li Z (2006) Biophysics of P2X receptors. Pflugers Arch Eur J Physiol 452:501–512. https://doi.org/10.1007/s00424-006-0078-1
Cheung G, Chever O, Rollenhagen A, et al (2023) Astroglial connexin 43 regulates synaptic vesicle release at hippocampal synapses. Cells 12:1–16. https://doi.org/10.3390/cells12081133
Kang J, Kang N, Lovatt D, et al (2008) Connexin 43 hemichannels are permeable to ATP. J Neurosci 28:4702–4711. https://doi.org/10.1523/jneurosci.5048-07.2008
Locovei S, Scemes E, Qiu F, et al (2006) Pannexin1 is part of the pore forming unit of the P2X(7) receptor death complex. Febs Lett 581:483–488. https://doi.org/10.1016/j.febslet.2006.12.056
Bravo D, Ibarra P, Retamal J, et al (2014) Pannexin 1: A novel participant in neuropathic pain signaling in the rat spinal cord. Pain 155:2108–2115. https://doi.org/10.1016/j.pain.2014.07.024
Chen Z, Zhang C, Song X, et al (2022) Bzatp activates satellite glial cells and increases the excitability of dorsal root ganglia neurons in vivo. Cells 11:1–17. https://doi.org/10.3390/cells11152280
Justet C, Chifflet S, Hernandez JA (2019) Calcium oscillatory behavior and its possible role during wound healing in bovine corneal endothelial cells in culture. Biomed Res Int 2019:11–16. https://doi.org/10.1155/2019/8647121

SupplementaryFigs12.pdf

Download PDF

Editor assigned by journal
11 Jul, 2023
First submitted to journal
11 Jul, 2023

You are reading this latest preprint version

Crosstalk between glial Cx43 and neuronal Panx1 and P2X7 receptors modulate presynaptic homeostatic plasticity

Status:

Version 1

Abstract

Figures

INTRODUCTION

RESULTS

P2X₇Rs control the size of the recycling pool of synaptic vesicles upon chronic inactivity

P2X₇Rs mediate the chronic inactivity-dependent increase in neuronal Ca²⁺ levels

Chronic inactivity triggers Cx43HC-dependent ATP release

Neuronal Panx1HCs cooperate with P2X₇Rs to control the activity-dependent increase in Ca²⁺ levels

The activity of neuronal Panx1HC is regulated during chronic inactivity

DISCUSSION

METHODS

Cell cultures

Pharmacology

Immunocytochemistry and live labeling

Calcium imaging

Image analysis

ATP measurement

Statistics

Declarations

References

Supplementary Files

Status:

Version 1

Crosstalk between glial Cx43 and neuronal Panx1 and P2X7 receptors modulate presynaptic homeostatic plasticity

Status:

Version 1

Abstract

Figures

INTRODUCTION

RESULTS

P2X7Rs control the size of the recycling pool of synaptic vesicles upon chronic inactivity

P2X7Rs mediate the chronic inactivity-dependent increase in neuronal Ca2+ levels

Chronic inactivity triggers Cx43HC-dependent ATP release

Neuronal Panx1HCs cooperate with P2X7Rs to control the activity-dependent increase in Ca2+ levels

The activity of neuronal Panx1HC is regulated during chronic inactivity

DISCUSSION

METHODS

Cell cultures

Pharmacology

Immunocytochemistry and live labeling

Calcium imaging

Image analysis

ATP measurement

Statistics

Declarations

References

Supplementary Files

Status:

Version 1

P2X₇Rs control the size of the recycling pool of synaptic vesicles upon chronic inactivity

P2X₇Rs mediate the chronic inactivity-dependent increase in neuronal Ca²⁺ levels

Neuronal Panx1HCs cooperate with P2X₇Rs to control the activity-dependent increase in Ca²⁺ levels