View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by Elsevier - Publisher Connector Current Biology Vol 19 No 7 R296 hypersensitivity system, an in vivo 2. Sun, J.C., Beilke, J.N., and Lanier, L.L. (2009). to parental marrow grafts in adult F1 hybrid Adaptive immune features of natural killer cells. mice. Nature 204, 450–453. reaction that involves poorly defined Nature 457, 557–561. 11. Glas, R., Franksson, L., Une, C., Eloranta, M.L., haptenated cell surface ‘antigens’, is 3. Cooper, M.A., Elliott, J.M., Keyel, P.A., Yang, L., Ohlen, C., Orn, A., and Karre, K. (2000). not ideal for defining novel recognition Carrero, J.A., and Yokoyama, W.M. (2009). Recruitment and activation of natural killer (NK) Cytokine-induced memory-like natural killer cells in vivo determined by the target receptors. cells. Proc. Natl. Acad. Sci. USA 106, cell phenotype: An adaptive component of NK The new data mark an evolution 1915–1919. cell-mediated responses. J. Exp. Med. 191, 4. Raulet, D.H. (2004). Interplay of natural killer 129–138. from the view that NK cells respond cells and their receptors with the adaptive 12. Gidlund, M., Orn, A., Wigzell, H., Senik, A., and de novo to each insult. The sustained immune response. Nat. Immunol. 5, Gresser, I. (1978). Enhanced NK cell activity in sensitization of NK cells as a result of 996–1002. mice injected with interferon and interferon 5. Dorshkind, K., Pollack, S.B., Bosma, M.J., and inducers. Nature 273, 759. cytokines or infection at the least Phillips, R.A. (1985). Natural killer (NK) cells are 13. Carson, W.E., Giri, J.G., Lindemann, M.J., constitutes a form of hazy, fairly present in mice with severe combined Linett, M.L., Ahdieh, M., Paxton, R., immunodeficiency (scid). J. Immunol. 134, Anderson, D., Eisenmann, J., Grabstein, K., short-term memory, wherein a previous 3798–3801. and Caligiuri, M.A. (1994). Interleukin (IL) 15 encounter ensures that NK cells will, 6. Lanier, L.L. (2008). Up on the tightrope: natural is a novel cytokine that activates human for a period of weeks or months, lash killer cell activation and inhibition. Nat. natural killer cells via components of the Immunol. 9, 495–502. IL-2 receptor. J. Exp. Med. 180, out vigorously when exposed to the 7. Murali-Krishna, K., Altman, J.D., Suresh, M., 1395–1403. same or a different insult. The clonal Sourdive, D., Zajac, A., Miller, J., Slansky, J., and 14. O’Leary, J.G., Goodarzi, M., Drayton, D.L., Ahmed, R. (1998). Counting antigen-specific and von Andrian, U.H. (2006). T cell- and expansion data and contact CD8 T cells: a reevaluation of bystander B cell-independent adaptive immunity hypersensitivity data indicate that more activation during viral infection. Immunity 8, mediated by natural killer cells. Nat. specific NK cell memories exist as well, 177–187. Immunol. 7, 507–516. 8. Kaech, S.M., Wherry, E.J., and Ahmed, R. but a detailed understanding of how (2002). Effector and memory T-cell this works, and how important it is, differentiation: implications for vaccine Department of Molecular and Cell development. Nat. Rev. Immunol. 2, Biology and Cancer Research Laboratory, awaits data that clarify whether novel 251–262. 489 Life Sciences Addition, University NK cell receptors, or specificities, exist. 9. Dokun, A.O., Kim, S., Smith, H.R., Kang, H.S., of California at Berkeley, Berkeley, Chu, D.T., and Yokoyama, W.M. (2001). CA 94720, USA. Specific and nonspecific NK cell activation References during virus infection. Nat. Immunol. 2, E-mail: [email protected] 1. Medzhitov, R. (2007). Recognition of 951–956. microorganisms and activation of the immune 10. Cudkowicz, G., and Stimpfling, J.H. (1964). response. Nature 449, 819–826. Induction of immunity and of unresponsiveness DOI: 10.1016/j.cub.2009.02.023 Synaptic Transmission: Excitatory sustained activation of certain Aplysia neurons during feeding behavior may Autapses Find a Function? be aided by excitatory autapses. In this context, then, an excitatory autapse makes perfect sense: autaptic An autapse is a synapse between a neuron and itself, a peculiar structure with self-excitation tips the neuron into an unclear function. A new study suggests that excitatory autapses contribute a hyperexcitable state that requires the to a positive-feedback loop that maintains persistent electrical activity in persistent firing of action potentials. neurons. Saada et al. [7] studied the buccal ganglia of Aplysia, focusing on John M. Bekkers Finallly, neurotransmission might B31/B32 neurons, motor neurons that occur at autapses. drive muscles involved in grasping Although the brain is complicated, An autapse is a self-synapse, food during a repetitive feeding we are told in undergraduate biology a specialized structure in which behavior in these animals [8]. When classes that the basic circuit element a neuron forms a synaptic connection briefly stimulated in intact ganglia, is simple: the axon of one neuron between its axon and its own dendrites B31/B32 neurons exhibit a prolonged forms synapses on the dendrites of [4,5]. Anatomical autapses are not depolarization and firing of action another neuron, and information is uncommon in neural circuits, but potentials (Figure 1), producing conveyed from cell to cell like the their purpose has remained puzzling. sustained muscle contraction. The baton in a relay race. The reality is, Inhibitory autapses — those made authors wished to understand the of course, more complex. The brain by a neuron that releases an inhibitory mechanism of the persistent activity has evolved other avenues of neurotransmitter such as g-amino- in B31/B32 cells. communication that seem to lack the butyric acid (GABA) — seem to make These cells are embedded in sequential, neuron-to-neuron sense because they could provide a surprisingly complex circuit discipline of synaptic transmission. a self-stabilizing influence (‘negative (Figure 1). They receive both fast For example, neurotransmitter might feedback’) [6]. But what about and slow excitatory synaptic inputs spill out of the synaptic cleft and excitatory autapses? Why should mediated by release of acetylcholine activate receptors on nearby neurons, a neuron want to destabilize itself by from presynaptic neurons, the B63 including the neuron that released the self-excitation in a positive-feedback cells. In addition, they are electrically transmitter [1,2]. Crosstalk could also loop? In a recent issue of Current connected to B63 cells via gap occur between neurons and Biology, Saada et al. [7] propose an junctions. These connections make it non-neuronal cells, such as glia [3]. answer to this question: that the difficult to establish with certainty the Dispatch R297 existence of a postulated excitatory autapse in a B31/B32 neuron (red in B31/B32 cell B31/B32 autapse Figure 1). If an action potential occurs starts firing is active in the B31/B32 cell it could pass through the gap junction into the B63 cell, exciting it to cause normal synaptic transmitter release. Such Repolarization a response would look just like EPSPs from an autapse: an action potential in B63 cell the B31/B32 neuron produces a 10 mV postsynaptic potential in the same cell. One would be fooled into 1 sec thinking an autapse is present when it is not. Saada et al. [7] elegantly resolved Increasing this ambiguity by simultaneously Stimulate depolarization recording from the soma and axon of a B31/B32 neuron. By stimulating the axon directly, they were able to show that a slow, muscarinic synaptic (or autaptic) response could be recorded in the B31/B32 neuron, even when the B63 synapse was clearly silent. Hence, B63 B31/B32 they tentatively concluded that Muscle B31/B32 cells do possess autapses, and they release acetylcholine onto themselves to produce a slow, muscarinic excitation that aids persistent activity. Next, Saada et al. [7] turned to a simplified culture system in which Current Biology they were able to demonstrate the presence of autaptic connections in Figure 1. Electrical activity in a B31/B32 neuron in an intact ganglion (top), and the postulated an isolated B31/B32 neuron, without circuit that gives rise to this behavior (bottom). the complications of the intact Following brief electrical stimulation to either the B63 or B31/B32 neuron, excitatory post- ganglion. This is a risky strategy. synaptic potentials (EPSPs), due to the release of acetylcholine from the B63 cell, are observed Mammalian neurons that do not form in the B31/B32 cell. These EPSPs accelerate and merge, causing increasing depolarization. autapses in vivo commonly establish Eventually the B31/B32 cell starts firing action potentials, which activates the slow, muscarinic profuse autaptic connections B31/B32 autapse (red connection, bottom panel). Synaptic release from the B63 cell can also occur due to electrical coupling between the B31/B32 cell and the B63 axon (zigzag line). when they are grown in cell culture, Repolarization is produced by inhibitory inputs that are not shown in this figure. presumably as an adaptation to the lack of axonal guidance cues [4,9]. Thus, the presence of autapses in vitro has shown are unlikely to contact Indeed, Saada et al. [7] suggest, by no means confirms their presence other neurons in the ganglion, may somewhat paradoxically, that the in vivo. Nevertheless, the authors be the sites of autaptic release of autapse might be more important for found that isolated B31/B32 neurons in acetylcholine. Strictly speaking, these turning off the sustained activity culture did form autaptic connections data are insufficient to formally define than maintaining it. This and other with properties similar to those seen an autapse, which is a structural questions could be resolved by in the intact ganglion. Importantly, specialization that is indistinguishable, finding a way to selectively block they also showed that some other at the ultrastructural level, from the autaptic input in the intact ganglion types of buccal ganglion neurons a synapse [10,11].