GABAB Receptor Isoforms Caught in Action at the Scene
2006; Cell Press; Volume: 50; Issue: 4 Linguagem: Inglês
10.1016/j.neuron.2006.05.005
ISSN1097-4199
Autores Tópico(s)Photoreceptor and optogenetics research
ResumoThe metabotropic GABAB receptors mediate slow synaptic inhibition and consist of heterodimers of GABAB1 and GABAB2 subunits. The only known molecular diversity of the GABAB receptors arises from the two GABAB1 isoforms, but its functional significance has been unclear. Two studies in this issue of Neuron now demonstrate that GABAB1a and GABAB1b show strategically distinct subcellular localization and physiological action. The metabotropic GABAB receptors mediate slow synaptic inhibition and consist of heterodimers of GABAB1 and GABAB2 subunits. The only known molecular diversity of the GABAB receptors arises from the two GABAB1 isoforms, but its functional significance has been unclear. Two studies in this issue of Neuron now demonstrate that GABAB1a and GABAB1b show strategically distinct subcellular localization and physiological action. Although GABA is the only major inhibitory neurotransmitter in the vertebrate brain, there are many different modes of inhibition, which act in concert to control synaptic integration, spike generation, and nearly all aspects of circuit activity. The GABAergic system seems to have deployed at least two strategies to greatly enrich the action of GABA. First, a different "flavor" of GABA is released by a rich array of interneuron subtypes at distinct spatial and temporal niches in the neural circuit (e.g., at different subcellular locations and precisely defined time windows during circuit operation). Second, different physiological effects of GABA are transmitted by a large variety of GABA receptors. For example, the fast component of GABAergic inhibition is mediated by the ionotropic GABAA receptors. The GABAA receptor family includes at least 17 genes. Each functional receptor consists of a pentamer of different subunits, which allows combinatorial coding of different biophysical and pharmacological properties. In addition, GABA also activates slow synaptic inhibition through the metabotropic GABAB receptors, which are coupled to hetertrimeric G proteins. Activation of presynaptic GABAB receptors located on GABAergic terminals (autoreceptors) or other nerve terminals (heteroreceptors) suppresses neurotransmitter release, whereas the stimulation of postsynaptic receptors produces a prolonged neuronal hyperpolarization. Although biochemical and pharmacological studies have long suggested the presence of diverse GABAB receptor subtypes (Kerr and Ong, 1995Kerr D.I. Ong J. Pharmacol. Ther. 1995; 67: 187-246Crossref PubMed Scopus (226) Google Scholar), molecular cloning has only identified two genes encoding receptor subunits: GABAB1 and GABAB2 (Bettler et al., 2004Bettler B. Kaupmann K. Mosbacher J. Gassmann M. Physiol. Rev. 2004; 84: 835-867Crossref PubMed Scopus (683) Google Scholar). It is now fairly well established that most functional GABAB receptors in the brain are formed as GABAB1 and GABAB2 heterodimers (Mohler and Fritschy, 1999Mohler H. Fritschy J.M. Trends Pharmacol. Sci. 1999; 20: 87-89Abstract Full Text Full Text PDF PubMed Scopus (85) Google Scholar). Therefore, the presumed diversity of native GABAB receptor subtypes in various in vivo preparations stands in contrast to the apparent simplicity of their basic molecular architecture. Two studies led by Bettler (Vigot et al., 2006Vigot R. Barbieri S. Bräuner-Osborne H. Turecek R. Shigemoto R. Zhang Y.-P. Luján R. Jacobson L.H. Biermann B. Fritschy J.-M. et al.Neuron. 2006; 50 (this issue): 589-601Abstract Full Text Full Text PDF PubMed Scopus (251) Google Scholar) and Larkum (Pérez-Garci et al., 2006Pérez-Garci E. Gassmann M. Bettler B. Larkum M.E. Neuron. 2006; 50 (this issue): 603-616Abstract Full Text Full Text PDF PubMed Scopus (208) Google Scholar) in this issue of Neuron provide insight to this conundrum and bring our understanding of the GABAB receptor system to a deeper level. The only firmly established molecular diversity in the GABAB system thus far arises from the two isoforms of the GABAB1 subunit: GABAB1a and GABAB1b (Kaupmann et al., 1998Kaupmann K. Malitschek B. Schuler V. Heid J. Froestl W. Beck P. Mosbacher J. Bischoff S. Kulik A. Shigemoto R. et al.Nature. 1998; 396: 683-687Crossref PubMed Scopus (1015) Google Scholar). However, these two isoforms seem to have very similar pharmacological and biophysical properties in vitro (Brauner-Osborne and Krogsgaard-Larsen, 1999Brauner-Osborne H. Krogsgaard-Larsen P. Br. J. Pharmacol. 1999; 128: 1370-1374Crossref PubMed Scopus (31) Google Scholar). Structurally, the only difference between the two isoforms is the presence of a pair of sushi repeats in the N-terminal ectodomain of GABAB1a (Bettler et al., 2004Bettler B. Kaupmann K. Mosbacher J. Gassmann M. Physiol. Rev. 2004; 84: 835-867Crossref PubMed Scopus (683) Google Scholar). Since sushi repeats have been shown to mediate protein interactions with a variety of cell adhesion molecules (Blein et al., 2004Blein S. Ginham R. Uhrin D. Smith B.O. Soares D.C. Veltel S. McIlhinney R.A. White J.H. Barlow P.N. J. Biol. Chem. 2004; 279: 48292-48306Crossref PubMed Scopus (57) Google Scholar), their presence in GABAB1a suggests the possibility that the two isoforms may associate with distinct auxiliary proteins for their localization and modification (Couve et al., 2004Couve A. Calver A.R. Fairfax B. Moss S.J. Pangalos M.N. Biochem. Pharmacol. 2004; 68: 1527-1536Crossref PubMed Scopus (41) Google Scholar, Mohler and Fritschy, 1999Mohler H. Fritschy J.M. Trends Pharmacol. Sci. 1999; 20: 87-89Abstract Full Text Full Text PDF PubMed Scopus (85) Google Scholar). However, most, if not all neurons co-express both GABAB1a and GABAB1b, and there has been no solid evidence for differential subcellular localization (e.g., pre- versus postsynaptic) of these isoforms. It is therefore not obvious whether and how GABAB1a and GABAB1b support the distinct physiological properties and functions in vivo. Clear answers to these important questions require a high-resolution examination of the cellular and subcellular localization and physiological action of GABAB1a (1a) and GABAB1b (1b) at specific connections in defined neural circuits, combined with specific inactivation of each isoform in vivo. Vigot et al. and Pérez-Garci et al. did exactly these experiments. GABAB1a and GABAB1b are generated by differential promoter usage of the GABAB1 gene; GABAB1b results from the presence of an alternative transcription initiation site within the fifth GABAB1a intron. Vigot et al. took advantage of this unusual genomic structure and converted the initiation codon of each isoform into a stop codon using a knockin approach. This elegant genetic manipulation inactivates one isoform at a time while leaving the other intact (the 1a−/− or 1b−/− mice). These mutant mice provide a unique opportunity to dissect the function of GABAB1a and GABAB1b isoforms. Vigot et al. focused their analysis at the CA3 to CA1 connections of the adult hippocampus. The lack of a suitable GABAB1a or GABAB1b specific antibody for ultrastructural studies has precluded the precise analysis of their subcellular localization. Using their isoform-specific mutants and a pan-GABAB1 antibody, Vigot et al. was able to quantify the subcellular distribution of each remaining isoform in either 1a−/− or 1b−/− mice by electron microscopy. They found that GABAB1b was mostly localized to dendritic spines opposite to glutamate release sites, while GABAB1a is predominantly found at glutamatergic terminals. Analysis of CA3-CA1 transmission indicated that 1a−/− but not 1b−/− mice lacked GABAB heteroreceptors on Schaffer collateral terminals; on the other hand, postsynaptic CA1 pyramidal neurons in 1b−/− but not 1a−/− mice showed greatly reduced slow inhibitory current induced by the GABAB agonist baclofen. To further examine the issue of subcellular localization, GFP-tagged GABAB1a or GABAB1b were expressed in CA1 pyramidal neurons of hippocampal organotypic cultures. Although both tagged isoforms were present in dendrites, GABAB1b–GFP was largely localized to spines while GABAB1a–GFP was largely excluded from this site. In addition, only GABAB1a-GFP was targeted to axons. Together these results strongly suggest that GABAB1a mainly assembles presynaptic heteroreceptors inhibiting glutamate release, while GABAB1b receptors mainly mediate postsynaptic inhibition. In addition, the sushi repeats in GABAB1a may contribute to heteroreceptor localization. To explore whether GABAB1a and GABAB1b play distinct roles in synaptic plasticity, long-term potentiation at CA3-CA1 synapses was measured in 1a−/− and 1b−/− mice. Although 1b−/− mice exhibited normal LTP, 1a−/− mice showed significantly impaired LTP. To examine the behavioral consequence of such LTP deficit, 1a−/− and 1b−/− mice were subjected to a hippocampal-dependent object recognition task. While 1b−/− mice showed similar performance as wild-type mice in discriminating between familiar or novel objects, 1a−/− mice were impaired in this task. Together, these studies provide compelling evidence that GABAB1a and GABAB1b indeed have distinct functions in synaptic physiology and behavior, and deficiency in one isoform cannot be compensated for by the other. On the other hand, since LTP can be induced even after acute pharmacological blockade of GABAB receptors, the LTP deficits in 1a−/− mice most likely result from adaptive changes following germ-line inactivation of this isoform. Consistent with this interpretation, The proportion of silent synapses was decreased in the hippocampus of 1a−/− mice. This finding therefore revealed an interesting compensatory mechanism by which developmental GABAB receptor deficit results in behavioral abnormality through altering a plasticity process. While Vigot et al. provided strong evidence for different functions of GABAB1 isoforms in hippocampus, the study by Pérez-Garci et al. in the accompanying paper nailed this issue at the cellular level with a particularly clean case, by demonstrating that GABAB1a and GABAB1b play strategically distinct physiological roles in neocortical neurons (Figure 1). The success of this study can be attributed both to the choice of studying synaptic inputs onto layer 5 (L5) pyramidal neurons in neocortex, where 1a and 1b turn out to have cleanly divided their physiological tasks, and to the elegant physiological paradigm that allowed interrogation of these neurons with whole-cell patch recording at multiple subcellular sites. L5 pyramidal neurons extend their axons and dendrites to all cortical layers and are characterized by their striking polarity and intrinsic compartmental architecture both at the anatomical and physiological levels. They receive convergent excitatory inputs from layer 1 (L1) fibers which carry top-down information from higher cortical areas, and information from thalamocortical pathways and other cortical areas. They are unusual in having both an axonal and a dendritic zone for the initiation of action potentials (Larkum et al., 2001Larkum M.E. Zhu J.J. Sakmann B. J. Physiol. 2001; 533: 447-466Crossref PubMed Scopus (289) Google Scholar, Figure 1). Distal dendritic inputs must evoke a calcium action potential (Ca2+-AP) at the dendritic initiation zone, which can propagate forwardly and generate a burst of axonal action potentials. On the other hand, back-propagating action potentials from the axon facilitates the initiation of these Ca2+-APs when it coincides with the distal inputs within a time window of several milliseconds (Larkum et al., 1999Larkum M.E. Zhu J.J. Sakmann B. Nature. 1999; 398: 338-341Crossref PubMed Scopus (777) Google Scholar). Such a temporal requirement of distal inputs with ongoing somatic spiking to facilitate dendritic Ca2+-APs has been suggested as a mechanism by which cortical neurons associate inputs arriving at different cortical layers. In addition, the apical tuft of L5 pyramidal neurons is innervated by a large number of inhibitory inputs with as yet unclear function (Figure 1). The interaction of excitatory and inhibitory actions on the distal apical dendrite is likely a rich source of computation possibilities. For example, Ca2+ spikes in L5 pyramidal neurons are exceedingly susceptible to local inhibitory inputs. A single action potential from an interneuron in L2/3 can abolish dendritic Ca2+ spikes without affecting the generation or back-propagation of axonally initiated sodium action potentials (Na+-APs) (Larkum et al., 1999Larkum M.E. Zhu J.J. Sakmann B. Nature. 1999; 398: 338-341Crossref PubMed Scopus (777) Google Scholar), thereby would decouple inputs which would otherwise associate in L5 pyramidal neurons. Surprisingly, the effect of inhibition can last for over 400 ms (Larkum et al., 1999Larkum M.E. Zhu J.J. Sakmann B. Nature. 1999; 398: 338-341Crossref PubMed Scopus (777) Google Scholar). The cellular and molecular mechanism underlying this powerful and long-lasting inhibition is unknown. In the current study, Pérez-Garci et al. demonstrate the role of GABAB receptors, specifically the GABAB1b isoform, in mediating the long-lasting inhibition of dendritic Ca spikes in L5 pyramidal neurons. Their experiments mainly involved using multiple patch electrodes to trigger and measure electrical signals from the apical tuft, dendrite, and soma of L5 pyramidal neurons, combined with calcium imaging at subcellular locations. Inhibition was generated from extracellular stimulation in L1. GABAB receptors were manipulated either by local puffing of agonists/antagonists onto defined subcellular regions or by the use of GABAB1 isoform-specific mutant mice. To substantiate previous findings, Pérez-Garci et al. first measured the entire time window of L1 triggered inhibition in blocking dendritic Ca2+ spikes and showed that it lasted for up to 450 ms. Importantly, such inhibition recruited by L1 stimulation is restricted to the distal dendrites and did not shunt the back-propagating Na+-APs, suggesting that either a dendritically targeted GABAergic input or a dendritic distribution of specific GABA receptors may mediate this form of slow inhibition in L5 pyramidal neurons. Using pharmacological methods, they further demonstrated that the inhibition consists of two components: GABAA receptors mediate the short ( 150 ms) component. GABAB receptors are known to activate postsynaptic K+ currents and also inhibit Ca2+ currents. By pharmacologically isolating the Ca2+ and K+ currents, Pérez-Garci et al. was able to show that GABAB activation directly inhibited Ca2+ conductances that participate in dendritic spiking. To examine whether a specific isoform of GABAB receptor is responsible for the inhibition of dendritic Ca2+ spike, Pérez-Garci et al. took full advantage of the 1a−/− and 1b−/− mice generated by Vigot et al. The result was particularly satisfying: L5 pyramidal neurons from 1a−/− mice were the same as those from wt mice, which displayed both short and long lasting components of inhibition. In contrast, L5 pyramidal neurons from 1b−/− mice completely lacked the long-lasting inhibitory component, while the short, GABAA-mediated component was intact. In addition, local puffing of baclofen only to the dendrites but not the soma elicited GABAB response. Therefore, GABAB1b is not only specifically involved in mediating the L1-triggered long lasting inhibition in L5 pyramidal neurons, but also is preferentially targeted to distal dendrites (Figure 1). An interesting surprise came when Pérez-Garci et al. took a closer look at the biphasic IPSPs evoked by L1 stimulation: GABAB antagonist in fact significantly increased the amplitude of the fast IPSP component. This is best explained by the disinhibition of GABAergic terminals by the blockage of presynaptic GABAB autoreceptors, resulting in more GABA release. Consistent with this notion, presynaptic inhibition was absent in 1a−/− but not 1b−/− mice, suggesting that GABAB1a exclusively makes up the presynaptic autoreceptors in the inhibitory terminals in this circuit (Figure 1). Therefore, when examined at exactly the right place and the right time, strategically distinct physiological roles of GABAB receptors can be revealed as a clean segregation of 1a and 1b isoforms at the cellular and subcellular levels. Is L5 pyramidal neuron the exception or the rule? Only more elegant studies like those of Pérez-Garci will tell. Together, these two studies provide compelling evidence for the cell biological, physiological, and functional distinctions of GABAB1 isoforms. They also raised many more questions. First, what is the mechanism that targets 1a and 1b to different subcellular locations? Since the only structural difference between GABAB1a and GABAB1b lies in the sushi repeats, extracellular interactions with putative auxiliary proteins may determine their subcellular localization (e.g., by selective transport or retention). In fact, the two sushi repeats in 1a have strikingly different structural properties and participate in protein interactions with multiple partners (Blein et al., 2004Blein S. Ginham R. Uhrin D. Smith B.O. Soares D.C. Veltel S. McIlhinney R.A. White J.H. Barlow P.N. J. Biol. Chem. 2004; 279: 48292-48306Crossref PubMed Scopus (57) Google Scholar), which may generate additional heterogeneity in the GABAB receptor system. It is possible that the extracellular domain of GABAB1 isoforms may interact with proteins not only on the same cell but also those on the synaptic partners or in the extracellular matrix to achieve proper subcellular localization. Second, are different isoforms of GABAB receptors at different subcellular locations (e.g., dendritic shaft versus spine of CA1 pyramidal neurons) also preferentially exposed to distinct subtypes of GABA terminals? There is evidence that only certain subtypes of interneurons activate GABAB receptors. Neurogliaform cells in the neocortex are such an example and appear to preferentially innervate GABAB receptor-containing dendritic spines (Tamas et al., 2003Tamas G. Lorincz A. Simon A. Szabadics J. Science. 2003; 299: 1902-1905Crossref PubMed Scopus (260) Google Scholar). Is it possible that GABAB1 isoforms might contribute to a matching of pre- and postsynaptic sites through organizing extracellular protein interactions? Third, the learning and memory deficits in 1a−/− mice may have resulted from altered developmental plasticity processes due to constitutive germ-line knockout. To further pin-point the precise physiological and behavioral functions of GABAB1 isoforms, conditional inactivation of 1a and 1b in specific neural circuits in the mature brain is necessary. Finally, although the molecular identities and functions of two distinct GABAB subtypes are finally recognized by these studies, it is still difficult to explain the apparently more diverse GABAB physiological responses in vivo (Kerr and Ong, 1995Kerr D.I. Ong J. Pharmacol. Ther. 1995; 67: 187-246Crossref PubMed Scopus (226) Google Scholar). It is possible that further functional variations of GABAB receptors may arise from the modification of these two "prototype" GABAB1 isoforms, for example, by auxiliary proteins and post-translational mechanisms. Identification of GABAB receptor-interacting proteins and characterization of their expression will undoubtedly provide further insight into the finer organization of the GABAB system. Differential Compartmentalization and Distinct Functions of GABAB Receptor VariantsVigot et al.NeuronMay 18, 2006In BriefGABAB receptors are the G protein-coupled receptors for the main inhibitory neurotransmitter in the brain, γ-aminobutyric acid (GABA). Molecular diversity in the GABAB system arises from the GABAB1a and GABAB1b subunit isoforms that solely differ in their ectodomains by a pair of sushi repeats that is unique to GABAB1a. Using a combined genetic, physiological, and morphological approach, we now demonstrate that GABAB1 isoforms localize to distinct synaptic sites and convey separate functions in vivo. Full-Text PDF Open ArchiveThe GABAB1b Isoform Mediates Long-Lasting Inhibition of Dendritic Ca2+ Spikes in Layer 5 Somatosensory Pyramidal NeuronsPérez-Garci et al.NeuronMay 18, 2006In BriefThe apical tuft of layer 5 pyramidal neurons is innervated by a large number of inhibitory inputs with unknown functions. Here, we studied the functional consequences and underlying molecular mechanisms of apical inhibition on dendritic spike activity. Extracellular stimulation of layer 1, during blockade of glutamatergic transmission, inhibited the dendritic Ca2+ spike for up to 400 ms. Activation of metabotropic GABAB receptors was responsible for a gradual and long-lasting inhibitory effect, whereas GABAA receptors mediated a short-lasting (∼150 ms) inhibition. Full-Text PDF Open Archive
Referência(s)