Chapter 16

Establishment of the Dual Whole Cell Recording Patch

Clamp Configuration for the Measurement of Gap
Junction Conductance
Richard D. Veenstra

Abstract
The development of the patch clamp technique has enabled investigators to directly measure gap junction
conductance between isolated pairs of small cells with resolution to the single channel level. The dual patch
clamp recording technique requires specialized equipment and the acquired skill to reliably establish
gigaohm seals and the whole cell recording configuration with high efficiency. This chapter describes the
equipment needed and methods required to achieve accurate measurement of macroscopic and single gap
junction channel conductances. Inherent limitations with the dual whole cell recording technique and
methods to correct for series access resistance errors are defined as well as basic procedures to determine
the essential electrical parameters necessary to evaluate the accuracy of gap junction conductance measure-
ments using this approach.

Key words Gap junction conductance, Transjunctional voltage, Patch clamp, Dual whole cell
configuration, Series resistance, Membrane potential, Channel conductance, Perforated patch

1 Introduction

The earliest quantitative estimates of the coupling resistance (Rc)

between electrically coupled cells were performed by measuring
the voltage response in a cell adjacent to a current injected cell or
æV ö R ´ (V m1 - V m 2 )
Rc = ç m1 - 1 ÷ ´ R2 = 2 , where Vm1 and Vm2 are the
è Vm 2 ø R2
membrane voltage responses in the injected cell and adjacent cell,
respectively, and R2 is the input resistance of the adjacent cell [1].
This arrangement requires separate voltage and current micro-­
electrodes impaled into each cell, four in total, and this technique
is only applicable to large cells. This same two-electrode voltage
clamp configuration was used to perform the first direct measure-
ment of gap junctional conductance (gj) by independently voltage
clamping two coupled cells and dividing the current change in the

Mathieu Vinken and Scott R. Johnstone (eds.), Gap Junction Protocols, Methods in Molecular Biology, vol. 1437,
DOI 10.1007/978-1-4939-3664-9_16, © Springer Science+Business Media New York 2016

213
214 Richard D. Veenstra

non-­injected cell by the membrane voltage difference between the

two cells [2]. This two-electrode two-cell voltage clamp technique
led to the first description of the kinetic and steady-state gating of
gj by transjunctional voltage (Vj) gradients [3, 4]. The develop-
ment of the whole cell patch clamp technique, where one patch
electrode serves as both the voltage and current electrode for each
cell, permitted the application of the two-cell voltage clamp tech-
nique to pairs of small cells (i.e., <20 μm in diameter) and led to
the first measurements of single gap junction channel conductances
(γj) within the next 5 years [5–7]. The cloning of the first two con-
nexins occurred shortly thereafter and the biophysical investigation
of connexin-­specific gap junctions rapidly evolved with the devel-
opment of exogenous expression systems of newly cloned connex-
ins using communication-deficient cells or connexin38 (Cx38)
antisense injected Xenopus oocytes [8–13]. This chapter focuses on
the dual whole cell patch clamp methods used to measure gj and γj
from paired cells with high input resistances with single channel
current resolution.

2 Materials

2.1 Solutions 1. Extracellular (bath) solution: Cultured cells must be thor-

oughly rinsed with a protein-free physiological saline solution
for effective gigaohm (GΩ) seal formation. Typically, this is
accomplished by rinsing the dish four to six times with a physi-
ological balanced salt solution (BSS) to remove the serum-
containing culture solution, letting the culture dish incubate at
room temperature for another 5–10 min to solubilize any
remaining serum proteins from the cells and culture dish sur-
face, rinsing the dish one to two times to remove any loose
cells, and transferring the culture dish to the microscope stage
for patch clamp procedures. The BSS composition is 140 mM
NaCl, 1.3 mM KCl, 4.0 mM CsCl, 2.0 mM tetraethyl ammo-
nium chloride (TEACl), 1.0 mM NaH2PO4, 1.8 mM CaCl2,
0.8 mM MgSO4, 5.5 mM dextrose, and 10 mM
4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES),
titrated to pH 7.4 with 1 N NaOH. The low KCl and addition
of CsCl and TEACl reduces any plasmalemmal background K+
currents to near zero and helps to maintain a high cellular
input resistance (Rin) upon formation of the whole cell patch
configuration, thus improving the gap junctional current sig-
nal-to-noise ratio. The BSS composition varies among gap
junction electrophysiologists but typically consists of 140 mM
monovalent salt (i.e., NaCl or CsCl) 1–3 mM divalent cations
(i.e., Ca2+ or Mg2+), 1 g/L (5.5 mM) d-glucose, 5–10 mM
HEPES, titrated to pH 7.2–7.4 with 1 N NaOH or CsOH.
 Dual Whole Cell Patch Clamping Techniques 215

2. Intracellular patch pipette solution (IPS): 140 mM KCl,

1.0 mM MgCl2, 5.0 mM 1,2-bis(2-aminophenoxy)ethane-
N,N,N′,N′-tetraacetic acid tetrapotassium salt (K4BAPTA),
3.0 mM CaCl2 (i.e., MAXCHELATOR [14] estimated free
Ca2+ about 200 nmol/L), and 25 mM HEPES, titrated to
pH 7.4 with 1 N KOH.

2.2 Dual Whole Cell 1. Patch clamp amplifiers: Controlling the Vj gradient between
Patch Clamp Setup two coupled cells requires two low-noise whole cell patch
clamp amplifiers. Examples include the Axopatch 200B or
computer-­ controlled Multipatch 700B amplifiers from
Molecular Devices, EPC800 USB or computer-controlled
EPC 10/2,3,4 USB amplifiers from Heka Elektronik, or
PC-505B from Warner Instruments. Automated patch clamp
systems cannot perform the dual whole cell patch clamp volt-
age clamp method. Each patch clamp amplifier consists of an
electronic rack mountable main amplifier and head stage ampli-
fier connected by a flexible shielded cable of 3–6 ft (i.e., 1–2 m)
in length and a patch electrode holder designed to fit onto the
front of the head stage amplifier. The patch electrode holder
will be equipped with a thin silver wire (i.e., 30 AWG, 0.010
in., 0.25 mm diameter), such as AGW1010 from World
Precision Instruments (WPI) with the outer half of the wire
electroplated with chloride (see Note 1).
2. XYZ micromanipulators: Stable positioning of each patch elec-
trode on a cell requires a three-axis (XYZ) coarse manual and
remote fine control micromanipulator sturdy enough to sup-
port the head stage/pipette holder of the selected patch clamp
amplifier with submicron resolution and negligible drift. There
are multiple choices in the type of fine/coarse control micro-
manipulators including hydraulic (i.e., preferably water, less
viscous drag), piezo-electric, stepper motor, and motorized
linear actuator models. Examples include Narishige MHW-3
or MHW-103 water hydraulic, ThorLabs Burleigh
5200/5300/5400 series piezo-electric, Sutter Instruments
MPC-200/MPC-285/MPC-225/MPC-265 series or
Narishige EMM-3NV stepper motor, Zaber M-LSM linear
actuator or Newport Corp. 462XYZ linear stage plus a variety
of linear actuator models (see Note 2). Each two-­cell patch
clamp setup will require one right-handed and one left-handed
version. The angle of the patch electrode should be between
40 and 60° when mounted in the micromanipulator.
3. Inverted microscope: For viewing cells in culture, an inverted
light microscope with 10× magnification oculars and 10×, 20×,
and 40× long working distance objective lenses is optimal.
Patching a cell, forming the GΩ seal, is usually performed
under 400× or even 600× magnification. To visualize the
216 Richard D. Veenstra

membrane, either phase contrast or Nomarski differential

interference contrast (DIC) imaging is necessary (see Note 3).
All major scientific microscope manufacturers offer inverted
light microscopes suitable for patch clamping. Examples
include the Olympus IX-73 or motorized IX-83, Nikon Eclipse
Ti, Leica DMi8, or Zeiss Axio Vert.A1 models.
4. Anti-vibration table: Stable patch clamp recordings require a
mechanical vibration-free environment. This requires that the
inverted light microscope and XYZ micromanipulators be
positioned on an anti-vibration table. Examples include the
Technical Manufacturing Corporation (TMC) Micro-G
63-500 series, Newport Integrity series, Kinetic Systems MK26
or MK52 series, ThorLabs Active-Air series, MinusK
Microscope vibration isolation, Electron Microscopy Solutions
(EMS) AMH series anti-­vibration table workstations.
5. Faraday cage: In addition to vibration isolation, isolation of
external electrical noise, such as 50–60 Hz AC power sources,
is critical since something as minor as room lights may transmit
alternating current signals in excess of the biological signal you
are attempting to record onto the patch clamp recording
chamber. Faraday cages are typically constructed of conductive
stainless steel or copper mesh screens or MuMetal (Magnetic
Shield Corporation) and cover the entire anti-vibration table
top or minimally the microscope stage and micromanipulators.
Some anti-vibration table manufacturers, including TMC,
Newport, Kinetic Systems, MinusK, ThorLabs, sell Faraday
cages for their anti-vibration tables as accessories or one may
custom-build one using metal screen wire, copper mesh, or
MuMetal to fit a specific patch clamp system.
6. Patch pipette puller: One cannot attain a live whole cell patch
clamp recording without preparing fresh patch electrodes on the
day of use. The most popular model of patch pipette puller in use
is the Sutter Instruments P-97 Flaming/Brown Micropipette
puller named after authors of method for its design [15]. Sutter
Instruments also markets P-1000 Next Generation and P-2000
laser-based micropipette pullers. Key factors in the development
of a suitable patch pipette for whole cell electrophysiological
recordings include the choice of capillary glass and program
design to pull a patch electrode with an appropriate geometry for
whole cell recording (i.e., low access resistance). Typically, a
1.5 mm outside diameter (OD) glass capillary with inside diam-
eters (ID) of 0.75–1.1 mm, 10 cm in length, without a filling
filament are used. References for the fabrication of patch pipettes
include Rae and Levis [16] and the Sutter Instrument P-97
pipette cookbook [17]. We use a 1.5/0.84 OD/ID borosilicate
glass from WPI without fire-polishing or coating with Sylgard®
182 after preparing the patch pipette.
 Dual Whole Cell Patch Clamping Techniques 217

7. Data Acquisition hardware/software: Analog signals of the

whole cell voltage clamp currents are recorded in real-time via
analog-to-digital (A/D) sampling and computer storage for
off-­line analysis. Molecular Devices markets its own data
Digidata 1550 digitizer (A/D converter) and pClamp10 and
Clampfit10 software for data acquisition and analyses using
the Axopatch 200B or 700B patch clamp amplifiers. Heka
Elektronik similarly markets InstruTECH LIH 8 + 8 and ITC-
18 data acquisition interfaces and Patchmaster and Fitmaster
software for data acquisition and analyses. We have Molecular
Devices Digidata 1440 and 1320 data acquisition interfaces
and pClamp8.2 and 10.1 versions of the software which rec-
ognize most commercially manufactured patch clamp ampli-
fiers for the purpose of telegraphing the amplifier gain and
filter settings.
8. Recording chamber and bath reference: The easiest recording
chamber to use is a 35 mm diameter culture dish filled with
3–4 mL of BSS. Glass bottom dishes or coverslips (#1 thick-
ness, 0.13–0.16 mm thick, 12 or 25 mm diameter) coated with
poly-­l-­lysine or an extracellular matrix protein, such as fibro-
nectin, laminin, will be required for quantitative fluorescence
measurements. The outer rim of the cell culture dish lid glued
to the insert of the microscope stage is cheapest way to affix a
culture dish to the microscope stage to prevent the dish from
moving during the experiment. Since an Ag/AgCl wire is used
to make electrical contact of the patch electrode with the patch
clamp amplifier, an Ag/AgCl junction should be used to make
electrical contact with the bath chamber filled with BSS. One
cannot place an Ag/AgCl wire directly in the bath solution,
since Ag2+ ions will leach into solution and kill the cells over a
period of minutes. Thus, we use an agar bridge fashioned from
a glass capillary tube, such as a hematocrit tube, filled with BSS
and 1 % agar and stored in BSS at 4 °C until the day of use.
One end of the bridge is placed in the cell culture dish or
recording chamber and the other end is placed into an external
reservoir filled with 1 mL of IPS containing an Ag/AgCl pellet
as a reference electrode. We use 2 mm × 4 mm diameter Ag/
AgCl2 pellets (WPI) with a 5 cm wire centrally located as our
reference electrode. The patch cords supplied with the ampli-
fier head stage are connected to the reference electrode to
establish the external ground reference for the patch clamp
recordings. Thus, the bath solution is connected to the exter-
nal (i.e., ground) reference electrode and the amplifier head
stage by IPS-Ag/AgCl wire junctions to minimize the voltage
offset between the bath and the patch electrodes.
9. Oscilloscope: A digital storage oscilloscope is somewhat
optional given that most acquisition software packages have an
218 Richard D. Veenstra

oscilloscope feature, but an independent oscilloscope is useful

when first measuring the patch pipette resistance in the bath,
forming the GΩ seal, and rupturing the membrane patch to
achieve the whole cell configuration. Once the patch clamp
recording is initiated, the experimenter will primarily be
involved with computerized data acquisition of the gap junc-
tion voltage clamp protocols being applied to the cell pair.
10. Voltage stimulator: Again, this is an optional feature with the
existence of today’s computerized data acquisition and soft-
ware packages, since the data acquisition interfaces also oper-
ate as a digital-to-analog (D/A) converter capable of converting
computer generated voltage clamp protocols into analog sig-
nals to be applied to the whole cell recording via the patch
clamp amplifier. We still rely on a custom-built voltage stimula-
tor with two independent voltage and transistor-transistor
logic (TTL) trigger outputs for our dual whole cell recordings
of gap junctional currents and conductance measurements.
Commercially manufactured multichannel voltage stimulators
are available, such as Panlab LE12000 series, WPI Pulsemaster
or A310 Accupulser signal waveform generator.
11. Fluorescence illumination system: In order to perform fluores-
cent dye transfer experiments or view fluorescent reporters for
transiently transfected cells, one will need an epifluorescence
illumination system. There are numerous fluorescence illumi-
nation systems available and the choice is usually determined
by the microscope manufacturer or compatibility with the
patch clamp electrophysiology software. Popular epifluores-
cence excitation (ex) and emission (em) filter sets (in nm) are
fluorescein isothiocyanate (green) 480/40ex–535/50em, tet-
ramethylrhodamine (red) 540/25ex–605/55em, enhanced
green fluorescent protein (green) 470/40ex–525/50em, and,
for dye transfer, Lucifer Yellow 425/40ex–540/50em.
12. Bessel filter: Unfiltered whole cell currents contain high fre-
quency components that will obscure any data for analytical
purposes. Thus, patch clamp current recordings are low-pass
filtered and 4-pole or 8-pole Bessel filters are usually used for
this purpose. Patch clamp amplifiers usually have a built-in fil-
ter with four or more settings. The Axopatch 200B has a 4-pole
Bessel filter with settings of 1, 2, 5, 10 or 100 kHz. Whole cell
recordings will typically be recorded at 1 or 2 kHz. This
requires a digital sample rate of 200 μs or faster to prevent
aliasing (i.e., distortion) of the original analog signal, resulting
in artifacts of the original current signal [18]. We digitally sam-
ple at 10 kHz for a 1 kHz low-pass 4-pole Bessel filtered signal.
For longer current recordings, we low-pass filter at 100, 200,
or 500 Hz using a Warner Instruments LPF202A 4-pole
Bessel filter and digital sample rates of 1, 2 or 4 kHz respectively.
 Dual Whole Cell Patch Clamping Techniques 219

To determine the response time of our whole cell recording

apparatus, we measured the rise times in response to an instan-
taneous voltage step using a model whole cell circuit at filter
settings of 100, 200, 500, 1000, 2000, and 5000 Hz and
selected the pClamp sample interval that was twice as fast as
the half-full amplitude rise time of the low-pass filtered signal.

2.3 Cultured Cells 1. Communication-deficient cell lines for exogenous connexin

expression: The two commonly used cell lines for the patch
clamp study of wild-type (WT) and disease mutation connexin-
specific gap junctions are the mouse neuro2A neuroblastoma
cell line (N2a cells) and HeLa cells (see Note 4) [12, 13].
HeLa cells tend to exhibit more endogenous coupling from
Cx45 expression than N2a cells and also form larger gap junc-
tional plaques suitable for immunocytochemical localization of
WT and mutant connexins, which translates into higher gj val-
ues that may limit the accuracy of the gj measurements.
Spherical cell geometry improves the speed and efficiency of
the whole cell patch clamp and N2a cells typically exhibit a
spherical cell morphology, less any neurite outgrowths, whereas
HeLa cells possess a flatter, squamous cell appearance. Both
cell types are easily amenable to whole cell patch clamp proce-
dures. Both of these cell lines are typically grown in culture
media consisting of minimum essential medium, 10 % fetal
bovine serum (FBS) (see Note 5), 1× nonessential amino acids,
2 mM l-glutamine, 100 U/mL penicillin, and 100 μg/mL
streptomycin, sterile-filtered upon preparation, stored at 4 °C
and kept sterile using aseptic cell culturing techniques. This
cell culturing method works well for the parental and tran-
siently transfected N2a and HeLa cells. Each T25 (i.e.,
12.5 cm2) flask is typically passaged once per week and patch
clamp dishes, such as 35 mm diameter culture dishes, prepared
by adding 1–2 × 105 cells to each culture dish for use within the
next 48 h. Culture media (i.e., 10 mL) is replaced as needed
on a daily basis, typically once per week, more frequently for
faster growing cells.
2. Primary cell cultures: To study the function of endogenously
expressed connexins requires preparation of native (i.e., pri-
mary) cell cultures from live animal tissue, such as heart, vascu-
lar, neuroendocrine, exocrine, neuronal, and liver tissue. The
dissociation of live tissues into its viable cellular components
varies depending on the tissue type and amount of connective
tissue requiring enzymatic digestion, such as collagenase, but
the requirements for whole cell patch clamp purposes are
essentially the same, namely a clean cell membrane preparation
free of connective tissue and glycocalyx that will interfere in
GΩ seal formation [19, 20]. Primary cells from embryonic or
220 Richard D. Veenstra

neonatal tissues are generally easier to dissociate and culture in

vitro since the tissues typically have less connective tissue to
digest and the isolated cells are more calcium-tolerant. For
neonatal mammalian cardiomyocyte cultures, we use M199
culture media supplemented with 10 % FBS and 100 U and
μg/mL penicillin and streptomycin [21].

3 Methods

1. Select an isolated cell pair: Under 100× or 200× magnification,

move the microscope stage containing the 35 mm cell culture
dish affixed to the microscope stage until a suitable cell pair is
identified and centered in the field of view.
2. Fill patch electrodes with IPS: Patch electrode glass does not
contain a capillary monofilament that helps to fill the tapered
tip of the patch electrode, so one will want to fill the tip of
patch electrode with a small volume of IPS while avoiding
trapping air bubbles and then backfill the barrel (i.e., shaft) of
the patch pipette to approximately halfway, being sure to not
backfill the patch electrode beyond the chloride-coated por-
tion of the Ag/AgCl wire of the patch pipette holder. If air
bubbles get trapped in the tip of the electrode during filling,
firmly hold the patch pipette tip down between the thumb and
finger of one hand and flick the barrel of the pipette with a
finger of the free hand to dislodge the bubbles until they rise
into the barrel of the pipette and disappear. This is best accom-
plished before backfilling the barrel of the pipette with IPS to
the midpoint. We use nonmetallic 28G (0.25/0.35 mm ID/
OD) MicroFil needles from WPI for filling our patch elec-
trodes since metal ions may leach from metallic syringe needles
into the IPS during use. Insert the backfilled patch electrode
onto the pipette holder, tighten and swing into position over
the culture dish. Apply positive pressure, approximately 4 in.
or 10 cm, using a manometer filled with colored water to mea-
sure the height of the air pressure.
3. Measure the patch electrode resistance (Rel): Carefully center
each patch electrode in the field of view just above the bath
surface and to each side of center to prevent the tips of elec-
trodes from colliding and breaking off the electrode tips.
Lower one electrode at a time into the bath while applying a
low voltage pulse to each electrode until electrical contact is
achieved. We use a 10 ms, +200 μV voltage step from 0 mV to
measure Rel. The ideal value of Rel is 4–5 MΩ, which corre-
sponds to 40–50 pA of current. Larger patch electrodes (i.e.,
2–3 MΩ and 60–100 pA) are more difficult to form stable GΩ
seals and smaller patch electrodes (i.e., >6 MΩ and <30 pA)
 Dual Whole Cell Patch Clamping Techniques 221

tend to have higher access resistances after patch break and

may be more difficult to rupture the membrane patch to
achieve the whole cell recording configuration. Null any patch
electrode offset using the pipette offset (i.e., potentiometer)
knob on the front panel of the amplifier prior to GΩ seal
formation.
4. Form the GΩ seal: Position the patch electrodes above and to
the right and left of the selected cell pair using the coarse and
then fine control XYZ micromanipulators and switch to 400×
magnification. While focused on the top of the cell pair, move
each patch electrode over the respective cell for each patch
electrode to record from, preferably just to the right and left of
center, respectively. Thus, the patch electrode should first
come into close contact with the cell by vertically lowering the
patch electrode towards the upper surface of the cell, prefera-
bly at a tangent to the spherical surface of the cell. The open-
ing at the tip of the patch electrode should actually not touch
the surface of the cell or anything else until the positive pres-
sure is released and negative pressure is applied in a single con-
tinuous motion to suck the cell membrane onto the end of the
patch electrode, forming the GΩ seal. With any luck, one will
be able to visualize the slight cupping of the cell membrane
resulting from the positive pressure on the patch electrode and
resulting IPS stream prior to reversing the pressure on the
patch electrode to form the GΩ seal. Use as little negative pres-
sure as necessary to form the GΩ seal, typically 2–5 cm H2O,
with a clean electrode and cell membrane. Excess pressure may
draw too much of the cell membrane into the tip of the patch
pipette, prematurely rupture the membrane patch covering the
opening of the electrode or blow the GΩ seal. Gently release
the negative pressure once the GΩ seal is formed. Repeat these
steps with the second electrode. While applying the 200 μV
pulse, the current trace will flatten out to a straight line. To
measure the GΩ seal, a 5, 10 or 20 mV will have to be applied
to observe the seal current. A value of 1 pA/mV equals 1 GΩ
of resistance.
5. Compensate the patch electrode capacitance: Compensate the
capacitive current transient during the mV voltage pulse used
to measure the seal resistance by setting the Bessel filter to
>10 kHz or wide band (i.e., bypass) and using the fast and
slow capacitance compensation circuits on the front panel of
the patch clamp amplifiers to minimize the capacitive transient
arising during the onset of the voltage pulse. The capacitive
transient will vary slightly between patch electrodes owing to
the thickness of the pipette tip submerged in the bath solution.
Typically, only minor adjustments of the fast capacitance com-
pensation circuit will be required between patch electrodes
provided the geometry does not vary significantly.
222 Richard D. Veenstra

6. Rupture the membrane patch: The whole cell recording con-

figuration is usually achieved by applying negative pressure to
the patch pipette until the cell membrane patch covering the
opening of the electrode ruptures, establishing access to the
cell interior and permitting voltage clamp control of the cell
membrane potential (Vm). An alternative to the negative pres-
sure (i.e., suction) ruptured patch approach to achieve the
whole cell configuration is the perforated patch technique
wherein a perforating agent is added to the patch electrode to
permeabilize the ­membrane patch to ions, thus establishing
electrical contact, but not small molecules like fluorescent dyes
and second messengers, such as cyclic adenosine monophos-
phate (see Note 6).
7. Measure the whole cell electrode resistance (Rel): Determining
the value of Rel is critical if one wants to apply corrective mea-
sures to the measurement of gj. Rel will increase as a result of
rupturing the cell membrane patch and is calculated by mea-
suring the time constant of the whole cell capacitive current
decay (τc) and cellular input capacitance (Cin) using the equa-
tion Rel = τc/Cin [22]. Calculation of the electrode series resis-
tance is an automatic feature of some data acquisition programs,
like pClamp, if one is using their built-in D/A converter to
generate voltage clamp protocols. We routinely use a 5 mV
step from a holding potential (Vh) of −40 mV to −35 mV for
10 ms and signal average 10 capacitive current (Ic) transient
signals to fit the transient with an exponential decaying func-
-t / t
tion I c = I 0 × e c + C , where I0 is the peak amplitude of the
capacitive current transient, τc is the decay constant and C is
the steady-state whole cell current value at the new voltage
(−35 mV). Since Cin = ΔQ/ΔV, integrating the area under the
capacitive transient curve (I = ΔQ/Δt, so I × Δt = ΔQ) provides
a measure of the amount of charge required to charge the cell
capacitance to the new Vm (ΔQ in pA/ms) and dividing that
value by 5 mV yields the value of Cin (Fig. 1). Once τc and Cin
have been calculated for each cell, then Rel is calculated for
each electrode. The 5 mV step in voltage must be simultane-
ous applied to both cells, otherwise a transjunctional voltage
(Vj) gradient will result and an unwanted junctional current
(Ij) component will be added to the whole cell current signal
(i.e., C will change in opposite directions in both cells).
8. Assessment of the whole cell input resistance (Rin) and gap
junction conductance (gj): Knowing the value of Rin is not nec-
essary to measure gj, but is helpful in determining the range of
Vm to use during the Vj voltage clamp protocols to be applied
during the experiment. A high Rin value keeps the nonjunc-
tional membrane currents (Im) to a minimum and improves the
Ij signal-to-noise ratio. Rin can be determined by any ΔVm step
 Dual Whole Cell Patch Clamping Techniques 223

Fig. 1 Measurement of whole cell patch electrode (access) resistance (Rel). After
rupturing the membrane patch of both GΩ-sealed electrodes, a +5 mV voltage
command (Vc) step is applied simultaneously to both cells of a coupled cell pair
from a common holding potential (Vh) of −40 mV and the whole cell capacitive
current transients (i.e., I1 and I2) are recorded using the wide-band (i.e., 100 kHz
or filter bypass) setting on the 4-pole Bessel filter. Fitting the decay phase with
an exponential function yields the decay time constant, τc, which is equal to the
product of the whole cell capacitance (Cin) × Rel. Thus, after integrating the area
under the I1 or I2 curve to obtain the value of Cin (= Q/V = (pA·ms)/5 mV), Rel is
calculated based on the expression Rel = τc/Cin [22]. These signals were digitally
sampled at 50 kHz

applied simultaneously to both cells, since Rin = ΔVm/ΔIm, but

this only determines the value of Rin at 1 Vm. Creating an Im–
Vm current–voltage relationship over a 200 mV range of poten-
tials requires only a few seconds and is easily accomplished by
running a Vm ramp from −140 mV to +60 mV with a slope of
20 ms/mV = 4000 ms total duration (Fig. 2). Applying the
same ramp to one cell of the pair provides a rapid assessment of
the gj value and whether Vj gating is observable or not (Fig. 3).
9. Perform the modulation of the gap junction conductance
experiment: The simplest approach to measuring gj is to apply
a voltage step (ΔV) to one cell while holding the Vm of the
partner cell ­constant at a common Vh (i.e., 0 or −40 mV). A
small amplitude pulse (i.e., 10 or 20 mV) is preferable because
higher amplitude ΔV steps may induce Vj gating, causing Ij and
gj to change during the pulse. The action of a pharmacological
modulator of gj, such as chemical gating by pH, Ca2+/calmod-
ulin, phosphorylation, gap junction blockers, like carbenoxo-
lone, or gap junction agonist, such as rotigaptide [23], is easily
monitored by applying the small amplitude ΔV step periodically
(i.e., every 15, 30 or 60 s) during application and washout, if
any, of the gj modulator being investigated. Fluorescent dye
224 Richard D. Veenstra

Fig. 2 Measurement of whole cell input resistance (Rin). Simultaneous application

of a ±100 mV, 20 ms/mV voltage ramp to both cells of a coupled cell pair pro-
duces a whole cell membrane current (i.e., I1 and I2)–voltage relationship illus-
trating a linear Rin at negative potentials in this pair of HEK293 cells. From Fig. 1,
the ratio of Rel/Rin × 100 is 0.45 % for cell 1 and 0.53 % for cell 2. Thus, while
operating within the linear Rin range for these cell pairs, the series resistance
error for controlling Vm of each cell is only 0.5 % (i.e., 500 μV/100 mV). These
signals were low-pass filtered at 500 Hz and digitally sampled at 4 kHz

transfer is readily assessed by adding a known concentration of

a gap junction permeable fluorescent dye to one patch pipette,
normally the electrode to which the ΔV step is being applied,
and monitoring change in fluorescence of the recipient cell.
Thus, dye permeability may be directly correlated with gj, pro-
vided that the dye concentrations in the donor and recipient
cells are known for each time step (ΔV pulse) [24]. A Vj = 0 mV
rest interval of at least equal to the duration of the ΔV pulse
(i.e., 50 % duty cycle) should be included between each test
pulse. The baseline whole cell currents (I1, I2) should remain
constant for the duration of the experiment if stable dual whole
cell recording conditions (i.e., constant Rin) are maintained. A
nonzero Vh, centered in the high Rin range of Vm determined
by the ­simultaneous Vm ramp is preferable since any change in
Rin will be detected as an increase in I1 or I2 during the
Vj = 0 mV baseline interval. Since the bath potential is 0 mV by
 Dual Whole Cell Patch Clamping Techniques 225

Fig. 3 Rapid assessment of junctional conductance (gj). Application of same volt-

age ramp illustrated in Fig. 2, while keeping Vh constant at −40 mV in cell 2
produces a ±100 mV Vj ramp. Now the I1 and I2 signals will have opposite sign. I1
this time will contain the same nonjunctional membrane current response shown
in Fig. 2 plus an Ij component resulting from V1 ≠ V2. Since V2 is constant at
−40 mV, the difference between the baseline I2 current value (dashed line, when
V1 = V2 = −40 mV) and the I2 curve represents ΔI2, the equivalent of the Ij current
component from cell 1 being subtracted out of cell 2 in order to maintain Vm2
constant. Thus, the unilateral application of the voltage ramp allows for a rapid
assessment of gj from the initial peak (asterisk) of I2 (≈450 pA) ÷ 100 mV ≅ 4.5 nS
and the observation that Vj-dependent gating is evident in this HEK293 clone
80-1 (i.e., stable Cx43 short hairpin-­based RNA interference knockdown [33])
cell pair. Note that the additional 450 pA of whole cell current at the start of the
V1 ramp will result in a 3 mV drop across each whole cell patch electrode, result-
ing in about 6 % error in Vj that was not present when ramping both cells simul-
taneously. High (GΩ) Rin cells with gj < 10 nS to keep are required to maintain
90 % accuracy in uncorrected gj measurements using the dual whole cell patch
clamp technique

definition (i.e., external reference), holding at Vh = 0 mV will

not result in a detectable change in the I1 or I2 baseline during
the Vj = 0 mV rest intervals between ΔV pulses, only during the
ΔV step during which gj is measured, which may lead to errors
in the gj calculation.
10. Measurement of gj: Since R = V/I, the simplest calculation of gj
is gj = Ij/Vj since g = 1/R. Transjunctional voltage (Vj) is defined
as the membrane potential (Vm) difference between the two
cells and the simplest calculation of Vj is to use the difference
in the voltage clamp steps (i.e., V1 and V2) applied to both
cells. Initially, V1 should equal V2 (= Vh) and a ΔV step is then
applied to one cell of the pair, which we will define as cell 1
(i.e., the prejunctional cell). Thus V1 = Vh + ΔV1 and V2 = Vh. Vj
may be defined as V1−V2 = ΔV1 or V2−V1 = −ΔV1. The ΔV1 pulse
226 Richard D. Veenstra

is the source of Ij, which will flow out of cell 1 into cell 2 (i.e.,
the postjunctional cell), resulting a change in I2 (ΔI2) during
the ΔV1 step. The ΔV1 step and ΔI2 response will always have
opposite sign, as will I1 and I2 (i.e., the equal magnitude and
opposite polarity criterion [7]), since, by design, the cell 2
patch clamp amplifier will subtract out the Ij current flowing
into cell 2 in order to keep V2 constant. Thus, ΔI2/ΔV1 = Ij/Vj
will result in a negative gj value unless the sign of one of the
values is reversed. Hence, if ΔI2 is taken as the value of Ij, then
Vj has to equal V2−V1 = −ΔV1. Conversely, if ΔV1 is taken as the
value of Vj, then Ij has to equal −ΔI2. Realistically, Vm does not
equal the command voltage (Vc) applied to each cell, since the
whole cell current flowing across the whole cell patch electrode
resistance for each cell will result in a voltage drop across the
electrode or Vm = Vc−Im·Rel for each cell [22]. Thus,
Vm1 = V1−I1·Rel1 and Vm2 = V2−I2·Rel2. If Ij is chosen to equal −
ΔI2 since Ij is being subtracted from the postjunctional cell to
maintain V2 constant, then the accurate gj calculation is
-DI 2
gj =
éë(V1 - I 1 × Rel1 ) - (V 2 - I 2 × Rel 2 ) ùû
-DI 2 -DI 2 -DI 2 -DI 2
= = @ »
V1 - I 1 × Rel1 - V 2 + I 2 × Rel 2 V m1 - V m 2 DV m1 DV1

after correcting for series resistance errors [25]. This correction

method makes significant differences in the calculation of the
half-inactivation voltage (V0 or V½) of gj in response to increas-
ing Vj, called fast Vj gating, defined by the steady state gj−Vj
relationship as described by the Boltzmann distribution
é ù
ê g j,max - g j,min ú+ g
g j, ¥ = j,min where A=(nq/kT)=(zF/RT)
(
ê 1 + exp é A V - V ù ú
ëê ë (
j 0 û ûú) )
or when calculating the electrical distance (δ) to the Vj-dependent
site of ionic block within a channel pore
æb ö æ -zF dV j ö
( )
K m V j = ç -1 ÷ × exp ç ÷ originally derived by Woodhull
è b1 ø è RT ø
[3, 26, 27]. Steady-state gj may be normalized by dividing gj,∞
by gj,max, yielding the expression
é ù
g j, ¥ 1 - g j,min ú + g j,min , where
Gj = =ê
êë (
g j,max ê 1 + exp é A V - V ù ú g j,max
ë j( 0 û ú
û
) )
g j,min
Gj = will vary between 0 and 1. To minimize the series
g j,max
 Dual Whole Cell Patch Clamping Techniques 227

resistance errors associated with the dual whole cell recording,

measurements of gj, I1 and I2 need to be minimized (i.e., the
lower the gj, the higher the accuracy).
These procedures will lead to accurate measurements of Vj-­
dependent processes of connexins and alleviate any confusion
about the assignment of the polarity of Vj gating based on the
conventions used to define gj.
11. Measuring single gap junction channel conductance (γj): The
conductance of a single gap junction channel (γj) is determined
by measuring the current amplitude (ij) of a single gap junction
channel opening and dividing this value by Vj. Resolving single
gap junction channel openings typically occurs in poorly cou-
pled cells with gj < 1 nS and fewer than six channels, so the cor-
rection methods defined above are not essential. The I1 and I2
currents will be small, tens of pA and so will the voltage drop
across each electrode (i.e., 10 pA × 10 MΩ = 100 μV and
100 μV/40 mV × 100 = 0.25 % error). It is usually customary to
determine the i−V relationship for an ion channel and develop-
ing an ij−Vj relationship provides the most accurate measure of
γj, especially when working with rectifying gap junction chan-
nels, such as connexin hemichannels or heterotypic gap junction
channels, or using asymmetric ionic solutions to measure the
relative ionic permeability of connexin-based channels [28].

4 Notes

1. Approximately 2 cm of bare silver wire should protrude from

the open end of the patch electrode holder. This segment of
the silver electrode wire will need to chloride electroplated
(chlorided) prior to use. Clean the bare wire (i.e., light sanding
with fine grit sand or emery paper to remove any tarnish) and
immerse the distal 2 cm of wire in 3 M KCl solution. One will
need a second wire or a Ag/AgCl2 pellet to serve as the cath-
ode (i.e., negative voltage terminal) for the electroplating pro-
cess. Attach the electrode wire to be chlorided to the positive
(i.e., red) terminal of a low amperage voltage source and the
Ag/AgCl2 pellet to the negative (i.e., black) terminal of the
voltage source. A DC power supply with variable current (i.e.,
mA) and voltage (i.e., 0–12 V) voltage outputs or a 9 V tran-
sistor battery will suffice as the voltage source. The current
flow should be <1 mA, such as 0.15 mA for 2 cm of 0.25 mm
diameter wire [29]). The chlorided wire should turn light gray
in color during the electroplating process after a few minutes.
Rinse with distilled water and reassemble the patch pipette
electrode holder for use. An alternative approach is to immerse
the clean Ag wire in Clorox bleach for a few minutes.
228 Richard D. Veenstra

2. Hydraulic micromanipulators are typically more prone to drift,

especially under heavy mechanical load than piezo-electric,
stepper-­motor, or direct-drive models. Many of the piezo-
electric, stepper-motor, or direct-drive models now have fine
submicron control of movement that closely approximates the
smooth continuous motion of hydraulic micromanipulators.
3. The bright field illumination required for phase contrast or
DIC imaging of live cells is powered by 50 or 60 Hz AC,
110 V power source and will transmit electrical noise onto the
microscope stage, the patch electrodes and the recording
chamber. Typically, one will have to use a low-noise DC power
source with variable 12 V output to provide bright field illumi-
nation for contrast imaging of the cells to be patch clamped.
4. HEK293 cells have been reported to be communication-defi-
cient although there are published reports to the contrary [30–
33]. After performing dual whole cell patch clamp experiments
on HEK293 cells from three different sources over the years
and attaining high saturating levels of gj each time, we con-
clude that HEK293 cells are not communication-deficient and
low levels of Cx43-based coupling are evident even after suc-
cessful stable short hairpin-based RNA interference knock-
down of Cx43 expression and gj by 75–90 % [33].
5. There are multiple vendors and choices of FBS for cell cultur-
ing purposes and many of them will permit stable cell growth
for culturing purposes. However, we have found that some
FBS lots that are perfectly acceptable for cell culturing needs
are less than optimal for patch clamp purposes (i.e., poor GΩ
seal formation, cells too fragile or rigid). Thus, we purchase
small samples of FBS lots and screen the different FBS lots by
patch clamping N2a cells to determine the ease of GΩ seal
formation and membrane rupture to achieve the whole cell
patch electrode recording configuration. Once a suitable FBS
lot is identified, we purchase a 3–5 year supply of that FBS and
utilize that lot until the entire supply is consumed or expires
after multiple years of storage at −80 °C.
6. There are several advantages of using perforated patch tech-
niques. Achieving the whole cell patch electrode recording
configuration by the conventional ruptured patch approach
allows for dialysis of the intracellular milieu, which is advanta-
geous for the introduction of a fluorescent dye or controlling
the intracellular ionic concentrations or pH, but may also
result in the washout of essential second messengers necessary
for a physiological response to neurotransmitters or hormones,
such as acetylcholine and adrenaline. Intracellular dialysis by a
conventional whole cell patch electrode will also dilute the
transfer of any fluorescent dye from the donor to recipient cell,
 Dual Whole Cell Patch Clamping Techniques 229

which may dramatically influence the dye permeability mea-

surements, especially when gap junction permeability is low.
Thus, using a perforated patch electrode on the recipient cell
of a dye transfer/gj experiment improves the accuracy of the
molecular permeability calculation as first performed by
Valiunas and colleagues [24]. Nystatin and amphotericin B are
polyene antibiotics that form small pores in cell membranes
small enough to allow ionic flow across the membrane patch,
thus permitting electrical whole recordings, but too small for
larger molecules such as fluorescent dyes and cyclic nucleotides
to pass through the membrane patch [34, 35]. They are not
water-soluble and a concentrated stock solution (i.e., 60 mg/
mL) is prepared in dimethylsulfoxide the day of use and diluted
250 times with IPS. One may attempt to keep the concen-
trated stock solution stored at −20 °C for the week and dilute
daily as needed. The tapered tip of the patch pipette has to be
filled with normal IPS prior to backfilling the shaft of the patch
electrode with the nystatin or amphotericin B IPS solution
since the perforated patch solutions will interfere with GΩ seal
formation. If the diffusion distance is too long, the permeabi-
lization of the membrane by the perforated patch solution will
take additional time beyond the ≤5 min usually required to
form a GΩ seal. β-escin is a water-soluble compound and is
stored at room temperature prior to dissolving in deionized
water, making β-escin easier to work with than nystatin or
amphotericin B [36]. When dissolved at 30–50 μM in IPS, the
patch electrode may be filled the conventional way without
concern about interfering with GΩ seal formation [24]. A
50 mM stock of β-escin in deionized water may be stored at
−20 °C for 1 week and diluted 1000–2000 times with IPS for
daily use to yield a working concentration of 50 μM or less.

Acknowledgements

This work was supported by NIH grant HL-042220 and a

Hendricks Fund grant to R.D.V. Xian Zhang and Dakshesh Patel
proofread the manuscript.

References

1. Auerbach AA, Bennett MVL (1969) A rectify- 3. Spray DC, Harris AL, Bennett MVL (1981)
ing electrical synapse in the central nervous Equilibrium properties of a voltage-dependent
system of a vertebrate. J Gen Physiol 53: junctional conductance. J Gen Physiol 77:
211–237 77–93
2. Spray DC, Harris AL, Bennett MVL (1979) 4. Harris AL, Spray DC, Bennett MVL (1981)
Voltage dependence of junctional conductance Kinetic properties of a voltage-dependent
in early amphibian embryos. Science 204: junctional conductance. J Gen Physiol 77:
432–434 95–117
230 Richard D. Veenstra

5. Hamill OP, Marty A, Neher E et al (1981) 19. Trube G (1983) Enzymatic dispersion of heart
Improved patch-clamp techniques for high- and other tissues. In: Sakmann B, Neher E
resolution current recording from cells and (eds) Single-channel recording. Plenum Press,
cell-free membrane patches. Pflugers Arch New York, pp 69–76
391:85–100 20. Spector I (1983) A primer in cell culture for
6. Neyton J, Trautmann A (1985) Single channel pathologists. In: Sakmann B, Neher E (eds)
currents of an intercellular junction. Nature Single-channel recording. Plenum Press,
317:331–335 New York, pp 77–90
7. Veenstra RD, DeHaan RL (1986) 21. Lin X, Gemel J, Beyer EC et al (2005) A
Measurement of single channel currents from dynamic model for ventricular junctional con-
cardiac gap junctions. Science 233:972–974 ductance during the cardiac action potential.
8. Paul DL (1985) Molecular cloning of cDNA Am J Physiol Heart Circ Physiol
for rat liver gap junction protein. J Cell Biol 288:H1113–H1123
103:123–134 22. Marty A, Neher E (1983) Whole-cell record-
9. Beyer EC, Paul DL, Goodenough DA (1987) ing. In: Sakmann B, Neher E (eds) Single-
Connexin43: a protein from rat heart homolo- channel recording. Plenum Press, New York,
gous to a gap junction protein from liver. pp 107–122
J Cell Biol 105:2621–2629 23. Lin X, Zemlin C, Hennan J et al (2008)
10. Dahl G, Miller T, Paul D et al (1987) Enhancement of ventricular gap junction cou-
Expression of functional cell-cell channels pling by rotigaptide. Cardiovasc Res 79:
from cloned rat liver gap junction complemen- 416–426
tary DNA. Science 236:1290–1293 24. Valiunas V, Beyer EC, Brink PR (2002) Cardiac
11. Barrio LC, Suchyna T, Bargiello T et al (1991) gap junction channels show quantitative differ-
Gap junctions formed by connexins 26 and 32 ences in selectivity. Circ Res 91:104–111
alone and in combination are differently 25. Veenstra RD (2001) Voltage clamp limitations
affected by applied voltage. Proc Natl Acad Sci of dual whole cell recordings of gap junction
U S A 88:8410–8414 current and voltage recordings. I. Conductance
12. Hennemann H, Suchyna T, Lichtenberg-Fraté measurements. Biophys J 80:2231–2247
H et al (1992) Molecular cloning and func- 26. Woodhull A (1973) Ionic blockage of sodium
tional expression of mouse connexin40, a sec- channels in nerve. J Gen Physiol 61:
ond gap junction gene preferentially expressed 687–708
in lung. J Cell Biol 117:1299–1310 27. Musa H, Veenstra RD (2003) Voltage-
13. Veenstra RD, Wang HZ, Westphale EM et al dependent blockade of connexin40 gap junc-
(1992) Multiple connexins confer distinct reg- tions by spermine. Biophys J 84:205–219
ulatory and conductance properties of gap 28. Veenstra RD (2001) Determining ionic per-
junctions in developing heart. Circ Res 71: meabilities of gap junction channels. In:
1277–1283 Bruzzone R, Giaume C (eds) Methods in
14. Bers DM, Patton CW, Nuccitelli R (2010) A molecular biology, vol 154, Connexin meth-
practical guide to the preparation of Ca2+ buf- ods and protocols. Humana Press, New Jersey,
fers. Methods Cell Biol 99:1–26 pp 293–311
15. Flaming DG, Brown KT (1982) Micropipette 29. Warner Instruments. (2004) Chloriding Ag/
puller design: form of the heating filament and AgCl electrodes. http://www.warneronline.
effects of filament width on tip length and com/Documents/uploader/ChloridingAg
diameter. J Neurosci Methods 6:91–102 AgClelectrodes(2004.02.02).pdf
16. Rae JL, Levis RA (2004) Fabrication of patch 30. Stong BC, Chang Q, Ahmad S et al (2006) A
pipets. Curr Protoc Neurosci 26:1–32 novel mechanism for connexin26 mutation
17. Sutter Instrument Company (2015) Pipette linked deafness: cell death caused by leaky gap
cookbook 2015. P-97 and P-1000 micropi- junction hemichannels. Laryngoscope
pette pullers. Rev E. http://www.sutter.com/ 116:2205–2210
PDFs/pipette_cookbook.pdf 31. McSpadden LC, Kirkton RD, Bursac N (2009)
18. Colquhoun D, Sigworth FJ (1983) Fitting and Electrotonic loading of anisotropic cardiac
statistical analysis of single-channel records. In: monolayers by unexcitable cells depends on
Sakmann B, Neher E (eds) Single-channel record- connexin type and expression level. Am
ing. Plenum Press, New York, pp 191–263 J Physiol Cell Physiol 297:C339–C351
 Dual Whole Cell Patch Clamping Techniques 231

32. Sroka J, Czyz J, Wojewoda M et al (2008) The 34. Horn R, Marty A (1988) Muscarinic activa-
inhibitory effect of diphenyltin on gap junc- tion of ionic currents measured by a new
tional intercellular communication in HEK293 whole-cell recording method. J Gen Physiol
cells is reduced by thioredoxin reductase 1. 92:145–159
Toxicol Lett 183:45–51 35. Rae J, Cooper K, Gates P et al (1991) Low access
33. Patel D, Zhang X, Veenstra RD (2014) resistance perforated patch recordings using
Connexin hemichannel and pannexin channel amphotericin B. J Neurosci Methods 37:15–26
electrophysiology: how do they differ? FEBS 36. Fan JS, Palade P (1998) Perforated patch recording
Lett 588:1372–1378 with beta-escin. Pflugers Arch 436:1021–1023