July 02, 2008


The main reason Achilles is running Parallels -- and thereby Windows XP -- is so that I can use software for analysing patch clamp data from the SICM experiments that are finally and very slowly starting to approach the point where they might conceivably bear some kind of fruit. This little snippet from yesterday's screen cap shows just a few seconds of one such (nearly half-hour) recording:

a few seconds of one such trace

What you see here is a trace of the electric current passing through a very tiny patch of a cell's surface. The current is mediated by ion channels -- single protein complexes that span the membrane and change conformation to allow selective access to particular ion species. Such channels have been discussed a few times before on WT, but I'm too lazy to link to those posts and they'd be pretty irrelevant to today's discussion anyway. The point is that the steps up and down in the trace represent the activity of, in effect, single molecules. These things are really incredibly small, and yet -- with a lot of effort but in some ways surprisingly clunky equipment -- we can make out what they do as plain as day. Neat, huh?

Since this particular recording -- like all those I'm concerned with these days, and probably for the next two and half years -- was performed using SICM, we can also see where it came from, albeit in somewhat less detail:

scanning ion conductance images of the source cell

The resolution of these images is noticeably lower than some of the others I've posted, but that's because these were obtained faster. In SICM, especially when dealing with big delicate cells like neurons, there's a significant tradeoff between speed and resolution, and when you're looking to patch you're better off not hanging around.

In any case, the upper scan shows the body and some of the processes of a cultured hippocampal neuron.1 In the lower, we've zoomed in on the dendrite at the top left. Since this looked vaguely promising -- and not too tricky to hit -- we picked a spot pretty much in the middle, carefully lowered our pipette to the surface, nanometre by nanometre, and -- thanks at least as much to luck as judgement -- formed a high-resistance seal, allowing the single-channel recording to take place.

One interesting thing about the trace fragment above is that it seems to clearly show the activity of two distinct channels, with two different conductances. Alternatively it could be a single channel that has several subconductance states. Determining what is actually going on from this sort of recording is not at all trivial, and there are various software tools and techniques available to help. The one you can see in yesterday's screenshot is ClampFit, part of the pClamp suite also used to capture the data in the first place.

While ClampFit is a handy thing to have at one's disposal, it wasn't the single channel analysis software that motivated me to sully my laptop with XP. That was the collection by cantankerous scientific luminary David Colquhoun, arch-nemesis of mountebanks and snake-oil salesmen everywhere. I have yet to try these programs, but I expect them to be wayward, ugly, user-hostile and tragically indispensable. Given that I'm going to be spending a week later this blogarific month learning ion channel analysis and the techniques underlying this software from DC himself, I confidently predict you'll be hearing more about this stuff...

1 Cultured neurons, obviously, are the ones that sit around sipping tea while listening to chamber music and discussing Susan Sontag.
Posted by matt at July 2, 2008 10:21 PM

Uh-oh. Guess who read this post before you...

Posted by: matt at July 4, 2008 12:23 AM

Something to say? Click here.