&fERfCAN
Vol.
JOURNAL
OP
229, No. 2, August
Water
PHYsroLocw
1975.
Pri?ztd
in U.S.A.
and electrolyte
transport
by rabbit
D. W. POWELL, S. M. MORRIS, AND D. D. BOYD Department of Medicine, University of North Carolina School of Medicine,
POWELL, D. W., S. M. MORRIS, AND D. D, BOYD. Water and eZecdrolyte transport by rabbit esophagus. Am. J. Physiol. 229(Z) : 438-443. 1975.-The nature of the transmural electrical potential difference and the characteristics of water and electrolyte transport by rabbit esophagus were determined with in vivo and in vitro studies. The potential difference of the perfused esophagus in vivo was - 28 k 3 mV (lumen negative). In vitro the potential difference was - 17.9 =I= 0.6 mV, the short-circuit current 12.9 + 0.6 pA/cm2, and the resistance 1,466 =k 43 ohm+cm2. Net mucosal-to-serosal sodium transport from Ringer solution in the short-circuited esophagus in vitro accounted for 77 yO of the simultaneously measured short-circuit current and net serosal-tomucosal chloride transport for 14 %. Studies with bicarbonate-free, chloride-free, and bicarbonate-chloride-free solutions suggested that the net serosal-to-mucosal transport of these two anions accounts for the short-circuit current not due to sodium absorption. The potential difference and short-circuit current were saturating functions of bathing solution sodium concentration and were inhibited by serosal ouabain and by amiloride. Thus active mucosal-to-serosal sodium transport is the major determinant of the potential difference and short-circuit current in this epithelium. epithelial
transport;
stratified
squamous
epithelium;
rumen;
eso-
phageal mucosal barrier
IT IS GENERALLY CONSIDERED that the only function of the esophagus is to serve as a propulsive conduit from the pharynx to the stomach. Over the past decade, the motor function of this organ has been extensively investigated in both normal animal and human subjects and in various disease states. These studies, at both the organ and cellular levels, have greatly extended knowledge of the mechanisms and control of muscle action and how muscular dysfunction can cause disease or symptoms. However, several investigators have explored another property of the esophagus-the presence of a transmural electrical potential difference (PD) (1, 2, 10, 11, 15-l 7, 20, 26). The origin of this PD has not been defined. We have studied the rabbit esophagus with in vivo and in vitro techniques in order to determine the nature of the PD and the characteristics of water and eIectrolyte transport in mammalian esophagus. METHODS
All studies were performed on rabbits weighing 2-3 kg* The Ringer solution of the following per liter: 140 Na, 5.2 K, 119.8 Mg, 2.4 HPO4, 0.4 HzPOd, mosmol/kg H 20). In Na-free
male, New Zealand albino basic solution used was a composition, in millimoles Cl, 25 HC03, 1.2 Ca, 1.2 and 10 mannitol (280 solutions, and those with
Ozafal Hill,
esophagus
North Carolina
27514
varying Na concentrations, Na was replaced with equimolar choline. Sodium isethionate replaced NaCl and/or NaHC03 in HCOs-free, Cl-free, and HCO&l free solutions. In Cl-free and HCO&I free solutions, the SO4 salts of Ca and Mg were used. HCOa-containing solutions were bubbled with 95 % 02-5 % (202, and HCOa-free solutions were bubbled with 100% 02. All solution pH’s were 7.357.45. In vim experiments. The rabbits were anesthetized with intravenous pentobarbital (20 “g/kg) and maintained with small periodic doses of the anesthetic. Body temperature was kept at 38°C throughout the experiment with a heating pad. A tracheostomy was performed and the cervical esophagus cannulated with an inflow cannula at a level just below the thyroid cartilage. A 4 % agar bridge made up with Ringer solution in polyethylene tubing (PE-205) was inserted with the inflow cannula so that its tip was at the midesophageal level. Following a midline abdominal incision, the gastroesophageal junction was ligated and an outflow cannula introduced. This resulted in a IO-cm segment of esophagus for perfusion. Similar Ringer-agar bridges were placed in the retrogastric area (brought out through a flank puncture wound) and behind the cervical esophagus to serve as reference bridges. The ends of the PD and reference bridges were placed in 3-M KC1 containers with calomel reference electrodes (Fisher Scientific Company, Pittsburgh, Pa.). The electrodes were connected to a high-impedance volt meter (Keithley Instruments, Inc., Cleveland, Ohio, model 61 OB) and the PD was recorded at lo- to 15-min intervals during perfusion. The PD value for each animal was taken as the average of these readings. The PD between the esophagus and cervical reference and between the esophagus and the intra-abdominal reference did not differ by more than l-2 mV. The Ringer perfusion solution contained 5 &i/liter [ 1., 2J4C]poIyethylene glycol (New England Nuclear Corp., Boston, Mass.) and 2 g/liter carrier polyethylene glycol 4,000 (PEG) to serve as a nonabsorbable water marker. The reliability of the [14C]PEG water marker technique has been previously established (5, 28). Twenty-five milliliters of this solution were recirculated through the esophagus at 8 ml/min with a peristaltic pump (Harvard Apparatus Co., Millis, Mass.). The bubbled reservoir was maintained at 38°C with a water bath, and the solution was passed through polyvinyl coils in the water bath just prior to entering the esophagus. Perfusion pressures measured at the entrance and exit from the esophagus were 2 cmHz0 with transient spikes to 6-7 cm during After 10 min perfusion, 2.0 ml of esophageal peristalsis. perfusion solution were removed from the reservoir (designated as time 0) and thereafter at 60, 120, and 180 min for
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 10, 2019.
TRANSPORT
BY
RABBIT
439
ESOPHAGUS
measurement of [i4C]PEG immediately, and frozen for Na, K, and Cl determinations at a later time. [14C]PEG was measured on duplicate 0.250-ml samples in Bray’s solution by liquid scintillation spectrometry (Searle Analytic, Inc., Des Plains, Ill.). Q uenching was monitored by the channels ratio and external standard ratio methods and was found to be uniform so that quench correction was not necessary. Sodium and K were measured by flame photometry, and chloride was measured by coulometric tritration on duplicate samples. Net water and electrolyte transport were calculated with the usual water marker formulas, as modified for the recirculation technique. Steady-state transport was demonstrated during the last two 60-min perfusion periods, indicating that approximately 1 h was necessary as an equilibration period. At the end of the experiment, the animal was killed with pentobarbital, the esophageal segment exposed, and its length measured in situ. Transport was expressed as microliters or microequivalents per hour times centimeter length, based on mean transport rates for the last two 60-min perfusion periods. At the end of each experiment, the volume of solution remaining in the perfusion system was measured to the nearest 0.1 ml for the calculation of [i4C]PEG recovery. Experiments where p4C]PEG recovery was less than 99 % were discarded. In vitro experiments. The experimental techniques described previously for in vitro transport studies in rabbit ileum (18, 22) were modified for use in the esophagus. Under pentobarbital anesthesia, the rabbit esophagus was excised from the level of the thyroid cartilage to the gastroesophageal junction. The esophagus was opened longitudinally and pinned, mucosal surface down, in a paraffin tray containing ice-cold oxygenated Ringer. The muscle layers were grasped with hemostats, lifted up, and dissected free of the underlying mucosa with a scalpel. This resulted in a preparation consisting of the stratified squamous epithelium, the muscularis mucosa, and a small amount of underlying connective tissue as shown in Fig. 1. The stripped preparation was cut into four equal pieces, and each piece was mounted between two Lucite half-chambers with an aperture of 1.13 cm2. The mucosal and serosal bathing solutions (maintained at 37°C) were connected to calomel
FIG. 1. Rabbit esophagus. Full thickness (A) and stripped preparation (B) from which longitudinal and circular muscular layers have been removed. Stripped preparation, which consists of stratified squamous epithelium, muscularis mucosae, and strands of residual connective tissue, was used for in vitro studies. This stratified squamous epithelium consists of lo-13 cell layers. (H & E Xl 11)
60 n -0
50
X z v
40
:
30
G -I = I
20
2 IO
TIME, MIN 2. Rate of appearance of r2Na in adjacent sections of esophagus after addition of isotope to either mucosal or serosal Ringer bath. Cumulative count rates are shown in serosal bath after isotope addition to mucosal bath (m + s), and in mucosal bath after addition to serosal bath (s + m). Linear part of curve after 45 min indicates that steady-state isotope transfer occurs after this time. FIG.
and Ag-AgCl electrodes with agar bridges made up with the solutions studied. The electrodes were connected to an automatic voltage clamp (Physiology Instrument Laboratory, Yale University School of Medicine), which automatically corrects for fluid resistance between the PD sensing bridges, for measurements of the PD and I,,. Tissues were continuously short circuited, except for 5- to 10-s intervals when the open-circuit PD was read. Radioactive isotopes were added to the mucosal and/or serosal bathing solution 5-10 min after mounting the tissues, and this was designated as time 0. It was found that 45 min were required before steady-state isotope transfer of Na or Cl occurred (Fig. 2) ; therefore three 30-min flux periods were measured from 45 to 135 min. The average of the three determinations was taken as the flux values for that tissue. Simultaneous bidirectional Na fluxes were determined with 22Na and 24Na. In separate experiments, on different tissues, oppositely directed unidirectional Cl fluxes were measured on adjacent pieces of esophagus with 36Cl. Net Cl transport was the difference in these two fluxes from the same esophagus. The methods of isotope counting have been previously described (18). A negative PD refers to the orientation of the mucosa with respect to the serosa, in keeping with the convention used in previous in vivo studies of esophagus (1, 2, 10, 11, 15-17, 20, 26). All esophageal PD’s in these experiments, both in vivo and in vitro, were mucosa negative with respect to the serosa. In preliminary experiments it was found that the current-voltage relationship was linear over a f 100 mV range, indicating that the esophageal epithelium behaves as an ohmic resistor. Therefore, resistance was determined by the ratio of the open-circuit PD to the I,,. A negative sign before a net flux infers net serosa-to-mucosal transport (secretion). N refers to the number of tissues in the Na experiments and number of tissue pairs in the Cl experi nents. Generally the number of tissues also equals the number of animals. The significance of differences in
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 10, 2019.
440
POWELL,
means was determined f test (23).
by the paired
or unpaired
MORRIS,
AND BOYD
r
Student
RESULTS
Ehctrical
Parameters
The mean esophageal PD in 10 rabbits perfused with Ringer solution for 150 min in vivo was - 28 =t 3 mV. The electrical parameters of the esophagus, stripped of its muscular and connective tissue layers and mounted in vitro, are shown in Fig. 3. The PD of 32 animals peaked at -20 mV, 30 min after mounting, with a gradual decline over the 135 min duration of the experiments. The I,, gradually declined from 18 to 11 PA/cm? The resistance reached a steady-state value of 1,500 ohm*cm2 within 1 h of mounting. There was little difference in the electrical parameters in vitro at different levels of esophagus as shown in Fig. 4. 30
FIG. 3 r
30
P. 0.
I 20
r
FIG. 4
01
1
1
20
40
I 60
[No]*
1 80
1 100
1
I
120
I40
mM
FIG. 5, Relationship between I sC and bathing solution Na concentration. Esophagi of 4 rabbits were studied simultaneously in 4 solutions with varying Na concentration (choline replacement). Means & 1 SE of I,, from 45 to 135 min after mounting are shown.
Only the PD of the most proximal esophagus (pharyngeal end) showed a statistically higher value, 22.4 & 1.7 mV, as opposed to 15*9-l 7.1 & 1.5 mV for the more distal segments. Although the Isc and resistance of the most proximal segments were slightly greater, these differences were, on the average, not significant.
I? D. f
In Vitro Transport
SC
r
20
I SC
r-
Ringer solution, Sodium and chloride fluxes and the electrical parameters in Ringer solution in vitro are shown in flux (Jf:) was 0.320 the top of Table 1. Net sodium peq/h cm2 from mucosa to serosa, which represented 77 % of the simultaneously measured Isc (expressed as units of ion flux). In separate experiments, a net chloride flux of -0.070 was determined in the opposite direction (from serosa to mucosa), and this value was significantly different from zero (P < 0.001). Having a negative value, the net chloride transport could be considered as accounting for 14 % of the measured I,,. Since net Na transport appeared to account for most of between bathing solution Na conthe L, the relationship centration and the I,, was investigated as shown in Fig. 5. The I,, was a saturating function of Na concentration with an apparent Km of 43 mM and a V,., of 19.5 pA/cm2. A similar relationship was found between bathing solution Na concentration and the PD with a Km of 43 mM and a V max Of 25 mV. The effect of 1W4 M ouabain on the I,, of the in vitro rabbit esophagus is shown in Fig. 6. This cardiac glycoside effectively inhibited the I sc within 45 min. Furthermore, it appeared that this agent was active only when added to the serosal bathing solution. The effects of ouabain on Na and Cl fluxes are shown in the bottom of Table 1. Jte: was inhibited 83 % by ouabain (P < 0.001 ), and the remaining net Na transport, 0.054, did not differ significantly from essentially to the L, 0.043 peg/h cm2. Jztt was reduced zero by ouabain. Another Na transport inhibitor, amiloride, was investigated as shown in Fig. 7. Large doses of this inhibitor, at lcasl KF4 M, were llecessal-y to achicvc a sigdimt inhibition of the I,, and PD. Of note is that 10-30 min were required for maximal inhibition, and in addition, this drug appeared to be more effective in the esophagus when added l
RES I STANCE
RESISTANCE
PHARY NGEAL END
GASfR I C END
(A) in in vitro rabbit esophaFIG. 3. PD @), Iac (O), and resistance gus in Ringer solution (n = 32). FIG+ 4. PD, I,,, and resistance at 4 levels of esophagus in vitro. Esophagi of 11 rabbits were divided into 4 segments from pharyngeal (1) to gastric end (4). Only PD of pharyngeal end was significantly greater than that of other 3 sections (P < 0.001). TABLE 1. Sodium and chloride trans@rt by rabbit esophagus in vitro: Ringer solution, without and with uuabain 1P4 A-4 Solution
(n) *
J ma
J am
J net
I BC
?TXV
peq/ h- cm2 Ringer Na
(25)
Cl (28)
Ringer-ouabain Na (13) Cl (13)
PD
R
oh*
cm?2
0.550 ztu. 035 0.260 zko.017
0.230 kO.017 0.330 zfzO.018
0.320 ho.027 -0 * 070 A0 .009
0.417 Eta rn028 0.508 zto.027
-17.7 ZLO.3 -18.0 +0.8
1,646 zt82 1,388 zt46
0.380 ztO.038 0.264 +o. 025
0.326 zto.033 0.258 zto.019
0.054 &0.023 0.006 ~3.024
0.043 &O .006 0.053 =t=o .004
-2.2 zto.3 -2.3 M.2
1,698 A300 1,442 *251
Values are means =tr SE. These ing the tissues . *n = Number experiments).
values were determined 45-l of tissues (Na experiments)
35 min after mountor tissue pairs {Cl
l
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 10, 2019.
TRANSPORT
BY
RABBIT
441
ESOPHAGUS
to the serosal bathing solution. The absolute degree of inhibition of the I,, with low2 M amiloride at 140 min, as compared to the control, was 65 % when added to the mucosal solution and 84 % when added to the serosal side. This inhibition was only partly reversible by changing the bathing solutions to fresh Ringer. Ion-repzacement solutions. The flux determinations in Ringer solutions (Table 1) suggested that net serosal-to-mucosal (s to m) anion transport accounted for 14-23 % of the I,,. Therefore, additional flux experiments were performed in Cl and/or HCOa-free solutions (Table 2) to further delineate the role of anion transport in this epithelium. These experiments revealed several important points. First, Na flux determinations in paired experiments utilizing four solutions (Ringer, HCOS free, Cl free, and HCOa-Cl free) showed that JF,“t was consistently less than the Isc when either Cl and/or HCOS were present in the solution. When both anions were replaced by isethionate, Jfz was slightly, but not significantly, greater than the I,,. Second, the sum of net Na and Cl transport in the HCOa-free solutions equaled the simultaneously measured I,,. Third, when both anions were replaced by isethionate, the PD and I,, of the esophagus were significantly inhibited. Finally, total tissue resistance was significantly increased by replacement of either HCO3 and/or Cl, indicating that these anions contribute to the total electrical conductance of this tissue.
0
30
60
90
I20
TIME,
150
180
El0
MIN
FIG. 6. Effect of ouabain 1W4 M on I,, of rabbit esophagus in vitro. Esophagi were mounted in Ringer solution alone or in Ringer containing ouabain in both m and s baths. In control group, ouabain was then added to either m or s baths at 140 min. Ouabain appears to be effective only when added to s bathing solution (n = 14 for both control and ouabain m + s). 100
AMILORIDE
K
m S } 10-6 CONTROL m s ) 10-4
0'
I 20
1 40
I
1
60
80
TIME,
MIN
I
I
I
I00
120
140
7. Effect of various concentrations of amiloride on I,, of rabbit in vitro. Values have been normalized to percent of . . esophagus initial I 6C, and SE’s of means omitted for clarity of presentation. At 30 min after mounting, amiloride 10-6, lo-“, or lUt2 IM was added to either m or s bathing solution (n = 18 for control, and 3 for each amiloride concentration and bathing solution addition). FIG.
2. Effect of anions on electrolyte trans;borf by rabbit esophagus in vitro
TABLE
SoIution
(fl>*
Jms
J sm
J net
IS,
peq/h+ cm2
PD
R
7nV
ohm 1,945 +I76
e cm2
Ringer Na
(10)
0,480 &O. 047
0.195 zto. 024
0.284 &O. 032
0.364 zko ,033
-18.3 Al.1
0.447 A0 .049 0.125 AO.013
0.171 AZ0 * 022 0.180 zto.019
0.260 410.042
*o
zfzo.012
0.298 IO33 0.298 zto. 033
-19.0 zt1.8 -17.7 Al.5
2,353 zk203t 2,365 ztx42t
0.536 zto.077
0.292 zto. 050
0.245 ~0.070
0.364 dEo.033
-22.6 zt1.q
2,541 *249 t
0.525 zto.037
0.228 AZ0 * 044
*o.
0.248 zt0.033t
-13.4 zt1.3t
2,626 zt307t
HCOa free Na
(10)
CI (11)
CI
-0.054
free
Na
(10)
HCOg-Cl ft,ee Na (IO)
0.297 030
Values
are means j, SE. Na ff uxes in the four solutions were performed simulon intestine from 10 animals. CI fluxes in the HiCOB-free experiments were performed separately. Values were determined from 45 to I35 min. * n = Number of tissues (Na experiments) or tissue pairs (Cl experiments). t Significance of difference from Ringer, P < 0.001. $ Significance of difference from Ringer, P < 0.05. taneously
In Viva Transport The net transport rates of H20, Na, K, and Cl measured in vivo in 10 animals by the recirculation-water marker technique. [14C]PEG recovery in these experiments was 99.7 & 0.3 %. The rate of water absorption was 10.2 & 3.8 pi/h *cm length, but was significantly difl’erent from zero and within the limits of sensitivity of this technique.’ Meaningful data for electrolyte absorption could not be obtained because of the propagated errors inherent in this experimental technique and because the low levels of absorption strain the accuracy of this technique. The rates of measured Na, K, and Cl transport were 0.41 h 0.38, - 0.03 AI 0.05, and 1.32 & 1.02 peq/h*cm length, respectively. It shouId be noted that Na absorption in vivo must take place against an adverse electrical gradient of 28 mV.
Although a transmural PD across mammalian esophagus was recorded as early a$ 1952 by Rovelstad, Owen, and Magath (ZO), little is known regarding the nature of this phenomenon. In fact, the magnitude and electrical orientation of this PD have not been conclusively established. Beck and colleagues (1, 2, 15, 26) and other investigators (11, 16, 17) h ave measured a variable PD generally oriented with the mucosa positive with respect to serosa. However, Grantham, Code, and Schlegel (10) have demonstrated that with the proper placement of the reference electrodes (intravenous rather than skin) the transesophageal PD in man is approximately - 10 mV, lumen negative. In the present experiments with an experimental animal, where the reference electrode was placed directly on the serosal surface of both the cervical and intra-abdominal esophagus, we have recorded a lumen negative PD (- 28 mV), which suggests this is the correct orientation of mammalian esophagus. l See ref. 19 for a discussion of the sensitivity perfusion technique in experimental animals.
of the water-marker
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 10, 2019.
442 Transmural PD’s can be divided into two general catephenomena, such as gories : those due to electrokinetic diffusion potentials or streaming potentials, and those due to active electrolyte transport. When the rabbit esophagus was stripped of its muscle and connective tissue layers and bathed in vitro in the absence of transepithelial ionic or hydrostatic pressure gradients, a similar lumen negative PD (-17.9 mV) was recorded. This indicates that the PD is related in some way to the active transepithelial transport of electrolytes. Measurements of ion fluxes in the short-circuited state allow determination of the ion species whose transport accounts for the I sc, and by inference, for the PD (25). Ion transport studies in the stripped esophagus in vitro (Table 1) revealed that the net transport of sodium from mucosa to serosa accounts for over 75 % of the simultaneously measured I Sc. The conclusion that net Na transport is the primary determinant of the PD and I,, in this epithelium is strengthened by the finding that the PD and ISc were saturable functions of bathing solution Na concentration; that the PD, I,,, and net electrolyte transport were inhibited by serosally placed ouabain; and that amiloride inhibited the PD and I,, in this epithelium. The fact that ouabain appears to be active only when added to the serosal bath is taken as evidence in other epithelia, such as frog skin, toad bladder, and intestine, that Na-K-ATPase (and the Na pump) is located at the serosal aspect of the transporting cells. However, the cells responsible for active Na transport in the esophagus may be located deep within this epithelium, and there could be a significant diffusion barrier to ouabain. Therefore, interpretation of the site of action of ouabain in this epithelium must be guarded, This may also account for the atypical effects of amiloride in the esophagus. This drug, which is thought to act by inhibiting Na entry into the transporting cells, has an essentially instantaneous onset of action and is effective in low concentration (10 -“lo-” M) when added to the mucosal bathing solution of toad bladder and frog skin (3, 4, 21). In the esophagus (Fig. 7), much larger concentrations of the drug and a longer time were necessary to achieve significant inhibition. Inhibition was also more effective when the drug was added to the serosal solution. The reasons for these differences are unclear but may be due to a significant diffusion barrier at the mucosal surface. In Ringer solution, net mucosal-to-serosal (m-to-s) Na transport accounted for only 77 % of the I,,, and net Cl transport was demonstrated from s to m. The difference between JfA and ISc is statistically significant as was the Jzi, significantly different from zero. These results suggest that net Cl (or perhaps Cl and HC03) transport may account for the remainder of the I,,. The ion-replacement experiments (Table 2) lend support to this idea. In paired experiments with Ringer, HCOs-free, Cl-free, and HCOS-Cl free solutions, Jfe: was consistently less than I,, when either HCO3 and/or Cl were present in the solution. When both HCOS and Cl were replaced by isethionate, Jzz was slightly but not significantly greater than the I,,. Pn addition, when HC03 and Cl were absent, both the PD and I,, in this solution were significantly less than when either anion or both were present. Finally, in HCOa-free solution, the sum of. Jr: and .J3, equaled the ISc. These results are
POWELL,
MORRIS,
AND BOYD
3. Comparison of variousstratzjkd squamous 8pithelia bathedin Ringersolution TABLE
Epithelium
PD, mV
LC , d/cm
2 R, ohm.cm2
Direction of Net Ion Transport, Short-Circuited Na+
Frog skin Rumen Goat (21) cow (21, 22) Sheep (23, 241 Esophagus rabbit Values
Cl-
50-100
40-70
1,000~2,500
m +s
0
7.8A4.1 lZ.Oztl .o 13*9&2.4
13+5 12+2 llzt2
610&19O 960~ 84 1,156&123
m+s m -3 m +8
m +s m+s
17.9zt0.6
13&l
1,466&
m +a
s--,m
are means
43
m-388
* SE.
compatible with a transport model in which net m-to-s sodium transport accounts for 77 % of the I,, and net anion transport from s to m for the remaining 23 %. When both HC03 and Cl are present, both would be transported, but one anion could substitute for the other in its absence. Our experiments were not designed to explore the exact manner by which these transport processes for Na (and anion) create a transmural PD. Electrolyte transport in this epithelium could be truly electrogenic or could involve ion exchange across cell borders with asymmetrical permeabilities, which could lead to a diffusion potential. The ability of a stratified squamous epithelium, such as esophageal mucosa, to transport electrolytes is not unusual, in view of the demonstrated transport in other epithelia of this type. In Table 3 are summarized the electrical parameters and direction of electrolyte transport in vitro of several stratified squamous epithelia. All are relatively high-resistance epithelia, and all transport Na from m to s. They differ in the direction in which they transport anion. Electrically, the esophagus closely resembles rumen epithelium, which is not surprising in that it is generally considered that the rumen is embryologically derived from esophagus. The cells responsible for electrolyte transport in these epithelia remain to be conclusively identified. In the frog skin, which is only three to five cell layers thick, and in rumen epithelium, which contains many cell layers like the esophagus, it is thought to be those cells at the midlevel of the epithelium called the stratum granulosum (9, 12-14, 27). The possibility that anion transport is accomplished by glandular structures must be considered, although we have not observed submucosal glands in several histological sections of rabbit esophagus. Although the demonstration of net m-to-s Na transport in vitro clarifies the nature of the PD in the rabbit esophagus, these studies raise another fundamental question: what is the physiological significance of electrolyte transport by the esophageal epithelium. In vivo, where active electrolyte transport must take place against a steep electrical gradient, significant net transport of sodium was not demonstrable by conventional techniques. The low rates of transport, the small surface area of this epithelium, and the short residence time of any fluid and electrolyte in the esophagus would all limit any important role in the conservation of water and electrolytes. Electrolyte transport by the esophagus could be vestigial in nature and serve no teleological purpose. On
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 10, 2019.
TRANSPORT
BY RABBIT
ESOPHAGUS
the other hand, epithelial cells, like all cells, maintain homeostasis by transporting sodium out of the cell. Specialization of epithelial cells into polar cells, which allows sodium transport in one direction only, may prevent disruption of epithelial integrity as a result of water flow in response to the transported solute. Therefore, transepithelial electrolyte transport could be necessary if epithelial cells are to maintain a low intracellular sodium concentration and yet still maintain the architecture of this barrier be-
443 tween the lumen and the blood. necessary to clarify these issues.
Additional
This investigation was supported by National Research Grant AM1 5350 from the National Metabolism, and Digestive Diseases. D. W. Powell is the recipient of Research Award AM70454 from the National Institute lism, and Digestive Diseases. Received
for publication A
31 December
1974.
lanthanum,
and
studies
will
be
Institutes of Health Institute of Arthritis, Career Development of Arthritis, Metabo-
REFERENCES 1. BECK, I. T., AND N. A. HERNANDEZ. Transmural potential difference in patients with hiatus hernia and esophageal ulcer. Gut 10: 469-476, 1969. 2. BECK, I. T., T. F. MCELLIGOTT, M. C. PATH, AND N. A. HERNANDEZ. Transmural potential difference at the level of the upper esophageal sphincter in man. Am. J. Digest. Diseases 14 : 456462, 1969. 3. BENTLEY, Pp. J. Amiloride: a potent inhibitor of sodium transport across toad bladder. J. Physiol., London 195 : 317-330, 1968, 4. BIBER, T. U. L. Effect of changes in transepithelial transport on the uptake of sodium across the outer surface of the frog skin. J. Gesz. Physiol. 58: 131-144, 1971. 5. BRIGHT&ARE, P., AND H. J. BINDER. Stimulation of colonic secretion of water and electrolytes by hydroxy fatty acids. Gasfroenterology 64 : 81-88, 1973. 6. CHIEN, W. J., AND C. E, STEVENS, Coupled active transport of Na and Cl across forestomach epithelium. Am. J. Physiol. 223: 997-1003, 1972. 7. FERREIRA, H. IL, F. A. HARRISON, AND R. D. KEYNES+ The potential and short-circuit current across isolated rumen epithelium of sheep. J. Physiol., London 187: 631-644, 1966. 8. FERREIRA, H. G., F. A. HARRISON, R. D. KEYNES, AND L. ZURICH. Ion transport across an isolated preparation of sheep rumen epithelium. J. Physiol., London 222: 77-93, 1972. 9. GEMMEL, R. T., AND B. D. STACY. Effects of ruminal hyperosmolality on the ultrastructure of ruminal epithelium and their relevance to sodium transport. Quart. J. Exjtl. Physiol. 58 : 315323, 1973. 10. GRANTHAM, R. N., C. F. CODE, AND J. F. SCHLEGEL. Reference electrode sites in determination of potential difference across the gastroesophageal mucosal junction. Mayo GZin. Pm. 45 : 265-274, 1970.
11. HELM,
W. J., J. F. SCHLEGEL, C. F. CODE, AND W. A. J. SUMIdentification of the gastroesophageal mucosal junction by transmucosal potential in healthy subjects and patients with hiatal hernia. Gastroenterology 48 : 25-35, 1965. 12. HENRIKSON, R. C. Mechanism of sodium transport across ruminal epithelium and histochemical localization of ATPase. Exptl. Cell Res. 68: 456458, 1971. 13. HENRIKSON, R. C. Ultrastructure of ovine run&al epithelium and localization of sodium in the tissue. J. Ultrastructure Res. 30: 385401, 1970. 14. HENRIKSON, R. C., AND B. D. STACY. The barrier to diffusion across ruminal epithelium: a studv bv electron microscosv using MERSKILL.
horseradish
peroxidase,
ferritin.
J.
Ultrastructure
Res. 34: 72-82, 1971. 15. HERNANDEZ, N. A., AND
I. T. BECK, Gastroesophageal transmural measured by a new constant infusion method. scarification on this potential difference. Am. J. Digest. Diseases 14 : 206-Z 16, 1969. 16. KATZKA, I, H. M. LEMON, AND F. JACKSON. The correlation of human gastric potentials with gastric physiology. Gastroenterology 28: 717-730, 1955. 17. MECKEI,ER, K. J. H., AND F. J. INGELFINGER. Correlation of electric surface potentials, intraluminal pressures, and the nature of tissue in the gastroesophageal junction of man. Gastroenterology potential difference The effect of skin
52: 966-971, 1967. 18. POWLL, D. W., H. J. BINDER, AND P. F. CURRAN. Electrolyte secretion by guinea pig ileum in vitro. Am. J. Physiol. 223 : 53 I537, 1972. 19. POWELL, D. W., S. J. MALA~ER, AND G. R. PLOTKKN. Secretion of electrolytes and water by the guinea pig small intestine itz viuo. Am. J. Physiol. 215: 1226-1233, 1968. 20. ROVELSTAD, R. A., C. A. OWEN, *JR., AND T. B. MAGATH. Factors influencing the continuous recording of in situ pH of gastric and duodenal contents. Gastroanterology 20 : 609-624, 1952. 21. SALAKO, L. A., AND A. J. SMITH. Effects of amiloride on active sodium transport by the isolated frog skin: evidence concerning site of action. Brit. J. Pharmacol. 38 : 702-7 18, 1970. 22. SCHULTZ, S. G., AND R. ZALUSKY. Ion transport in isolated rabbit ileum. I. Short-circuit current and Na fluxes. J. Gen. Physiol. 47 : 567-584, 1964. G. W., AND W. F. COCHRAN. Statistical Methods (6th 23. SNEDECOR, ed.) Ames : Iowa State Univ. Press, 1967. C. Transport of sodium and chloride by isolated rumen 24. STEVENS, epithelium. Am. J. Physiol. 206 : 1099-l 105, 1964. Active transport of sodium as the 25. USSING, H. H., AND K. ZERAHN. source of electric current in the short-circuited isolated frog skin. Acta
Physiol.
&and.
23 : 110-127,
1951.
E, I., J. A. E. Fox, AND I. T. BECK. Transmural potential 26. VIDINS, difference (P.D.) in the body of the esophagus in patients with esophagitis, Barrett’s epithelium and carcinoma of the esophagus. Am. J. Digest. Diseases 16 : 991-999, 197 I. G. Electrical potential profile of the toad skin 27. WHTTTEMBURY, epithelium. J. Gen. Physiol. 47 : 795-808, 1964. D. L., R. J. SANDBERG, AND S. J. PHILLIPS. A compari28. WINGATE, son of stable and 14C-labeled polyethylene glycol as volume indicators in the human jejunum. Gut 13: 812-815, 1972.
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 10, 2019.
Vol.
JOURNAL
OP
229, No. 2, August
Water
PHYsroLocw
1975.
Pri?ztd
in U.S.A.
and electrolyte
transport
by rabbit
D. W. POWELL, S. M. MORRIS, AND D. D. BOYD Department of Medicine, University of North Carolina School of Medicine,
POWELL, D. W., S. M. MORRIS, AND D. D, BOYD. Water and eZecdrolyte transport by rabbit esophagus. Am. J. Physiol. 229(Z) : 438-443. 1975.-The nature of the transmural electrical potential difference and the characteristics of water and electrolyte transport by rabbit esophagus were determined with in vivo and in vitro studies. The potential difference of the perfused esophagus in vivo was - 28 k 3 mV (lumen negative). In vitro the potential difference was - 17.9 =I= 0.6 mV, the short-circuit current 12.9 + 0.6 pA/cm2, and the resistance 1,466 =k 43 ohm+cm2. Net mucosal-to-serosal sodium transport from Ringer solution in the short-circuited esophagus in vitro accounted for 77 yO of the simultaneously measured short-circuit current and net serosal-tomucosal chloride transport for 14 %. Studies with bicarbonate-free, chloride-free, and bicarbonate-chloride-free solutions suggested that the net serosal-to-mucosal transport of these two anions accounts for the short-circuit current not due to sodium absorption. The potential difference and short-circuit current were saturating functions of bathing solution sodium concentration and were inhibited by serosal ouabain and by amiloride. Thus active mucosal-to-serosal sodium transport is the major determinant of the potential difference and short-circuit current in this epithelium. epithelial
transport;
stratified
squamous
epithelium;
rumen;
eso-
phageal mucosal barrier
IT IS GENERALLY CONSIDERED that the only function of the esophagus is to serve as a propulsive conduit from the pharynx to the stomach. Over the past decade, the motor function of this organ has been extensively investigated in both normal animal and human subjects and in various disease states. These studies, at both the organ and cellular levels, have greatly extended knowledge of the mechanisms and control of muscle action and how muscular dysfunction can cause disease or symptoms. However, several investigators have explored another property of the esophagus-the presence of a transmural electrical potential difference (PD) (1, 2, 10, 11, 15-l 7, 20, 26). The origin of this PD has not been defined. We have studied the rabbit esophagus with in vivo and in vitro techniques in order to determine the nature of the PD and the characteristics of water and eIectrolyte transport in mammalian esophagus. METHODS
All studies were performed on rabbits weighing 2-3 kg* The Ringer solution of the following per liter: 140 Na, 5.2 K, 119.8 Mg, 2.4 HPO4, 0.4 HzPOd, mosmol/kg H 20). In Na-free
male, New Zealand albino basic solution used was a composition, in millimoles Cl, 25 HC03, 1.2 Ca, 1.2 and 10 mannitol (280 solutions, and those with
Ozafal Hill,
esophagus
North Carolina
27514
varying Na concentrations, Na was replaced with equimolar choline. Sodium isethionate replaced NaCl and/or NaHC03 in HCOs-free, Cl-free, and HCO&l free solutions. In Cl-free and HCO&I free solutions, the SO4 salts of Ca and Mg were used. HCOa-containing solutions were bubbled with 95 % 02-5 % (202, and HCOa-free solutions were bubbled with 100% 02. All solution pH’s were 7.357.45. In vim experiments. The rabbits were anesthetized with intravenous pentobarbital (20 “g/kg) and maintained with small periodic doses of the anesthetic. Body temperature was kept at 38°C throughout the experiment with a heating pad. A tracheostomy was performed and the cervical esophagus cannulated with an inflow cannula at a level just below the thyroid cartilage. A 4 % agar bridge made up with Ringer solution in polyethylene tubing (PE-205) was inserted with the inflow cannula so that its tip was at the midesophageal level. Following a midline abdominal incision, the gastroesophageal junction was ligated and an outflow cannula introduced. This resulted in a IO-cm segment of esophagus for perfusion. Similar Ringer-agar bridges were placed in the retrogastric area (brought out through a flank puncture wound) and behind the cervical esophagus to serve as reference bridges. The ends of the PD and reference bridges were placed in 3-M KC1 containers with calomel reference electrodes (Fisher Scientific Company, Pittsburgh, Pa.). The electrodes were connected to a high-impedance volt meter (Keithley Instruments, Inc., Cleveland, Ohio, model 61 OB) and the PD was recorded at lo- to 15-min intervals during perfusion. The PD value for each animal was taken as the average of these readings. The PD between the esophagus and cervical reference and between the esophagus and the intra-abdominal reference did not differ by more than l-2 mV. The Ringer perfusion solution contained 5 &i/liter [ 1., 2J4C]poIyethylene glycol (New England Nuclear Corp., Boston, Mass.) and 2 g/liter carrier polyethylene glycol 4,000 (PEG) to serve as a nonabsorbable water marker. The reliability of the [14C]PEG water marker technique has been previously established (5, 28). Twenty-five milliliters of this solution were recirculated through the esophagus at 8 ml/min with a peristaltic pump (Harvard Apparatus Co., Millis, Mass.). The bubbled reservoir was maintained at 38°C with a water bath, and the solution was passed through polyvinyl coils in the water bath just prior to entering the esophagus. Perfusion pressures measured at the entrance and exit from the esophagus were 2 cmHz0 with transient spikes to 6-7 cm during After 10 min perfusion, 2.0 ml of esophageal peristalsis. perfusion solution were removed from the reservoir (designated as time 0) and thereafter at 60, 120, and 180 min for
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 10, 2019.
TRANSPORT
BY
RABBIT
439
ESOPHAGUS
measurement of [i4C]PEG immediately, and frozen for Na, K, and Cl determinations at a later time. [14C]PEG was measured on duplicate 0.250-ml samples in Bray’s solution by liquid scintillation spectrometry (Searle Analytic, Inc., Des Plains, Ill.). Q uenching was monitored by the channels ratio and external standard ratio methods and was found to be uniform so that quench correction was not necessary. Sodium and K were measured by flame photometry, and chloride was measured by coulometric tritration on duplicate samples. Net water and electrolyte transport were calculated with the usual water marker formulas, as modified for the recirculation technique. Steady-state transport was demonstrated during the last two 60-min perfusion periods, indicating that approximately 1 h was necessary as an equilibration period. At the end of the experiment, the animal was killed with pentobarbital, the esophageal segment exposed, and its length measured in situ. Transport was expressed as microliters or microequivalents per hour times centimeter length, based on mean transport rates for the last two 60-min perfusion periods. At the end of each experiment, the volume of solution remaining in the perfusion system was measured to the nearest 0.1 ml for the calculation of [i4C]PEG recovery. Experiments where p4C]PEG recovery was less than 99 % were discarded. In vitro experiments. The experimental techniques described previously for in vitro transport studies in rabbit ileum (18, 22) were modified for use in the esophagus. Under pentobarbital anesthesia, the rabbit esophagus was excised from the level of the thyroid cartilage to the gastroesophageal junction. The esophagus was opened longitudinally and pinned, mucosal surface down, in a paraffin tray containing ice-cold oxygenated Ringer. The muscle layers were grasped with hemostats, lifted up, and dissected free of the underlying mucosa with a scalpel. This resulted in a preparation consisting of the stratified squamous epithelium, the muscularis mucosa, and a small amount of underlying connective tissue as shown in Fig. 1. The stripped preparation was cut into four equal pieces, and each piece was mounted between two Lucite half-chambers with an aperture of 1.13 cm2. The mucosal and serosal bathing solutions (maintained at 37°C) were connected to calomel
FIG. 1. Rabbit esophagus. Full thickness (A) and stripped preparation (B) from which longitudinal and circular muscular layers have been removed. Stripped preparation, which consists of stratified squamous epithelium, muscularis mucosae, and strands of residual connective tissue, was used for in vitro studies. This stratified squamous epithelium consists of lo-13 cell layers. (H & E Xl 11)
60 n -0
50
X z v
40
:
30
G -I = I
20
2 IO
TIME, MIN 2. Rate of appearance of r2Na in adjacent sections of esophagus after addition of isotope to either mucosal or serosal Ringer bath. Cumulative count rates are shown in serosal bath after isotope addition to mucosal bath (m + s), and in mucosal bath after addition to serosal bath (s + m). Linear part of curve after 45 min indicates that steady-state isotope transfer occurs after this time. FIG.
and Ag-AgCl electrodes with agar bridges made up with the solutions studied. The electrodes were connected to an automatic voltage clamp (Physiology Instrument Laboratory, Yale University School of Medicine), which automatically corrects for fluid resistance between the PD sensing bridges, for measurements of the PD and I,,. Tissues were continuously short circuited, except for 5- to 10-s intervals when the open-circuit PD was read. Radioactive isotopes were added to the mucosal and/or serosal bathing solution 5-10 min after mounting the tissues, and this was designated as time 0. It was found that 45 min were required before steady-state isotope transfer of Na or Cl occurred (Fig. 2) ; therefore three 30-min flux periods were measured from 45 to 135 min. The average of the three determinations was taken as the flux values for that tissue. Simultaneous bidirectional Na fluxes were determined with 22Na and 24Na. In separate experiments, on different tissues, oppositely directed unidirectional Cl fluxes were measured on adjacent pieces of esophagus with 36Cl. Net Cl transport was the difference in these two fluxes from the same esophagus. The methods of isotope counting have been previously described (18). A negative PD refers to the orientation of the mucosa with respect to the serosa, in keeping with the convention used in previous in vivo studies of esophagus (1, 2, 10, 11, 15-17, 20, 26). All esophageal PD’s in these experiments, both in vivo and in vitro, were mucosa negative with respect to the serosa. In preliminary experiments it was found that the current-voltage relationship was linear over a f 100 mV range, indicating that the esophageal epithelium behaves as an ohmic resistor. Therefore, resistance was determined by the ratio of the open-circuit PD to the I,,. A negative sign before a net flux infers net serosa-to-mucosal transport (secretion). N refers to the number of tissues in the Na experiments and number of tissue pairs in the Cl experi nents. Generally the number of tissues also equals the number of animals. The significance of differences in
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 10, 2019.
440
POWELL,
means was determined f test (23).
by the paired
or unpaired
MORRIS,
AND BOYD
r
Student
RESULTS
Ehctrical
Parameters
The mean esophageal PD in 10 rabbits perfused with Ringer solution for 150 min in vivo was - 28 =t 3 mV. The electrical parameters of the esophagus, stripped of its muscular and connective tissue layers and mounted in vitro, are shown in Fig. 3. The PD of 32 animals peaked at -20 mV, 30 min after mounting, with a gradual decline over the 135 min duration of the experiments. The I,, gradually declined from 18 to 11 PA/cm? The resistance reached a steady-state value of 1,500 ohm*cm2 within 1 h of mounting. There was little difference in the electrical parameters in vitro at different levels of esophagus as shown in Fig. 4. 30
FIG. 3 r
30
P. 0.
I 20
r
FIG. 4
01
1
1
20
40
I 60
[No]*
1 80
1 100
1
I
120
I40
mM
FIG. 5, Relationship between I sC and bathing solution Na concentration. Esophagi of 4 rabbits were studied simultaneously in 4 solutions with varying Na concentration (choline replacement). Means & 1 SE of I,, from 45 to 135 min after mounting are shown.
Only the PD of the most proximal esophagus (pharyngeal end) showed a statistically higher value, 22.4 & 1.7 mV, as opposed to 15*9-l 7.1 & 1.5 mV for the more distal segments. Although the Isc and resistance of the most proximal segments were slightly greater, these differences were, on the average, not significant.
I? D. f
In Vitro Transport
SC
r
20
I SC
r-
Ringer solution, Sodium and chloride fluxes and the electrical parameters in Ringer solution in vitro are shown in flux (Jf:) was 0.320 the top of Table 1. Net sodium peq/h cm2 from mucosa to serosa, which represented 77 % of the simultaneously measured Isc (expressed as units of ion flux). In separate experiments, a net chloride flux of -0.070 was determined in the opposite direction (from serosa to mucosa), and this value was significantly different from zero (P < 0.001). Having a negative value, the net chloride transport could be considered as accounting for 14 % of the measured I,,. Since net Na transport appeared to account for most of between bathing solution Na conthe L, the relationship centration and the I,, was investigated as shown in Fig. 5. The I,, was a saturating function of Na concentration with an apparent Km of 43 mM and a V,., of 19.5 pA/cm2. A similar relationship was found between bathing solution Na concentration and the PD with a Km of 43 mM and a V max Of 25 mV. The effect of 1W4 M ouabain on the I,, of the in vitro rabbit esophagus is shown in Fig. 6. This cardiac glycoside effectively inhibited the I sc within 45 min. Furthermore, it appeared that this agent was active only when added to the serosal bathing solution. The effects of ouabain on Na and Cl fluxes are shown in the bottom of Table 1. Jte: was inhibited 83 % by ouabain (P < 0.001 ), and the remaining net Na transport, 0.054, did not differ significantly from essentially to the L, 0.043 peg/h cm2. Jztt was reduced zero by ouabain. Another Na transport inhibitor, amiloride, was investigated as shown in Fig. 7. Large doses of this inhibitor, at lcasl KF4 M, were llecessal-y to achicvc a sigdimt inhibition of the I,, and PD. Of note is that 10-30 min were required for maximal inhibition, and in addition, this drug appeared to be more effective in the esophagus when added l
RES I STANCE
RESISTANCE
PHARY NGEAL END
GASfR I C END
(A) in in vitro rabbit esophaFIG. 3. PD @), Iac (O), and resistance gus in Ringer solution (n = 32). FIG+ 4. PD, I,,, and resistance at 4 levels of esophagus in vitro. Esophagi of 11 rabbits were divided into 4 segments from pharyngeal (1) to gastric end (4). Only PD of pharyngeal end was significantly greater than that of other 3 sections (P < 0.001). TABLE 1. Sodium and chloride trans@rt by rabbit esophagus in vitro: Ringer solution, without and with uuabain 1P4 A-4 Solution
(n) *
J ma
J am
J net
I BC
?TXV
peq/ h- cm2 Ringer Na
(25)
Cl (28)
Ringer-ouabain Na (13) Cl (13)
PD
R
oh*
cm?2
0.550 ztu. 035 0.260 zko.017
0.230 kO.017 0.330 zfzO.018
0.320 ho.027 -0 * 070 A0 .009
0.417 Eta rn028 0.508 zto.027
-17.7 ZLO.3 -18.0 +0.8
1,646 zt82 1,388 zt46
0.380 ztO.038 0.264 +o. 025
0.326 zto.033 0.258 zto.019
0.054 &0.023 0.006 ~3.024
0.043 &O .006 0.053 =t=o .004
-2.2 zto.3 -2.3 M.2
1,698 A300 1,442 *251
Values are means =tr SE. These ing the tissues . *n = Number experiments).
values were determined 45-l of tissues (Na experiments)
35 min after mountor tissue pairs {Cl
l
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 10, 2019.
TRANSPORT
BY
RABBIT
441
ESOPHAGUS
to the serosal bathing solution. The absolute degree of inhibition of the I,, with low2 M amiloride at 140 min, as compared to the control, was 65 % when added to the mucosal solution and 84 % when added to the serosal side. This inhibition was only partly reversible by changing the bathing solutions to fresh Ringer. Ion-repzacement solutions. The flux determinations in Ringer solutions (Table 1) suggested that net serosal-to-mucosal (s to m) anion transport accounted for 14-23 % of the I,,. Therefore, additional flux experiments were performed in Cl and/or HCOa-free solutions (Table 2) to further delineate the role of anion transport in this epithelium. These experiments revealed several important points. First, Na flux determinations in paired experiments utilizing four solutions (Ringer, HCOS free, Cl free, and HCOa-Cl free) showed that JF,“t was consistently less than the Isc when either Cl and/or HCOS were present in the solution. When both anions were replaced by isethionate, Jfz was slightly, but not significantly, greater than the I,,. Second, the sum of net Na and Cl transport in the HCOa-free solutions equaled the simultaneously measured I,,. Third, when both anions were replaced by isethionate, the PD and I,, of the esophagus were significantly inhibited. Finally, total tissue resistance was significantly increased by replacement of either HCO3 and/or Cl, indicating that these anions contribute to the total electrical conductance of this tissue.
0
30
60
90
I20
TIME,
150
180
El0
MIN
FIG. 6. Effect of ouabain 1W4 M on I,, of rabbit esophagus in vitro. Esophagi were mounted in Ringer solution alone or in Ringer containing ouabain in both m and s baths. In control group, ouabain was then added to either m or s baths at 140 min. Ouabain appears to be effective only when added to s bathing solution (n = 14 for both control and ouabain m + s). 100
AMILORIDE
K
m S } 10-6 CONTROL m s ) 10-4
0'
I 20
1 40
I
1
60
80
TIME,
MIN
I
I
I
I00
120
140
7. Effect of various concentrations of amiloride on I,, of rabbit in vitro. Values have been normalized to percent of . . esophagus initial I 6C, and SE’s of means omitted for clarity of presentation. At 30 min after mounting, amiloride 10-6, lo-“, or lUt2 IM was added to either m or s bathing solution (n = 18 for control, and 3 for each amiloride concentration and bathing solution addition). FIG.
2. Effect of anions on electrolyte trans;borf by rabbit esophagus in vitro
TABLE
SoIution
(fl>*
Jms
J sm
J net
IS,
peq/h+ cm2
PD
R
7nV
ohm 1,945 +I76
e cm2
Ringer Na
(10)
0,480 &O. 047
0.195 zto. 024
0.284 &O. 032
0.364 zko ,033
-18.3 Al.1
0.447 A0 .049 0.125 AO.013
0.171 AZ0 * 022 0.180 zto.019
0.260 410.042
*o
zfzo.012
0.298 IO33 0.298 zto. 033
-19.0 zt1.8 -17.7 Al.5
2,353 zk203t 2,365 ztx42t
0.536 zto.077
0.292 zto. 050
0.245 ~0.070
0.364 dEo.033
-22.6 zt1.q
2,541 *249 t
0.525 zto.037
0.228 AZ0 * 044
*o.
0.248 zt0.033t
-13.4 zt1.3t
2,626 zt307t
HCOa free Na
(10)
CI (11)
CI
-0.054
free
Na
(10)
HCOg-Cl ft,ee Na (IO)
0.297 030
Values
are means j, SE. Na ff uxes in the four solutions were performed simulon intestine from 10 animals. CI fluxes in the HiCOB-free experiments were performed separately. Values were determined from 45 to I35 min. * n = Number of tissues (Na experiments) or tissue pairs (Cl experiments). t Significance of difference from Ringer, P < 0.001. $ Significance of difference from Ringer, P < 0.05. taneously
In Viva Transport The net transport rates of H20, Na, K, and Cl measured in vivo in 10 animals by the recirculation-water marker technique. [14C]PEG recovery in these experiments was 99.7 & 0.3 %. The rate of water absorption was 10.2 & 3.8 pi/h *cm length, but was significantly difl’erent from zero and within the limits of sensitivity of this technique.’ Meaningful data for electrolyte absorption could not be obtained because of the propagated errors inherent in this experimental technique and because the low levels of absorption strain the accuracy of this technique. The rates of measured Na, K, and Cl transport were 0.41 h 0.38, - 0.03 AI 0.05, and 1.32 & 1.02 peq/h*cm length, respectively. It shouId be noted that Na absorption in vivo must take place against an adverse electrical gradient of 28 mV.
Although a transmural PD across mammalian esophagus was recorded as early a$ 1952 by Rovelstad, Owen, and Magath (ZO), little is known regarding the nature of this phenomenon. In fact, the magnitude and electrical orientation of this PD have not been conclusively established. Beck and colleagues (1, 2, 15, 26) and other investigators (11, 16, 17) h ave measured a variable PD generally oriented with the mucosa positive with respect to serosa. However, Grantham, Code, and Schlegel (10) have demonstrated that with the proper placement of the reference electrodes (intravenous rather than skin) the transesophageal PD in man is approximately - 10 mV, lumen negative. In the present experiments with an experimental animal, where the reference electrode was placed directly on the serosal surface of both the cervical and intra-abdominal esophagus, we have recorded a lumen negative PD (- 28 mV), which suggests this is the correct orientation of mammalian esophagus. l See ref. 19 for a discussion of the sensitivity perfusion technique in experimental animals.
of the water-marker
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 10, 2019.
442 Transmural PD’s can be divided into two general catephenomena, such as gories : those due to electrokinetic diffusion potentials or streaming potentials, and those due to active electrolyte transport. When the rabbit esophagus was stripped of its muscle and connective tissue layers and bathed in vitro in the absence of transepithelial ionic or hydrostatic pressure gradients, a similar lumen negative PD (-17.9 mV) was recorded. This indicates that the PD is related in some way to the active transepithelial transport of electrolytes. Measurements of ion fluxes in the short-circuited state allow determination of the ion species whose transport accounts for the I sc, and by inference, for the PD (25). Ion transport studies in the stripped esophagus in vitro (Table 1) revealed that the net transport of sodium from mucosa to serosa accounts for over 75 % of the simultaneously measured I Sc. The conclusion that net Na transport is the primary determinant of the PD and I,, in this epithelium is strengthened by the finding that the PD and ISc were saturable functions of bathing solution Na concentration; that the PD, I,,, and net electrolyte transport were inhibited by serosally placed ouabain; and that amiloride inhibited the PD and I,, in this epithelium. The fact that ouabain appears to be active only when added to the serosal bath is taken as evidence in other epithelia, such as frog skin, toad bladder, and intestine, that Na-K-ATPase (and the Na pump) is located at the serosal aspect of the transporting cells. However, the cells responsible for active Na transport in the esophagus may be located deep within this epithelium, and there could be a significant diffusion barrier to ouabain. Therefore, interpretation of the site of action of ouabain in this epithelium must be guarded, This may also account for the atypical effects of amiloride in the esophagus. This drug, which is thought to act by inhibiting Na entry into the transporting cells, has an essentially instantaneous onset of action and is effective in low concentration (10 -“lo-” M) when added to the mucosal bathing solution of toad bladder and frog skin (3, 4, 21). In the esophagus (Fig. 7), much larger concentrations of the drug and a longer time were necessary to achieve significant inhibition. Inhibition was also more effective when the drug was added to the serosal solution. The reasons for these differences are unclear but may be due to a significant diffusion barrier at the mucosal surface. In Ringer solution, net mucosal-to-serosal (m-to-s) Na transport accounted for only 77 % of the I,,, and net Cl transport was demonstrated from s to m. The difference between JfA and ISc is statistically significant as was the Jzi, significantly different from zero. These results suggest that net Cl (or perhaps Cl and HC03) transport may account for the remainder of the I,,. The ion-replacement experiments (Table 2) lend support to this idea. In paired experiments with Ringer, HCOs-free, Cl-free, and HCOS-Cl free solutions, Jfe: was consistently less than I,, when either HCO3 and/or Cl were present in the solution. When both HCOS and Cl were replaced by isethionate, Jzz was slightly but not significantly greater than the I,,. Pn addition, when HC03 and Cl were absent, both the PD and I,, in this solution were significantly less than when either anion or both were present. Finally, in HCOa-free solution, the sum of. Jr: and .J3, equaled the ISc. These results are
POWELL,
MORRIS,
AND BOYD
3. Comparison of variousstratzjkd squamous 8pithelia bathedin Ringersolution TABLE
Epithelium
PD, mV
LC , d/cm
2 R, ohm.cm2
Direction of Net Ion Transport, Short-Circuited Na+
Frog skin Rumen Goat (21) cow (21, 22) Sheep (23, 241 Esophagus rabbit Values
Cl-
50-100
40-70
1,000~2,500
m +s
0
7.8A4.1 lZ.Oztl .o 13*9&2.4
13+5 12+2 llzt2
610&19O 960~ 84 1,156&123
m+s m -3 m +8
m +s m+s
17.9zt0.6
13&l
1,466&
m +a
s--,m
are means
43
m-388
* SE.
compatible with a transport model in which net m-to-s sodium transport accounts for 77 % of the I,, and net anion transport from s to m for the remaining 23 %. When both HC03 and Cl are present, both would be transported, but one anion could substitute for the other in its absence. Our experiments were not designed to explore the exact manner by which these transport processes for Na (and anion) create a transmural PD. Electrolyte transport in this epithelium could be truly electrogenic or could involve ion exchange across cell borders with asymmetrical permeabilities, which could lead to a diffusion potential. The ability of a stratified squamous epithelium, such as esophageal mucosa, to transport electrolytes is not unusual, in view of the demonstrated transport in other epithelia of this type. In Table 3 are summarized the electrical parameters and direction of electrolyte transport in vitro of several stratified squamous epithelia. All are relatively high-resistance epithelia, and all transport Na from m to s. They differ in the direction in which they transport anion. Electrically, the esophagus closely resembles rumen epithelium, which is not surprising in that it is generally considered that the rumen is embryologically derived from esophagus. The cells responsible for electrolyte transport in these epithelia remain to be conclusively identified. In the frog skin, which is only three to five cell layers thick, and in rumen epithelium, which contains many cell layers like the esophagus, it is thought to be those cells at the midlevel of the epithelium called the stratum granulosum (9, 12-14, 27). The possibility that anion transport is accomplished by glandular structures must be considered, although we have not observed submucosal glands in several histological sections of rabbit esophagus. Although the demonstration of net m-to-s Na transport in vitro clarifies the nature of the PD in the rabbit esophagus, these studies raise another fundamental question: what is the physiological significance of electrolyte transport by the esophageal epithelium. In vivo, where active electrolyte transport must take place against a steep electrical gradient, significant net transport of sodium was not demonstrable by conventional techniques. The low rates of transport, the small surface area of this epithelium, and the short residence time of any fluid and electrolyte in the esophagus would all limit any important role in the conservation of water and electrolytes. Electrolyte transport by the esophagus could be vestigial in nature and serve no teleological purpose. On
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 10, 2019.
TRANSPORT
BY RABBIT
ESOPHAGUS
the other hand, epithelial cells, like all cells, maintain homeostasis by transporting sodium out of the cell. Specialization of epithelial cells into polar cells, which allows sodium transport in one direction only, may prevent disruption of epithelial integrity as a result of water flow in response to the transported solute. Therefore, transepithelial electrolyte transport could be necessary if epithelial cells are to maintain a low intracellular sodium concentration and yet still maintain the architecture of this barrier be-
443 tween the lumen and the blood. necessary to clarify these issues.
Additional
This investigation was supported by National Research Grant AM1 5350 from the National Metabolism, and Digestive Diseases. D. W. Powell is the recipient of Research Award AM70454 from the National Institute lism, and Digestive Diseases. Received
for publication A
31 December
1974.
lanthanum,
and
studies
will
be
Institutes of Health Institute of Arthritis, Career Development of Arthritis, Metabo-
REFERENCES 1. BECK, I. T., AND N. A. HERNANDEZ. Transmural potential difference in patients with hiatus hernia and esophageal ulcer. Gut 10: 469-476, 1969. 2. BECK, I. T., T. F. MCELLIGOTT, M. C. PATH, AND N. A. HERNANDEZ. Transmural potential difference at the level of the upper esophageal sphincter in man. Am. J. Digest. Diseases 14 : 456462, 1969. 3. BENTLEY, Pp. J. Amiloride: a potent inhibitor of sodium transport across toad bladder. J. Physiol., London 195 : 317-330, 1968, 4. BIBER, T. U. L. Effect of changes in transepithelial transport on the uptake of sodium across the outer surface of the frog skin. J. Gesz. Physiol. 58: 131-144, 1971. 5. BRIGHT&ARE, P., AND H. J. BINDER. Stimulation of colonic secretion of water and electrolytes by hydroxy fatty acids. Gasfroenterology 64 : 81-88, 1973. 6. CHIEN, W. J., AND C. E, STEVENS, Coupled active transport of Na and Cl across forestomach epithelium. Am. J. Physiol. 223: 997-1003, 1972. 7. FERREIRA, H. IL, F. A. HARRISON, AND R. D. KEYNES+ The potential and short-circuit current across isolated rumen epithelium of sheep. J. Physiol., London 187: 631-644, 1966. 8. FERREIRA, H. G., F. A. HARRISON, R. D. KEYNES, AND L. ZURICH. Ion transport across an isolated preparation of sheep rumen epithelium. J. Physiol., London 222: 77-93, 1972. 9. GEMMEL, R. T., AND B. D. STACY. Effects of ruminal hyperosmolality on the ultrastructure of ruminal epithelium and their relevance to sodium transport. Quart. J. Exjtl. Physiol. 58 : 315323, 1973. 10. GRANTHAM, R. N., C. F. CODE, AND J. F. SCHLEGEL. Reference electrode sites in determination of potential difference across the gastroesophageal mucosal junction. Mayo GZin. Pm. 45 : 265-274, 1970.
11. HELM,
W. J., J. F. SCHLEGEL, C. F. CODE, AND W. A. J. SUMIdentification of the gastroesophageal mucosal junction by transmucosal potential in healthy subjects and patients with hiatal hernia. Gastroenterology 48 : 25-35, 1965. 12. HENRIKSON, R. C. Mechanism of sodium transport across ruminal epithelium and histochemical localization of ATPase. Exptl. Cell Res. 68: 456458, 1971. 13. HENRIKSON, R. C. Ultrastructure of ovine run&al epithelium and localization of sodium in the tissue. J. Ultrastructure Res. 30: 385401, 1970. 14. HENRIKSON, R. C., AND B. D. STACY. The barrier to diffusion across ruminal epithelium: a studv bv electron microscosv using MERSKILL.
horseradish
peroxidase,
ferritin.
J.
Ultrastructure
Res. 34: 72-82, 1971. 15. HERNANDEZ, N. A., AND
I. T. BECK, Gastroesophageal transmural measured by a new constant infusion method. scarification on this potential difference. Am. J. Digest. Diseases 14 : 206-Z 16, 1969. 16. KATZKA, I, H. M. LEMON, AND F. JACKSON. The correlation of human gastric potentials with gastric physiology. Gastroenterology 28: 717-730, 1955. 17. MECKEI,ER, K. J. H., AND F. J. INGELFINGER. Correlation of electric surface potentials, intraluminal pressures, and the nature of tissue in the gastroesophageal junction of man. Gastroenterology potential difference The effect of skin
52: 966-971, 1967. 18. POWLL, D. W., H. J. BINDER, AND P. F. CURRAN. Electrolyte secretion by guinea pig ileum in vitro. Am. J. Physiol. 223 : 53 I537, 1972. 19. POWELL, D. W., S. J. MALA~ER, AND G. R. PLOTKKN. Secretion of electrolytes and water by the guinea pig small intestine itz viuo. Am. J. Physiol. 215: 1226-1233, 1968. 20. ROVELSTAD, R. A., C. A. OWEN, *JR., AND T. B. MAGATH. Factors influencing the continuous recording of in situ pH of gastric and duodenal contents. Gastroanterology 20 : 609-624, 1952. 21. SALAKO, L. A., AND A. J. SMITH. Effects of amiloride on active sodium transport by the isolated frog skin: evidence concerning site of action. Brit. J. Pharmacol. 38 : 702-7 18, 1970. 22. SCHULTZ, S. G., AND R. ZALUSKY. Ion transport in isolated rabbit ileum. I. Short-circuit current and Na fluxes. J. Gen. Physiol. 47 : 567-584, 1964. G. W., AND W. F. COCHRAN. Statistical Methods (6th 23. SNEDECOR, ed.) Ames : Iowa State Univ. Press, 1967. C. Transport of sodium and chloride by isolated rumen 24. STEVENS, epithelium. Am. J. Physiol. 206 : 1099-l 105, 1964. Active transport of sodium as the 25. USSING, H. H., AND K. ZERAHN. source of electric current in the short-circuited isolated frog skin. Acta
Physiol.
&and.
23 : 110-127,
1951.
E, I., J. A. E. Fox, AND I. T. BECK. Transmural potential 26. VIDINS, difference (P.D.) in the body of the esophagus in patients with esophagitis, Barrett’s epithelium and carcinoma of the esophagus. Am. J. Digest. Diseases 16 : 991-999, 197 I. G. Electrical potential profile of the toad skin 27. WHTTTEMBURY, epithelium. J. Gen. Physiol. 47 : 795-808, 1964. D. L., R. J. SANDBERG, AND S. J. PHILLIPS. A compari28. WINGATE, son of stable and 14C-labeled polyethylene glycol as volume indicators in the human jejunum. Gut 13: 812-815, 1972.
Downloaded from www.physiology.org/journal/ajplegacy by ${individualUser.givenNames} ${individualUser.surname} (129.081.226.078) on January 10, 2019.
