Wednesday, April 1, 2009

What is a junction potential?

It's clear to me that a number of visitors to this humble blog arrive each day via a Googly search for the term "junction potential." I can only imagine that some must be fellow electrophysiologists, perhaps in their formative larval stages, searching for more information about this important topic. So, as a service to these folks, I thought a post or two about junction potentials, would be in order. First, what is a liquid junction potential? Then, How do you measure and correct for them?

So, what is a liquid junction potential? Sure, maybe you could look in some of the Electrophysiology Bibles. Or maybe you could even hit up an electrochemistry textbook. But it's 2009, and you've got two things on your side: Google, and me. So forget that, and allow me to regale you with the story of the liquid junction potential:

Long ago, in a galaxy far far away, there was a Gedanken experiment...
Figure 1: Set up of the Gedanken. No, it ain't to scale, though aspartate is bigger than potassium. Not shown is the impermeable wall separating the two solutions. Hey, it's my Gedanken thank you very much.

And in this Gedanken experiment there was a pipette filled with your typical pseudo-intracellular solution: You know the drill, high potassium (light blue), low calcium, and an anion species that's usually not chloride. This anion could be something like methanesulfonate, gluconate, or my own personal favorite, aspartate. The main thing to note is that all of these are bigger than chloride, and bigger than potassium. Thus, they have a lower mobility, meaning they don't diffuse as quickly as the accompanying cation.

Now, what happens when we stick this pipette into a bath solution that has your typical extracellular saline, made to mimic extracellular fluid (i.e., mostly sodium chloride)? Well, the chemical gradients favor the pipette constituents diffusing into the bath, and the bath constituents diffusing into the pipette. But remember, the aspartate is big, so it doesn't diffuse as quickly as any of the other ionic species. That slower diffusion of the anion leaves a net negative charge in the pipette. This charge separation across the junction between two solutions is THE LIQUID JUNCTION POTENTIAL!!!11!!!1!


Figure 2*: The Gedaken imposed barrier is removed, and ions are diffusing down their electrochemical gradients. The bigger, slower aspartate can't keep up relative to the smaller, faster potassium, sodium and chlorides. They get left behind in the pipette, generating an excess of negative charge.

Note that a liquid junction potential would also occur if the bath solution has cations and anions with significantly different mobilities. It just turns out that sodium and chloride have pretty similar mobilities, so that their contribution to the liquid junction potential is much smaller. But if you have N-methyl-d-glucamine (NMDG) as the main cation in your pipette solution, you'll have an excess of positive charge in the pipette solution, and a corresponding slightly positive junction potential.

Next up, how to measure the liquid junction potential.

---

*-Note that these figures were created using Inkscape, a very cool and usable opensource vector
graphics drawing program (a la Illustrator). Check it out, download it, play around with it!

23 comments:

Comrade Physioprof said...

Junction potential!?!?!?

HAHAHAHAHAHAH!! Quit fucking with n00bs on April Fool's Day, dude! You know there's no such thing as a "junction potential". HAHAHAHAH!

Nat Blair said...

Goddammit CPP! Why's you gotta ruin a perfectly good nefarious plot by speaking your truth to power.

Dr. Jekyll and Mrs. Hyde said...

Here's my question: how come a lot of labs state in their methods that they have NOT corected for the JP? Isn't that....rude?

(uh oh, here's where it turns out that you do this)

Nat Blair said...

Not correcting for their junction potential is for a bunch of lazy, know-nothing MFers who deserve to be ridiculed until they lay down their pipettes or shape up. Seriously folks, it's not that friggin hard. I can only imagine these idiots don't know how to work their analysis program, so they can't figure out how to apply a constant offset to their command voltage.

And don't worry, yours truly is not in that ignominious group. In fact, I alter my voltage protocols prior to the experiment to account for the particular internal solution junction potential. Again, it's not that hard, it just takes someone willing to put in a tiny bit of extra effort for a slightly more elegant output. But, I'm anal retentive that way.

It's easy to tell when somebody does this, because there's no goofy -68 mV, and steps from -68 to +22 mV. Uh, no, you stuck -60 to +30 in your program. Still though, if you do it this way, at least correct for it post-hoc.

Anonymous said...

ha i'm one of those larval staged pipette manipulators and i'm very thankful

Nat Blair said...

Hang in there Anony. It will work out, but you gotta keep the faith!

Anonymous said...

Thank you!

Biochemistry noob here, came across this simple and useful explanation while studying potentiometric analytical methods.

Unknown said...

We wrote an open source program for the calculation of the liquid junction potential. It can be downloaded for free
here. I hope it can be useful!

Nat Blair said...

Cool Doriano, I'll have to check it out.

Where did you get constants for all the cations and anions? That's always my main worry when using the calculations. How good are those constants for things like Tris, NMDG, aspartate, methansulfonate, etc.?

Neulectro said...

Hi, first thanks for that post! I have a (maybe...) stupid question. If I offset my pipete (in CC), and zeroing it in the bath before approaching the cell. Why do I need to add the correction afterwards?...
Thanks!

Nat Blair said...

Hey Neulectro,

The issue is that although there is a liquid junction potential when the open pipette is first in the bath, there is no longer junction potential once you have a gigaohm seal. So, when you are initially zeroing the pipette in the bath, part of what is in that amount of mV is the junction potential. But then when you have a seal, that junction potential is gone, though the pipette offset potentiometer hasn't moved. Does that help?

Neulectro said...

Hey Nat,
Yes it did, thanks!

Unknown said...

Bitch you showed that you are really a fool on april. I think you are an history student who is commenting about chemistry.

Nat Blair said...

Thanks for the insightful comment Unknown. I think it probably helped people here quite a lot!

Joe said...

Hi Nat,

Thanks for this handy explanation. I am still a little confused as to your answer to the above question from Neulectro. I understand that you offset the junction potential in the bath, and this offset remains constant though your conditions change when you go gigaseal, necessitating the correction.

However, when you go whole cell, I'm at a bit of a loss as to what happens, as you are dealing with a completely unknown LJP between the cells endogenous intracellular solution and the pipette intracellular solution.

Presumably this "endogenous" intracellular solution is more similar to the pipette intracellular solution than the bath solution, so won't any correction based on the measured potential between bath and pipette ICS become meaningless? Also, as the experiment progresses and the pipette ICS dialyzes into the cell, will the LJP change slightly over time?

Thanks for your help with this, maybe I have confused myself :D

Cheers,
Joe

Nat Blair said...

Hi Joe!

I don' think there's too much to worry about with respect to the contribution of the cytoplasm. First, the volume is so small relative to the giant pipette, so that over time the internal solution will dominate.

And even in cells where it takes some time for the internal solution to replace the cytoplasm, the internal and the cytoplasm usually have similar amounts of cations/anions, each with similar mobilities.

Struggling Postdoc said...

These comments are cracking me up. Thanks for the post, I am a mostly self-taught electrophysiologist and struggle with these lesser-known nuggets without a mentor!

Nat Blair said...

Glad it's helpful! If you have other questions, feel free to ask away. Electrophysiology has lots of t icky things that make it easy to get tripped up.

Casey V said...

Thanks everyone for the post and comments. Similar to Joe, I have a similar confusion. If the LJP between pipette and cytoplasm is negligible, why must be correct the LJP in the bath? If we correct in the bath, and this correction is maintained throughout our experiment, then when we break into whole cell, aren't we correcting when a correction is no longer necessary?

Thanks

Nat Blair said...

The real issue Casey is that when you first put the pipette in the bath, and zero the current on the amplifier, the liquid junction potential is part of that. (and that's the junction between the pipette solution and the bath solution).

Then, when you break into the cell, there is no longer a liquid junction potential (the cell membrane blocks free diffusion of the ions).

BUT...your pipette offset still is accounting for the potential that was there. So the correction we're doing is essentially getting rid of the liquid junction potential portion that shouldn't be there. That help at all?

Casey V said...

Yes, that's very clear. Thanks!

Unknown said...

Hi Nate!
As a patch clamper myself (but yet only elementary), I've been having fun reading through your blog.
Recently, I've been trying to record chloride currents on cell lines by using NMDG based bath and internal solution.
The thing is, it seems almost impossible for me to make a good giga seal when using NMDG based bath solution. So as a plan B, I am making giga seals in normal extracellular solution (140 NaCl, 5 KCl, 2 CaCl2,...) and then change the bath to NMDG to make whole cell configuration.
But is this okay? I see that pipette offset (probably due to different liquid junction potential?) changes considerably when my pipette is exposed to normal ECS and then to NMDG based bath.
And, if possible, could I get some advice on making giga seal in NMDG based bath solution? I originally used NMDG based bath with 1.2 CaCl2, and to see if there are any changes, raised calcium concentration to 2.5 CaCl2, but it didn't make any difference.
Thanks!

Nat Blair said...

Hi, and welcome to the world of patch clamping. Don't forget, we all started at an elementary level, so we've gone through some of the confusion you're experiencing. Hang in there and you'll get it.

As for your junction potential question, yes, it's ok to first make the seal in normal extracellular solution, and then switch to something else. The correct junction potential to use for the offset is that between the internal and external solution when you first put the pipette in the bath.

Note that when a seal is formed, the cell membrane prevents the free diffusion of internal to external solutions so there is no longer a liquid junction potential. That's why we do the offset. The junction potential is present when we offset the voltage after putting the pipette in the bath. Then the seal is formed, there is no junction potential, but we leave the offset potentiometer at the same setting.

As for NMDG in the bath during sealing, I've done it before without much difference compared to a more typical, sodium based external solution. So if your cells are happy in NMDG after making the seal and break-in, I would think it's ok. But it could be cell-type specific, and your cells may be different.

Good luck and keep at it!