Consciousness is a nervous process that handles only a limited amount of information. Therefore the nervous system needs to select the most relevant input for aware processing. For the visual system, it has been suggested that recurrent excitation from the cortex to neurones in the lateral geniculate nucleus provides a "spotlight of attention", that selectively enhances the relay of information to the cortex. Such feedback excitation could be supplied by corticogeniculate neurones in layer 6 of the primary visual cortex. The corticogeniculate synaptic strength increases with neuronal firing frequency. From this property it can be hypothesised that the feedback excitation would function as a variable neuronal amplifier for boosting the information transfer in the attentive state. The general aim of this thesis was to study the synaptic mechanisms that make the corticogeniculate synapse frequency sensitive and evaluate this property in relation to the proposed neuronal amplifier function.

Experiments were performed with whole-cell patch-clamp recordings from principal cells in a slice preparation of the rat dorsal lateral geniculate nucleus. Ex citatory postsynaptic currents (EPSCs) evoked by stimulation of corticogeniculate axons consistently displayed paired pulse facilitation. The ratio EPSC₂ I EPSC₁ was 3.7 ± 1.6 (mean ± standard deviation) for two pulses separated 40 ms. The paired pulse facilitation comprised a fast and slow component, evident from its double exponential decay. EPSCs evoked in the same cells by stimulating axons from the retina displayed paired pulse depression. The two types of EPSCs differed in their response to alterations in the extracellular calcium ion concentration ([Ca²⁺]_o). The paired pulse depression at retinogeniculate synapses was attenuated by decreasing [Ca²⁺]_o, apparently from lowering the level of transmitter release. At the corticogeniculate synapse, paired pulse facilitation was optimal at physiological [Ca²⁺]_o. The facilitation was presynaptic in origin since the facilitated EPSC2 quantal size (q = - 5.2 ± 0.8 pA) was essentially the same as for EPSC₁ (q = - 4.9 ± 0.9 pA). Each corticogeniculate axon terminated with 1 - 2 functional synapses (n_syn) per principal cell and the basal transmitter release probability was low (p_syn = 0.09 ± 0.04) but increased with facilitation (p_syn = 0.25 ± 0.10).

When short trains of pulses were used for stimulation of corticogeniculate axons, the EPSCs rapidly increased in amplitude with the first 2 - 3 stimuli followed by a more gradual growth. A double exponential function, likely to represent the fast and slow components of facilitation could describe the EPSC build-up in amplitude. The time constant of fast facilitation was dependent on [Ca²⁺]_o , presumably representing Ca²⁺ binding to a saturable intraterminal Ca²⁺ buffer. When pulse trains were repeated at 1 - 10 s intervals, EPSC₁ in each train was progressively enhanced by augmentation, leaving late EPSCs unaffected. When [Ca²⁺]_o was altered, augmented EPSCs changed in proportion to the basal EPSC amplitude, i.e. EPSC_1:n / EPSC_1,1 remained the same. The results indicate that augmentation is determined by a Ca²⁺ residue in the presynaptic terminal after repetitive spike firing, competing with the mechanism of the fast component of facilitation.

The two components of facilitation and augmentation at the corticogeniculate synapse define the function of the suggested neuronal amplifier. The low basal synaptic strength ascertains that single random spikes will be virtually ineffective at the target cell, which protects the ex citatory feedback system from self-generated cyclic activity. Since the different forms of synaptic enhancement are presynaptic, the neuronal amplifier will be strictly stimulus specific in increasing synaptic strength. Furthermore, the different components seem to interact to increase EPSC amplitudes on a linear scale to firing frequency, that will increase the dynamic range of neuronal firing without distorting the basic characteristics of thalamic relay. Fast facilitation would account for most of the gain of the neuronal amplifier, while augmentation primarily reduces the time required to reach an effective level of synaptic strength. Thus it might serve to preserve the gain of the neuronal amplifier during attentive visual exploration, when the gaze may return repeatedly to the same fixation point.