ATP as a multi-target danger signal in the brain

ATP is released in an activity-dependent manner from different cell types in the brain, fulfilling different roles as a neurotransmitter, neuromodulator, in astrocyte-to-neuron communication, propagating astrocytic responses and formatting microglia responses. This involves the activation of different ATP P2 receptors (P2R) as well as adenosine receptors upon extracellular ATP catabolism by ecto-nucleotidases. Notably, brain noxious stimuli trigger a sustained increase of extracellular ATP, which plays a key role as danger signal in the brain. This involves a combined action of extracellular ATP in different cell types, namely increasing the susceptibility of neurons to damage, promoting astrogliosis and recruiting and formatting microglia to mount neuroinflammatory responses. Such actions involve the activation of different receptors, as heralded by neuroprotective effects resulting from blockade mainly of P2X7R, P2Y1R and adenosine A2A receptors (A2AR), which hierarchy, cooperation and/or redundancy is still not resolved. These pleiotropic functions of ATP as a danger signal in brain damage prompt a therapeutic interest to multi-target different purinergic receptors to provide maximal opportunities for neuroprotection.

The variety of purinergic receptors and their widespread region-and cell-specific expression pattern and actions places purinergic signaling as a major system for integration of functional activity between neurons, glial and vascular cells in the brain as heralded by the role of purines (ATP and adenosine) in neuron-neuron, astrocyte-neuron, oligodendrocyte-neuron and/or microglia/neuron bi-directional communication (Fields and Burnstock, 2006;Butt, 2011). Moreover, the different sensitivities of the different receptors to their different ligands (ATP, ADP, adenosine) displaying spatial and temporal finetuned gradients Cunha, 2008), endows purinergic signaling with unique features adapted to control brain networks. Not surprisingly, the dysfunction of this purinergic system is closely associated with brain disorders and we will now exploit the concept that ATP acts as a danger signal, implying an abnormal and sustained elevation of extracellular ATP levels in brain dysfunction and the involvement of purine receptors, namely P2X7R (ATP), P2Y1R (ADP) and A 2A R (adenosine), in brain damage.

Purinergic Receptors in Brain Pathology
The concept of ATP as a danger signal implies the release of ATP but also the involvement of purinergic receptors in brain disorders, which has mostly been documented for P2X7R, P2Y1R, and A 2A R.
P2X7R up-regulation has been mainly associated with microgliosis, since P2X7R promote neuronal death through microglia-derived interleukin-1β (IL-1β) (Ferrari et al., 1996;Chakfe et al., 2002;Skaper et al., 2006;Bernardino et al., 2008;Takenouchi et al., 2009) or production of reactive oxygen species (Parvathenani et al., 2003;Skaper et al., 2006;Lee et al., 2011). In FIGURE 1 | Integrated view of the purinergic signaling in brain disorders. In addition to the leakage of ATP through damaged cell membrane from injured or dying cells, evolution has assured multiple mechanisms from different sources to place ATP in the extracellular milieu as a danger signal in the brain. Interestingly, this increase is self-sustained: activation of P2X7R induces the release of ATP either directly through its channel or by exocytotic or non-exocytotic mechanisms (e.g., hemichannels); P2Y1R induces the release of ATP from astrocytes; A 2A R controls the release of ATP from microglia and presynaptic terminals. Once in the extracellular millieu, ATP seems to contribute to neurotoxicity through an integrated action through P2X7R, P2Y1R, and A 2A R. P2X7R: it is well-established that P2X7R antagonism is beneficial by preventing the neurotoxic processing and release of IL-1β from microglia; yet a deleterious action through astrocytes namely through the regulation of glutamate levels or pro-inflammatory cytokines, or a direct neurotoxic action cannot be discarded. P2Y1R: the contribution of P2Y1R to brain demises has been mainly associated to astrocytic reactivity through Ca 2+ -waves and through an astrocytic-driven release of glutamate; this may be further promoted by direct actions on neuronal and synaptic function. A 2A R: there is gain of function of A 2A R particularly targeted to synapses in different brain disorders, where A 2A R either with a presynaptic or postsynaptic locus of action, has been associated to synaptic dysfunction/loss; the precise mechanisms remain to be identified.
In summary, the observed gain of function of P2X7R in pathological conditions, suggests that P2X7R may essentially act as a danger sensor shared by different brain disorders, contributing to the progression of brain diseases through a combined neurotoxic overactivation of microglia, also involving astrocytic-mediated or direct neurotoxic actions (Figures 1, 2).

P2Y1 Receptor
P2Y1R is a metabotropic receptor preferentially activated by ADP, which pharmacological or genetic blockade affords neuroprotection in ischemic conditions (Sun et al., 2008;Kuboyama et al., 2011;Chin et al., 2013;Carmo et al., 2014b) or trauma (Choo et al., 2013). P2Y1R have a widespread cellular distribution and modulate neurons (Bowser and Khakh, 2004;Guzman et al., 2010), astrocytes (Fam et al., 2003;Fumagalli et al., 2003;Zheng et al., 2013) and microglia (Boucsein et al., 2003;Ballerini et al., 2005;Bianco et al., 2005). However, the pathological role of P2Y1R has been predominantly associated to reactive astrocytes since P2Y1R play a key role in entraining the propagation of calcium waves throughout the astrocyte network (Fam et al., 2003;Neary et al., 2003;Bowser and Khakh, 2007) and promote astrocytic hyperactivity and astrogliosis upon mechanical injury , ischemic conditions (Sun et al., 2008) or AD (Delekate et al., 2014), which is known to interfere with neuronal repair and regeneration (McKeon et al., 1999;Tian et al., 2006). The neuroprotection resulting from P2Y1R blockade might also involve the ability of P2Y1R to control GABA uptake (Jacob et al., 2014) and glutamate release (Domercq et al., 2006) impacting on synaptic function (Jourdain et al., 2007;Santello et al., 2011), and to regulate inflammatory/trophic factors expression in astrocytes (Kuboyama et al., 2011). However, in line with the existence of multiple populations of P2Y1R with different functions in astrocytes operating different transducing pathways (Fam et al., 2003;Sun et al., 2008;Kuboyama et al., 2011;Zheng et al., FIGURE 2 | Schematic diagram of the actions of P2X7R, P2Y1R and A 2A R in brain pathologies. Extracellular ATP, both directly through the activation of P2X7R and indirectly through the activation of P2Y1R and A 2A R upon its extracellular catabolism into ADP and adenosine, seems to be a key signal in brain pathologies, being endowed with the unique capacity to promote and integrate neuroinflammation, reactive astrogliosis, synaptic dysfunction/loss, and increased susceptibility of neurons to damage. Here, it is summarized the different mechanisms reported for each receptor that are or may be contributing to neurodegeneration. The knowledge of the precise mechanisms and the challenging characterization of the temporal and spatial hierarchy of these different actions, perhaps as a common neurodegenerative pathway to different brain disorders, will most likely unravel an opportunity for multi-drug target therapeutics. Frontiers in Neuroscience | www.frontiersin.org 2013), the blockade or the stimulation of P2Y1R in astrocytes can cause paradoxical effects; thus, the exogenous overactivation of P2Y1R can prevent astrocytic damage (Shinozaki et al., 2006) and protect against neuronal damage induced by oxidative stress through IL-6 release (Fujita et al., 2009). This apparently paradoxical effect might also result from the up-regulation of P2Y1R in pathological conditions, such as epilepsy Padrão et al., 2011), mechanical injury , ischemia (Kuboyama et al., 2011) or AD (Moore et al., 2000), which might trigger a time-dependent gain of noxious function of P2Y1R under non-acute pathological conditions. Neuronal P2Y1R may also directly affect brain function and damage (Carmo et al., 2014b). P2Y1R are located in central synapses, where they control glutamate release (Mendonza-Fernández et al., 2000;Rodrigues et al., 2005) and NMDA receptors (Luthardt et al., 2003). P2Y1R also control calcium and potassium conductances (Gerevich et al., 2004;Filippov et al., 2006;Coppi et al., 2012) and inhibitory transmission (Bowser and Khakh, 2004;Kawamura et al., 2004), but it is unclear how these different effects impact on the functioning and viability of neuronal networks; in fact, brain insults trigger an up-regulation of neuronal P2Y1R (Moore et al., 2000) coupled to a noxious gain of function, as heralded by the selective ability of P2Y1R to inhibit cortical LTD only in hypoxic conditions (Guzman et al., 2010) and to normalize neurotransmission upon anoxic depolarization (Traini et al., 2011). Finally, microglia P2Y1R are also expected to be involved in the neuroprotection associated with P2Y1R blockade since P2Y1R modulate neuroinflammatory responses (Ballerini et al., 2005). Thus, the role of P2Y1R in neurodegeneration is likely to involve a trans-cellular network, as illustrated by the evidence that activated microglia is capable to modulate synaptic function through ATP release, which in turn stimulates astrocytic P2Y1R controlling glutamatergic gliotransmission that feeds-back to impact on synaptic activity (Pascual et al., 2012) (Figures 1, 2).
In summary, it seems that, in addition to P2X7R, P2Y1R also contribute to brain dysfunction and damage, further arguing for the role of extracellular ATP as a danger signal in brain pathology. This is further heralded by the neurotoxicity of exogenously added ATP (Ryu et al., 2002;Amadio et al., 2005;Resta et al., 2005) and by the neuroprotection afforded by non-selective P2R antagonists Lämmer et al., 2006), supporting that P2R might be valuable targets for neuroprotection Franke et al., 2006).
Notably, it has been established that the adenosine activating A 2A R is derived from the activity of ecto-5 ′nucleotidase (Cunha et al., 1996b;Rebola et al., 2008;Augusto et al., 2013), the final step in the ATP catabolism into adenosine. Furthermore, unpublished work from our group has documented that the blockade of ecto-5 ′nucleotidase or of A 2A R affords comparable neuroprotection, further heralding the concept that A 2A R activation is part of the signaling operated by extracellular ATP as a danger signal.

P2X7R-P2Y1R-A 2A R: an Hazardous Orchestra
The sustained increase of extracellular ATP levels upon brain dysfunction/damage together with the compelling evidence that the pharmacological blockade or genetic deletion of P2X7R or P2Y1R or A 2A R prevents or attenuates neuronal injury or the onset/evolution of brain diseases, supports a role for ATP both as a warning and harmful signal in the brain. It will now be important to understand the time-dependent involvement of these three purinoceptors and their inter-play. In fact, the activation of A 2A R or P2X7R may constitute an auto-stimulatory loop (Verderio and Matteoli, 2001;Cunha et al., 2012) since they can trigger ATP release from astrocytes, neurons or microglia (George et al., 2015), either directly through the P2X7R pore (Duan and Neary, 2006), through interaction with pannexin channels (Locovei et al., 2007;Iglesias et al., 2008;Bennett et al., 2012), or by exocytotic release (Gutiérrez-Martín et al., 2011). Furthermore, P2X7R synergistically regulate P2Y1R activation (Locovei et al., 2006), particularly in pathological conditions (Traini et al., 2011;Vessey et al., 2011;Choo et al., 2013). Finally, emerging evidence indicates a synergic interplay between ATP and its metabolite adenosine (Gerwins and Fredholm, 1992;Neary et al., 1998;Chevrier et al., 2006;Färber et al., 2008;Koizumi et al., 2013;George et al., 2015), namely between A 2A R and P2X7R (Chen et al., 2004;Pellegatti et al., 2011) and P2Y1R (Stafford et al., 2007;Doengi et al., 2008;Suzuki et al., 2011), which highlights the possible key role of ectonucleotides in regulating the integration of purinergic responses. Thus, the action of individual purinergic receptors may be part of a time-dependent orchestrated response triggered by the increase of extracellular ATP levels in brain pathology (Figure 2).
The understanding of the hierarchy and integration/redundancy of their actions will be paramount to develop multi-target therapeutics to exploit this role of ATP as a danger signal in the brain.