I see a face—a happy face - Mike W. Oram and Barry J. Richmond

Aug 29, 2026

Face-selective neurons in primate inferior temporal cortex encode distinct stimulus attributes during early and late phases of responding to the same image. The text contrasts the prevailing assumption that evoked activity maintains a constant relationship to stimulus identity throughout a response period. It highlights findings by Sugase and colleagues using information-theoretic analysis, showing that information levels shift over time from more global to more specific stimulus discrimination. The discussion frames implications for how sensory processing evolves during perception.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages3

Reading Time8 min

AccessFree to download

News and Views

Persistent neural and behavioral changes in chronic drug exposure

Addiction research and molecular regulation of reward circuits

Face-selective coding in inferior temporal cortex

Information theory and time-varying stimulus discrimination

© 1999 Nature America Inc.• [http://neurosci.nature.com](http://neurosci.nature.com) news and views atonigral neurons6. Kelz and colleagues 1 tested the hypothesis that ∆FosB could mediate some of the persistent neural and behavioral changes accompanying chronic drug exposure. By expressing ∆fosB under the control of a tetracycline-regulated promoter, the authors were able to investigate the effects of this protein without the multifarious homeostatic adaptations produced in nucleus accumbens and other striatal neurons by repeated administration of cocaine. In behavioral tests,∆fosB-expressing animals found cocaine more rewarding than did appropriate control mice. The transgenics also had greater sensitivity to the locomotor stimulant effects of cocaine, without demonstrating differences in sensitization. Because ∆FosBis a transcription factor (Fig. 1), Kelz and colleagues1 have correlated the expression of ∆FosB with increased expression of a potential target, the GluR2 subunit of AMPA receptors. A likely role for GluR2 in the ∆FosB-induced enhancement of the rewarding properties of cocaine was demonstrated by separately expressing GluR2 by virally mediated gene transfer1. The results of these studies are surprising and interesting because ∆FosB was initially thought to represent a homeostatic response to cocaine. Indeed one might have predicted opposite results based on the fosB knockout mouse7, which also exhibits increased sensitivity to the rewarding and locomotor stimulant effects of cocaine. The approach taken by Kelz and colleagues overcomes some of the limitations of analyzing a knockout mouse that lacks FosB and ∆FosB throughout development. The mouse described by Kelz and colleagues could serve as a model of enhanced vulnerability for cocaine use, and at the same time suggests that cocaine, acting via ∆FosB, might increase its own rewarding properties without inducing locomotor sensitization. Investigation of the physiological effects of increased GluR2 expression and identification of other alterations in gene expression in the ∆FosB-expressing line should provide useful insights into regulation of brain reward circuits. It will also be interesting to see what happens when this line of mice receives chronic cocaine; it is possible that effects of extra ∆FosB expression, either on homeostasis or on associative learning, will emerge. In addiction research to date, there have been major efforts to investigate homeostatic adaptations to excess neurotransmitter stimulation4. This has led to the search for molecular species that undergo longlived up-or downregulation in response to repeated drug administration. Indeed, the isoforms of ∆fosB that were studied by Kelzand colleagues 1 were initially interesting precisely because of their prolonged upregulation by cocaine6. Although homeostatic changes are clearly significant, the long-term relapse risk that characterizes addiction is undoubtedly related to associative learning mechanisms8. Mechanistically, it is clear that short-term molecular events that produce long-term changes in synaptic structure9 are likely to be critical actors and deserve at least as much attention as the search for long-lived molecular species. It will be interesting to determine the impact of ∆FosB on the induction of molecular changes that might be involved in synaptic plasticity. This paper also whets our intellectual appetites for additional lines of mice expressing additional drug- inducible genes with steadily improving spatial and temporal control. 1. Kelz, M. B. et al. Nature 401 (in press) . 2. Self, D. W. & Nestler, E. J. Annu. Rev. Neurosci. 18, 463–495 (1995) . 3. Schultz, W., Dayan, P. & Montague, P. R. Science 275, 1593–1599 (1997) . 4. Koob, G. F. & Le Moal, M. Science 278, 52–58 (1997) . 5. Robinson, T. E. & Berridge, K. C. Brain Res. Rev. 18, 247–291 (1993) . 6. Hope, B., Kosofsky, B., Hyman, S. E. & Nestler, E. J. Proc. Natl. Acad. Sci. USA 89, 5764–5768 (

FAQs

What do face-selective neurons in inferior temporal cortex encode over time?

They encode different stimulus attributes early and late in their response to the same image, with information changing across the response period.

How does this work challenge the traditional view of sensory processing?

It disputes the assumption that evoked activity maintains a constant relationship to stimulus information throughout the entire response period.

What method do Sugase and colleagues use to analyze neuronal responses?

They apply information theory to quantify how the information carried by neuronal responses evolves from global to more specific discrimination.

Aug 29, 2026

Uploaded by Chumphorn