Extracting whole-cell capacitance for isopotential cells is more complex that a lot of people give credit. This is because, in reality, nearly all cells do not behave as if you have a single resistor and capacitor in parallel, but instead many resistors and capacitors (see typical cable / transmission line diagram). There are several papers out there that discuss how to deal with this problem. I would have a read of "Membrane Capacitance Measurements Revisited: Dependence of Capacitance Value on Measurement Method in Nonisopotential Neurons" by Golowasch et al. which is a relatively readable treatment of this moderately complex subject.

That said, the simple approach is to fit an exponential to the decaying phase the current in response to a square voltage step, and extracting the time constant (when you have not corrected for whole-cell capacitance/series resistance). You can extract the series resistance by back extrapolating the exponential to the start of your voltage step (i.e. to get the peak current) and using Ohms law. Then an approximation of the whole cell capacitance can be made by using the formulate Time_constant = Whole_cell_capacitance * Series_resistance.

Though, assuming you have a non-isopotential cell, you might as well just perform whole-cell correction (i.e. set the series resistance and whole-cell capacitance using the dials on the front of your 200B) and read the values off there.

You may also look at the following paper: Palmer et al., JNeurosci., 23:4831, 2003.

See their Figure 1, which follows the analysis proposed by William Connelly (above).

Here the capacitance, series resistance and membrane time constant are calculated from measurements of the voltage clamped capacitive current transient. An isopotential nerve terminal has a good fit with a single exponential but a bipolar neuron requires a double exponential fit. From the two time constants you can figure out the respective capacitance of the nerve terminal + axon & soma.

Boa sorte!

-Henrique

First, you need to acquire a capacitive spike minimizing interference with any membrane currents. The best way to achieve this is to apply as low stimulation pulse as possible for adequate current resolution (e.g. Ep = 5 – 10 mV from holding potential). The duration of such recording also should provide suitable resolutions of capacitive transition (5 – 20 ms depending on a cell size) and steady-state current following the termination of capacitive spike (Io) at maximal bandwidth (1 kHz or more) with compensation of the stray electrode capacitive spikes. Then the procedure for calculation is quite simple:

1)    Export your recording to any data analyzing package: Exel, SigmaPlot, Origin etc.

2)    By subtracting Io from each Ic (values corresponding capacitive transition) integrate all current differences and multiply the sum by time of integration (t). This provide you with the total charge stored on a cell capacitance Q=Sum(Ic-Io) t.

Cell capacitance Cm=Q/Em, where Em is a membrane potential, which can be defined as:

Em = Ep Rm/Rt, where Rt = Ra + Rm = Ep/Io, where

Rm, membrane resistance;

Ra, access resistance;

Taking into consideration that a characteristic time of capacitive transition tau = Ra Cm:

Access resistance Ra = tau Ep/Q

Total resistance Rt=Ep/Io

Membrane resistance Rm = Rt – Ra

Cell capacitance Cm = Q Rt/Ep/Rm

3)    In order to accomplish these calculations you will need to define tau - a steepness of exponential decay characterizing measured capacitive relaxation. Using Exel or SigmaPlot you may draw the capacitive transition in log scale. In such plot, exponential decay (or its substantial portion) will be linear. Define the steepness of this linear part (it must have dimension of time) and use it for calculations.

Thank you so much!