The first GPCR translatome published was the one of the ET₁R induced by endothelin-1 (ET1) in cardiomyocytes (22). Cardiomyocyte hypertrophy is associated with increased cell size and myofibrillar content and the authors investigated the ET₁-dependent relationships between the early signaling/gene expression (IEG) and the late gene expression in the established phenotype. To evaluate how the variations in early transcriptional response to ET1 were reflected by mRNA translation, the profiles of total and polysomal RNAs have been compared. It appeared that 67% of variations in mRNA content were also reflected at the translational level. Hence, most of the mRNA transcribed from IEG appeared also translated in response to ET1. However, 17% of mRNA were increased to a greater extent in the polysome-associated pool than in the total transcriptome whereas some others were excluded, which is indicative of regulations of the translational machinery by ET1R-induced signaling to promote mRNA-selective translation, as the authors confirmed later (23).

More recently, the unfolded protein response (UPR), a mechanism that maintains protein quality in secretory cells, has been explored in the LβT2 gonadotrope cell line derived from the anterior pituitary. This study has revealed that GnRHR activation enhances the translation of selective mRNA, such as the one encoding Dusp1, which is assumed to participate in the decoding of GnRH pulsatility in gonadotrope cells (24). Simultaneously, ligand-bound GnRHR also appeared to induce a pause in the translation of several mRNA involved in reproduction, such as the ones encoding the LH β and α chains of gonadotropins, in the endoplasmic reticulum-associated polysomes (25). By these combined means, GnRHR stimulation would fine-tune both protein quality and quantity in secretory gonadotrope cells.

By using the TRAP technology, the effect of FSH and LH on the regulation of spermatogenesis by somatic gonadal cells has been assessed (26). AMH-cre and Cyp17i-cre mice have been crossed with mice expressing HA-tagged RpL22, in Sertoli cells and in Leydig cells, respectively. Polysome-associated mRNA were identified on Affymetrix microarrays in both cell types in isolation. While no increase in the FSH translatome was observed, LH altered the basal Leydig cell translatome. For example, as soon as after 1 h, the LH signal enhanced the translation of the Nr4a1 et Egr1 transcription factors, and of the Rgs2 cell cycle regulator. After 4 h, the number of translated mRNA still increased, showing that, in the adult, LH regulates mRNA translation in the seminiferous tubules.

The translational control by GPCRs has been the most frequently studied in neurons. One of the reasons for this interest is that the subcellular localization of translation in these highly polarized and organized cells is a critical aspect of GPCR physiological function in the nervous system. Hence it deserves consideration, particularly in synaptic plasticity (see below), such as long-term depression (LTD). LTD is an activity-dependent decrease in synaptic tone that is mediated notably by ionotropic or metabotropic glutamate receptors (iGluR and mGluR, respectively). mGluRs, but not iGluRs, are GPCRs, and at the hippocampal synapses, the specific activation of group I mGluRs (mGluR₁ and mGluR₅) induces LTD via local protein neosynthesis (27). Furthermore, the phosphorylation of the eIF2α translation initiation factor is a major effector of mGluR-induced LTD in these neurons (28). By combining polysome profiling and deep-sequencing, the authors showed that 3,5-dihydroxyphenylglycine (DHPG), the selective agonist for mGluR₁ and mGluR₅, induced eIF2α phosphorylation that correlated with not only a general silencing of translation but also significantly enhanced the selective translation of several mRNA such as the one encoding oligonephrin 1 (Ophn1). Injection of Ophn1 shRNA in mouse hippocampus prevents LTD formation and impairs mouse performance in a hippocampal learning task. Altogether, mGluR₁/₅ activation, eIF2α phosphorylation, and Ophn1 translation contribute to remove AMPA-type Glutamate receptors (AMPARs) from the cell surface, leading to depression of AMPAR-mediated excitatory postsynaptic current (28). mGluR-LTD-dependent local translation of activity-regulated cytoskeleton-associated protein (Arc) in the dendrites is also involved in AMPAR endocytosis (29).

Mouse lines have been genetically engineered to express a GFP-tagged ribosomal protein L10a under the control of a defined locus, in a specific cell type, by using BAC vectors (30). Following anti-GFP immunoprecipitation, ribosome-bound transcripts have been sequenced. This TRAP approach has been initially designed to discriminate the translatome of different neuronal subtypes that are morphologically indistinguishable, namely striatonigral and striatopallidal medium spiny neurons using 2 GPCR loci: cell-type selectivity of L10a-GFP expression is controlled either by dopamine D₁ (striatonigral) or D₂ (striatopallidal) receptor locus (30). By these means, specific mRNA associated with ribosomes of either striatopallidal (e.g., Adk, Plxdc1, BC004044, et Hist1h2bc) or in striatonigral (e.g., Slc35d3, Zfp521, Ebf1, Stmn2, Gnb4, et Nrxn1) neurons have been identified. Hence, the TRAP method has paved the way for in vivo studies of mRNA translation profiling in selective cell-types of native tissues, in physiological or pathological conditions.

The discovery of GPCR translatome in increasing number should help to identifying new secondary structures in mRNA UTRs. By these means, new advances should be made on how some mRNA subpopulations bearing similar binding motifs for common cargos may be transported in the intracellular space, to be cotranslated locally.

For site-specific translation to occur, mRNAs need to be transported from the nucleus to the specific intracellular location where they will be ultimately translated, while preventing premature translation during their transport. In the nucleus, a ribonucleoprotein granule (RNP) gathers ribonucleoproteins (RBPs) and mature mRNAs (Figure 1). RBPs may recognize a structural motif, a so-called zipcode, in the 3′ or 5′ UTRs of the mRNAs, through a canonical RNA-binding domain (37). Recognition of mRNA sequences by proteins with low-complexity domains has emerged as a means to dynamically integrate mRNA into RNP granules to form an mRNP (38, 39). Cross-linking immunoprecipitation (CLIP) analyses have revealed that some RBPs, such as fragile X mental retardation protein (FMRP), are able to bind hundreds of different mRNAs, while others only bind one. mRNA passage through the nucleopores is facilitated by export receptors such as the TAP-15 complex that associate directly with mRNA-bound RBP such as Aly/REF adaptors, and dissociate from the RNP granule once in the cytoplasm (40). Then, the silent mRNP complex is transported by motor proteins along the actin or microtubule cytoskeleton tracks toward the site of translation. Among others, the Staufen protein could be a general regulator of mRNA transport (41), in Drosophila embryo as well as in neurons (42). Another well-described RBP is the ZBP1 protein that silences the β-actin mRNA to avoid its premature translation in the cytoplasm, prior to its arrival at fibroblast protrusions (43). Upon arrival at the site of translation, the Src protein phosphorylates ZBP1, which leads to its dissociation from the mRNP, hence making the mRNAs accessible to the translational machinery (44). Extracellular signaling events are presumably the key step that unmasks the mRNAs enabling their local translation. Besides budding yeast, Xenopus oocyte, and Drosophila embryo, cells of the nervous systems have become the paradigm to study localized mRNA translation in Mammals. Notably, the mRNA content is qualitatively different not only in dendrites, axons, and cell body but also among several dendritic spines of the same dendrite. Likewise, the distribution of neuronal GPCR content differs in the plasma membrane of the soma, the dendrites or the axon terminals (45). For example, in the nucleus accumbens, D₁ and D₅ dopaminergic receptors are located mainly in dendritic spines and axonal terminals, respectively (46). Their distribution determines their postsynaptic function, their responsiveness to local neurotransmitter release and, more generally, their role in regulating neuronal activity (Figure 1). In synapses, silencing by components of the miRNA processing machinery have also been invoked in translational repression. Interestingly, neuronal activity could relieve this silencing by degrading miRNA locally, as shown in Drosophila (47).

The hypothesis that mRNA might be transported to specific subcellular compartments in neurons has emerged from the observation that the mRNA content of dendrites and axons differs. mRNA prelocalization would enable rapid production of proteins at the activated synapse, independently of the inactive synapses of the same neuron, which is important for synaptic plasticity and memory. For example, the mRNA encoding microtubule-associated protein 2 (MAP2) is transported to the dendrites, whereas the tubulin mRNA is located in the axons and dendrites (48). Accordingly, the presence of polyribosomes, translation factors and mRNA at the base of dendritic spines has been reported (49–51). In addition, protein synthesis inhibitors have revealed the central role of local translation in synaptic plasticity, in the dendrites of CA1 pyramidal neurons (52). The mRNA content of the dendrites is dynamically regulated. mGluRs are the GPCRs the most extensively reported to control local translation in neurons. Their stimulation by the DHPG mGluR₁/₅ agonist leads to actin-related protein mRNA redistribution to dendritic spines, that permits spine remodeling upon synaptogenesis. This process is in part mediated by the RNA-binding protein TLS/FUS (53, 54). TLS/FUS harbors a low complexity sequence domain involved in the dynamic aggregation of RNA-binding proteins to form RNP granules (38, 55). These regulatory events lead to the rapid withdrawal of excitatory synapses in the neurons of the hippocampus and striatum. Furthermore, in the hippocampus, dopamine D₁/D₅ receptor signaling alters synaptic plasticity via the translation of the GluR1 subunit of AMPA iGluR, which leads to local activity at synapses that were otherwise silent (56). Likewise, mRNA translation of AMPAR is stimulated during β1-adrenergic receptor-primed long-term potentiation in the CA1 hippocampal region, in a PKA and ERK-dependent manner (57), ultimately leading to its enhanced location at the plasma membrane and synaptic incorporation.

Protein synthesis could serve as one of the gates for synaptic plasticity that physiologically operates when mGluRs are activated. FMR1 knock-out mice models are useful tools to investigate activity-dependent translation in synaptic plasticity. In the fragile X syndrome, a monogenic form of autism spectrum disorder, the loss-of-function of FMRP encoded by the FMR1 gene leads to excessive translation of proteins, that would otherwise be rate-limiting for synaptic remodeling. Examination of the phenotype reveals defects in synaptic plasticity associated with an exacerbated mGluR-LTD (58–60). These observations are consistent with a direct silencing role of FMRP (Fragile X mental retardation protein) on neuronal translation (61). For example, the Chrm4 mRNA encoding the muscarinic acetylcholine receptor 4 (M₄) is excessively translated in Fmr1^−/y mice occluding mGluR₁/₅ activation (62). Unexpectedly, enhancement of M₄ activation is required to correct the excessive translation and mGluR-LTD pathological phenotype of these mice.

FMRP is an RNA-binding protein involved in mRNA trafficking and translation. The FMR1 knock-out mice model has also highlighted the association of aberrant synthesis of proteins involved in mGluR-dependent synaptogenesis such as PSD-95, MAP1b or CAMKIIα and defective long-term plasticity (63–65).

In contrast to local translation in dendrites, much less is known about local translation in axons. Both dendritic spines and axonal terminals are specialized subcellular compartments located far away from the cell body. In growing axons, local mRNA translation might be involved in collapse or expansion of the growth cone during axon guidance whereas in mature axons, it might support regeneration of the nervous fiber upon injury. Most of these responses are dynamically regulated by signals emanating from non-GPCR receptors. However, in cortical neurons of the developing brain, it has been reported that glutamate enhances local protein synthesis, by interacting with both iGluRs and mGluRs in axons (66). In addition, in an in vitro model of Aplysia sensory/motor neurons, the local translation of the eEF1A mRNA has been involved in maintaining newly grown synapses, a prerequisite for long-term facilitation required for memory storage. Serotonin stimulation supports this site-specific targeting of the eEF1α mRNA if it is applied at the synaptic site solely (67).

Cells of the nervous systems are highly interconnected and local translation can be viewed as one means whereby they communicate. This is the case at synaptic junctions, including synaptic communications between neurons and oligodendrocytes. These glial cells synthesize the myelin sheath, to wrap the axons, in order to increase the propagation speed of the nervous influx. The mRNA that encodes myelin basic protein (MBP), a major component of myelin, is transported to distal regions of oligodendrocytes for local translation and delivery to the membrane of the adjacent axon (68). Vesicular release of glutamate from activated axons of mouse dorsal root ganglion neurons has been shown to stimulate the local translation of MBP, by using a photoconvertible fluorescent protein. The glutamate signal is sensed by both AMPA iGluR and mGluRs on oligodendrocytes (69). Recently, local translation in distal perisynaptic processes of astrocytes has also been observed (70), suggesting that the proteins these cells secrete are produced locally prior to release. However, the sensitivity to neurotransmitters of this process has not been explored yet.

More than 1,000 mRNAs have been shown to exhibit site-specific translation in migrating fibroblasts (33), which supports a broader role of this process in cellular monitoring of the proteome in time and space than initially appreciated. Beside promoting the migration of a wide spectrum of motile cells of hematopoietic type, the CXCR₄ chemokine receptor also triggers the migration of non-hematopoietic cells, such as fibroblasts. Local translation in migrating fibroblasts is exemplified by the β-actin mRNA, which displays an asymmetric distribution at the leading edge where its mRNA is translated upon ZBP1 release, as described above, in order to enable cell motility. Recently, CXCR₄ has been shown to interact with the eukaryotic initiation factor 2B (eIF2B) (71). This protein is an exchange factor that negatively regulates translational efficacy. Binding of the CXCR4 agonist, SDF1/CXCL12, leads to the dissociation of eIF2B from the receptor. This observation suggests that the local release of this translation factor could then be utilized locally, close to the plasma membrane, to enable local translation of the β-actin mRNA. Early work already showed that eIF2B also interacts with the β₂-adrenergic receptors in protruding regions of the cell membrane (72).

Evidence is lacking that other GPCRs might be involved in local translation of actin-related proteins needed for cell motility. But, noteworthy, β-arrestins, major effectors that regulate the efficacy, duration and location of GPCR-responsive signaling pathways (73, 74), are involved in chemotaxis (75). Despite their ability to redistribute signaling components to selective subcellular compartments, a role of these adaptor proteins in local translation in migrating cells has not been addressed yet.

As discussed above, it appears clearly that most studies to decipher the control of site-specific mRNA translation by GPCR have been undertaken in neurons. However, local translation processes virtually take place in any cell type, such as epithelial cells from the kidney, the gut, the skin, provided that they exhibit some degree of polarization and functional specialization. For example, Sertoli cells represent an ideal model for studying local mRNA translation. They are polarized cells that constitute the seminiferous tubules of the male gonad and the intimate association they undergo with the successive steps of spermatogenesis has prompted investigators to refer to this histological architecture as the “testicular synapse”. One of their major regulatory GPCR is the FSHR that stimulates Sertoli cell anabolic activity to provide each spermatogenic cell with the adapted complement of paracrine factors. Recently, the FSHR has been reported to promote mRNA-selective translation of some mRNAs such as the vegfa mRNA (19) that is involved in spermatogonia renewal occurring at their basal region. However, despite many analogies to neurons, hormone-induced local translation has never been demonstrated in these cells.