In the Scottish highlands, hikers risk being caught in a sudden mist that obliterates all visible landmarks. People in this situation have two options: attempt to retrace their steps by estimating how far and in which direction they have traveled, or wait until the mist lifts and the landmarks return. The neural basis of these spatial abilities in humans is not clear. In rodents, however, both types of navigation may rely on hip- pocampal neurons called ‘place cells’ 1 , which encode spatial information defined by self- motion cues 2,3 or by visual landmarks 4–6 . One influential theory is that together, the population of place cells provides a ‘cogni- tive map’ of an animal’s environment 7 . In this issue, Rosenzweig et al. 8 compare these hippocampal cognitive maps and spa- tial behavior in young adult and aged rodents. They find that the ability of both young adult and aged rats to find a reward in the environment is correlated with the ability of place cells in the hippocampus to switch between two different cognitive maps, one based on self-motion cues that are irrelevant to solving the task, and another based on relevant landmark cues. Intriguingly, they observe that old rats are impaired relative to young adult rats, both in switching from the irrelevant to the rele- vant map and in finding the reward (Fig. 1). Since the discovery of hippocampal place cells 1 in the early 1970s, researchers have sought to understand which aspects of the environment control their spatial activity. Early studies demonstrated that the location in which a place cell fires—its ‘place field’—is controlled by visual landmarks in the environ- ment 4–6 . Thus, if one rotates the landmarks in an environment by 90° while the rat is else- where, then when the rat returns, its place fields will also rotate by 90° to agree with the landmarks. Visual landmarks are not the whole story, however. Place cells are also controlled by path integration—deriving the direction and distance traveled from self-motion informa- tion, including motor propioception, the vestibular system and perhaps optic flow 2,3 . E.R.W. is in the Division of Neuroscience and Centre for Neuroscience Research, University of Edinburgh, 1 George Square, Edinburgh, EH8 9JZ, Scotland. P.A.D. is in the Department of Psychology, University of Stirling, Stirling FK9 4LA, Scotland. e-mail: emma.wood@ed.ac.uk Aging, spatial behavior and the cognitive map Emma R Wood & Paul A Dudchenko Hippocampal neurons are thought to form a cognitive map of the environment based on multiple cues. A new study shows that young animals switch between cues more easily than aged animals and also perform better on a spatial learning task. NEWS AND VIEWS cells within one or more region of the brain. The preliminary results of van Swinderen and Greenspan suggest that 20–30 Hz activity is phase-locked across regions of the central brain, thus showing spatiotemporal synchrony or coherence—a hallmark of early perception in verte- brates 9 . Increases in power within the 20–30 Hz bandwidth may reflect changes in the recruitment, firing rate and synchrony of a subpopulation of cells within the medial protocerebrum or an adjacent region. Future experiments using micro- arrays of extracellular electrodes combin- ed, when possible, with intracellular recordings, should be able to test these various hypotheses. Although Drosophila will never win high marks as an electrophysiological model, it does permit extraordinarily pre- cise manipulations of gene expression in both time and space. In particular, by clever implementation of the GAL4/UAS expression system of yeast 10 , it is possible to express a gene of interest (including ones not found in the fly’s genome) ectopically within specific regions of the nervous system. For example, the gene for tetanus toxin or an allele of shibire can be introduced to block synaptic transmis- sion. Such techniques have been used to examine mechanisms of sensory discrimi- nation 11 , sensorimotor integration 12 and learning and memory 13 . Until now, how- ever, physiological links between gene products and behavior have been elu- sive—especially at the systems level. By systematically targeting reversible temperature-sensitive mutations in mem- brane conductance and synapse function to specific regions of the brain, van Swinderen and Greenspan spatially local- ized the origins of the 20–30 Hz response—at least in part—to synaptic output from the mushroom bodies. Biochemical, physiological and behavioral evidence implicates these structures in odor-mediated learning in flies 14 . It remains to be shown how visual feedback is integrated with mushroom body out- put. Postsynaptic targets of mushroom body efferents such as those within the lateral horn are probably involved 15 . The discovery of structure–function relationships mediating higher-order brain processes such as perception, atten- tion and learning will be accelerated by integrating molecular-genetic, physiologi- cal and behavioral approaches. Toward these ends, the experimental power of cou- pling spatially localized and reversible genetic manipulations with robust physio- logical recording preparations in behaving flies is difficult to overstate. Armed with a physiological assay for object salience, the next challenge is to find out how this sig- nal is used to structure fly motor behaviors such as foraging or courtship. Given the remarkable complexity even within one of nature’s smaller brains, this challenge is not an easy one. As Donald Hebb noted, “the brain…may not be able to do simple things in a simple way.” 1. van Swinderen, B. & Greenspan, R.G. Nat. Neurosci. 6, 581–588 (2003). 2. Laurent, G. Science 286, 723–728 (1999). 3. Egelhaaf, M. & Kern, R. Curr. Opin. Neurobiol. 12, 699–706 (2002). 4. Trimarchi, J.R., Jin, P. & Murphey, R.K. Intl. Rev. Neurobiol. 43, 241–264 (1999). 5. Ernst, R. & Heisenberg, M. Vision Res. 39, 3920–3933 (1999). 6. Götz, K.G. Basic Life Sci. 16, 391–407 (1980). 7. Frye, M.A., Tarsitano, M. & Dickinson, M.H. J. Exp. Biol. 206, 843–855 (2003). 8. Brett, M., Johnsrude, I.S. & Owen, A.M. Nat. Rev. Neurosci. 3, 243–249 (2002). 9. Singer, W. Nature 397, 391–393 (1999). 10. Brand, A. & Perrimon, N. Development 118, 401–415 (1993). 11.Keller, A., Sweeney, S.T., Zars, T., O'Kane, C.J. & Heisenberg, M. J. Neurobiol. 50, 221–223 (2002). 12. Strauss, R. Curr. Opin. Neurobiol. 12, 633–638 (2002). 13. deBelle, J.S. & Heisenberg, M. Science 263, 692–695 (1994). 14. Strausfeld, N.J., Hansen, L., Li, Y.S., Gomez, R.S. & Ito, K. Learn. Mem. 5, 11–37 (1998). 15. Ito, K. et al. Learn. Mem. 5, 52–77 (1998). 546 VOLUME 6 | NUMBER 6 | JUNE 2003 NATURE NEUROSCIENCE © 2003 Nature Publishing Group http://www.nature.com/natureneuroscience