Glutamate Inputs to the Nucleus Accumbens: Does Source Matter?
2012; Cell Press; Volume: 76; Issue: 4 Linguagem: Inglês
10.1016/j.neuron.2012.11.008
ISSN1097-4199
Autores Tópico(s)Neurotransmitter Receptor Influence on Behavior
ResumoHow the nucleus accumbens integrates information from multiple upstream regions has been a central question for decades. In this issue of Neuron, Britt et al., 2012Britt J.P. Benaliouad F. McDevitt R.A. Stuber G.D. Wise R.A. Bonci A. Neuron. 2012; 76 (this issue): 790-803Abstract Full Text Full Text PDF PubMed Scopus (480) Google Scholar photostimulate glutamatergic axons from the amygdala, prefrontal cortex, and hippocampus in the nucleus accumbens, characterizing the functional role of each pathway both in vivo and ex vivo. How the nucleus accumbens integrates information from multiple upstream regions has been a central question for decades. In this issue of Neuron, Britt et al., 2012Britt J.P. Benaliouad F. McDevitt R.A. Stuber G.D. Wise R.A. Bonci A. Neuron. 2012; 76 (this issue): 790-803Abstract Full Text Full Text PDF PubMed Scopus (480) Google Scholar photostimulate glutamatergic axons from the amygdala, prefrontal cortex, and hippocampus in the nucleus accumbens, characterizing the functional role of each pathway both in vivo and ex vivo. The nucleus accumbens (NAc) has been described as a crucial convergence point for information about environmental contexts and cues before the selection and execution of a final motor output and has long been known to be important in the processing of reward-related behaviors (Cardinal et al., 2002Cardinal R.N. Parkinson J.A. Hall J. Everitt B.J. Neurosci. Biobehav. Rev. 2002; 26: 321-352Crossref PubMed Scopus (1635) Google Scholar; Carelli, 2002Carelli R.M. Behav. Cogn. Neurosci. Rev. 2002; 1: 281-296Crossref PubMed Scopus (107) Google Scholar), specifically in the context of cocaine-induced plasticity (Thomas et al., 2001Thomas M.J. Beurrier C. Bonci A. Malenka R.C. Nat. Neurosci. 2001; 4: 1217-1223Crossref PubMed Scopus (514) Google Scholar; Boudreau and Wolf, 2005Boudreau A.C. Wolf M.E. J. Neurosci. 2005; 25: 9144-9151Crossref PubMed Scopus (392) Google Scholar). What happens at this last stop? Three of the most robust glutamatergic inputs to the NAc are the basolateral amygdala (Amyg), medial prefrontal cortex (PFC), and the ventral hippocampus (vHipp), each probed by Britt et al., 2012Britt J.P. Benaliouad F. McDevitt R.A. Stuber G.D. Wise R.A. Bonci A. Neuron. 2012; 76 (this issue): 790-803Abstract Full Text Full Text PDF PubMed Scopus (480) Google Scholar using optogenetic methods (Figure 1). This characterization revealed many novel insights: while Britt et al., 2012Britt J.P. Benaliouad F. McDevitt R.A. Stuber G.D. Wise R.A. Bonci A. Neuron. 2012; 76 (this issue): 790-803Abstract Full Text Full Text PDF PubMed Scopus (480) Google Scholar confirmed some assumptions about these limbic systems, they challenged the dogma surrounding NAc information integration. The most provocative implication of this paper is that Britt et al., 2012Britt J.P. Benaliouad F. McDevitt R.A. Stuber G.D. Wise R.A. Bonci A. Neuron. 2012; 76 (this issue): 790-803Abstract Full Text Full Text PDF PubMed Scopus (480) Google Scholar raise "the possibility that the specific pathway releasing glutamate is not as important as the amount of glutamate that is released." This is indeed a new concept that would change the way much of the field thinks about the way that the NAc integrates information: what if the complex computations are actually much simpler than we thought? What if projection origin matters less than projection target? More than half a century ago, intracranial self-stimulation (ICSS) was first used to identify several fiber tracts, including putative hippocampal outputs, as neural substrates for reward or reinforcement (Olds and Milner, 1954Olds J. Milner P. J. Comp. Physiol. Psychol. 1954; 47: 419-427Crossref PubMed Scopus (1801) Google Scholar). However, these seminal studies used electrical stimulation—nonspecifically activating multiple cell types and axons of passage—making it difficult to determine the critical neural circuit element with confidence. In another seminal study from the 1990s, elegant in vivo intracellular recordings in anesthetized animals first characterized the role of hippocampal, prefrontal cortical, and amygdalar inputs to the NAc, demonstrating distinct properties of electrical stimulation in each upstream region (O'Donnell and Grace, 1995O'Donnell P. Grace A.A. J. Neurosci. 1995; 15: 3622-3639PubMed Google Scholar). O'Donnell and Grace established the unique ability of hippocampal inputs to the NAc to induce changes in membrane potential, commonly referred to as "up and down states"—medium spiny neurons were pushed into step-function-like states in which the cells were slightly depolarized and more excitable in response to prefrontal cortical inputs (O'Donnell and Grace, 1995O'Donnell P. Grace A.A. J. Neurosci. 1995; 15: 3622-3639PubMed Google Scholar). Distinct from the bistable responses elicited by fornix stimulation, electrical stimulation of the amygdala produced longer-lasting depolarization with greater onset latency, and electrical stimulation of the prefrontal cortex elicited a fast, but transient, depolarization (O'Donnell and Grace, 1995O'Donnell P. Grace A.A. J. Neurosci. 1995; 15: 3622-3639PubMed Google Scholar). Until the development of optogenetic projection-specific targeting approaches, we did not have the ability to manipulate axons originating in specific regions during freely moving behaviors nor to stimulate axons arriving from a known source in acute slice preparations (Tye et al., 2011Tye K.M. Prakash R. Kim S.-Y. Fenno L.E. Grosenick L. Zarabi H. Thompson K.R. Gradinaru V. Ramakrishnan C. Deisseroth K. Nature. 2011; 471: 358-362Crossref PubMed Scopus (844) Google Scholar; Stuber et al., 2011Stuber G.D. Sparta D.R. Stamatakis A.M. van Leeuwen W.A. Hardjoprajitno J.E. Cho S. Tye K.M. Kempadoo K.A. Zhang F. Deisseroth K. Bonci A. Nature. 2011; 475: 377-380Crossref PubMed Scopus (596) Google Scholar). Optogenetic-mediated projection-specific targeting leverages the genetically encodable capability of these light-sensitive proteins and allows for the selective activation of specific populations of cells and axons. However, caveats still include the possibility of depolarizing axons of passage that do not form synapses in the illumination field or the induction of backpropagating action potentials (Petreanu et al., 2007Petreanu L. Huber D. Sobczyk A. Svoboda K. Nat. Neurosci. 2007; 10: 663-668Crossref PubMed Scopus (665) Google Scholar), also known as antidromic stimulation, which may scale with stronger illumination parameters, opsin expression levels, and the specific characteristics of the preparation. These early studies in optogenetic projection-specific targeting used local pharmacological manipulations, blocking glutamate receptors in the postsynaptic target region to demonstrate that the behavioral changes observed were indeed due to local effects—ruling out the possible contribution of axons of passage or antidromic activation to the light-induced behavioral change (Tye et al., 2011Tye K.M. Prakash R. Kim S.-Y. Fenno L.E. Grosenick L. Zarabi H. Thompson K.R. Gradinaru V. Ramakrishnan C. Deisseroth K. Nature. 2011; 471: 358-362Crossref PubMed Scopus (844) Google Scholar; Stuber et al., 2011Stuber G.D. Sparta D.R. Stamatakis A.M. van Leeuwen W.A. Hardjoprajitno J.E. Cho S. Tye K.M. Kempadoo K.A. Zhang F. Deisseroth K. Bonci A. Nature. 2011; 475: 377-380Crossref PubMed Scopus (596) Google Scholar). Stuber and colleagues investigated two of the same projections, specifically testing the ability of amygdalar and prefrontal cortical inputs of the NAc to support ICSS, by expressing channelrhodopsin-2 (ChR2), a light-activated cation channel, in glutamatergic pyramidal neurons of the amygdala or prefrontal cortex and implanting an optical fiber into the medial shell of the NAc. They observed that amygdalar, but not prefrontal cortical, inputs to the NAc supported ICSS (Stuber et al., 2011Stuber G.D. Sparta D.R. Stamatakis A.M. van Leeuwen W.A. Hardjoprajitno J.E. Cho S. Tye K.M. Kempadoo K.A. Zhang F. Deisseroth K. Bonci A. Nature. 2011; 475: 377-380Crossref PubMed Scopus (596) Google Scholar). In this issue of Neuron, Britt et al., 2012Britt J.P. Benaliouad F. McDevitt R.A. Stuber G.D. Wise R.A. Bonci A. Neuron. 2012; 76 (this issue): 790-803Abstract Full Text Full Text PDF PubMed Scopus (480) Google Scholar put forth an article of impressive breadth, characterizing three pathways from anatomical, electrophysiological, and behavioral perspectives (Figure 1). Anatomically, Britt et al., 2012Britt J.P. Benaliouad F. McDevitt R.A. Stuber G.D. Wise R.A. Bonci A. Neuron. 2012; 76 (this issue): 790-803Abstract Full Text Full Text PDF PubMed Scopus (480) Google Scholar examined the patterns of axons expressing a fluorescent protein in the NAc from the Amyg, PFC, and vHipp, revealing the unique distribution of axons throughout the NAc in exquisite detail across multiple animals (Britt et al., 2012Britt J.P. Benaliouad F. McDevitt R.A. Stuber G.D. Wise R.A. Bonci A. Neuron. 2012; 76 (this issue): 790-803Abstract Full Text Full Text PDF PubMed Scopus (480) Google Scholar), largely consistent with earlier studies (Voorn et al., 2004Voorn P. Vanderschuren L.J.M. Groenewegen H.J. Robbins T.W. Pennartz C.M. Trends Neurosci. 2004; 27: 468-474Abstract Full Text Full Text PDF PubMed Scopus (882) Google Scholar). They also investigated the properties of synaptic transmission from each of these pathways using ex vivo whole-cell patch-clamp recording techniques in acute slice preparations of different animals expressing ChR2 in one of the upstream regions (Amyg, PFC, or vHipp). These experiments revealed new insights about the relative strength of light-evoked excitatory postsynaptic currents (EPSCs), showing that vHipp inputs evoked the greatest EPSC amplitudes in the NAc shell, with the PFC inputs evoking the smallest EPSC amplitudes of the three (Britt et al., 2012Britt J.P. Benaliouad F. McDevitt R.A. Stuber G.D. Wise R.A. Bonci A. Neuron. 2012; 76 (this issue): 790-803Abstract Full Text Full Text PDF PubMed Scopus (480) Google Scholar). This was not a result of varying sensitivity or composition of postsynaptic AMPARs for each input, as demonstrated by the nearly identical amplitudes of quantal release and indistinguishable current-voltage relationships across synapses, respectively (Britt et al., 2012Britt J.P. Benaliouad F. McDevitt R.A. Stuber G.D. Wise R.A. Bonci A. Neuron. 2012; 76 (this issue): 790-803Abstract Full Text Full Text PDF PubMed Scopus (480) Google Scholar). However, Britt et al., 2012Britt J.P. Benaliouad F. McDevitt R.A. Stuber G.D. Wise R.A. Bonci A. Neuron. 2012; 76 (this issue): 790-803Abstract Full Text Full Text PDF PubMed Scopus (480) Google Scholar did observe that the vHipp-NAc synapses showed greater NMDAR-mediated inward currents, which could explain the unique ability of this input to induce the stable depolarization seen in "up and down states" of NAc MSNs (O'Donnell and Grace, 1995O'Donnell P. Grace A.A. J. Neurosci. 1995; 15: 3622-3639PubMed Google Scholar). Electrophysiologically, there is a unique feature that Britt et al., 2012Britt J.P. Benaliouad F. McDevitt R.A. Stuber G.D. Wise R.A. Bonci A. Neuron. 2012; 76 (this issue): 790-803Abstract Full Text Full Text PDF PubMed Scopus (480) Google Scholar identified of vHipp-NAc synapses: they were exclusively potentiated after cocaine treatment. In contrast to Pascoli et al., 2012Pascoli V. Turiault M. Lüscher C. Nature. 2012; 481: 71-75Crossref Scopus (286) Google Scholar, they did not observe a cocaine-induced potentiation of PFC inputs to the NAc (Pascoli et al., 2012Pascoli V. Turiault M. Lüscher C. Nature. 2012; 481: 71-75Crossref Scopus (286) Google Scholar). This might be explained by the fact that Pascoli and colleagues investigated only infralimbic inputs to D1 receptor-expressing medium spiny neurons (MSNs) in the NAc, while Britt et al., 2012Britt J.P. Benaliouad F. McDevitt R.A. Stuber G.D. Wise R.A. Bonci A. Neuron. 2012; 76 (this issue): 790-803Abstract Full Text Full Text PDF PubMed Scopus (480) Google Scholar expressed ChR2 throughout the mPFC including both prelimbic and infralimbic regions and recorded from all MSNs. Given the opposing functions observed in both prelimbic and infralimbic cortices as well as D1 and D2 receptor-expressing neurons, this may have resulted in a "zero-sum" effect when pooled together. Perhaps vHipp inputs to the medial shell of the NAc preferentially formed synapses on D1-type MSNs, though testing this hypothesis would require additional experiments. Behaviorally, the inhibition of vHipp axons in the NAc reduced, while activation increased, cocaine-induced locomotion (Britt et al., 2012Britt J.P. Benaliouad F. McDevitt R.A. Stuber G.D. Wise R.A. Bonci A. Neuron. 2012; 76 (this issue): 790-803Abstract Full Text Full Text PDF PubMed Scopus (480) Google Scholar). Britt et al., 2012Britt J.P. Benaliouad F. McDevitt R.A. Stuber G.D. Wise R.A. Bonci A. Neuron. 2012; 76 (this issue): 790-803Abstract Full Text Full Text PDF PubMed Scopus (480) Google Scholar also demonstrated that illumination of vHipp axons in the NAc supported ICSS and real-time place preference (RTPP). While they replicated the finding that photostimulation of Amyg axons in the NAc could support reward-related behaviors (Stuber et al., 2011Stuber G.D. Sparta D.R. Stamatakis A.M. van Leeuwen W.A. Hardjoprajitno J.E. Cho S. Tye K.M. Kempadoo K.A. Zhang F. Deisseroth K. Bonci A. Nature. 2011; 475: 377-380Crossref PubMed Scopus (596) Google Scholar), in contrast to earlier work from this group, they found that illuminating ChR2-expressing PFC neurons could also support ICSS. This discrepancy can be reconciled by several experimental details; Britt et al., 2012Britt J.P. Benaliouad F. McDevitt R.A. Stuber G.D. Wise R.A. Bonci A. Neuron. 2012; 76 (this issue): 790-803Abstract Full Text Full Text PDF PubMed Scopus (480) Google Scholar performed a more robust activation of PFC axons in the NAc by using bilateral stimulation and illumination parameters at a 50% higher frequency and train duration. This difference highlights the importance of titrating optogenetic experimental parameters in much the same way as pharmacological experiments, using light and/or viral "dose-dependent curves." Finally, yet another surprising result emerged from this study with their ability to support ICSS with nonspecific MSN activation (Britt et al., 2012Britt J.P. Benaliouad F. McDevitt R.A. Stuber G.D. Wise R.A. Bonci A. Neuron. 2012; 76 (this issue): 790-803Abstract Full Text Full Text PDF PubMed Scopus (480) Google Scholar). In the NAc (Lobo et al., 2010Lobo M.K. Covington 3rd, H.E. Chaudhury D. Friedman A.K. Sun H. Damez-Werno D. Dietz D.M. Zaman S. Koo J.W. Kennedy P.J. et al.Science. 2010; 330: 385-390Crossref PubMed Scopus (612) Google Scholar), D1 and D2 receptor-expressing cells showed opposing effects on reward-related behaviors. However, when examining the data from these studies, the degree to which activation of D1 receptor-expressing neurons was positively reinforcing may have overpowered the aversive properties of D2 receptor-expressing neuronal activation in the NAc, leading to a net effect of positive reinforcement. This finding led Britt et al., 2012Britt J.P. Benaliouad F. McDevitt R.A. Stuber G.D. Wise R.A. Bonci A. Neuron. 2012; 76 (this issue): 790-803Abstract Full Text Full Text PDF PubMed Scopus (480) Google Scholar to suggest that perhaps the source of glutamatergic innervation was less important than the bulk amount of glutamate released into the medial shell of the NAc. While this might not be true in physiological settings, where glutamate release is governed by the natural spiking of neurons rather than robust trains at frequencies only seen in bursting pyramidal neurons, Britt et al., 2012Britt J.P. Benaliouad F. McDevitt R.A. Stuber G.D. Wise R.A. Bonci A. Neuron. 2012; 76 (this issue): 790-803Abstract Full Text Full Text PDF PubMed Scopus (480) Google Scholar certainly put forth a host of new questions. The subtleties of this study need to be explored, particularly given the caveats that the Amyg, vHipp, and PFC are all robustly and reciprocally connected to each other. While they may provide direct input to the NAc, further experiments are needed to confirm that monosynaptic input from each of these inputs is sufficient to support reward-related behaviors. An important caveat to note for nearly all optogenetic studies published to date is that the use of cylindrical optical fibers with blunt-cut tips creates a relatively narrow and small cone of light that may not capture all of the axon terminals expressing ChR2—particularly in large structures such as the NAc, which is organized spherically rather than cylindrically. Here, Britt et al., 2012Britt J.P. Benaliouad F. McDevitt R.A. Stuber G.D. Wise R.A. Bonci A. Neuron. 2012; 76 (this issue): 790-803Abstract Full Text Full Text PDF PubMed Scopus (480) Google Scholar looked only at the medial shell of the NAc, but other recent studies in the NAc core or lateral shell could have different effects, as recently suggested (Lammel et al., 2012Lammel S. Lim B.K. Ran C. Huang K.W. Betley M.J. Tye K.M. Deisseroth K. Malenka R.C. Nature. 2012; 491: 212-217Crossref PubMed Scopus (900) Google Scholar). Another possibility raised by Lammel and colleagues is that multiple distinct experiential qualities could support ICSS, including salience, alertness, motivation, and hedonic pleasure in addition to general reward and reinforcement (Lammel et al., 2011Lammel S. Ion D.I. Roeper J. Malenka R.C. Neuron. 2011; 70: 855-862Abstract Full Text Full Text PDF PubMed Scopus (520) Google Scholar). It would also be interesting to characterize the ultrastructural organization across the NAc of axonal terminals arriving from the vHipp, PFC, and Amyg—how often do these axon terminals synapse onto the same cell, and how are these interactions assembled (axoaxonal synapses, on the same dendritic arbor, etc.)? To conclude, even with the recent flood of insights toward causal relationships between the brain and behavior facilitated by optogenetic approaches (Tye and Deisseroth, 2012Tye K.M. Deisseroth K. Nat. Rev. Neurosci. 2012; 13: 251-266Crossref PubMed Scopus (551) Google Scholar), there is still much to do. The paper from Britt et al., 2012Britt J.P. Benaliouad F. McDevitt R.A. Stuber G.D. Wise R.A. Bonci A. Neuron. 2012; 76 (this issue): 790-803Abstract Full Text Full Text PDF PubMed Scopus (480) Google Scholar in this issue of Neuron makes an important contribution to the field by providing multiple new insights, raising provocative new questions, and opening the floodgates even wider than before to invite more research in this exciting new arena of systems neuroscience. Synaptic and Behavioral Profile of Multiple Glutamatergic Inputs to the Nucleus AccumbensBritt et al.NeuronNovember 21, 2012In BriefBritt et al. characterize hippocampal, amygdala, and prefrontal cortex input to the nucleus accumbens. Hippocampal input is shown to be predominant in the medial accumbens shell, yet photostimulation of axons from each pathway is sufficient to reinforce instrumental behaviors. Full-Text PDF Open Archive
Referência(s)