Electrophysiology protocols for LTP and LTD: Recording synaptic plasticity with whole-cell patch clamp

Long-term potentiation (LTP), first discovered in hippocampal neurons by Bliss, Lømo and Gardner-Medwin ^1,2, has been described by many as the canonical form of synaptic plasticity. Long-term depression (LTD), its logical partner, was identified in the following two decades – first in the cerebellum³, and then in the hippocampus ⁴. The basic principle of both is this: synapses are able to adjust their strength in response to the magnitude of inputs they receive, shaping network dynamics and storing information – and as such they represent a type of learning that occurs at the level of the synapse. In LTP, when high-frequency stimulation is applied to presynaptic cells, their postsynaptic partners show an elevation in amplitude of excitatory postsynaptic potentials (EPSPs)¹. In LTD, application of low-frequency stimulation triggers a decrease in amplitude of EPSPs⁴. Under certain experimental conditions, LTP and LTD can be stimulated in the same cell sequentially⁵.

Often, LTP/LTD experiments are performed using extracellular field recordings⁶, which are easier to maintain for a prolonged period than a patch. However, field recordings output fEPSPs (field EPSPs), which represent the summed synaptic activity experienced by many cells. If you want to record LTP/LTD within an isolated cell type occurring within heterogeneous tissue, you will need whole-cell patch clamp, which allows you to record synaptic events from within specific individual cells.

Trying to record LTP or LTD from acute slices in the whole-cell patch configuration? Here’s your guide.

When patching, it is most straightforward to record synaptic events in voltage clamp, meaning that they are recorded not as EPSPs, but as excitatory postsynaptic currents (EPSCs). You can record EPSCs under voltage clamp at -60mV, using a K⁺-gluconate internal, or for a better space-clamp use a Cs⁺-gluconate internal and clamp the voltage at -70mV. In both configurations, EPSCs will register as inward or negative-going events with the classic shape of rapid onset and slow decay time. Note that some plasticity protocols rely on a post-synaptic action potential - in these cases, potassium based internal solutions must be used⁸, as a caesium-based internal will alter AP waveforms and impede high-frequency firing. In addition, the cell should be help in current clamp mode instead of voltage clamp, for either the full experiment or the induction period, to allow APs to alter the membrane potential and initiate downstream effectors.

There are two options as to how you go about recording synaptic events – spontaneous, or evoked. The most straightforward option is to record spontaneous synaptic events (spontaneous EPSCs, aka sEPSCs), particularly if you know that the vast majority of inputs received by your cell will be from your presynaptic partner cell of interest. The more reliable option, however, is to use evoked excitatory postsynaptic currents (eEPSCs), as in this case you know the provenance of the events you evoke. In the olfactory bulb, for example, an effective way to do this is to place your extracellular stim electrode on axon bundles in the olfactory nerve layer – this way you know that the events you evoke come from the olfactory sensory neuron axons you stimulate. Typically, you will already need a stim electrode placed for the LTP/LTD induction itself, so this option should require little extra work, and in the absence of other counterindications is perhaps therefore a no-brainer. Note that eEPSCs, like fEPSPs, are also summed events, but they are a composite of many inputs to the same cell, whereas the latter is the composite of many inputs into different cells. A benefit of using eEPSCs is that they are larger than sEPSCs, and therefore less likely to be obscured by noise. If using eEPSCs, use the same intensity for every test-stim throughout the recording duration, at least on a within-cell basis.

Record eEPSCs at consistent intervals of 60s to allow the synapses to recover between stims and to reduce the likelihood of inducing any unintended plasticity. Make sure that the pulse from your stim electrode is very brief, so as to not obscure the recorded event, and the monosynaptic EPSC should occur after a very brief latency (<10ms). Repeat for approx. 15-minutes to establish a clear baseline.

A good practice is to play around before initiating your actual recording with the amplitude of the stim so that you get a moderately sized EPSC to begin within (perhaps around 500pA), to give it room to either grow or shrink according to your induction protocol. This means that individual cells receive different nominal stimulation intensities, but intensity remains consistent within-cell, once the actual recording starts. A more in-depth method for doing this is to map a basic input-output (I-O) curve for your cell, so that you can identify the stim intensity at which the EPSC amplitude reaches its maximum. For LTD, it is clear that you need a starting size of EPSC that is large enough to allow for shrinkage with plasticity. However, it has been argued that the reverse is not true for LTP⁹. At first thought, it seems clear that choosing a test-stim intensity in the middle of the I-O dynamic range curve allows space for growth. However, movement up along the I-O curve in this context may reflect better the number of axon bundles being stimulated rather than the strength at which individual synapses are stimulated. Thus, picking a higher test-stim intensity may mean more axon bundles stimulated, and more synapses participating in the recording which have potential to undergo LTP, resulting in a larger and clearer result.

Lastly, when using eEPSCs, it’s very important to confirm that your stim electrode is stimulating the presynaptic axons, but not stimulating the postsynaptic cells directly. This way you can be sure that the events you measure are entirely synaptic in nature. This can be done easily by performing some test recordings in the presence of TTX, which, if no direct stimulation is present, should totally abolish the eEPSC (by inhibiting transmission in the presynaptic axon). It is a good practice to establish a working distance between the stim electrode and postsynaptic dendrites (in olfactory bulb this is the distance between the stim electrode and glomerulus), validate this distance, and then stick to it.

If you keep in mind all of the above, with luck, your experiment may well be successful! For more information on LTP and the methodologies that have historically been used to investigate it, see these amazing reviews by Jiang and colleagues12, and Glasgow and colleagues13. Best of luck!