The discovery that receptors from all families can establish allosteric receptor–receptor interactions and variably associate to form receptor complexes operating as integrative input units endowed with a high functional and structural plasticity has expanded our understanding of intercellular communication. Regarding the nervous system, most research in the field has focused on neuronal populations and has led to the identification of many receptor complexes representing an important mechanism to fine-tune synaptic efficiency. Receptor–receptor interactions, however, also modulate glia–neuron and glia–glia intercellular communication, with significant consequences on synaptic activity and brain network plasticity.
In the early 1980s, however, in vitro and in vivo experiments  provided indirect evidence that GPCRs may also establish structural receptor–receptor interactions (RRI), leading to the formation at the cell membrane of multimeric receptor complexes (see  for reviews) that operate as integrative input units . In the years that followed, direct evidence for the existence of this structural organization was provided by several groups , and the amount of data supporting the existence of GPCR oligomers further increased when biophysical techniques capable of detecting the spatial proximity of protein molecules became available .
These findings demonstrated that GPCRs can signal both as monomers and as part of receptor complexes and indicated that oligomeric organization represents a quite common feature in the different receptor families, with the ion channel receptors (where multimerization is needed) lying at one end of the spectrum and GPCRs at the other . Receptor channels, indeed, are constitutively multimeric , the majority of nuclear hormone receptors operate as homo- or hetero-dimers  and, with few exceptions , receptor tyrosine kinases need dimerization for their activation . Thus, as pointed out by Changeux and Christopoulos in a detailed review , oligomerization emerges as an efficient mechanism for tuning the functionality of receptor proteins, including those able to signal as monomers, such as GPCRs. In this respect, recently reported evidence for receptor complexes involving protomers from different families  is also of significant interest.
RRI at the cell membrane have expanded our understanding of intercellular communication and they appeared to play a major role in the physiology and pathology of many districts of the body (see  for a review). Examples include the regulation of vascular homeostasis through the angiotensin II AT 1 receptor and its heterodimers , the chemoreceptor function of the carotid body  and the endocrine system, where a growing number of reports suggested receptor oligomerization as a significant aspect of endocrine regulation . The possibility of pharmacological strategies targeting receptor heteromers has also been proposed in oncology . However, the largest body of available data concerns the central nervous system (CNS). The formation of receptor complexes, indeed, is considered of key importance in neurophysiology (see  for more specific reviews), since the integration of input signals already at the level of the plasma membrane significantly helps to tune synaptic efficiency. Furthermore, increasing evidence indicates receptor complexes as potential targets for the treatment of serious diseases of the CNS .
In this context, glial cells, the non-action potential generating cells in the CNS, received less attention. More recently, however, the increased evidence that glial cells are not merely a support to neuronal life, but are actively involved in neuronal development, function and synaptic plasticity , generated an intense research interest focused on the mechanisms of glia–neural communication with significant new findings on the role played by receptor–receptor interactions in this process. Thus, after a brief recapitulation of the basic aspects concerning the structural biology of receptor complexes and their signaling, the available data on the role RRI play in the intercellular communication involving glial cells will be the focus of the present review article.
It is well known that receptors can interact in a functional sense by sharing signaling patterns or by mechanisms of transactivation, even without coming into physical contact with each other . The term RRI, on the contrary, indicates a type of interaction requiring direct physical contact between the receptors involved, leading to the formation of receptor complexes at the cell membrane. In this respect, a more detailed definition was provided in 2010 by a specific international consensus workshop : “Receptor-receptor interactions: when the binding of a ligand to the orthosteric or allosteric sites of one receptor causes, via direct allosteric interactions, a change in the ligand recognition, decoding and trafficking processes of another receptor”. On this basis, it is also possible to provide an operational definition, in which the term RRI is translated into a set of experimental procedures leading to unambiguous numerical descriptions of the phenomenon . In this respect, it is possible to maintain that two receptors are involved in an RRI process when the binding of one receptor causes detectable changes in the biochemical characteristics of the partner and the two receptor molecules are located in close proximity (<10 nm). In the last few decades, several biophysical techniques have been developed to detect the spatial proximity of protein molecules (see  for more details). They include energy transfer-based methods, bimolecular luminescence or fluorescence complementation, total internal reflection fluorescence microscopy, fluorescence correlation spectroscopy, coimmunoprecipitation, assays based on bivalent ligands and in situ proximity ligation assays.
Structural plasticity, however, is important not only to allow intra-receptor interactions and conformational fluctuations, but also to enable the formation of receptor complexes and their dynamics. When protomers, indeed, establish direct RRI leading to a quaternary structure, energy perturbations occurring at some site of one protomer can propagate over the interface between receptors into the nearby protomers, changing their conformational and functional properties, thus allowing the cooperative behavior of the complex .
The establishment of these supramolecular assemblies is considered of particular importance because it allows the emergence of integrative functions performed by a receptor complex as a whole . In fact, owing to allosteric RRI, a configuration change of a given protomer will change the probability of changing the configuration for the adjacent receptors in the complex and the effect will propagate throughout the cluster, leading to complex collective behavior and to an integrated regulation of multiple effectors. These concepts have been well illustrated by mathematical models of cooperativity in receptor assemblies , based on discrete dynamics  or on thermodynamics-based approaches . In the former case, receptors are supposed to assume a limited number of configurations (e.g., only two: “active” or “inactive”) changing in the time according to a “switching rule” based on the pattern of interactions each receptor establishes with the partners in the complex. In the latter case, the transition is stochastic and depends on the estimated energy of each protomer in the complex (see  for more details). Mathematical models indicated that receptor complexes can be described as possessing “emergent properties”, i.e., biochemical and functional features that cannot be fully anticipated on the basis of the characteristics of the single receptor partners .
In a variety of receptor complexes, the modulation of the binding sites has been reported as a consequence of allosteric RRI. Examples include the heterodimer between adenosine A 2A and dopamine D 2 GPCRs , where reciprocal antagonism occurs, and the human insulin RTK , a glycoprotein existing in two dimeric isoforms that exhibit significant differences in affinity for insulin-like growth factors. Changes in the decoding of signals reaching protomers represent a second mechanism induced by allosteric RRI. This aspect seems of particular importance in GPCRs, as illustrated by the heterodimer formed by dopamine D 1 and histamine H 3 receptors , in which the D 1 receptor changes its coupling from the G s to the G i protein, or by the switch from G protein to β-arrestin signaling  documented after κ-μ and κ-δ opioid receptor oligomerization. A final relevant aspect of receptor complex formation is the possibility that novel specific allosteric sites suitable for the binding of some modulators could appear in the quaternary structure resulting from the assemblage of protomers . Thus, ligands specific to the receptor complex as such may also exist.
Depression or enhancement of synaptic plasticity may also result from cannabinoid receptor-mediated astrocyte activation and the release of gliotransmitter ATP/adenosine, as suggested by studies on the basolateral amygdala . In this respect, the identification by proximity ligation assay of cannabinoid CB 2 and GPR55 receptor complexes in the astrocytes of the dorsolateral prefrontal cortex of the human brain  is of potential interest. From the functional point of view, the results revealed an association between the expression levels of this heteromer and mood disorders, but no data are still available on the signaling features specific to this receptor complex.
Of interest in the study of neurological disorders with cognitive decline is the recent demonstration by proximity ligation assay of receptor complexes involving fibroblast growth factor receptor 1 (FGFR1) and serotonin 5HT 1A receptor in hippocampal astrocytes . The FGFR1–5HT 1A heteroreceptor complex may allow astroglial modulation of the hippocampal neurons’ gamma oscillations, a pattern of electrical activity (30–80 Hz) playing an important role in cognitive processes, such as memory storage and recall.
A quite large set of receptors allows them to detect molecular patterns associated with tissue damage, to modulate the release of cytokines and to facilitate phagocytosis. In this context, of potential interest for the present discussion are P2X (ligand-gated cationic channels) purinergic receptors. As a matter of fact, P2X 4 and P2X 7 are the dominant forms of P2X receptors expressed in microglia . Although still a matter of debate (see ), the possible occurrence of P2X 4–P2X 7 heteromers has been reported in these cells , probably allowing a more sophisticated regulation of cytokine production and early inflammatory gene expression .
In the peripheral nervous system, Schwann cells are the myelinating cells and neuregulins represent an example of axonally derived ligands interacting with cognate receptors in Schwann cells to regulate their development and proliferation . The receptors involved are erb2 and erb3, which become tyrosine phosphorylated and form erb2–erb3 heterodimers upon ligand binding .
Intercellular communication represents a key feature of living organisms, and in the nervous system it determines virtually all aspects of its function. The main mechanism of communication in biological tissues involves the interaction of chemicals and/or energy forms released from a source with specific receptors expressed by the target cells. In the last few decades, the emerging evidence that receptors from all families can establish allosteric RRI and variably associate to form receptor complexes  indicated RRI as a basic mechanism modulating and tuning intercellular communication . In a receptor complex, indeed, the configuration of each single receptor is shaped by a network of electrostatic interactions (hydrogen bonds and Van der Waals forces) defined by the presence of receptor partners, thereby enabling the complex to operate as an integrative input unit .
Several additional lines of future research, however, can be identified. In the CNS, indeed, chemical transmitters are released in two distinct transmission modes: wiring transmission and volume transmission (see  for reviews). Wiring transmission (WT) is intercellular communication mediated via physically defined connection structures. Synapses and related glial processes represent the typical example. Volume transmission (VT) occurs by the release and diffusion of chemical signals in the extracellular space defined by the intricate morphological organization of neurons, glial cells and extracellular matrix . It is primarily mediated by simple diffusion, but also by pressure waves due to the arterial pulses, thermal gradients and local electric fields . VT signals can be released from any type of brain cells and can be sensed by a relatively large number of cells, including microglia  and astrocytes . VT mainly employs the same set of signals (transmitters, peptides, ions, gases) as WT (see  for a summary table). An important finding, however, was that non-synaptic receptors are usually characterized by high affinity for the signal . Thus, an interesting topic for future research in the field could be a differential analysis of the glial and neural receptor complexes involved in the two forms of intercellular communication to assess possible differences in their signaling features. The analysis of such an issue could likely require a more detailed description of the cellular localization of the receptor complexes. As summarized before (see Table 1 ), this aspect has been addressed only to a limited extent, with the majority of available studies being aimed only at demonstrating the presence of receptor complexes in the cells. In some cell populations, however, this issue could be of significant physiological importance, as indicated by the increasing number of studies revealing the existence of functional microdomains in astrocytes (see  for a recent review). The term “microdomains” describes Ca 2+ events that are restricted to small portions of individual astrocyte territories and can either remain restricted locally or eventually propagate to the main processes and to the soma of the cells. The characterization of the panel of receptors and receptor complexes associated with these sub-cellular functional domains could, therefore, represent a key step to increase our understanding of the astrocyte role in brain function. This issue, however, poses some methodological challenges. Important techniques currently used to demonstrate receptor complexes, such as proximity ligation assays, also provide morphological information on their location. However, the obtainable resolution at light microscopy may be a limitation and the development of more suitable imaging techniques would be beneficial. In this respect, procedures based on 3D super-resolution microscopy , electron microscopy , and atomic force microscopy  supported by specific image analysis methods  have been suggested and may represent topics for further methodological development.
|Glial Cell Population||Receptor Complex||Cellular Localization||Reference|
|Astrocytes||A2A–D2||Striatal astrocyte processes|||
|GABAB–SSTR4 (probable)||Cortical astrocyte processes|||
|5HT1A–D2||Mainly cell soma|||
|mGluR3–mGluR5 (putative)||Not reported|||
|Myelinating cells||GABAB1–GABAB2||OPC–neuron contacts|||
Receptor complexes are also of interest from a pharmacological standpoint, and their pharmacology certainly represents a significant line of future research. RRI, indeed, may provide new opportunities to optimize existing pharmacological treatments or to develop completely new pharmacological strategies. In this respect, the use of agonists/antagonists of single protomers in the receptor complexes has been, to some extent, successfully explored . However, the search for receptor heteromers’ selective compounds would be of key importance to fully exploit their properties. At least three approaches could be followed to achieve this goal. The first is based on the fact that, due to a different pattern of allosteric RRI, the conformational state of a given protomer may change according to the type of complex in which it is involved . Thus, the pharmacology of some agonists/antagonists of a given protomer in terms of affinity and efficacy may show substantial differences among various types of receptor complexes. A second approach to identify receptor complex selective compounds is based on the possibility that, when the complex forms, the quaternary structure could display novel specific allosteric sites suitable for the binding of some modulators. The abovementioned effect of homocysteine on astrocytic A 2A –D 2 receptor complexes  provides an example. The use of bivalent ligands constitutes a third possible approach for targeting receptor heteromers (see  for a review). A bivalent ligand consists of two pharmacophoric entities linked by an appropriate spacer. In this way, it should be possible to target GPCR heteromers by adequate, potent, and receptor-selective pharmacophores. The work of Portoghese and collaborators on opioid receptor complexes (see ) provided a proof of principle. In this research effort, bioinformatics can be of help and MD simulations appear of particular importance in the field, since they allow the analysis of the conformational dynamics of receptors and receptor complexes in a realistic model of their biological environment, including the lipid bilayer and the extra- and intra-cellular water spaces . MD methods, however, are in general computationally demanding and require specific software and expertise. On-line resources, however, are becoming available to facilitate MD data acquisition and analysis, and some of them are specifically designed to support studies on receptor proteins .
When applied to glial receptor complexes, these pharmacological research lines can represent a topic of particular interest from a therapeutical standpoint. Indeed, as suggested by some of the available studies discussed here, they open the possibility to explore novel, glia-mediated strategies to address neurodegenerative  and functional  CNS disorders.