Departments of Medicine and Physiology, Tulane University School of
Medicine, and Veterans Administration Medical Center, New Orleans,
Louisiana 70112-2699
 |
INTRODUCTION |
THE ESOPHAGUS IS LINED
BY a moist stratified squamous epithelium, which actively
transports Na+ from lumen to serosa. Active Na+
transport accounts for the main portion (85%) of the short-circuit current developed across this tissue (~10 µAmp/cm2),
whereas the majority of the remaining current is due to the transport
of Cl
from serosa to mucosa (20). Although
the contribution of Cl to the short-circuit current is
very small, cellular transport of Cl, similar to other
epithelia, is fundamental for the maintenance of the intracellular
milieu. Regulation of cellular Cl transport is
particularly important in the esophageal epithelium, a tissue regularly
exposed to wide fluctuations in luminal content including high levels
of NaCl (salt) in food and HCl (gastric acid) due to the reflux of
gastric content (14, 16). Because very little information
is available about Cl transport in esophageal epithelial
cells, the purpose of this study was to examine its transport pathways
across both the basolateral and the apical cell membranes of this
epithelium. This was done using microelectrodes to measure
intracellular Cl activity and basolateral and apical
membrane potential of esophageal cells within the intact epithelium.
Our study indicates that the a
in this tissue is
maintained above electrochemical equilibrium independently of the
presence of bicarbonate. We have also identified a
Ca2+-dependant conductive pathway for Cl
transport across the basolateral membrane that is sensitive to the
nonsteroidal anti-inflammatory aromatic Cl-channel
blocker flufenamate. A component of Cl transport in the
serosal cells is sensitive to bumetanide. Moreover, there is little
capacity for the transport of Cl across the apical
membrane, which likely accounts for the minimal contribution of
Cl to net active transport of ions across this epithelium.
| |
METHODS |
Animal and Tissue Preparation
New Zealand rabbits were killed by administration of an
intravenous overdose of pentobarbital sodium (60 mg/ml). The esophagus was excised, opened longitudinally, and pinned mucosal side down in a
paraffin tray containing ice-cold oxygenated Ringer. The muscle layers
were lifted up with forceps, and the underlying mucosa were dissected
free with a scalpel. The sheet of mucosa thus obtained was cut, and a
section was mounted horizontally serosal side up (for basal cells
impalements) or mucosal side up (for luminal cell impalements) in a
modified Ussing chamber with an aperture of 1.13 cm2. The
chamber allows independent and continuous perfusion of the apical and
the serosal side of the tissue (1). The fluid for the
perfusion of the tissue is delivered by gravity. The perfusion solutions can be switched quickly and with minimal dead space by means
of a combination of rotary and slider valves (Rainin, Emeryville, CA)
that allow one of six experimental solutions to flow to each side of
the chamber. The volume of fluid in the upper and in the lower sides of
the chamber was ~0.25 ml. The rate of flow of the solution to the
upper chamber was 2 ml/min, which yielded a turnover rate (i.e., time
for a solution to be washed out completely from the chamber) of ~8 s.
Perfusion fluid to the lower part of the chamber was delivered at 2 ml/min; however, only part of this fluid (~1 ml/min) entered the
chamber, the rest leaving through a drain that was kept at a fixed
level. The turnover rate in the lower side of the chamber was ~15 s.
All solutions were placed at the same height from the chamber. The
solutions were prewarmed and delivered to the chamber at 37°C.
Electrodes
Transepithelial potential (VTE) was
measured as the voltage difference between a free-flowing KCl (tip <10
µm) electrode placed in the bath fluid of the luminal side and a
similar electrode placed in the bath fluid of the serosal side (luminal
electrode potential 
serosal electrode potential). Both
electrodes were fitted with an Ag-AgCl wire, and the leads were
connected to the amplifier of a voltage clamp (model VCC 600 Physiologic Instruments, San Diego, CA). The voltage clamp allowed
automatic compensation for voltage errors due to the resistance of the
fluid and was also used to deliver a direct-current (DC) pulse (I) of
5-15 µA via platinum wires located in each side of the chamber.
This allowed us to determine the transepithelial resistance
(RTE) from the voltage deflection [change in
VTE (
VTE)] as follows
|
(1)
|
The esophagus possesses relatively large basal cells (20-40
µm) that allow stable impalements with double-barreled
microelectrodes. Double-barreled microelectrodes were used in basal
cells to measure simultaneously and in the same cell the cell membrane
potential difference [basolateral membrane potential
(VmBL)] and the intracellular activity of
Cl (a). Luminal cells are
exceptionally long (100-250 µm) and slender (5-10 µm) so
that single-barreled microelectrode impalements proved more
stable. Thus, for apical membrane studies, two adjacent luminal
cells were impaled with two single-barreled microelectrodes, one a
Ling-Gerard microelectrode for cell membrane potential measurement and
the other a Cl-sensitive microelectrode for intracellular
Cl measurements.
The following is a brief description of microelectrode preparation:
double-barreled microelectrodes were prepared from two 1.2-mm OD
borosilicate glass fiber capillaries (A-M Systems, Everett, WA) that
were held together with shrinkable tubing and twisted 360° over an
open flame. The electrodes were then pulled on a vertical
microelectrode puller (David Kopf, Tujunga, CA) to a tip smaller than
0.2 µm. One barrel was exposed to hexamethyldisilazane (Sigma, St.
Louis, MO) vapor for 30 min, after which the electrodes were baked at
100oC for 2 h. The Cl exchanger (WPI,
Sarasota, FL.) was then introduced into the tip of the silanized barrel
by means of very fine glass capillaries. The silanized barrel was
further backfilled with 0.5 M KCl, and the reference barrel was filled
with 1 M KCl. Each barrel was connected to an Ag-AgCl half cell and
connected to one of the probes of a high-input impedance electrometer
(WPI). The resistance of the reference barrel ranged from 50 to 100 M
, and the tip potential was <5 mV. Single-barreled
Cl-selective microelectrodes were pulled from 1.2-mm-OD
borosilicate glass capillaries dried in the oven at 200°C for 2 h. Ten microliters of tri-n-butyl-chlorosilane were then
introduced in a closed vessel (300 ml) that contained the
microelectrodes, after which the silane fumes were vented and the
electrodes were left in the oven for an additional 30 min. The
electrodes were then filled and backfilled as described for the
ion-sensitive barrel above. Ling-Gerard microelectrodes were pulled
from 1.2-mm-OD borosilicate glass fiber capillaries and filled with 3 M
KCl. Their resistance ranged from 20 to 30 M, and the tip potentials
were <5 mV.
The slope (S) of the Cl microelectrode (or
ion-sensitive barrel) was determined from the equation
|
(2)
|
where V100mM NaCl and
V10mM NaCl denote the electrode potential in the
solutions as noted, 0.94 equals the logarithm (base 10) of the
Cl activity ratio of pure 100 mM to pure 10 mM NaCl. The
S of the electrodes averaged 55 mV/decade of change in
activity of Cl. The selectivity coefficient
kClHCO3 for Cl over
HCO
was calculated from the equation
|
(3)
|
where V100HCO3 is the voltage of the
Cl electrode in 100 mM NaHCO3 and
V100Cl is the voltage of the Cl electrode in
100 mM NaCl.
The intracellular ionic activities (a) were
calculated from the potential readings of cellular impalements of the
Cl sensitive barrel of the double-barreled microelectrode
or of the single barreled Cl sensitive microelectrode.
The total potential (ViCl) of the ion-sensitive
electrode was measured as the voltage difference between the
ion-sensitive microelectrode and the free flowing reference electrode
in the bath. The pure ionic potential was obtained by subtracting
electronically VmBL or apical membrane potential
(VmC) from the total potential of the
ion-sensitive electrode.
The intracellular ionic activity was calculated according to the
equation
|
(4)
|
where a
indicates the activity of
Cl in the bathing solution, the activity coefficient
being 0.76 at 37°C, Vcell is the cell membrane
potential difference (VmBL or
VmC), and V is the reading of
the electrode in the bathing solution. In 50 Cl
electrodes used for the study, the mean
Cl-to-HCO selectivity ratio was 10:1. Because of the low estimated intracellular activity of bicarbonate in
the cell and the good selectivity of our electrodes, the correction factor (KCl-HCO
· 
) for
the selectivity was dropped from the calculations of
a (2).
When a basal cell was impaled, the basolateral membrane potential
(VmBL) was read as the voltage difference
between the reference barrel of the double-barreled microelectrode and
the reference electrode in the serosal bath. When a luminal cell was
impaled, the VmC was read as the voltage
difference between the single-barreled microelectrode (Ling-Gerard) and
the reference electrode in the luminal bath.
The apparent ratio (23) of the apical to the basolateral
membrane resistance (Ra/Rb) was determined from
the ratio of the voltage deflections produced by the transepithelial DC
pulse across the apical and the basolateral membranes [change in
VmC (VmC) and VmBL (VmBL),
respectively] according to the equation
|
(5)
|
when a serosal cell was impaled, VmBL
produced by the transepithelial current pulse was measured directly and
VmC was calculated as
VTE VmBL.
When a luminal cell was impaled, VmC was
measured directly and VmBL was calculated as
VTE VmC . This calculation relies on the assumption that
VTE = VmBL + VmC, which was verified only in simple tissues.
To validate that in this multilayered tissue this calculation is a
reasonable approximation we did the following. We used the signal from
the bath voltage electrode at either the serosal or the luminal side of
the tissue to drive the circuit ground of the intracellular measuring
electrometer, thus allowing the intracellular potential (from the
voltage microelectrode) to be referenced to either the serosal or the
mucosal side of the tissue. We used this approach to compare during
serosal impalements a calculated VmC (from
VmBL VTE) to a
presumed VmC measured as cell potential
referenced to the luminal electrode. On the other hand, during luminal
impalements, we compared a calculated VmBL (from
VmC VTE) with a
presumed VmBL measured as cell potential referenced to the serosal electrode. The differences between the calculated and measured values of VmBL or of
VmC were within an acceptable range close to the
standard error of the measurement.
Readings were recorded on a three-channel strip chart recorder (Kipp & Zonen, Bohemia, NY).
The Nernst equilibrium potential for Cl was calculated
from the equation
|
(6)
|
where R is the gas constant, T the absolute temperature, z is
the valence, F is the Faraday constant, a is the
intracellular Cl activity, and a
indicates the activity of Cl in the luminal or serosal bath.
Solutions
The composition of Ringer solutions is given in Table 1. We have
measured ionized Ca2+ in gluconate-containing solutions and
found that ionized Ca2+ concentration was reduced by almost
50% compared with control Ringer. Consequently, we have doubled the
concentration of Ca2+ in all solutions containing
gluconate. Chelation of Ca2+ in 0 Cl solution
was corrected in a similar manner by Boron and Boulpaep (3). Indanyloxyacetic acid 94 (R(+)-IAA-94) and
R(+)-butylindazone [R(+)-DIOA] were purchased from RBI-Sigma (St
Louis, MO), diphenylamine-2-carboxylate (DPC) was from Fluka
(Milwaukee, WI), 5-nitro-2-(3-phenylpropylamino) benzoic acid (NPPB)
was from Calbiochem (La Jolla, CA), and DIDS, SITS, and all other
chemicals were from Sigma. Inhibitors insoluble in aqueous solutions
were dissolved in a small volume of dimethyl sulfoxide and added to the
solution. The concentrations used were based on the concentrations
required to achieve maximal inhibition of the transporters as reported
previously in other preparations (7, 18). The
concentration of dimethyl sulfoxide never exceeded 0.5% of the final
solution.
Statistical Analysis
The results are presented as means ± SE. Data were
analyzed using the two-tailed paired Student's t-test
unless otherwise indicated. n Is the number of observations.
| |
RESULTS |
Basolateral Membrane Studies in Basal Cells
Measurements in control bicarbonate Ringer.
Intracellular measurements of basal cells obtained in tissues from 10 rabbits, bathed bilaterally with control HCO Ringer
solution (Table 1, solution 1)
yielded the following values: VmBL = 70 ± 4.2 mV, VTE = 12.9 ± 0.7 mV, a = 16.2 ± 2.0 mM,
RTE = 2,807 ± 147 · cm2 and Ra/Rb=
4.74 ± 0.6 (n = 16).
On the basis of the measured a, the electrochemical
equilibrium for Cl (ECl), as calculated from
Eq. 6, is 46 mV, a value significantly smaller than the
electrical potential difference of 70 mV observed across the cell
basolateral membrane. Thus, in the basal cells, a is
higher than the a predicted by electrochemical
equilibrium across the basolateral membrane.
Effect of removal of serosal Cl in the presence of
CO2/HCO.
When the serosal bath solution was switched to a Cl-free
solution (solution 2, Table 1), a
decreased from 14.3 ± 1.9 to 5.7 ± 1.1 mM
(n = 12, P < 0.001), and this decrease
was accompanied by a rapid (<1 min) initial depolarization of
VmBL as shown in Fig.
1, segment ab, from 66 ± 3.0 to 61 ± 3.5 mV (n = 21, P < 0.001). This initial depolarization was then
followed by a slower (~5-8 min) repolarization of
VmBL (segment bc) until the membrane
potential reached 67 ± 3.3 mV, a value not significantly
different from control. The switch to Cl-free solution
also resulted in the rapid depolarization of VTE from 12.8 ± 0.6 to 4.9 ± 0.6 mV and then partial
repolarization to 8.9 ± 0.8 mV. RTE
increased from 2,530 ± 162 · cm2 to
2,787 ± 170 · cm2, whereas
Ra/Rb decreased from 4.32 ± 0.85 to
3.53 ± 0.73 (n = 10, P < 0.001).
Data from this and similar experiments are summarized in Fig.
2. When Cl was returned to
the serosal solution, all changes were completely reversed.
VmBL and VTE
hyperpolarized rapidly (<1 min.) and transiently (Fig. 1,
segment cd) to 74 ± 5.8 and 16.1 ± 0.9 mV,
respectively, before going back to their control values.
RTE and Ra/Rb returned to
normal, and a monotonically increased to its control
value over the course of 5-8 min.
View larger version (19K):
[in this window]
[in a new window]
|
Fig. 1.
Tracing from an experiment showing the effect of serosal bath
removal of Cl in a basal cell. In control bicarbonate
Ringer, removal of Cl from the serosal bath caused an
initial transient depolarization (ab) of the basolateral membrane
potential (VmBL) that was followed by a quick
recovery (bc) to a value not significantly different from control. The
transepithelial potential (VTE) also depolarized
and partially recovered, whereas intracellular Cl
activity (a) decreased substantially. The changes
were reversible on restoring bath Cl.
|
|
View larger version (25K):
[in this window]
[in a new window]
|
Fig. 2.
Effect of Cl removal from the serosal bath in
HCO Ringer in a basal cell, on
VmBL (A), VTE
(B), a (C), transepithelial
resistance (RTE; D), and
apical-to-basolateral membrane resistance ratio
(Ra/Rb; E). Removal of serosal bath
Cl caused an initial depolarization in
VmBL of 5.2 ± 0.8 mV, which then recovered
to a value not significantly different from control (n = 21; P < 0.001). VTE
depolarized initially by 7.9 ± 0.5 mV and recovered partially to
a value 4.3 ± 0.5 mV more positive than control.
a decreased by 8.6 ± 1.4 mM (n = 12; P < 0.001). RTE increased
by 257 ± 35 · cm2 (n = 21; P < 0.001), and Ra/Rb
decreased by 0.79 ± 0.14 (n = 10;
P < 0.001). *Significantly different from control,
paired data (P < 0.05).
|
|
Effect of SITS or DIDS on removal of serosal Cl in
CO2/HCO Ringer.
Cl/HCO exchange and other
bicarbonate-dependent transporters are inhibited by the stilbene
derivatives SITS and DIDS in a variety of tissues. To check the
contribution of such a mechanism to Cl exit on removal of
serosal Cl, we treated tissues from four different
animals with 0.5 mM SITS for 15 min before the removal of extracellular
Cl. SITS did not have an effect on the steady-state
parameters in the treated tissues. Moreover, when Cl was
removed in the presence of SITS, we observed a transient depolarization
of VmBL of 8.6 ± 0.6 mV, after which
VmBL hyperpolarized to a value not significantly
different from the control value in the presence of SITS. Also,
a decreased from 20.8 ± 5.0 to 9.7 ± 2.0 mM. The changes in VmBL observed on removal of
serosal Cl in the presence of SITS are not significantly
different (n = 4, P > 0.05) from those
observed in its absence (see above). The rate of change in
a on removal and readdition of serosal
Cl were not different in the absence and presence of SITS
(Table 2).
View this table:
[in this window]
[in a new window]
|
Table 2.
Rates of change in a of basal cells caused by
removal and readdition of bath Cl
under different conditions
|
|
In a different set of experiments, the esophageal tissue from 4 different rabbits were exposed to DIDS (0.1 mM) for 15 min before the
removal of serosal Cl. DIDS had no significant effect on
the steady-state parameters. The initial depolarization of
VmBL on removal of serosal Cl was
4.3 ± 1.9 and 3.3 ± 1.9 mV in the absence and presence of DIDS, respectively. This initial depolarization was followed by a
hyperpolarization of VmBL and a reduction
in the a by 7.4 ± 1.2 mM in the absence and
6.2 ± 3.05 mM (n = 4, P > 0.05) in the presence of DIDS. All the observed changes in cell membrane potential difference, resistance, and transepithelial parameters were
not statistically different in the absence or presence of DIDS
(n = 4, P > 0.05). The rates of change
in a on removal and readdition of serosal
Cl were not different in the absence and presence of
DIDS. (Table 2).
Measurements of intracellular Cl in the nominal
absence of bicarbonate.
The lack of inhibition by SITS or DIDS of the changes in
a and the depolarization of
VmBL associated with Cl removal
suggest that a conductive pathway for Cl exists across
the basolateral membrane of basal cells. If this is the case, then the
cellular changes induced by removal of serosal Cl are
expected to persist in the absence of
CO2/HCO. Tissues from 11 rabbits
were therefore perfused with a bicarbonate-free HEPES Ringer solution
(solution 3, Table 1), and the following values were
obtained: VmBL was 73 ± 4.3 mV,
VTE was 12.6 ± 1.3 mV, and
a was 16.4 ± 2.1 mM.
RTE was 2,357 ± 263 · cm2, and Ra/Rb was
5.47 ± 1.1 (n = 13). From these data, the
ECl, as calculated from the Nernst equation (Eq. 6), was 48 mV, which is significantly smaller than the
electrical potential difference of 73 mV observed across the
basolateral cell membrane. Thus the a remains higher
than the activity of Cl predicted by electrochemical
equilibrium across the basolateral membrane even in the absence of
bicarbonate. This indicates that the mechanism for maintenance of
Cl above electrochemical equilibrium is essentially
bicarbonate independent.
Effect of removal of serosal Cl in the absence of
bicarbonate.
When the serosal bath HEPES solution was switched to a
Cl-free HEPES solution (solution 4, Table 1)
the observed changes in cellular and transepithelial parameters were
very similar to the changes observed on removal of Cl in
the presence of HCO. On removal of bath
Cl, a decreased significantly from
16.4 ± 2.1 to 9.9 ± 1.5 mM (n = 13, P < 0.001). This was accompanied by a very rapid (<1
min) initial depolarization of VmBL (Fig. 3, segment ab) from 68 ± 2.5 to 63 ± 2.8 mV (n = 30, P < 0.001) followed by a slow (~5-8 min)
repolarization (segment bc) until the membrane potential
reached 66 ± 2.5 mV, a value not significantly different from
control. VTE depolarized rapidly from
11.5 ± 1.0 to 2.9 ± 1.1 mV and then repolarized to
5.5 ± 1.2 mV (n = 32, P < 0.001). RTE increased from 1,995 ± 144 to
2,278 ± 158 · cm2 (n = 26, P < 0.001), whereas Ra/Rb
decreased from 5.47 ± 1.1 to 3.56 ± 0.6 (n = 13, P < 0.001). Data from this and similar experiments are summarized in Fig. 4. All
changes were fully reversed when Cl was restored to the
bath. There was a rapid (<1 min) and transient hyperpolarization of
VmBL and VTE (Fig. 3,
segment cd) by 6.1 ± 0.5 and 8.9 ± 0.3 mV,
respectively, before going back to their control values (segment
de). RTE and Ra/Rb
returned to normal values, whereas a monotonically
increased to its control value over the course of 5-8 min. The
rates of change of a on removal and restoration of
serosal Cl were not different in the presence and absence
of bicarbonate (Table 2).
View larger version (24K):
[in this window]
[in a new window]
|
Fig. 3.
Tracing from an experiment showing the effect of removal of serosal
Cl in a basal cell, on a,
VmBL, and VTE in the
absence of CO2/HCO (HEPES Ringer).
Removal of Cl from the serosal bath caused an initial
transient depolarization (ab) of VmBL that was
followed by a quick hyperpolarization and recovery (bc) to a value
slightly lower but not significantly different from control.
VTE also quickly depolarized and partially
recovered, whereas a decreased substantially. The
changes were reversible on restoring bath Cl.
|
|
View larger version (26K):
[in this window]
[in a new window]
|
Fig. 4.
Effect of removal of Cl from the serosal bath in the
absence of CO2/HCOin a basal cell, on
VmBL (A), VTE
(B), a (C),
RTE (D), and
Ra/Rb (E). Removal of serosal
Cl caused an initial depolarization of
VmBL of 5.0 ± 0.8 mV, which then recovered
to a value not significantly different from control (n = 30; P < 0.001). VTE
depolarized initially by 8.6 ± 0.6 mV and recovered partially to
a value 6.3 ± 0.7 mV more positive than control
(n = 32; P < 0.001).
a decreased by 6.5 ± 0.94 mM
(n = 13; P < 0.001).
RTE increased by 283 ± 29 · cm2 (n = 26;
P < 0.001), and Ra/Rb decreased
by 1.91 ± 0.54 (n = 13; P < 0.001). *Significantly different from control, paired data
(P < 0.05).
|
|
Removal of serosal Cl in the presence of anion
transport inhibitors.
To further investigate the ion transport processes involved in the
cellular responses to removal of serosal Cl, we tested
the effect of two known inhibitors of Cl cotransport. The
experiments were run in the absence of bicarbonate to minimize the
effect of putative bicarbonate transporting mechanisms. Bumetanide
blocks Na-K-2Cl cotransport in a large variety of tissues, and
DIOA blocks K-Cl cotransport (6). In these
experiments, esophageal tissues were exposed to the inhibitors for
~15 min, after which the solution was switched to a chloride-free
solution also containing the inhibitor. Application of bumetanide (0.1 mM) or DIOA (0.1 mM), each in three different rabbits, had no effect on
steady-state VmBL or a. The
change in a (a) on removal of
serosal Cl was 4.6 ± 2.6 mM in the presence of
bumetanide, which is slightly but not significantly different from
a of 7.4 ± 1.7 mM in the absence of
bumetanide (n = 5, P > 0.05). However,
the rate of decrease in a on removal of serosal
Cl was 0.96 ± 0.3 in the presence of bumetanide,
a value significantly smaller than 2.67 ± 0.79 mM/min
in its absence. Moreover, the rate of increase in a
on restoration of serosal Cl was 0.97 ± 0.3 mM/min
in the presence of bumetanide, a value significantly smaller than
1.93 ± 0.63 mM/min in its absence (n = 5, P < 0.05; Table 2). In the presence of DIOA,
a was 9.0 ± 2.3 mM, a value not
significantly different from a of 9.1 ± 1.7 mM in the absence of DIOA (n = 5, P > 0.05). Also, the VmBL and
VTE were similar in the absence and presence
of bumetanide or DIOA (P > 0.05). The rate of change
in a on removal and addition of serosal
Cl was 3.3 ± 0.28 and 2.02 ± 0.3 mM/min,
respectively, in the absence and 2.83 ± 0.44 and 1.09 ± 0.36 mM/min, respectively, in the presence of DIOA. Although the rates
were slightly reduced in the presence of the inhibitor, the values were
not statistically different. These data indicate that although under
steady-state conditions, bumetanide or DIOA do not alter
a, bumetanide decreased the rate of change of
a on removal and addition of serosal
Cl without affecting the accompanying changes in cell
membrane potential difference. This finding is consistent with the
presence of an electroneutral transport mechanism for Cl
transport at the basolateral membrane, likely Na-K-2Cl.
Removal of serosal Cl in the presence of
Cl-channel blockers.
Because the depolarization of VmBL on removal of
serosal Cl indicates that a conductive pathway for
Cl exists at the basolateral membrane of basal cells, the
following four compounds reported to inhibit different types of
Cl channels in other preparations were investigated:
anthracene-9-carboxylate (A-9-COOH), DPC, NPPB, and R(+)-IAA 94 (7, 12, 13). Application of each of these compounds,
A-9-COOH (1 mM), DPC (0.1 mM), NPPB (0.1 mM), or R(+)-IAA94 (20 µM)
in at least three tissues from different animals had no effect on
steady-state VmBL or a in
either bicarbonate- or HEPES-containing solutions.
VmBL, VTE, a, and the rate of decrease in
a on the removal of serosal Cl were
not significantly different in the absence and presence of each of
these four inhibitors (Table 3).
Flufenamic acid, at concentrations ranging from 0.1 to 0.5 mM, is an
inhibitor of Cl conductance, particularly the
Ca+2-activated Cl channel (27,
29). Similar to A-9-COOH, DPC, NPPB, and R(+)-IAA 94, serosal
application of the Cl-channel inhibitor flufenamate (0.5 mM) did not cause significant changes in steady-state values of
VmBL or a. Moreover,
in the presence of flufenamate, removal of serosal Cl
decreased a by 5.9 ± 0.74 mM, a value not
significantly different from the decrease in a (6.8 ± 1.8 mM) in the absence of the inhibitor. However, in the presence of flufenamate, VmBL depolarized by
only 2.9 ± 0.5 mV (Fig. 5,
segment a'b') compared with 7.4 ± 1 mV in control
(Fig. 5, segment ab), and the repolarization was only
partial (Fig. 5, segment b'c'). Similarly, the initial
depolarization of VTE was only 5.8 ± 0.6 mV compared with 9.0 ± 0.7 mV in control, thus the peak
depolarizations of VmBL and
VTE caused by Cl removal in the
presence of flufenamate were significantly smaller than those in its
absence (P < 0.01, n = 14). The rate
of decrease of a was markedly reduced from
2.9 ± 0.8 to 0.9 ± 0.2 mM/min in the presence of
flufenamate (n = 5, P < 0.05). These
results are summarized in Table 3.
View larger version (15K):
[in this window]
[in a new window]
|
Fig. 5.
Tracing from an experiment showing the effect of removal of serosal
Cl in a basal cell on a,
VmBL, and VTE in the
absence and in the presence of flufenamate. Removal of Cl
from the serosal bath in the presence of flufenamate caused a marked
reduction in the initial transient depolarization of
VmBL (a'b' vs. ab) and a marked reduction in the
depolarization of VTE. It can also be noted that
the initial rate of decrease in a is reduced in the
presence of flufenamate. When Cl was given back to the
bath, the hyperpolarization of VmBL (cd)
observed under control conditions is also markedly reduced (c'd').
|
|
Effect of removal of serosal Cl in the absence
of Ca2+.
The effect of flufenamate is indicative of the presence of
Ca2+-activated Cl channels in the basolateral
membrane (5, 27). Prolonged exposure of the tissue to a
Ca2+-free solution in the presence of a Ca2+
chelator is likely to result in reduction of intracellular
Ca2+. Moreover, studies in other epithelial cells have
indicated that Cl channels can be regulated by external
and internal Ca2+ (24, 28). We, therefore,
investigated the effect of Cl removal in the absence of
Ca2+ from the bathing solutions. In seven esophageal
tissues from six different rabbits, Ca2+ removal was
achieved by perfusing the esophagus bilaterally for ~20 min with a
Ca2+-free HEPES solution to which 3 mM EDTA was added
(solution 5). Figure 6 is a
representative tracing showing the results of those experiments. The
removal of Ca2+ caused a slow depolarization of
VmBL from 57 ± 3.5 to 42 ± 1.9 mV and of VTE from 11.9 ± 0.6 to
8.6 ± 0.5 mV (n = 12, P < 0.001), whereas a, RTE, and
Ra/Rb did not change significantly. The removal
of Cl in the absence of Ca2+ (solution
6) caused an initial small depolarization of
VmBL, which was followed by a repolarization
(segment a'b'c'). In the absence of Ca2+, the
initial depolarization caused by Cl removal
(segment a'b') averaged 2.9 ± 0.8 mV, which was
significantly smaller than the depolarization of 6.1 ± 0.6 mV in
control (segment ab; n = 10, P < 0.006). Subsequent to this depolarization, in the
absence of Ca2+, VmBL repolarized to
a value of 9.3 ± 1.0 mV (n = 10, P < 0.001), more negative than the initial value
(segment b'c') whereas in the presence of Ca2+,
it repolarized to a value not significantly different from control (segment bc). The transient hyperpolarization of
VmBL (segment cd; 5.8 ± 0.9 mV) on the return of Cl to the bath was also inhibited
(2.0 ± 0.4 mV) in the absence of Ca2+
(segment c'd'). VTE depolarized
initially by 8.6 ± 0.6 mV on removal of Cl in the
absence of Ca2+, a value slightly smaller than in control
(10.1 ± 0.6). The changes in RTE and
Ra/Rb were similar in the absence and presence of
Ca2+. a decreased by 4.6 ± 1.8 mM
in control, a value not significantly different from 7.8 ± 0.7 mM
in the absence of Ca2+ (n = 3, P > 0.05). The inhibition of the depolarization on
removal of Cl both in the presence of flufenamate and in
the absence of Ca2+ strongly supports the presence of
Ca2+-sensitive Cl channels on the basolateral
membrane.
View larger version (14K):
[in this window]
[in a new window]
|
Fig. 6.
Tracing from an experiment showing the effect of removal of serosal
Cl in a basal cell on VmBL and
VTE in the presence and absence
Ca2+. Removal of Cl from the serosal bath in
the absence of Ca2+ caused a marked reduction in the
initial transient depolarization of VmBL (a'b'
vs. ab). When Cl was given back to the bath, the
hyperpolarization of VmBL (cd) observed under
control conditions is also markedly reduced (c'd').
|
|
Cyclic AMP.
To examine the effect of a Cl-channel activator on
Cl transport in the basal esophageal cell, we studied the
effect of cAMP on the changes induced by the removal of
Cl. We exposed the serosal side of the tissue to a Ringer
solution containing the permeable analog of cAMP, 8-bromo-cAMP
(104 M), for several minutes. 8-bromocyclic AMP did not
have an effect on the steady-state values of
VmBL, VTE, or
a. When Cl was removed from the
serosal bath, VmBL depolarized transiently by
4.8 ± 3.0 mV, after which it repolarized to a value not
significantly different from control, whereas a
decreased by 10.2 ± 1.5 mM, with a rate of change in
a that was not different from control (Table 2).
VTE depolarized transiently by 4.0 ± 2.3 mV and then repolarized to a value 2.4 ± 0.4 mV more positive
than control. Those values were not significantly different from the
values observed in the absence of 8-bromo-cAMP in the same experiments
(n = 4, P > 0.05).
Apical Membrane Studies in Luminal Cells
For apical membrane studies, two adjacent luminal cells were
impaled with two single-barreled microelectrodes, one a
Ling-Gerard microelectrode for cell membrane potential measurement and
the other a Cl-sensitive microelectrode for intracellular
Cl measurements. For these measurements to be valid, the
membrane potentials recorded by the two electrodes should be identical and ionic substitutions in the luminal bath should cause equal changes
in the two cell membrane potential differences. To validate the
measurements thus obtained, experiments were performed in which two
different luminal cells in the same tissue were impaled using a
Ling-Gerard microelectrode and a dummy microelectrode identical to the
ion-selective microelectrode but filled with 3 M KCl. As shown in Fig.
7, cells 1 and 2
have similar membrane potential differences, and the potential changes
produced by the removal of Na+ or Cl from the
luminal bath are the same in the two cells. This experiment was
repeated in tissues from three different rabbits, and the luminal cell
membrane potential difference was 62 ± 1.9 and 61 ± 2.1 mV in the simultaneously impaled cells, cells 1 and
2 (n = 9, P > 0.05). In the
absence of Cl from the luminal bath those values were
60 ± 3.3 and 59 ± 2.3, mV respectively
(n = 4, P > 0.05).
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 7.
Tracing from an experiment in which 2 different luminal cells in
the same tissue were impaled using a Ling-Gerard microelectrode and a
dummy microelectrode identical to the ion-selective microelectrode but
filled with 3 M KCl. Cells 1 and 2 indicate 2 adjacent cells a few millimeters apart. The 2 cells have similar
membrane potential differences (VmC), and the
potential changes produced by the removal of Na+ or
Cl from the luminal bath are the same in the 2 cells.
|
|
Measurements of intracellular Cl in the presence of
HCO Ringer.
In tissues from six rabbits perfused with control
HCO Ringer (Table 1, solution 1),
VmC was 53 ± 5.1 mV,
VTE was 11.1 ± 2.5 mV, and
a was 26.7 ± 3.3 mM.
RTE was 2,237 · cm2, and
Ra/Rb was 1.75 ± 0.37. On the basis of the
measured value of a, the calculated electrochemical
equilibrium potential in luminal cells for Cl
(ECl) was 34 mV, which is significantly smaller than the
luminal cell potential difference VmC of 53 mV
observed in these cells. Thus, in the presence of
CO2/HCO, intracellular Cl
is higher than predicted from electrochemical equilibrium.
Measurements of intracellular Cl in the nominal
absence of bicarbonate.
In tissues from five rabbits perfused with HEPES Ringer (solution
3, Table 1), VmC was 65 ± 3.6 mV,
VTE was 14.5 ± 1.5 mV, and
a was 16.3 ± 2.2 mM.
RTE was 2,413 ± 177 · cm2, and Ra/Rb was
2.27 ± 0.2 (n = 12). ECl, calculated
from Eq. 6, was 49 mV in the absence of
CO2/HCO, a value significantly
smaller than the electrical potential difference of 64 mV observed
across the apical cell membrane. These experiments indicate that
a is higher than the activity of Cl
predicted by electrochemical equilibrium across both the apical and the
basolateral membranes both in the presence or absence of
CO2/HCO.
Effect of removal of luminal and basolateral Cl in
the absence of CO2/HCO.
To investigate Cl transport at the apical membrane of
luminal cells of EE, we conducted the experiment depicted in Fig.
8. When the luminal bath solution was
switched to a Cl-free solution in the nominal absence of
CO2/HCO, VmC depolarized from 65 ± 3.0 to
57 ± 2.7 mV (segment abc), VTE hyperpolarized from 14.5 ± 1.5 to
21.8 ± 1.1 mV, and a decreased from
16.3 ± 2.2 to 13.8 ± 1.9 mM. As shown in Fig. 8, the
changes in VmC and VTE
were not transient, which is different from the changes in
VmBL and VTE when
Cl was removed from the basolateral bath.
RTE increased from 2,413 ± 177 to
3,100 ± 247 · cm2, and
Ra/Rb increased from 2.27 ± 0.2 to
2.63 ± 0.2 (n = 12, P < 0.02).
The hyperpolarization of VTE on luminal
Cl removal could be explained by a diffusion potential
for Cl across the paracellular shunt, a fact also
supported by the observed increase in RTE.
View larger version (19K):
[in this window]
[in a new window]
|
Fig. 8.
Tracing from an experiment showing the effect of removal of luminal
Cl in an apical cell on a,
VmC, and on VTE. When
luminal Cl was removed, there was a sustained
depolarization of VmC (ab), a hyperpolarization
of VTE, and a small decrease in
a. The changes were reversible when Cl
was returned to the luminal bath.
|
|
When Cl was subsequently removed from the basolateral
solution in the continuous absence of luminal Cl, there
was an initial transient hyperpolarization of
VmC from 55 ± 2.2 to 63 ± 2.4 mV
after which VmC depolarized to 51 ± 3.6 mV and VTE depolarized transiently from
23.6 ± 1.4 to 13.2 ± 1.4 mV to recover to 18.5 ± 2.2 mV. RTE increased from 3,292 ± 264 to 3,675 ± 323 · cm2. The
a in the luminal cell decreased from 15.2 ± 3.4 to 8.8 ± 2.4 mM (n = 6, P < 0.001) in response to basolateral Cl removal. Thus the
decrease in a in the apical cell is significantly
larger (6.3 ± 1.0 mM) when Cl is removed from the
basolateral side than when removed from the luminal side (2.6 ± 0.9 mM; unpaired <
TOP
ABSTRACT
INTRODUCTION
