click on a figure to see a larger view

1. Defining the circuit necessary for the expression of trace eyelid conditioning

What forebrain regions are necessary for trace eyelid conditioning? To determine this, we are implanting cannulae in forebrain regions hypothesized to be necessary for trace eyelid conditioning. Then, using a dual delay-trace conditioning paradigm, we can test which forebrain regions are necessary for trace conditioning using reversible inactivationIn contrast to lesion studies which irreversibly damage the brain region of interest, we infuse neuroactive substances (e.g. muscimol which mimics inhibitory transmission), which impairs behavioral performance for a few trials. with the infusion of neuroactive compounds. Using this strategy we have successfully identified the medial prefrontal cortex (mPFC) as necessary for trace eyelid conditioning (Kalmbach et al., 2009).  In addition to identifying necessary forebrain regions, we can elucidate the role of neuromodulators within those brain regions to the expression of trace eyelid conditioned responses.
For all of our behavioral experiments, we train rats in both delay conditioning (which only requires the cerebellum) and trace conditioning (which requires forebrain regions, including mPFC). This provides us with an important control to ensure that our manipulations are not impairing the animals’ overall ability to respond. Rat behavior

2. Identifying the electrophysiological features of Layer 5 mPFC neurons in vitro
A main source of output from the mPFC are Layer 5 pyramidal neurons. Some of these neurons project to specific regions of the pontine nuclei that in turn provide the trace information to the cerebellum. Using vital retrograde tracersRetrograde tracers travel from axons/terminals back to the cell body. Infusing retrograde tracers into a brain region (e.g. the pons) will label neurons that project to the infusion site. We utilize Lumafluor beads, which are fluorescently-labeled latex microspheres that give the soma of a labelled neuron a fluorescently punctate appearance (Katz, 1984). The beads offer several distinct advantages. They are retrogradely transported quickly (in 2-3 days), and label the neurons for at least several months. Importantly, they are “vital tracers” because they are non-toxic, allowing us to record labeled neurons’ electrophysiological properties.  (see Lumifluor retrobeads and Katz, 1984), we are identifying the electrophysiological properties of both neurons that may contribute directly (by projecting to the pons) and neurons that may contribute indirectly (by projecting to the other hemisphere of mPFC). Additionally, we are determining how the electrophysiological properties of these neurons change with neuromodulation and after the induction of intrinsic plasticityA neuron's input/output properties are determined, in part, by its repertoire of voltage-gated ion channels. Intrinsic plasticity refers to enduring changes in the conductances mediated by these channels and the resulting changes in input/output properties..
Bead-filled neurons
Strikingly, we have found that mPFC Layer 5 neuron that project to the pons have distinct electrophysiological properties and respond to certain neuromodulatorsNeurotransmitters (the chemical compounds that transfer signals between neurons) can act by directly changing the voltage of a neuron by binding to a ligand gated ion channel, or indirectly by binding to a receptor that in turn leads to an internal signal that alters ion channels. Neuromodulation generally occurs via this indirect signalling, which changes the way neurons will integrate and transfer information. Neuromodulators in the brain (such as dopamine, noradrenaline and acetylcholine) are believed to play a critical role in attentiveness and may help working-memory tasks. differently than neurons that project to the contralateral cortex! This work has recently been accepted for publication in the Journal of Neuroscience (Dembrow et al. 2010).  In addition to these perisomatic differencesthe subregion of the neuron containing the nucleus and surrounding cell machinery. This region connects the dendrites and the axon, and is the region electrophysiological parameters are most commonly recorded from., we will also examine whether the dendrites of these neurons integrate information differently.

The dendrites of neurons were once viewed as passive integrators of synaptic inputs, but pioneering work by the Johnston lab and others has demonstrated that dendrites contain a myriad of active conductances that enhance a neuron’s computational power. Because voltage fluctuations in the dendrites are electrically distant from the cell body, understanding dendritic contributions and how they are altered with learning requires direct dendritic recordings. Thus far no such recordings have been made from identified projection neuron dendrites…..until now!

Simultaneous soma and dendrite recording in a CPn neruon

3. Determining the necessary inputs that support persistent activity in vivo during trace eyelid conditioning
Neurons in the prefrontal cortex increase activity in response to predictive stimuli in order to guide future or ongoing behavior – even beyond the offset of the stimuli. These neurons are thought to contribute to working memory by maintaining a persistently active state (a memory trace of the stimulus) to bridge time until a behavioral response is made. We and others observe tone-evoked persistent activity in the deep layers of mPFC (and the pons) that spans the trace interval during trace eyelid conditioning. But what are the necessary inputs/signals that induce persistent activityPersistently active neurons are often called “delay cells” because they were initially discovered in the context of match-to-sample delay tasks. These tasks are procedurally similar to trace eyelid conditioning in that they impose a delay between the presentation of a stimulus and the outcome the stimulus predicts. in vivo?  And, exactly which neurons display persistent activity – those that project to other brain regions and/or those that project to the pons?

Using high-density in vivo recording techniques combined with pharmacological manipulations and/or antidromic stimulation in mPFC-associated brain regions we are working to determine the inputs that play a role in  persistent activity, and to determine the target brain regions that receive that output during trace eyelid conditioning.

4. Mapping the brain circuitry of
trace eyelid conditioning

To accurately map out the circuits underlying trace eyelid conditioning requires a carefully derived anatomical atlas of the neurons and specific brain regions that participate. To make our research truly integrated between the projects listed above, we are building a combined atlas to map out which regions pharmacological manipulations, in vivo recordings and in vitro recordings are made. We are engaging in a careful, iterative process by which we will hone in on the necessary structures and reveal the brain circuitry and neurochemicals necessary for the working memory-like mechanisms that support trace conditioning. For example, having identified retrogradely labeled mPFC neurons by infusion of Lumafluor beads in the pons, we then target those labeled mPFC regions with anterograde tracers to confirm the specificity of our labeling. To see our anatomical atlas thus far, click here.