H I G H L I G H T S

NATURE REVIEWS | NEUROSCIENCE VOLUME 5 | JANUARY 2004 | 7

If a magician gave you a list of words to memorize, and was thenable to predict precisely which words you would be able to recall,you would no doubt be impressed. Reporting in the Journal ofNeuroscience, Sederberg and colleagues show that this feat can beachieved without resorting to magic or ‘mind-control’ tricks.Using intracranial recordings from the cortex, they identifiedspecific patterns of rhythmic neuronal activity that correlatedwith successful memory formation.

Sederberg et al. used a group of 10 patients, aged from 8–17,each of whom had up to 128 electrodes implanted into theirbrains to monitor epilepsy. The patients were asked to memorizea list of 20 common nouns while the authors recorded theircortical and subcortical brain activity. The patients were thenasked to recall as many words as possible, after being distractedbriefly with an arithmetical task to reduce bias towards the mostrecently presented words.

The authors observed increases in both low-frequency thetaoscillations (4–8 Hz) and high-frequency gamma oscillations(28–64 Hz) during the presentation of words that weresubsequently recalled successfully. The theta oscillations weredetected largely in the right temporal and frontal cortex, whereasthe gamma oscillations were more widespread.

What is the importance of gamma and theta power formemory formation? Both types of activity have been implicatedin processes that are crucial to memory encoding — gammaoscillations in attention, and theta oscillations in long-termpotentiation. Approaches such as that of Sederberg et al. shouldhelp us to understand how different cortical regions arerecruited to these processes as new memories are laid down, andto clarify the precise role of this rhythmic activity in memoryformation.

Heather WoodReferences and links

ORIGINAL RESEARCH PAPER Sederberg, P. B. et al. Theta and gamma oscillationsduring encoding predict subsequent recall. J. Neurosci. 23, 10809–10814 (2003)FURTHER READING Paller, K. A. & Wagner, A. D. Observing the transformation ofexperience into memory. Trends Cogn. Sci. 6, 93–102 (2002)WEB SITEEncyclopedia of Life Sciences: http://www.els.net/Learning and memory

Antibiotic yields anxiolytic

N E U R O P H A R M A C O LO G Y

Benzodiazepines such as diazepamare commonly prescribed to controlanxiety, and act by enhancing theaction of the neurotransmitterGABA (γ-aminobutyric acid) at thebrain GABA

Areceptor. But drugs of

this class have sedative side effects,an undesirable feature that hasspurred a search for alternatives byKelvin Gee and colleagues. In aNature Medicine advanced onlinepublication, the team describe afirst-in-class anxiolytic that is free ofsedative side effects.

The search strategy devised byGee et al. was based on the hypothesisthat mitigation of side effects might be achieved by targeting non-benzodiazepine-binding sites ofGABA

Areceptors. Compounds

related to the fluoroquinolone anti-biotic norfloxacin were chosen asscreening candidates for two reasons:first, a small proportion of patientstaking this antibiotic experience anxiety as a side effect; second,norfloxacin is known to antagonizeGABA

Areceptors at an undefined

site. So, chemical modification of norfloxacin could generate compounds with the opposite,GABA-enhancing anxiolytic effect.

In a standard assay for GABAmodulation, several norfloxacin-related chemotypes inhibited thebinding of the convulsant TBPS to GABA

Areceptors. Analysis of

the most active product, dubbedcompound 4, showed that this compound does not act directly onthe GABA

Areceptor-binding sites for

TBPS, norfloxacin, benzodiazepinesor GABA.

To confirm the GABA-modulatoryaction of compound 4, electrophysio-logical studies were conducted onHEK293 cells expressing differenthuman GABA

Areceptor subtypes.

Compound 4 enhanced GABA-mediated currents in a dose-dependent manner, but only in thosecells that expressed the α

2subunit

of the receptor. Failure of the benzo-diazepine binding-site antagonist flumazenil to block this potentiation confirmed that compound 4 binds to a non-benzodiazepine site. So struc-tural manipulation of norfloxacin didproduce a compound that binds to an alternative GABA

Areceptor site, but

did this modification result in anxi-olytic rather than anxiety-inducingeffects?

In two rodent models of anxiety,treatment with compound 4 relievedanxiety-related behaviour. Notably, themagnitude of this effect was compara-ble to that afforded by diazepam. Butthe real breakthrough came fromassessments of motor performanceusing the rotarod test — diazepamcaused motor impairment whereascompound 4 did not, even at the maximal soluble dose. Compound 4might therefore be at the forefront of anew approach to the design of robustanxiolytics with improved side-effectprofiles.

Suzanne FarleyReferences and links

ORIGINAL RESEARCH PAPER Johnstone, T. B. C.et al. Modifying quinolone antibiotics yields newanxiolytics. Nature Med. 30 November 2003 (doi:10.1038/nm967)FURTHER READING Christopoulos, A. Allostericbinding sites on cell-surface receptors: noveltargets for drug discovery. Nature Rev. DrugDiscov. 1, 198–210 (2002)

Oscillations predict recall

M E M O R Y

COOH

O

NN

F

C2H5

HN

HN

Cl

OH

O O

N

Compound 4

Norfloxacin

© 2004 Nature Publishing Group