Fluorescent membrane potential. A good surrogate for E-phys?


Fluorescent membrane potential. A good surrogate for E-phys?
The key for every screen is to effectively predict efficacy in disease. Practically higher-throughput screens tend to sacrifice disease relevance for throughput. These then filter the compounds through to the lower throughput- more disease relevant assays. With ion channels it is often too expensive and/or throughput is too low to run large screens via electrophysiology and therefore fluorescent assays are often used as a higher-throughput and cheaper surrogate. The assumption is there is a good correlation between the two, but generally how good this correlation is only know after many compounds have been screened and the doubt is always there as to how many active compounds are discarded as false negatives.
In a recent paper Ghisdal et al., (http://jbx.sagepub.com/content/19/3/462.long) compared a set of ~41 compounds identified as GABAA binders by 3H flunitrazepam binding to rat brain membranes and compared their functional effects using a validated fluorescence membrane potential (FMP) assay vs. automated patch clamp. The correlation between 3H flunitrazepam binding and the fluorescent membrane potential assay was good. However, when the relative efficacy of these compounds in the FMP assay was compared to automated patch, there was no correlation and some compounds that produced significant potentiation in patch clamp were inactive in the FMP. This suggests FMP may not a good surrogate to measure compound efficacy and whether FMP should be used for high-throughput screening to identify novel chemistry.
A selection of fluorescent dyes are now available for measuring Na+ (sodium green, SBFI) and K+ ions (PBFI, FluxOR) and yellow fluorescent protein as a halide dye. It would be interesting to hear others experience with these dyes and whether they are more suitable surrogates for than a fluorescent membrane potential dye.
Key words
Fluorescent membrane potential
GABA

Advertisements

2 thoughts on “Fluorescent membrane potential. A good surrogate for E-phys?

  1. Is automated patch that expensive? I think the costs may look reasonable compared to flexstation/flipr if you have a look.

    In my years working with ion channels, fluorescent assays were considered acceptable for Ca channels but only because you could do 384 well plates. Everything had to be followed up on electrophysiology assay before anyone took compounds seriously, and my experience chimes with the paper you have posted. Quite a lot of people simply did not trust the potassium dyes. I never heard anyone ever talk about Na channel dyes ever.

  2. Indirect assays of ion channel activity have a reputation for being ‘error’ prone, with membrane potential (MP) being one of the most challenging – this relates to membrane potential being a non-linear function of channel activity. As such there are a number of factors that dictate how good (or bad) an MP assay will be – arguably most important are the cell background and the level of target channel expression, the latter determining assay sensitivity. That said I have had some good experiences with MP assays in hit finding but they don’t correlate well with absolute potency/efficacy measures in electrophysiology. Automated electrophysiology looks expensive as the equipment is costly and datapoint cost is relatively high, but it can be definitive. For me automated electrophysiologys ‘sweetspot’ is in i) triaging output of fluorescent HTS assays, ii) focussed screens with small compound sets, iii) providing benchmark data as a frontline assay to support lead optimisation.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s