N. A. Tobey, C. M. Argote, X. C. Vanegas, W. Barlow, and R. C. Orlando

Transcription

1 Am J Physiol Gastrointest Liver Physiol 293: G264 G270, First published April 12, 2007; doi: /ajpgi Electrical parameters and ion species for active transport in human esophageal stratified squamous epithelium and Barrett s specialized columnar epithelium N. A. Tobey, C. M. Argote, X. C. Vanegas, W. Barlow, and R. C. Orlando Department of Medicine Tulane University Health Sciences Center and Veterans Administration Hospital, New Orleans, Louisiana Submitted 24 January 2007; accepted in final form 6 April 2007 Tobey NA, Argote CM, Vanegas XC, Barlow W, Orlando RC. Electrical parameters and ion species for active transport in human esophageal stratified squamous epithelium and Barrett s specialized columnar epithelium. Am J Physiol Gastrointest Liver Physiol 293: G264 G270, First published April 12, 2007; doi: /ajpgi The human esophagus is lined by stratified squamous epithelium (ESSE), and in some subjects with reflux disease the distal esophagus becomes lined by Barrett s specialized columnar epithelium (BSCE). ESSE and BSCE differ both histologically and functionally, the latter evident by differences in their in vivo transmural electrical potential difference (PD), ESSE averaging 15 mv and BSCE being greater than 25 mv. In this report we examine the basis for this difference in PD. This is done by mounting endoscopic biopsies of ESSE from 25 subjects without esophageal disease and BSCE from 19 with Barrett s esophagus in mini-ussing chambers for electrical recordings basally and after bathing solution ion replacement. The results show that the PD of human ESSE reflects a low level of active ion transport ( A/cm 2 ) combined with a high level of tissue (electrical) resistance ( cm 2 ) and that of BSCE reflects a high level of active transport ( A/cm 2 ) combined with a low level of resistance (69 8 cm 2 ). Furthermore, active transport in ESSE was principally due to sodium absorption whereas in BSCE it was equally divided between sodium absorption and anion (chloride/ bicarbonate) secretion, the latter through an apical membrane, 4- acetamido4 -isothiocyano-2,2 -stilbenedisulfonic acid-sensitive anion channel. As an anion-secreting tissue with bicarbonate secretory capacity more than fivefold greater than ESSE, BSCE is better suited than ESSE for defense of the esophagus against reflux disease. mini-ussing chambers; ouabain; potential difference; short-circuit current; electrical resistance; gastroesophageal reflux disease; ethoxyzolamide; 4-acetamido4 -isothiocyano-2,2 -stilbenedisulfonic acid THE HUMAN ESOPHAGUS IS LINED by stratified squamous epithelium (ESSE), and in 10% of subjects with gastroesophageal reflux disease the ESSE of distal esophagus is destroyed and replaced to a varying extent by a lining of Barrett s specialized columnar epithelium (BSCE) (19). ESSE and BSCE differ not only structurally, as the names imply, but also functionally, and one marker of this functional difference is reflected by differences in their in vivo transmural electrical potential difference (PD). The in vivo PD for ESSE averages 15 5 mv and the in vivo PD for BSCE is typically greater than 25 mv (11, 22, 24, 30, 39). Since the PD, according to Ohm s law, is the product of current times (electrical) resistance, such differences in PD for ESSE and BSCE indicate that these epithelia have fundamental differences in ion transport and/or barrier properties. The nature of these differences, however, remains elusive, principally because of limitations in electrical recordings in humans in vivo and in obtaining suitable human specimens for electrical recordings in vitro. For instance, in vivo studies require passage of oral or nasal catheters into the esophagus for recordings, a technique that is inherently uncomfortable and restricts access to only one (luminal) side of the tissue. And in vitro studies require access to esophagectomy specimens so that a large (2 cm 2 cm) section of human esophageal epithelium could be obtained for mounting in standard Ussing chambers. Esophagectomies, however, are not common procedures in most centers and when they are done (for cancer) specimen quality is often uncertain owing to preoperative chemotherapy and/or radiotherapy. Nonetheless, recent modifications to this old technology, i.e., Ussing chamber, have provided another means to investigate digestive tract epithelia, and that is through development of a minichamber system. The mini-ussing chamber has a Lucite ring with aperture small enough to mount an endoscopic biopsy so that each side of the tissue is independently perfused and in contact with electrodes for recording and current generation. In this report, the minichamber was applied to study human esophageal epithelium with two goals in mind. One was to measure and compare for ESSE and BSCE the electrical properties that form the basis for the in vivo PD, and the other was to investigate the nature of the ions responsible for active ion transport. The results demonstrate that the low in vivo PD of human ESSE is the product of a low level of active ion transport and high level of tissue (electrical) resistance, whereas the high in vivo PD of BSCE is the product of a high level of active ion transport and low level of tissue resistance. Furthermore, active ion transport in human ESSE is predominantly due to sodium absorption, whereas active transport in BSCE is equally divided between sodium absorption and anion secretion, anion secretion reflecting the movement of chloride and/or bicarbonate through an apical membrane 4-acetamido4 -isothiocyano-2,2 -stilbenedisulfonic acid (SITS)-sensitive channel. The implications of these findings with respect to defense of the esophagus against acid reflux disease are presented in the discussion. MATERIALS AND METHODS Subjects. Using jumbo biopsy forceps (Radial Jaw-3 Maximum Capacity, Boston Scientific, Natick, MA; length when opened 3.3 mm), we obtained esophageal biopsies from adult subjects, ages Address for reprint requests and other correspondence: N. A. Tobey, Tulane Univ. Health Sciences Center, SL-35, 1430 Tulane Ave., New Orleans, LA ( ntobey@tulane.edu). The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. G /07 $8.00 Copyright 2007 the American Physiological Society

2 ELECTRICAL PARAMETERS AND ION TRANSPORT IN HUMAN ESOPHAGEAL EPITHELIUM G265 Fig. 1. Mini-Ussing chamber setup. Left: small (2 mm diameter) aperture of slide on which the endoscopic biopsy is positioned. A dime is shown for size reference. Middle: 2 halves of the Lucite chamber into which the biopsy holder is inserted. Right: complete Ussing chamber setup with electrical connections for recording and tubing for gassing luminal and serosal perfusates warmed to body temperature yr old, undergoing upper endoscopy for clinical reasons. Healthy human ESSE was retrieved from 25 subjects with no history of esophageal symptoms or disease and a grossly normal esophagus on endoscopy. The health of retrieved ESSE was further verified by histological review of the biopsy after staining with hematoxylin and eosin. BSCE was retrieved from 19 subjects with Barrett s esophagus, and this was verified histologically by review of the biopsy for goblet cells after staining with Alcian blue, ph 2.5. This protocol was approved by the Internal Review Board for Human Subjects at the Tulane University Health Sciences Center, New Orleans, LA, and all subjects provided written, informed consent. Mini-Ussing chamber. After retrieval from the forceps, biopsies were immersed in ice-cold oxygenated Ringer solution and immediately transported to the laboratory. Biopsies were mounted mucosal side up in mini-ussing chambers with Lucite rings whose aperture diameter was 2 mm and square area was cm 2 (Fig. 1). Biopsies were bathed on both sides with 5 ml of normal Ringer solution (composition in mmol/l): Na 140, Cl 119.8, K 5.2, HCO 3 25, Ca 2 1.2, Mg 2 1.2, HPO , H 2PO 4 0.4, 268 mosmol/kgh 2O, ph 7.4 when gassed with 95% O 2-5% CO 2 at 37 C. Two sets of electrodes connected the solutions in the chambers to voltage clamps (Voltage Current Clamp, MC6; Physiologic Instruments, San Diego, CA) that permitted the direct recording of the transmural electrical PD and determination of short-circuit current (I sc) by passage of current. [Note: I sc is recorded in A/cm 2 and converted to eq cm 2 h 1 by multiplying by ] Total electrical resistance (R T) was calculated by using Ohm s law, where PD I sc R T. All experiments were conducted under open-circuit conditions except when periodically switched to the short-circuit state for recording of I sc. The n value for each experiment represents the values obtained from one tissue per subject. After equilibration for min, basal electrical readings of PD, I sc, and R T were obtained and repeated at 15 and 30 min following replacement of luminal and serosal bathing solutions with Na-free or Cl-free Ringer solution, bubbled with 95% O 2-5% CO 2,orHCO 3-free Ringer bubbled with 100% O 2. The specific compositions of the ion-substituted bathing solutions were as follows (unless otherwise specified, chemicals were obtained from Sigma, St. Louis, MO). Na-free Ringer (in mmol/l): 115 N-methyl-D-glucamine, Cl, 5.2 K, 25 choline, 25 HCO 3, 1.2 Ca 2, 1.2 Mg 2, 2.4 HPO 4 2, 0.4 H 2PO 4, 295 mosmol/kgh 2O, ph 7.5 when gassed with 95% O 2-5% CO 2 at 37 C; Cl-free Ringer (in mmol/l): 140 Na, 115 gluconate, 5.2 K,25HCO 3, 1.2 Ca 2, 1.2 Mg 2, 2.4 SO 4 2, 2.4 HPO 4 2, 0.4 H 2PO 4, 294 mosmol/kgh 2O, ph 7.5 when gassed with 95% O 2-5% CO 2 at 37 C; and HCO 3-free Ringer (in mmol/l): 140 Na, Cl, 5.2 K, 25 HEPES, 1.2 Ca 2, 1.2 Mg 2, 2.4 HPO 4 2, 0.4 H 2PO 4, 288 mosmol/kgh 2O, ph 7.5 when gassed with 100% O 2 at 37 C. Between exposures to different ion-substituted bathing solutions, bathing solutions were returned to normal Ringer solution for 30 min to ensure reversibility, and tissue viability was assessed at the end of each experiment by addition of ouabain, 1 mm, to the serosal bath. After exposure to ouabain for 30 min, tissues in which I sc fell by 60% were utilized for data analysis. [Note: similar ouabain responsiveness was used to establish viability in minichambered duodenal epithelium (28) and standard-chambered ESSE in rabbit (26)]. In other experiments, following equilibration and basal electrical readings, ESSE or BSCE were exposed for 30 min to serosal ethoxyzolamide, a carbonic anhydrase inhibitor, or to luminal or serosal SITS, an inhibitor of bicarbonate-dependent anion exchange and some anion (chloride or bicarbonate) channels. Data analysis. Data are reported as the mean SE. Statistical significance was determined by Student s t-test for paired and unpaired samples. RESULTS After mounting and equilibration, tissues remained electrically stable for upward of 2 h. The range and means for basal PD, I sc, and R T for ESSE and BSCE are shown in Table 1. ESSE and BSCE differed significantly for all parameters, with the differences in PD reflecting a high I sc combined with low R T for BSCE and a low I sc combined with high R T for ESSE. These patterns for I sc and R T were sufficiently distinctive that the nature of the esophageal tissue mounted in the chamber could be predicted from basal electrical data. In addition, the Table 1. In vitro basal transmural electrical PD, I sc, and R T of human ESSE and BSCE PD, mv A/cm 2 RT, cm 2 ESSE (n 22) Range 0.2 to to to to 635 BSCE (n 12) * * * 69 8* Range 0.3 to to to to 114 Values are means SE; n no. of tissues (1 per subject). PD, potential difference; I sc, short-circuit current; R T, electrical resistance; ESSE, esophageal stratified squamous epithelium; BSCE, Barrett s specialized columnar epithelium. *P 0.05 compared to ESSE.

3 G266 ELECTRICAL PARAMETERS AND ION TRANSPORT IN HUMAN ESOPHAGEAL EPITHELIUM Table 2. Effect of Na-free, Cl-free, or HCO 3 -free Ringer solution on the transmural electrical PD, I sc, and R T of human ESSE Treatment n PD, mv Amps/cm 2 Normal Ringer baseline Na-free Ringer 15 min * * Na-free-Ringer 30 min * Normal Ringer baseline Cl-free Ringer 15 min * Cl-free Ringer 30 min * * Normal Ringer baseline HCO 3-free Ringer 15 min * HCO 3-free Ringer 30 min * Values are means SE; n no. of tissues (1 per subject). *P 0.05 vs. time 0. P 0.05 vs. 15 min. RT, cm 2 in vitro PD for BSCE was more than twofold higher than the PD for ESSE paralleling that in vivo, and each in vitro PD value was 10-fold lower than its in vivo counterpart (11, 22). Also, the in vitro I sc for BSCE was eightfold greater than that for ESSE, whereas the in vitro R T for ESSE was fivefold greater than that for BSCE (Table 1). After basal PD, I sc, and R T were recorded, the ions responsible for active transport were sought in ESSE and BSCE by replacement of normal Ringer solution with either Na-free, Cl-free, or HCO 3 -free Ringer solution. For both ESSE and BSCE, Na-free and Cl-free solutions yielded lower values for PD and I sc and higher values for R T, whereas HCO 3 -free solution for ESSE and BSCE yielded lower values for PD and I sc but without change in R T (Tables 2 and 3). The magnitude of the changes in I sc and R T for ESSE and BSCE are illustrated in absolute values in Fig. 2, A and B and compared as percent change from basal values in Fig. 3, A and B. Notably, BSCE, with its high I sc, responded to ion replacement with greater declines in I sc than ESSE whereas ESSE, with its high R T, responded to ion replacement (for Na-free and Cl-free) with greater increases in R T than BSCE. Notably, however, on a percent basis, Na-free solution almost abolished the I sc of ESSE (decline of 94 4%) whereas Na-free solution produced only a moderate decline in I sc of BSCE (decline of 50 15%) (Fig. 3A). These declines in I sc were reflected in a rise in R T for ESSE of 27 19%, a value comparable to that produced by ouabain (32 4%) in ESSE, and a rise in R T for BSCE of 13 10%, which was 50% of that produced by ouabain (27 5%) in BSCE. As shown in Fig. 3A, replacement with Cl-free or HCO 3 -free solution resulted in a decline in I sc for ESSE and BSCE that was essentially equivalent on a percent basis; yet on an absolute basis, the decline in I sc for ESSE was far smaller (each 0.05 eq cm 2 h 1 ) (Table 2, Fig. 2A) than that for BSCE ( 0.9 eq cm 2 h 1 for Cl-free and 0.28 eq cm 2 h 1 for HCO 3 -free) (Table 3, Fig. 2A). Notably, despite the greater (absolute) declines in I sc with Cl-free or HCO 3 -free solution for BSCE, there was only a modest change in R T with Cl-free and no change in R T with HCO 3 -free solution (Figs. 2B and 3B). This observation and the fact that the sum of the residual currents in Na-free, Cl-free, and HCO 3 -free solutions (I sc values in Tables 2 and 3) are greater than basal I sc for BSCE, but not ESSE, indicate that in BSCE chloride and bicarbonate can share the same secretory pathway. Given the potential importance of anion secretion in BSCE, and particularly as it relates to the transport of HCO 3 for defense against acid reflux in vivo, we assessed in BSCE (and for comparison ESSE) the effect on I sc of ethoxyzolamide, 100 M, a carbonic anhydrase inhibitor, and SITS, 4 mm, an inhibitor of both anion exchange and (some) anion channels (5, 36). [Note: SITS was used at 4 mm to ensure blockade of all anion transporters in ESSE, with SITS, 4 mm, required to protect ESSE against serosal acid injury, a process mediated by activity of a basolateral membrane HCO 3 -dependent transporter (36).] In ESSE, neither ethoxyzolamide nor luminal or serosal SITS had any effect on the I sc. Similarly, neither ethoxyzolamide nor serosal SITS had any effect on the I sc of BSCE. However, luminal SITS significantly reduced the I sc of BSCE from eq cm 2 h 1 to eq 1; Table 3. Effect of Na-free, Cl-free, or HCO 3 -free Ringer solution on the transmural electrical PD, I sc, and R T of BSCE Treatment n PD, mv Amps/cm 2 RT, cm 2 Normal Ringer baseline Na-free Ringer 15 min * * Na-free-Ringer 30 min * Normal Ringer baseline Cl-free Ringer 15 min * Cl-free Ringer 30 min * * Normal Ringer baseline HCO 3-free Ringer 15 min * HCO 3-free Ringer 30 min * Values are means SE; n no. of tissues (1 per subject). *P 0.05 vs. time 0. P 0.05 vs. 15 min.

4 ELECTRICAL PARAMETERS AND ION TRANSPORT IN HUMAN ESOPHAGEAL EPITHELIUM hm cm 2 h 1 for an overall decline of 38 7% (n 4, P 0.05 vs. initial value) (Fig. 4). G267 DISCUSSION The healthy esophagus is lined by ESSE, although in some patients with reflux disease it is replaced with BSCE. BSCE differs from ESSE both structurally and functionally, presumably because BSCE is a protective adaptation as mucosal defense against (further) reflux injury to the esophagus (18). The PD is a marker of epithelial function that is known to differ in vivo for ESSE and BSCE. The in vivo PD of ESSE is relatively low at approximately 15 mv compared with that of the in vivo PD of BSCE which is typically greater than 25 mv (11, 22). On the basis of Ohm s law [PD current resistance], these differences in PD imply differences in current and/or resistance, with differences in current being due to differences in net ion flow across the epithelium and differences in resistance being due to barriers (resistors) to net ion flow created by epithelial structures such as cell membranes and tight junctions. Information on the parameters that form the basis for the PD in humans is limited for ESSE, knowledge based on two preliminary reports (see below) and nonexistent Fig. 3. A: percent decline in I sc for human ESSE and BSCE when switched from normal Ringer solution to Na-free, Cl-free, or HCO 3-free Ringer solutions. Values are means SE at 30 min. *P 0.05 for BSCE, n 5, vs. ESSE, for Na-free, n 5, for Cl-free, n 10, and for HCO 3-free, n 12. Note that on a percent basis, the decline in I sc in Na-free solution for ESSE is significantly greater than for BSCE whereas the decline for both Cl-free and HCO 3-free are approximately the same. B: percent increase in transepithelial R T for ESSE and BSCE when switched from normal Ringer solution to Na-free, Cl-free, or HCO 3-free Ringer solutions. Values are means SE at 30 min. BSCE, n 5, vs. ESSE, for Na-free, n 5, for Cl-free, n 10, and for HCO 3-free, n 12. Note that on a percent basis, the increases in R T for ESSE and BSCE are similar for all solutions. for BSCE (Note: some data in this manuscript on BSCE have appeared in abstract form; see Refs. 32 and 33). In this report we expanded knowledge about ESSE and BSCE by mounting endoscopic biopsies in mini-ussing chambers for recording of their electrical parameters both basally and after either bathing solution ion replacement or exposure to Fig. 2. A: decline in short-circuit current (I sc) for human esophageal stratified squamous epithelium (ESSE) and Barrett s specialized columnar epithelium (BSCE) when switched from normal Ringer solution to Na-free, Cl-free, or HCO 3-free Ringer solutions. Values are means SE at 30 min. *P 0.05 for BSCE, n 5, vs. ESSE; for Na-free, n 5, for Cl-free, n 10, and for HCO 3-free, n 12. Note that on an absolute basis the decline in I sc is significantly greater for BSCE than ESSE for each ion substitution experiment. B: increase in transepithelial electrical resistance (R T) for ESSE and BSCE when switched from normal Ringer solution to Na-free, Cl-free, or HCO 3-free Ringer solutions. Values are means SE at 30 min. *P 0.05 for BSCE, n 5, vs. ESSE, for Na-free, n 5, for Cl-free, n 10, and for HCO 3-free, n 12. Note that R T increases for both ESSE and BSCE in Na-free and Cl-free but remains unchanged when cells are exposed to HCO 3-free Ringer solution. Fig. 4. Decline in I sc for human ESSE and BSCE when exposed to serosal ethoxyzolamide, serosal 4-acetamido4 -isothiocyano-2,2 -stilbenedisulfonic acid (SITS), or luminal SITS. Note the significant decline in I sc for BSCE with luminal, but not serosal, SITS. Values are means SE; *P 0.05 for BSCE, n 4, vs. ESSE, n 4.

5 G268 ELECTRICAL PARAMETERS AND ION TRANSPORT IN HUMAN ESOPHAGEAL EPITHELIUM ethoxyzolamide or SITS. The results indicate that the known differences in the in vivo PD for ESSE and BSCE are reflected by differences in their in vitro PDs, differences that reflect distinctive electrical patterns. Specifically, the PD of ESSE is a product of a low rate of active transport combined with high electrical resistance whereas the PD of BSCE is a product of a high rate of active transport combined with low electrical resistance. For human ESSE, the present findings differ in some ways from those previously reported in preliminary form (Table 4). For instance, Pompa et al. (25), using a comparable technique (i.e., endoscopic biopsies mounted in mini-ussing chambers), found higher values for I sc and lower values for R T than reported here, whereas Orlando and Powell (21) found higher values for PD and R T and lower values for I sc. These differences likely reflect significant differences in both technique and patient selection; this is particularly true when comparing the results obtained using endoscopic biopsies in mini-ussing chambers with esophagectomy sections in standard Ussing chambers, an advantage going to the latter technique because of a greater tissue area-to-circumference ratio reducing the impact of edge damage on electrical readings. A bias in support of these higher values as being a more accurate reflection of these parameters in vivo is provided by the values reported for ESSE of rabbits in standard Ussing chambers, with I sc ranging from eq cm 2 h 1 and R T ranging from 1,500 2,800 cm 2 (23, 26). In contrast to human ESSE, the electrical values for I sc and R T reported here for BSCE are, to our knowledge, the first in manuscript form, and consequently there are no others for comparison. Nonetheless, BSCE bears morphological similarity to duodenum, including the presence of goblet cells (10), and these electrical findings for BSCE conform generally with those reported for human duodenum (Table 4). For instance, Larsen et al. (13) and Pratha et al. (27, 28), using a comparable technique, i.e., endoscopic biopsies of human duodenum mounted in mini-ussing chambers, reported values for I sc averaging 1 2 eq cm 2 h 1 and for R T cm 2. After basal electrical recordings, the ions responsible for active transport were sought in ESSE and BSCE by replacement of normal Ringer solution with either Na-free, Cl-free, or HCO 3 -free Ringer solution. The results showed that almost all of the small basal current in ESSE was due to active sodium absorption, a finding consistent with bidirectional sodium and chloride flux results reported for both human and rabbit ESSE (20, 21, 26). Interestingly, and unlike the prior results of Orlando and Powell (21), a small but significant (different from zero) amount of current compatible with anion (chloride/ bicarbonate) secretion was noted in human ESSE, a similar phenomenon being noted in at least one study in rabbit ESSE (20, 26). In contrast to the low current in human ESSE, basal current in BSCE was high and eightfold greater than that of ESSE. Furthermore, only one-half of this current in BSCE was due to sodium absorption, with the other one-half ( 0.8 eq cm 2 h 1 ) due to anion secretion. On the basis of Cl-free and HCO 3 -free experiments, anion secretion in BSCE was a reflection of both chloride and bicarbonate movement. Notably, addition of residual currents in Na-free, Cl-free, and HCO 3 -free for BSCE (but not for ESSE) exceeded total basal current (Tables 2 and 3), suggesting that the basal conductive sodium absorption and basal conductive secretion of chloride and bicarbonate noted in Table 3 do not occur simultaneously but vary depending on the conditions and stimulus as is known to occur in human duodenum (27, 28). In the present experiments, BSCE exhibited a chloride secretion that was twofold greater than bicarbonate secretion, yet given the high rate of basal secretion bicarbonate output in BSCE still exceeds that for ESSE greater than fivefold. Moreover, this higher level of bicarbonate secretion in BSCE is likely to be clinically significant, potentially representing a superior means for defense against acid injury to the esophagus. Support for this concept can be found in a preliminary report by Abdulnour-Nakhoul et al. (2). In this report ph microelectrodes were used to measure the lumen-to-surface ph gradient in vitro on endoscopic biopsies of BSCE and human ESSE mounted in modified mini-ussing chambers. When luminal ph was lowered with HCl to 3.5, BSCE was capable of maintaining a surface ph of 5.0 whereas that of ESSE could only maintain a surface ph of 4.0. Thus BSCE exhibited a 10-fold greater capacity than ESSE to neutralize acid as it backdiffuses toward epithelium, an observation consistent with its greater capacity for bicarbonate secretion noted above. Considerable data exist from human and animal studies on the nature of the ion transporters in mammalian ESSE. For instance, the apical membranes of squamous cells contain Na channels that are unusual in that they are both nonselective, accepting Na,K, and Li almost equally, and amiloride Table 4. Comparison of basal transmural PD, I sc, and R T of human ESSE and BSCE in the present report with that of prior publications Treatment n PD, mv RT, cm 2 Human ESSE Biopsy* Biopsy (Ref. 12) Esophagectomy (Ref. 13) Rabbit ESSE Surgical section (Ref. 8) , Human BSCE Biopsy* Human Duodenum Biopsy (Ref. 16) 48 NA Biopsy (Ref. 7) Values are means SE; n no. of tissues. *Data from Table 1 of present manuscript. NA, not available. Note that BSCE is compared with duodenum for lack of prior reports on BSCE.

6 ELECTRICAL PARAMETERS AND ION TRANSPORT IN HUMAN ESOPHAGEAL EPITHELIUM insensitive (4). These channels coupled with basolateral membrane Na-K-ATPase and K channels sensitive to barium, tetraethylammonium, and quinine are responsible for the major component ( 80 90% of I sc ) of active transport in ESSE, a predominantly Na-absorbing tissue (12). In addition to the Na-K-ATPase and K channels, the basolateral membranes of squamous cells contain two cotransporters, an acid-activated, bumetanide-sensitive NaK2Cl cotransporter that causes cell swelling at acidic intracellular ph and a volume-activated, R -butylindazone-sensitive, KCl cotransporter that mediates cell shrinkage during hypoosmolar stress (17, 34). Hypoosmolar stress also activates volume-sensitive barium- and quinineinhibitable K channels and volume-sensitive dihydro DIDSand indanyloxyacetic acid 94-inhibitable Cl channels (9, 17). These channels together with the KCl cotransporter account for the ability of squamous cells to undergo regulatory volume decrease for protection against persistent hypoosmolar stress (17, 31). The basolateral membrane of squamous cells also possess a flufenamate-sensitive Cl channel that is insensitive to SITS and DIDS (1) and three mechanisms for regulation of intracellular ph: an amiloride-sensitive Na/H exchanger; a DIDSsensitive, Na-dependent Cl/HCO 3 exchanger; and a DIDS-sensitive, Na-independent Cl/HCO 3 exchanger. The former two exchangers are cytoplasmic acid extruders and the latter a cytoplasmic alkali extruder that under conditions of low extracellular ph serves as the mechanism for pathological acid loading of squamous cells (14, 15, 35, 37, 38). Given the nature of the transporters and the fact that ESSE is predominantly Na absorbing (18, 23), it was not surprising then that in the present study neither ethoxyzolamide nor luminal or serosal SITS had an effect on I sc in ESSE. Nonetheless, the findings are of interest since ESSE is known to possess a variety of carbonic anhydrases (5, 6) and, as noted above, at least two types of Cl channels, one of which, the volume-sensitive Cl channel, is known to be sensitive to disulfonic stilbene derivatives (17). In contrast to ESSE, very little is known about the transporters in BSCE. However, data support, as in ESSE, the presence in BSCE of an amiloride-sensitive Na/H exchanger that serves as a cytoplasmic acid extruder (8) and a DIDSsensitive, Na-independent Cl/HCO 3 exchanger that serves as an acid loader at low extracellular ph (29). The former was identified both in biopsies containing BSCE and in TE7 cells, a Barrett s cancer cell line, and the latter was identified in SEG-1 cells, also a Barrett s cancer cell line. In neither study, however, was it clear whether the identified exchangers were localized to the apical and/or basolateral membrane. In the present study neither serosal ethoxyzolamide nor serosal SITS had an effect on the I sc in BSCE. This suggests that active anion secretion in BSCE is dependent neither on HCO 3 generated by carbonic anhydrase nor on Cl or HCO 3 transport through a basolateral membrane, SITS-sensitive pathway. In contrast to serosal SITS, however, there was a significant decline in I sc for BSCE with exposure to luminal SITS. This suggests that conductive anion (largely Cl) transport in BSCE occurs through an apical membrane, SITS-sensitive pathway consistent with a Cl channel. Furthermore, since the magnitude of the decline in I sc with luminal SITS (38%) was almost comparable to that of Cl-free solution (Fig. 3A), and both Cl and HCO 3 may share the same pathway (see discussion above), this apical Cl channel appears capable of transporting both Cl and, to a lesser extent, HCO 3. Interestingly, the presence in G269 BSCE of a prominent apical membrane, SITS-sensitive Cl channel makes it, from the perspective of the secreting digestive tract, more like colonic epithelium than that of small intestine (or gallbladder), whose most prominent apical Cl channel is the SITS-insensitive cystic fibrosis transmembrane regulator (3, 7, 16). In summary, this report establishes both the electrical basis for the differences in the in vivo PD for ESSE and BSCE and the nature of ions responsible for their active transport. In so doing, it demonstrates that BSCE has a far greater capacity for anion secretion than ESSE and that this process is dependent on an apical membrane, SITS-sensitive anion channel. Moreover, and as part of this process, BSCE has a greater capacity than ESSE for the secretion of bicarbonate. Since bicarbonate is an effective buffer for luminal acid or acid as it backdiffuses toward epithelium, the superior capacity of BSCE for bicarbonate secretion gives it a protective advantage over ESSE. Consequently, the replacement of reflux-damaged ESSE by BSCE in the distal esophagus, although unproven experimentally, can be viewed as a form of protective adaptation since BSCE is better suited than ESSE for defense of the esophagus against reflux disease. GRANTS This research was funded in part by National Institute of Diabetes and Digestive and Kidney Diseases Grants RO1 DK and 5R37DK REFERENCES 1. Abdulnour-Nakhoul S, Nakhoul NL, Caymaz-Bor C, Orlando RC. Chloride transport in rabbit esophageal epithelial cells. Am J Physiol Gastrointest Liver Physiol 282: G663 G675, Abdulnour-Nakhoul S, Orlando RC. Lumen-to-surface ph gradient in Barrett s esophagus. Gastroenterology 126: A235, Alrefai WA, Ramaswamy K, Dudeja PK. Mechanism(s) of chloride transport in human distal colonic apical membrane vesicles (Abstract). Dig Dis Sci 46: , Awayda MS, Bengrine A, Tobey NA, Stockand JD, Orlando RC. Nonselective cation transport in native esophageal epithelia. Am J Physiol Cell Physiol 287: C395 C402, Christie KN, Thomson C, Morley S, Anderson J, Hopwood D. Carbonic anhydrase is present in human oesophageal epithelium and submucosal glands. Histochem J 27: , Christie KN, Thomson C, Xue L, Lucocq JM, Hopwood D. Carbonic anhydrase isoenzymes I, II, III, and IV are present in human esophageal epithelium. J Histochem Cytochem 45: 35 40, Clarke LL, Grubb BR, Yankaskas JR, Cotton CU, McKenzie A, Boucher RC. Relationship of a non-cystic fibrosis transmembrane conductance regulator-mediated chloride conductance to organ-level disease in Cftr (-/-) mice. Proc Natl Acad Sci USA 91: , Fitzgerald RC, Omary MB, Triadafilopoulos G. Altered sodium-hydrogen exchange activity is a mechanism for acid-induced hyperproliferation in Barrett s esophagus. Am J Physiol Gastrointest Liver Physiol 275: G47 G55, Goldstein JL, Fogelson BG, Snow JC, Schmidt LN, Mozwecz H, Layden TJ. Rabbit esophageal cells possess K channels: effect of hyposmotic stress on channel activity. Gastroenterology 104: , Hammersen F. Sobotta/Hammersen: Histology, Color Atlas of Microscopic Anatomy. Baltimore, MD: Urban and Schwarzenberg, 1985, p Herlihy KJ, Orlando RC, Bryson JC, Bozymski EM, Carney CN, Powell DW. Barrett s esophagus: clinical, endoscopic, histologic, manometric, and electrical potential difference characteristics. Gastroenterology 86: , Khalbuss WE, Alkiek R, Marousis CG, Orlando RC. Potassium conductance in rabbit esophageal epithelium. Am J Physiol Gastrointest Liver Physiol 265: G28 G34, Larsen R, Mertz-Nielsen A, Hansen MB, Poulsen SS, Bindslev N. Novel modified Ussing chamber for the study of absorption and secretion in human endoscopic biopsies. Acta Physiol Scand 173: , 2001.

7 G270 ELECTRICAL PARAMETERS AND ION TRANSPORT IN HUMAN ESOPHAGEAL EPITHELIUM 14. Layden TJ, Agnone LM, Schmidt LN, Hakim B, Goldstein JL. Rabbit esophageal cells possess an Na,H antiport. Gastroenterology 99: , Layden TJ, Schmidt L, Agnone L, Lisitza P, Brewer J, Goldstein JL. Rabbit esophageal cell cytoplasmic ph regulation: role of Na -H antiport and Na -dependent HCO 3 transport systems. Am J Physiol Gastrointest Liver Physiol 263: G407 G413, Meyer G, Guizzardi F, Rodighiero S, Manfredi R, Saino S, Sironi C, Garavaglia ML, Bazzini C, Botta G, Portincasa P, Calamita G, Paulmichl M. Ion transport across the gallbladder epithelium. Curr Drug Targets Immune Endocr Metabol Disord 5: , Orlando GS, Tobey NA, Wang P, Abdulnour-Nakhoul S, Orlando RC. Regulatory volume decrease in human esophageal epithelial cells. Am J Physiol Gastrointest Liver Physiol 283: G932 G937, Orlando RC. Mucosal factors in Barrett s esophagus. In: Barrett s Esophagus and Esophageal Adenocarcinoma, edited by Sharma P and Sampliner R. Oxford, UK: Blackwell, Orlando RC. Pathogenesis of reflux esophagitis and Barrett s esophagus. Med Clin North Am 89: , vii, Orlando RC, Bryson JC, Powell DW. Effect of cigarette smoke on esophageal epithelium of the rabbit. Gastroenterology 91: , Orlando RC, Powell D. Studies of esophageal epithelial electrolyte transport and potential difference in man. In: Mechanisms of Mucosal Protection in the Upper Gastrointestinal Tract, edited by Allen A, Flemstrom G, Garner A, Silen W, and Turnberg L. New York: Raven, 1984, p Orlando RC, Powell DW, Bryson JC, Kinard HB 3rd, Carney CN, Jones JD, Bozymski EM. Esophageal potential difference measurements in esophageal disease. Gastroenterology 83: , Orlando RC, Powell DW, Carney CN. Pathophysiology of acute acid injury in rabbit esophageal epithelium. J Clin Invest 68: , Orlando RC, Powell DW, Croom RD, Berschneider HM, Boucher RC, Knowles MR. Colonic and esophageal transepithelial potential difference in cystic fibrosis. Gastroenterology 96: , Pompa DJSZ, Paulus P, Isenberg JI, Mittal R, Pratha VSK. Human esophageal epithelium in vitro: a novel technique using endoscopic biopsies (Abstract). Gastroenterology 122: A112, Powell DW, Morris SM, Boyd DD. Water and electrolyte transport by rabbit esophagus. Am J Physiol 229: , Pratha VS, Hogan DL, Martensson BA, Bernard J, Zhou R, Isenberg JI. Identification of transport abnormalities in duodenal mucosa and duodenal enterocytes from patients with cystic fibrosis. Gastroenterology 118: , Pratha VS, Thompson SM, Hogan DL, Paulus P, Dreilinger AD, Barrett KE, Isenberg JI. Utility of endoscopic biopsy samples to quantitate human duodenal ion transport. J Lab Clin Med 132: , Sarosi GA Jr, Jaiswal K, Herndon E, Lopez-Guzman C, Spechler SJ, Souza RF. Acid increases MAPK-mediated proliferation in Barrett s esophageal adenocarcinoma cells via intracellular acidification through a Cl /HCO 3 exchanger. Am J Physiol Gastrointest Liver Physiol 289: G991 G997, Scarpignato C, Micali B, Galmiche JP. Transmucosal potential difference as an index of esophageal mucosal integrity. Digestion 56, Suppl 1: 51 60, Snow JC, Goldstein JL, Schmidt LN, Lisitza P, Layden TJ. Rabbit esophageal cells show regulatory volume decrease: ionic basis and effect of ph. Gastroenterology 105: , Tobey NA, Argote CM, Kav T, Barlow W, Orlando RC. Comparison of electrical properties and ion transport in human esophageal stratified squamous epithelium and Barrett s esophagus (Abstract). Gastroenterology 126: A931, Tobey NA, Argote CM, Kav T, Vanegas X, Atug O, Sepmprun-Prieto L, Barlow W, Orlando RC. Anion transport in human squamous and Barrett s esophageal epithelium (Abstract). Gastroenterology 128: A234, Tobey NA, Cragoe EJ Jr, Orlando RC. HCl-induced cell edema in rabbit esophageal epithelium: a bumetanide-sensitive process. Gastroenterology 109: , Tobey NA, Koves G, Orlando RC. Human esophageal epithelial cells possess an Na /H exchanger for H extrusion. Am J Gastroenterol 93: , Tobey NA, Orlando RC. Mechanisms of acid injury to rabbit esophageal epithelium. Role of basolateral cell membrane acidification. Gastroenterology 101: , Tobey NA, Reddy SP, Keku TO, Cragoe EJ Jr, Orlando RC. Mechanisms of HCl-induced lowering of intracellular ph in rabbit esophageal epithelial cells. Gastroenterology 105: , Tobey NA, Reddy SP, Khalbuss WE, Silvers SM, Cragoe EJ Jr, Orlando RC. Na( )-dependent and -independent Cl-/HCO 3 exchangers in cultured rabbit esophageal epithelial cells. Gastroenterology 104: , Turner KS, Powell DW, Carney CN, Orlando RC, Bozymski EM. Transmural electrical potential difference in the mammalian esophagus in vivo. Gastroenterology 75: , 1978.