Thursday, July 22, 2010

Calcium imaging sucks...

...unless you’re interested in calcium. If you’re using it as a proxy for channel activation then it sucks.

There’s a tendency among those of us who study calcium permeable channels to use fluorescence imaging of Ca sensitive dyes as a way to assess channel activity. And sure, the degree of channel activation will determine the resulting fluorescence signal, and generally speaking more activation will lead to more signal. Plus it’s just easy to do: in a day you can try a huge range of manipulations, you’ll get hundreds of cells as a result (making those statistics sure seem impressively significant). Compare that with a good day of patch clamp where 10s of cells is a GREAT day. I get it. I’ve been tempted by that dark side as well.

Still, frankly, it sucks. There are so many uncontrolled and/or untested variables present in your typical Ca imaging run that if you’re making dose-response curves, or inhibition curves, or whatever, then really, you’re deluding yourself if you think you’re only looking at the channel. And don’t even get me started on those people who treat a given delta ratio as equivalent over the whole range of ratios. GRRR.

In that case, it’s time to suck it up and do the electrophysiology. Yeah, it’s hard, it’s slow, blah blah waa waa whatever. If you want to characterize a channel, do electrophysiology. If you want to see how that channel affects intracellular [Ca2+], do imaging. But then really, WTF are you doing overexpressing that channel in some poor cell line, sitting in the incubator, just minding it’s own business?

10 comments:

Anonymous said...

Might I refer you to 18599484 ?

Anonymous said...

that PMID, sorry.

Nat Blair said...

Help me out here Anony: what's your point here? That Ca activated Cl current is a better indicator of [Ca2+]? Sometimes I'm a little slow on the uptake.

Anonymous said...

No, not superior, just simply that there is a way to measure [Ca2+] electrophysiologically. Not the dichotomy you present.
I think the point of the paper that struck me was that it was about the direct submembrane [Ca2+] as opposed to global concentrations. 'Course it probably only works in xenopus oocytes, unless you want to co-transfect the Icacl, too (and then ChR2 to calibrate for that matter, oh and set up a laser). No not better, just different...

Nat Blair said...

Yeah, I guess so. But the idea of using the ICl(Ca) in oocytes as an indicator gives me the willies. We still don't know the protein(s) that generate it, right? Sure, it might be useful in oocytes, and maybe helps you with the local [Ca2+] angle, but as you said, its applicability to other cell types is limited, to say the least.

As for that paper, I wouldn't objected to the authors doing actual calcium imaging, because they were interested in the effect of calcium.

Though you could do it with just electrophysiology alone, using pipette solns with [Ca] buffered to varying levels and assessing the GIRK response to mGluR activation (assuming you could get all those proteins in your target cells). Or you could have done caged calcium.

Daimo said...

Agree entirely.

Minor edit:
"If you want to characterize a channel, do electrophysiology. If you want to see how that channel affects intracellular [Ca2+], do imaging AND ELECTROPHYIOLOGY."

Not necessarily at the same time, although I've heard that some really awesome people do that, too.

Nat Blair said...

simultaneous ca2+ imaging and electrophysiology is totally awesome and really reserved for those with stunning technical ability.

:)

Austin Elliott said...

Ole Petersen's lab have spent many yrs using Ca-act'd K+ and (latterly) Cl- currents as a "proxy indicator" of local [Ca]. However, the technique still has a LOT of unknowns, as Nat says - especially w IClCa as we don't really know the channel activation mechanism.

Of course, TIRF microscopy has been heavily plugged as an "in between" stage where [Ca] signals are recorded from the submembrane space and therefore people claim they are "seeing" the Ca entering through the channels more directly.

Any views of TIRF used this way, Nat? I imagined the Clapham lab would probably have a TIRF rig.

Nat Blair said...

Austin, thanks for coming by. It's always nice to see other people interested in calcium and ion channels!

I'll be honest, I haven't given a lot of thought to TIRF measurements for local calcium, but my gut tells me it's a bit dodgy. And yes, here in the Clapper Land, we do have a TIRF set up. It was even used in a few papers! :p

There's some technical issues: usually you need a decent amount of excitation light, so lasers are often used (so fura is out the window). Also, multiple excitation wavelengths can be tricky to get in good spatial register (and might they fall off differently in intensity away from the glass surface? Can't recall off the top of my head).

Also, for chemical fluorophores, wouldn't the Ca-indicator diffuse away quickly during the excitation time (and faster than unbound Ca2+)? That might abrogate one's ability to really sample only the local submembrane [Ca2+].

Perhaps a membrane tethered protein fluorophore indicator would be better? Anyone ever see this being used?

Austin Elliott said...

Yeah, I have sort of automatic misgivings about TIRF for similar reasons. I wonder if one could do TIRF with Indo? Indo was designed for excitation with an Hg line at 365 nm, but there's probably a fancy laser that does 365 nm somewhere, if you're a lab with plenty of ££/$$.

Not sure about the TIRF indicator diffusing problem, though I guess the TIRF fiends must have some sort of standard answer. It's actually a problem a bit like the kind of things Erwin Neher used to model to calculate [Ca] "active zones" round the mouth of channels, so I would guess someone must have crunched the nos somewhere in the TIRF lit.

I always thought a membrane tethered/targetted Ca-dep fluorophore would be the solution, but I can't recall one being used. I suspect now that you would need a small molecule rather than a protein-based indicator, so maybe that is the stumbling block.

About a dozen years ago I think we all thought GFP-based Ca indicators like the Cameleons were going to be the solution to all this stuff. The first generation Cameleons were (by repute, not used them myself) a bit "un-bright", and low on dynamic range, so maybe that was part of the problem. Anyway, they never really caught on big time, for some reason, and the papers I do remember were all "multiple single cell photometry" rather than subcellular imaging. I am pretty sure there are 2nd generation ones now, but again I don't recall many papers. I used to be into measuring [Ca]ER, and the targetable Cameleons were hotly awaited for that, but then the whole field kind of went cold about a decade ago.

Anyway, I guess Fura et al are just so easy to use that you have to have a really compelling reason to accept less bright and less "easy" indicators.