> Imre, my explanation was extremely simplified. Thick boooks are written
> on the subject. Also see R Norman's reply. You are both right.
Lots of other good replies already, but whenever I try to explain it to
someone I find myself getting confused and have to think through it
again from the beginning. So here's another try, just to keep myself in
practice. Here's how I think about it, which is probably the same as
everybody else, but in slightly different words:
1) Recording fEPSPs extracellularly in the dendrites:
This is the simplest case to understand. Your electrode is near where
the synapses are, but outside the cell. When the synapse activates,
positive charge flows from outside to inside, which is away from your
pipette. The local potential near your pipette therefore gets a little
negative from the temporary reduction in positive charge near it. By
convention, negative potential goes "down" on an oscilloscope, so the
fEPSP recording in the dendrites goes down (negative).
2) Recording Spikes extracellularly near the soma:
This is equally easy to understand, and has exactly the same
explanation as #1 above. Except now, you are near the soma so the
largest path of charge flow from the outside to the inside is via
Na-channels underlying the spike (caused by the propagating EPSP
reaching the soma - I'll get to that in a sec), not open glutamate
receptors at the synapse, which are far away in the dendrites.
3) Recording fEPSPs extracellularly near the soma:
This is the weird one. Unlike the fEPSP recorded in the dendrites, the
one at the soma goes up instead of down. This means that there is a
temporary wave of *excess* positive charge near your electrode, instead
of a decrease, opposite to the two other explanations above. Where is
this excess positive charge coming from?
Imre Vida said in a previous post that :
"distally the transmembrane current is an outward current (passive
consisting of a capacitive and a resistive component.
This latter component is largely but not exclusively K+ current, as
a "leak" current. "
I agree with this.
Then Richard Norman said:
"Field potentials don't care what carries the current nor what causes
it. They only care about the current density in their vicinity and
its direction. So whether the transmembrane current is active (caused
by ions moving through newly opened channels) or passive (caused by
ions moving through already open channels including "leak" current)
or capacitative (caused by a changing electric field across the
membrane but no ion translocation) doesn't matter. "
I agree with this too. But I explain the upward fEPSP as follows (and
this is not mutually exclusive with anything anyone else has said):
The synapse is activated in the dendrites, and (ignoring the spike
because we're just talking about the EPSP now) this is where most
positive charge is flowing into the cell. IF you could record this
charge flow from your pipette near the SOMA, it would go downward (look
like a negative wave) because charge is flowing toward the dendrite
away from your pipette at the soma. But this isn't what you see, so
this dendritic sink isn't what makes the fEPSP at the soma. The
conductance of the dendrite is relatively low, except out where the
EPSC occurred, so I expect a relatively small contribution (but not
zero) from resistive currents as the EPSP propagates toward you down
the dendrite. Instead, where I expect a strong contribution is from the
capacitive component. As the EPSP travels down the dendrite, it
depolarizes each local region of dendrite. That is, it "charges the
capacitance' by causing positive charge to line up against the *inside*
of the membrane. This occurs exactly simultaneously with negative
charge lining up against the *outside* of the membrane (by definition
of capacitance). Where did this negative charge come from? It came from
the solution near your pipette! (Alternatively you could think of
charging the membrane as knocking positive charge off the *outside* of
the membrane - same thing). So it's this deficit of negative charge
(excess of positive charge) near your pipette, caused by the capacitive
charging of the membrane by the propagating EPSP, that makes the
somatic fEPSP go up.
Well, what d'ya all think? Do you buy it?
By the way, I had a laugh recently while trying for the first time to
use a patch clamp amplifier to record "subthreshold" somatic fEPSPs in
dentate gyrus. Instead of a popSpike, there were all these tiny little,
very fast spikelets riding on top of the fEPSP. I infer that the
spikelets were single neurons spiking right near my pipette. Not enough
to give a population response, but enough to see the individual spikes.
Pretty funky. I can just hear the single-unit folks out there thinking
"Well, duh! What did you think was causing the popSpike, moron?"
Still, it was cool to actually see it.