I think both perforated patch and cell-attached patch are good techniques, but they each have some drawbacks (as does whole cell).

Not only does perforated patch take a while to stabilize, but series resistance in a perforated patch configuration can be fairly high. That could make it difficult to get good voltage clamp recordings. Also, in some cases you do want to dialyze the cell quickly to introduce a peptide or alter intracellular ionic concentrations.

Cell-attached patch is especially valuable if the channels of interest are not uniformly distributed and especially if they're along the dendrite (i.e., far from the site of a somatic recording). But currents from a small patch can be small and noisy; you have don't access to the cell's interior, to measure Vm or alter intracellular ionic concentrations; and you can't see how channels in your patch interact with all the other channels (e.g., to generate spikes), as in a whole-cell current clamp recording.

Anyway, that's my two cents. I'm also curious to hear what others think.

Interesting question! I agree with Niraj S. Desai. Yes there are many situations where perforated patch is the best method, the most evident one is for the study of Cl- driven ion currents without modifying the cells intracellular [Cl-] (e.g. Sauer et al 2012) . But whole cell also has its place and it is more straightforward.

In regards to the specific question "Why would one use whole-cell configuration when investigating the signalling events that activate a specific channel?".

In my experience whole-cell configuration is useful for this as you can change your intracellular solution (one factor at a time per cell - then compare across experimental groups)... to address very specific mechanistic questions. For example, whether the activation of your channel is Ca2+ dependent (using BAPTA or EGTA which could address fast/slow dynamics respectively, e.g. Mercer et al 2011), to assess the role of GTP eg. by using the nonhydrolyzable analog GTP-γ-S instead of GTP (e.g. Toth et al 1996), by using very specific exogenous peptides that interfere with protein-protein interactions to analyse ion channel regulation e.g. by kinases (e.g. Liu et al 2008) , or even include an active or inactive form of a recombinant protein in your patch pipette (e.g. Yang et al 2012).

Most of the ionic channel modulation events I'm referring to would survive the ~1hr recording and you would not be dialysing the entire contents of the cell.

On the other hand, yes it is known that there is some washout of intracellular components required for LTP after whole-cell breakthrough. For this reason most researchers stick to 5-10 min baseline in whole cell configuration for LTP measurements.

Your question is important and is as old as the invention of the technique itself. Sakmann, Neher and many others addressed this question in length (many books are out there; Just google 'Patch clamp technique books'). This limitation is inherent to the configuration of the whole-cell mode but can be useful in some situations and annoying in others as discussed by Mariana and Niraj (above). 'Small' molecules such as ATP, GTP, phosphocreatine that are crucial for the metabolism (well-being!) of the cell are usually added in mM amounts in the pipette to prevent their depletion. But the depletion of ATP is an important question and was examined in W-C mode simply by omitting ATP. Other molecules like cyclic nucleotides (cAMP, cGMP...etc) could be added to the pipette media to manipulate the activity of channels (CNGs and HCN...etc) that require cyclic nucleotides binding. You could also add permeable cyclic nucleotides to the bath and reproduce the same effect as seen when you added them from the pipette. Elegant experiments using an adequate flash light to instantaneously increase cytosolic amounts of Calcium for example or cyclic nucleotides or other molecules such Inositol tris- and hexakis-Phosphates..etc were used in W-C mode. Bottom line, All techniques have some sort of drawbacks and the main important thing is to know them and do all the required controls to validate (or may be not...) your findings.

Ok,

a big thank you to all of you! Very helpful! I am going to have more fun now reading ephys papers, watching out for this "little" details why and how they measured their specific channel and this specific way.