In vivo, presentation of a novel sensory stimulus is associated with a characteristic biphasic response in human electroencephalograph recordings, consisting of an initial λ frequency (30-80 Hz) component and subsequent β frequency (15-30 Hz) component (Hanschel et al. 2000). The β component is thought to be associated with mnemonic processes related to the stimulus because the β component of the response habituates with continued presentation of the ‘novel’ stimulus. An analogous phenomenon has been observed in vitro. Applying 2-site strong tetanic stimulation to CA1 of rat hippocampus causes an initial gamma frequency response, followed by a later β frequency response (Whittington et al. 1997). Computer modelling suggests that enhanced recurrent excitation between pyramidal cells combined with recovery of the pyramidal cell spike afterhyperpolarisation (AHP) could promote synchronous ‘beat skipping’ by pyramidal cells, despite ongoing gamma frequency interneuronal activity, leading to a λ/β spectral shift (Bibbig et al. 2001).

We present here pharmacological data that may help to elucidate the mechanism underlying the λ/β shift in vitro. Hippocampal slices (450 µm) were obtained from rats, which had been anaesthetised humanely with isoflurane, prior to injection with xylazine (> 100 mg kg^-1) and ketamine (> 10 mg kg^-1) intramuscularly. Brains were excised quickly and placed in ice-cold sucrose/ACSF. λ oscillations were elicited by application of the metabotropic glutamate receptor agonist (s)-3,5 dihydrophenylglycine (DHPG). Subsequent application of 5-hydroxytryptamine (5-HT) caused a reversible spectral shift in extracellularly recorded field potentials from λ to β frequency (mean gamma frequency = 33.2 ± 0.52 Hz, mean beta frequency = 20.9 ± 1.58 Hz, shift highly significant; P < 0.001, n = 16, RMS test). Intracellular recordings from pyramidal cells demonstrated a 2-fold increase in excitatory postsynaptic potential (EPSP) amplitude (increase highly significant; P < 0.001, n = 2900 events from 5 cells, RMS test), and halving of EPSP frequency. Inhibitory postsynaptic potential (IPSP) frequency remained within the λ band during β activity. Additionally, an decrease in pyramidal cell slow AHP amplitude was observed during β activity, which was associated with spike ‘beat skipping’. Application of 5-HT to the quiescent slice did not evoke synaptic activity and tended to cause long-term depression of stimulus-evoked EPSP amplitude, suggesting a role for intrinsic network mechanisms in promoting λ to β shifts.

The mechanism by which 5-HT causes an increase in EPSP amplitude may represent a common mode of action of SSRIs and AMPAkines used to treat depression.

This work was generously funded by the MRC and GSK plc.

All procedures accord with current UK legislation.