Science sometimes makes me feel like...

Let's just hope I don't die some rock and roll death. You know, the "Nat didn't show up in the lab for the big experiment, we found him having choked on vomit, but we don't know whose vomit it was, cause you can't dust for vomit." It's either that or a bizarre gardening accident, and it is spring time.

(Seriously though, I love science, right beneath the family, but sometimes it's a heartbreaker. Goddamned unrequited love.)

Sometimes, the little things are what keep you going

When you're having a crazy busy morning, and experiments are going kinda crappy (I'm at the "let's remake the solutions" stage), it can be nice to get a Friday morning email alert to 3 fresh citations to your papers! Sure, it's a little thing, but at this stage I'll take what I can get.

People are out there reading, so there must be people out there who care. Keep 'em coming folks, keep 'em coming!

By the way, what's with variability between the citation search engines? I know this has been covered before somewhere, but really.

Searching on my first paper from grad school, we get:

90 cites in ISI from the J Neurosci site.
93 from ISI itself
98 from Scopus (I got a free preview for reviewing a paper. It's pretty cool, but I haven't used it enough to say much substantive. I do like the display of citations by year, Journal, author).
102 from Google Scholar

Now, so of these I'd cut slack for (e.g., J Neurosci probably only pulls cite numbers from ISI infrequently, and who know how reliable Google Scholar is, but what's the difference between ISI and Scopus? Anyone look deeper into this?

If citations, h-indices and impact factors have traction as important metrics, shouldn't they be, oh I dunno, accurate?

Electrophysiology isn't a technique you add to your CV; it's a state of being!

Neuropharma's comment on my last post contained something that stuck in the craw of this old electrophysiologist. Some grad student she knew thought he could waltz over and learn some electrophysiology right quick, and include it in his thesis.

Then reality struck this student, rapidly disabusing them of this conceit:
"He was shocked to discover that it would take him such a long time to learn the technique (he's starting from level 0) and said that it seemed so easy when reading it from some published paper!"

Every newb thinks that a technique they haven't mastered is easy, until they actually try it. And in fact, the bare bones mechanics of patching are pretty straightforward. I've taught a lot of novices how to patch, and by and large they can get to the point of gigaohm seals in a week or two (ok, we're talking transfected HEK cells here). Hell, I'm thinking any primate above lemurs could learn to get seals. (Not a bad idea actually; screw those automated patch systems, gimme an army of squirrel monkeys and an old warehouse, and I'll screen your chemical library right quick! It'd be like the nut shelling squirrels in Willy Wonka. And they'd literally be DrugMonkeys! LOLZ.)

But there's a huge distinction between the currents you're recording at that point, and 'good' currents. These first currents are crap, e.g. the leak is terrible, the series resistance is awful, the throughput stinks, the solution applications kill cells or generate huge noise, you've got visible 60 Hz pickup. At best they're barely interpretable. But Electophysiologicus will dole out a decent cell here, a nice recording there. That'll be just enough to keep you coming back for more, and to keep the image of good recordings in your mind's eye.

The transition between the crappy recordings of the apprentice and the regular good recordings of the master takes a long, long time, on the order of a year I'd say.
These are the dark times, where the progress is non-existent, perhaps to a greater extent than an analogous part of the curve for other technical subspecialties. Most electrophysiologists I've talked with had this time in their training, typically falling into the 2nd year of graduate school.

And yet, there's very little useful advice the masters can give their apprentices during this time, other than "keep at it". Sure, there are suggestions to try this, or don't do that. In the end though, everyone just has to put in their time, slowly perfecting each requisite skill, and evolving their own personal technique. It sucks, for sure, but it does end.

It's just not gonna end before your rotation or your last few months before you finish your thesis.

A pictorial presentation of pipette pulling

In the interests of both responding to Dr. A's request for pipette pulling related pics, and appeasing the apparently still irked electrophysiology gods, I present to you a brief montage of the glorious task of patch pipette fabrication!

First, what the heck are we even doing? Well, we're gonna pull a glass needle, fill it with salt solution, stick it on a plastic holder with a wire inside, maneuver it to a cell, apply a little suction, and let the magic of "seal formation" occur. Next, assuming we're doing whole cell voltage clamp, we break the seal membrane with more suction, gaining control of the voltage across the cells' remaining membrane, while recording the current (also filling the cell with our pipette solution). Sheesh, when you distill it down to two sentences, it pretty much trivializes what I spent years learning and do everyday.

The opening of the pipette tip will be ~1 µm, while the cell is on the order of ~10 µm. Obviously, if the pipette tip is too big, then we'll just suck up the entire cell. Not good. But, as the pipette tip gets smaller, the resistance between the pipette interior and the cell interior gets larger. Also not good. In fact, that causes a whole host of problems that are left as an exercise for the reader to derive (ok, just kidding. There's a series of posts reserved for this, with current working title: "Dr. RseriesLove, or, How I learned to stop worrying and love the fact that my currents are all wrong)

I start by cutting the capillary glass, by scoring it with a diamond pencil and breaking it off to the correct length (so it'll fit in my particular set up, given the headstage position, etc.).

Then I smooth the ends of the capillary glass with a bunsen burner flame, because jagged end (even how it comes from the factory) will scrape off the silver chloride on the wire that transmits the current from the ions in solution to the electrons in the amplifier circuitry (as well as tearing up the O-ring in the headstage holder).


Then we move onto the puller itself. There are many different kinds of pullers, but having been in a number of electrophysiology labs, I would say most people use pullers made by Sutter Instruments. The basic puller operation is to melt the glass capillary while pulling on either end, drawing the ends apart. Now to get a nice wide tip patch pipette, we use computer controlled application of the heat, allowing you to stop the heating a certain time after the capillary begins to pull apart. Over repeated heating/cooling cycles, you can make the perfect pipette.
Here's the puller, a P-97, and if you unscrewed the 5 screws on the font panel, you could peer in and see the brushless super quiet 92 mm fan we installed (way in the back of course, a real pain in arse to reach). The smoked plexiglass cover opens to reveal:


The business end of the puller. The circle thumbscrews clamp down on the ends of the capillary and maintain tension. The capillary feeds through the box filament, which gets hot when the puller is activated (sorry for the flash glare here). When the glass separates, we're left with a pair of pipettes, which we fire polish by bringing them close to a red hot wire (observing under the microscope). Finally, we're ready to patch!




The pipette is filled with intracellular solution, stuck in the polycarbonate holder (which has the silver wire in it), and stuck into the headstage of the amplifier. The suction tube allows you to provide postive pressure while you're approaching the cell, or negative pressure to form the seal and to breakthrough. The cells are sitting in an extracellular-like solution in the chamber, and the pipette approaches under micromanipulator control (here, a piezoelectric based Sutter MP-285), all the while watching through the microscope. In fact this pipette in this picture is making a GOhm seal on a little HEK cell. Of course, when I applied suction to break through, this cell was terribly leaky (again, Electrophysiologicus, patron of patchers, I'm like so over that hubris- could we maybe move on now?).

If you look closely, the tip of the pipette is wrapped with a thin strip of Parafilm. This helps reduce the capacitance of the pipette, but isn't nearly as time consuming or messy as using Sylgard. A requirement for setting the series resistance compensation. All of which are good topics for future posts!

Hope this was at least mildly useful to some people out there, and marginally enjoyable to others. If anything's not clear, just fire up the comments and lemme know.

We got monkeys running all over the house!

Ok, well not so much running, though those little legs are always moving! There's plenty of room to grow into it, as the young'un is still pretty mini. But she's a happy one, especially while modeling DrugMonkey schwag! So what are you waiting for? Got get yours here. Do it, do it now!

In other news, I turned 35 yesterday. It was a great day (thanks to the wife and family!), but frankly it's a crappy age. First time I actually feel old on my birthday. *sigh*

2 grand? How about 20 bucks?

Ha!

I've been having a recent bit of difficulty with our pipette pullers in the lab. Well, pulling program parameter searching is no news to any of the l33t electrophysiologists who frequent this blog. We've all been there, and we'll all surely revisit that terrible state of being.

But not too long ago one of the pullers started making a horrible noise when switched on, as the cooling fan must have lost a ball bearing. The company service folks were in the area for the recent Biophysical meeting (how was it Dr. Samways?), and made a swing through the lab. They offered to refurbish the whole thing and replace the fan for $2000. Not a pressing issue, but apparently as the heat builds up inside the puller, it would slowly dim the display, making it hard to edit programs.

Economic times being what they are, we demurred. Which is a good thing, because when we pulled the front panel off that sucker, the fan was just a 92 mm fan like you'd have in your computer. So we got one of those, pulled out the old dead one, and slipped in the new. Voila! And while we were at it, we got rid of the decade plus layer of dust in that thing.

Still though, I miss the old P-84 I used in my thesis lab.

Hubris - and that ain't a circumscision in Whoville

Alas, I fear I have angered the Gods of Electophysiology with my last post. 

For now in their puckish ways, they have sentenced me to several hours fidding with the puller, in a vain sisyphusian search for something resembling a stable program. Or really anything that might give a useable pipette.

I've been burned bad. So what does one do when burned like Icarus in the tale of old?

WHY, IT'S TIME FOR MAIDEN!

(I'm digging the bass in this one!)

And it's back to the puller...

It's a beautiful....pair of sodium currents!

I was recently looking back over one of my papers from my thesis work, and came across a recording that I frankly find just beautiful. I had it in my mind to post it, and then Dr. J's post about their recent beautiful gel spurred me on to finally do it (another of the benefits of non-pseudonymity):

Figure 1: Sodium currents during action potential waveforms in nociceptive sensory neurons.

Look at that goddamn subtraction! And that outward transient current in TTX-R? Gating. Current. Sure it might be from a 6 year old paper, but it still thrills me a little. Some of this nostalgia might be related to the comparative lack of beautiful TRP channel currents I encounter these days. Oh, for a TTX of TRP channels. Oh, for closing your channels with negative potentials. *sigh*

As I commented over at Dr. J's place, beautiful data don't necessarily equal meaningful data (same with the converse), but I do think there's likely some correlation between the two.