Visual Classical Conditioning in Wood Ants.

Several species of insects have become model systems for studying learning and memory formation. Although many studies focus on freely moving animals, studies implementing classical conditioning paradigms with harnessed insects have been important for investigating the exact cues that individuals learn and the neural mechanisms underlying learning and memory formation. Here we present a protocol for evoking visual associative learning in wood ants through classical conditioning. In this paradigm, ants are harnessed and presented with a visual cue (a blue cardboard), the conditional stimulus (CS), paired with an appetitive sugar reward, the unconditional stimulus (US). Ants perform a Maxilla-Labium Extension Reflex (MaLER), the unconditional response (UR), which can be used as a readout for learning. Training consists of 10 trials, separated by a 5-minute intertrial interval (ITI). Ants are also tested for memory retention 10 minutes or 1 hour after training. This protocol has the potential to allow researchers to analyze, in a precise and controlled manner, the details of visual memory formation and the neural basis of learning and memory formation in wood ants.

Comparison of Langevin and Markov channel noise models for neuronal signal generation.

The stochastic opening and closing of voltage-gated ion channels produce noise in neurons. The effect of this noise on the neuronal performance has been modeled using either an approximate or Langevin model based on stochastic differential equations or an exact model based on a Markov process model of channel gating. Yet whether the Langevin model accurately reproduces the channel noise produced by the Markov model remains unclear. Here we present a comparison between Langevin and Markov models of channel noise in neurons using single compartment Hodgkin-Huxley models containing either Na+ and K+, or only K+ voltage-gated ion channels. The performance of the Langevin and Markov models was quantified over a range of stimulus statistics, membrane areas, and channel numbers. We find that in comparison to the Markov model, the Langevin model underestimates the noise contributed by voltage-gated ion channels, overestimating information rates for both spiking and nonspiking membranes. Even with increasing numbers of channels, the difference between the two models persists. This suggests that the Langevin model may not be suitable for accurately simulating channel noise in neurons, even in simulations with large numbers of ion channels.

Shaker K(+)-channels are predicted to reduce the metabolic cost of neural information in Drosophila photoreceptors.

Shaker K(+)-channels are one of several voltage-activated K(+)-channels expressed in Drosophila photoreceptors. We have shown recently that Shaker channels act as selective amplifiers, attenuating some signals while boosting others. Loss of these channels reduces the photoreceptor information capacity (bits s(-1)) and induces compensatory changes in photoreceptors enabling them to minimize the impact of this loss upon coding natural-like stimuli. Energy as well as coding is also an important consideration in understanding the role of ion channels in neural processing. Here, we use a simple circuit model that incorporates the major ion channels, pumps and exchangers of the photoreceptors to derive experimentally based estimates of the metabolic cost of neural information in wild-type (WT) and Shaker mutant photoreceptors. We show that in WT photoreceptors, which contain Shaker K(+)-channels, each bit of information costs approximately half the number of ATP molecules than each bit in Shaker photoreceptors, in which lack of the Shaker K(+)-channels is compensated by increased leak conductance. Additionally, using a Hodgkin-Huxley-type model coupled to the circuit model we show that the amount of leak present in both WT and Shaker photoreceptors is optimized to both maximize the available voltage range and minimize the metabolic cost.

Spike width reduction modifies the dynamics of short-term depression at a central synapse in the locust.

Short-term synaptic depression is an important component of computation within neural networks, but little is known of its contribution to information processing during synaptically generated spike trains. We analyzed short-term synaptic depression at a synapse between two identified motoneurons innervating the hind leg of the locust: the FETi-FlTi synapse (fast extensor tibiae-flexor tibiae). Brief electrical stimulation of a single hind leg proprioceptor, the lump receptor (LR), led to prolonged sequences of spikes in FETi, similar in number and frequency to those during natural kicking movements. Depression at the FETi-FlTi synapse during LR-evoked spike bursts was compared quantitatively to that during antidromic spike trains evoked by electrical stimulation of FETi in the extensor tibiae muscle, and by modeling. The magnitude of the short-term depression was significantly greater during LR-evoked spike trains. On the basis of the model parameters required to fit the depression, the FETi-FlTi synapse is predominantly used for transmitting the timing of the onset of FETi spiking rather than its spike rate. During LR-evoked spike trains, there was a rapid reduction in presynaptic spike width that did not occur during antidromic spike trains under physiological calcium concentrations. This produced a concomitant reduction in the amplitude of the FlTi EPSP, suggesting that it contributed to the differences between the two stimulation regimes. Differences in the short-term depression between synaptically evoked and antidromic spike trains emphasize that the properties of synaptic information transfer are dependent on the in vivo conditions at the synapse and may not be reproduced by in vitro spike trains.