Networks of neurones in mammalian cortex have the propensity to adopt synchronous oscillations in the gamma frequency range (~20-80 Hz) due to definite stimuli. The appearance of these oscillations during sensory stimulation has led to the suggestion that they are involved in cognitive functions such as memory formation. The medial temporal lobe appears to be crucial in the formation of long-term memories. Indeed, consolidation of information in associational cortical areas is believed to be important. One such associative area is the medial entorhinal cortex (mEC). Previously, in vivo experiments have demonstrated that the entorhinal cortex can generate gamma frequency oscillations in response to various stimuli (Chrobak et al. 2000). Using the combined entorhinal-hippocampal slice we can report that application of the specific AMPA/kainate receptor agonist kainic acid (400 nM; n = 16) induces gamma activity in the mEC.

Combined entorhinal-hippocampal slices (~450 mm), were taken from adult (200-250 g) Wistar rats after terminal anaesthesia using ketamine-xylazine (administered intramuscularly), and intracardial perfusion with artificial cerebrospinal fluid (ACSF) in which NaCl was replaced with sucrose. Slices were maintained at an interface of oxygenated ACSF and humidified (95 % O₂-5 % CO₂) gas at 36 °C. Extracellular field recordings were made in the mEC, using glass microelectrodes containing ACSF (resistance < 3 MΩ). Intracellular recordings were made in neurones, using glass microelectrodes (resistance 100-120 MΩ) containing 1.5 M KCH₃SO₄.

Once initiated, oscillations were stable for several hours. Whilst activity within lamina (< 1 mm) was highly synchronous and with no apparent phase lag, activity recorded across deep and superficial layers showed a 180 deg phase angle shift. A laminar profile of the mEC illustrated a distinct phase shift between layer II and layer III. In order to clarify this phase reversal issue we carried out intracellular recordings simultaneously with field recordings. These recordings revealed IPSPs at depolarised holding potentials (▓ge│ -40 mV), occurring at frequencies within the gamma range. Bath application of bicuculline (2 mM) abolished both IPSPs and field gamma. IPSPs in layer III were phase locked, and in antiphase with field activity in layer III. Conversely, intracellular events in layer III were phase locked and in phase with field activity in layer II. Finally, IPSPs in layer II were phase locked but antiphase with field activity in deep layers.

GABAergic activity was crucial to network activity. IPSP trains were observed in both deep and superficial layers with no phase reversal. Thus the phase shift observed in the mEC field recordings was probably due to a current sink/source phenomenon.

This work was supported by the MRC and GSK.

All procedures accord with current UK legislation.