Marine teleosts can absorb imbibed seawater (SW) to maintain water balance, with esophageal desalination playing an essential role. NaCl absorption from luminal SW was enhanced 10-fold in the esophagus of SW-acclimated eels, and removal of Na+ or Cl− from luminal SW abolished the facilitated absorption, indicating coupled transport. Mucosal/serosal application of various blockers for Na+/Cl− transporters profoundly decreased the absorption. Among the transporter genes expressed in eel esophagus detected by RNA-seq, dimethyl amiloride-sensitive Na+/H+ exchanger (NHE3) and 4,4′-diisothiocyano-2,2′-disulfonic acid-sensitive Cl−/ exchanger (AE) coupled by the scaffolding protein on the apical membrane of epithelial cells, and ouabain-sensitive Na+-K+-ATPases (NKA1α1c and NKA3α) and diphenylamine-2-carboxylic acid-sensitive Cl− channel (CLCN2) on the basolateral membrane, may be responsible for enhanced transcellular NaCl transport because of their profound upregulation after SW acclimation. Upregulated carbonic anhydrase 2a (CA2a) supplies H+ and for activation of the coupled NHE and AE. Apical hydrochlorothiazide-sensitive Na+-Cl− cotransporters and basolateral Na+- cotransporter (NBCe1) and AE1 are other possible candidates. Concerning the low water permeability that is typically seen in marine teleost esophagus, downregulated aquaporin genes (aqp1a and aqp3) and upregulated claudin gene (cldn15a) are candidates for transcellular/paracellular route. In situ hybridization showed that these upregulated transporters and tight-junction protein genes were expressed in the absorptive columnar epithelial cells of eel esophagus. These results allow us to provide a full picture of the molecular mechanism of active desalination and low water permeability that are characteristic to marine teleost esophagus and gain deeper insights into the role of gastrointestinal tracts in SW acclimation.
- seawater acclimation
- epithelial transport
- digestive tracts
- teleost fish
marine teleosts are exposed to a constant threat of dehydration due to the high environmental osmotic pressure (33). To compensate for the unavoidable osmotic water loss, they drink a large amount of seawater (SW) (20, 42, 43) and can absorb most of this water from the hypertonic SW by the action of the digestive tract (17, 26, 40, 48). This is in contrast to worsened dehydration after SW intake in most terrestrial vertebrates (42). To achieve this absorption, the esophagus of marine teleosts can remove NaCl specifically from SW with minimal water loss. This unique phenomenon is called esophageal desalination, which is known to contribute greatly to the efficient absorption of water by the distal digestive tract. In fact, NaCl absorption across the esophagus is enhanced, while water permeability is suppressed after SW acclimation of euryhaline teleosts (21, 27, 38). The desalination decreases osmotic pressure of ingested SW to half when it enters the stomach, which facilitates water absorption with NaCl in the intestine (40). The excess NaCl absorbed is excreted actively by the gill ionocytes (22, 23).
The molecular mechanisms of NaCl absorption have been well documented in the intestine of marine teleosts because of its importance in SW adaptation (17). In the intestine of SW-acclimated eels, Anguilla spp. and flounder, Pseudopleuronectes americanus, water moves in parallel with Na+ and Cl− absorption via Na+-K+-2 Cl− cotransporter 2 (NKCC2) and/or Na+-Cl− cotransporter 1 (NCC1) on the apical membrane of epithelial cells (10, 11, for nomenclature of NCC see Ref. 36) and via Na+-K+-ATPase (NKA) and Cl− channel (CLC) on the basolateral membrane. Effluxes of Cl− into blood via CLC and K+ back into the lumen via K+ channel produce a serosa-negative transepithelial potential that is unique to marine teleost intestine (2, 32, 39). However, molecular identity of NKA, CLC, and K+ channel has not yet been determined (17, 48). In marine teleosts, Cl−/ anion exchanger (AE), which may be coupled with Na+/H+ exchanger (NHE), on the apical membrane is also involved in Cl− absorption and excretion, the latter of which precipitates Ca/MgCO3 to decrease luminal fluid osmolality for further water absorption (16, 28). In contrast to the intestine, however, transporter molecules involved in esophageal desalination are almost unknown.
There have been two pharmacological studies to assess transporters involved in esophageal desalination in euryhaline teleosts (37, 38). In these in vitro studies, serosal application of ouabain (a NKA blocker) and mucosal application of amiloride (an epithelial Na channel (ENaC) blocker) decreased both Na+ and Cl− fluxes by 60–70% and 30–40%, respectively, when the mucosal side was SW and the serosal side was a Ringer solution. Mucosal bumetanide (NKCC blocker) was also effective in SW-acclimated eels but not in flounder. An AE blocker (DIDS) and an NCC blocker (hydrochlorothiazide, HTCZ) also decreased NaCl fluxes by 30% in SW-acclimated eels. However, molecular identity of the transporters remains undetermined. Recently, possible involvement of NHE2 in the desalination has been suggested in the toadfish esophagus (15), in which amiloride was ineffective but EIPA, a more NHE-specific amiloride derivative, inhibited Na+ uptake by 30%.
In the present study, we attempted to identify transporter molecules responsible for esophageal desalination in SW-acclimated eels, Anguilla japonica. In addition, the molecular basis of low water permeability in SW eel esophagus was examined in both paracellular and transcellular pathways. We initially confirmed the basic characteristics of esophageal desalination using sac preparations from SW-acclimated eels. As only Na+ and Cl−, but not Mg2+, Ca2+, and , which exist at high concentrations in SW, were transported in the sac, various blockers for Na+ and/or Cl− transporters were applied to the mucosal or serosal side to narrow the list of candidate transporters. Then, to further limit the candidates, we performed a transcriptomic analysis (RNA-seq) to identify the genes that are expressed in the esophagus and upregulated or downregulated after SW acclimation. Changes in the expression levels were confirmed by real-time quantitative PCR (qPCR) using gene-specific primers. In situ hybridization was performed to further select the genes expressed in the epithelial cells. Some cDNAs were cloned to increase the accuracy of qPCR and in situ hybridization, because teleosts have more paralogs of highly similar sequences due to an additional whole genome duplication (10, 11, 44, 49, 50). Finally, we suggest a new molecular mechanism of esophageal desalination and low water permeability via both transcellular and paracellular pathway using euryhaline eels as a model.
MATERIALS AND METHODS
Animals and Drugs
Pond-raised Japanese eels (Anguilla japonica) weighing around 200 g were purchased from a local dealer and acclimated in freshwater (FW) tanks in the aquarium room of the Atmosphere and Ocean Research Institute for 1 wk at 18°C. Some eels were then transferred to SW tanks and kept for 1–4 wk before the experiment, as fish fully acclimated to SW in 1 wk. Eels were not fed after purchase. All experiments were approved by the Animal Experiment Committee of the University of Tokyo and were performed according to the guidelines prepared by the committee.
Bumetanide (Daiichi-Sankyo, Tokyo, Japan), ouabain (Merck, Kenilworth, NJ), DIDS (Sigma-Aldrich, Tokyo, Japan), dimethyl amiloride (DMA; Sigma-Aldrich), diphenylamine-2-carboxylic acid (DPC; Wako Pure Chemical Industries, Tokyo, Japan), HCTZ (Sigma-Aldrich) were obtained from commercial sources. DPC and bumetanide were dissolved in ethanol, and the final concentration of ethanol was diluted by eel Ringer to <1%. HCTZ was dissolved in DMSO (Sigma-Aldrich) at 1 M, and the final concentration of DMSO was <0.1%. These solvents (1% ethanol or 0.1% DMSO) alone had no effects on the NaCl uptake in the esophageal sac. DIDS and ouabain were dissolved in eel Ringer solution.
Esophageal Sac Experiments
After decapitation of eels (n = 38), the whole length of esophagus was quickly excised, and the longitudinal and circular muscle layers on the serosal side were carefully stripped off as described for the intestine (3). The anterior side of esophagus was tied with a silk suture around the posterior half of 2.5 ml polyethylene syringe (ID: 9 mm, Terumo, Tokyo, Japan), and the posterior end was closed with a silk suture. The sac was set to a 5-ml test tube containing 3 ml of eel Ringer (Table 1), and 1 ml SW was added to the lumen. The height of luminal fluid was adjusted to that of the outside Ringer. The outside Ringer was continuously aerated by 95% O2 and 5% CO2, as used previously (21, 37). Although lower CO2 concentrations have been used for aeration in other teleost esophagus (15, 38), electrophysiological properties were not different from those with aeration of lower CO2 concentrations in the eel intestine (3). The area of the esophageal lumen was calculated from the luminal fluid volume (1 ml) and its height after the experiment.
Initially, the difference in the ability of desalination was compared between FW and SW eel esophagus. For this purpose, 140 μl of outside Ringer was collected every 15 min, and Na+, K+, Mg2+, Ca2+ concentrations were measured by atomic absorption spectrophotometer (Z5300, Hitachi, Tokyo, Japan) and Cl– concentration by digital chloridometer (Labconco, Kansas City, MO). The volume was replaced after each collection. As K+, Mg2+, and Ca2+ uptake by the esophagus were minimal and unchanged after SW acclimation (data not shown), we examined only the effects on Na+ and Cl− uptake after alteration of luminal Na+ or Cl− concentration (Table 1) using SW eel esophagus. In this experiment, 140 μl of external medium was collected every 10 min.
Transcriptomic (RNA-seq) Analyses
The esophagus was isolated as above from a FW- or SW-acclimated eel and snap frozen in liquid nitrogen. Total RNA was extracted from the esophagus using Isogen (Nippon Gene, Toyama, Japan), according to the manufacturer’s protocols. RNA quality was monitored using an Agilent 2100 bioanalyzer system (Agilent Technology, Santa Clara, CA), and only RNA samples with RNA integrity number >7.0 were used for sequencing. The cDNA libraries were prepared from the RNA samples using TruSeq RNA Sample Preparation v2 and were sequenced by Illumina HiSeq 2500 (Illumina, San Diego, CA).
The sequenced reads were mapped to the Japanese eel genome (19) using TopHat (version 2.0.9). We selected read pairs that were properly and uniquely mapped using samtools (version 1.2) “view&” command with the options “-q 50 -f 0x2”. HTSeq (version 0.5.3p9) was used to count the number of reads for each gene model with parameters “-m union -i gene_id–stranded=no”. Before normalization, genes with low read counts were removed if there were less than 10 reads in each sample. Read counts were normalized by DEGES/EdgeR methods using an R package “TCC” (version 1.8.5). The normalized read counts were then used to search for differentially expressed genes between the FW and SW conditions using the EdgeR package (version 3.10.5) with a false discovery rate of <0.05. From the list of upregulated or downregulated genes, the ion transporters, water channels, and membrane proteins were manually short-listed for real-time PCR analysis, according to their known and potential functions in relation to the changes in transporting properties. The short-listed genes were manually curated to check for possible isoforms from the eel genome before real-time PCR analysis.
Quantitative Analyses of Gene Expression
The esophagi of FW and SW eels (n = 6 for each) were dissected out and frozen as above. Total RNA was extracted with Isogen after homogenization by Micro Smash MS100 (Tomy, Tokyo, Japan). After treatment with DNase I (Life Technologies, Carlsbad, CA) to remove genomic DNA contamination, 1 μg of RNA was reverse-transcribed with Iscript cDNA synthesis kit (Bio-Rad, Hercules, CA), according to the manufacturer’s protocols. Real-time qPCR was performed using Kappa SYBR 2× PCR mix (Kappa Biosystems, Wilmington, MA) and ABI 7900HT Fast real-time PCR system (Life Technologies). The amplification of a single amplicon was confirmed by analyzing the melting curve after the real-time cycling. Elongation factor 1α (eef1a) was used as an internal control to normalize the gene expressions among different samples. We also measured the expression of other transporter genes known to be upregulated in SW eel intestine to confirm the reliability of the data. Primer sequences are listed in Table 2.
In Situ Hybridization
In situ hybridization was performed for the candidate genes using both sense and antisense probes. The esophagi of FW and SW eels were fixed in 4% paraformaldehyde in 0.1 M phosphate buffer at pH 7.4 for 1 day at 4°C. Tissues were dehydrated, embedded in paraplast (Leica Microsystems, Wetzlar, Germany), and cut at 5 μm. Sections were deparaffinized, rehydrated, treated with proteinase K (5 μg/ml) for 10 min, postfixed, and equilibrated in hybridization buffer (5× SSC, 50% formamide) at 58°C for 2 h. A partial sequence of each target gene was cloned on the PGEM-T Easy vector (Promega, Madison, WI) using primer information listed in Table 2. Sense and antisense probes were prepared using a digoxigenin (DIG) RNA labeling kit (Roche Applied Science, Indianapolis, IN), diluted in hybridization buffer containing calf thymus DNA (40 μg/ml), and denatured at 85°C for 10 min. Denatured RNA probes were spread on the sections and incubated at 58°C for >40 h depending on the expression level in a moist chamber. Specific signals were developed using a DIG nucleic acid detection kit (Roche Applied Science), according to the manufacturer’s protocol. Some sections were stained with hematoxylin and eosin or Periodic acid–Schiff to know the basic structure of epithelial cells.
Statistical analyses of data obtained by the sac experiment were performed by two-way ANOVA for time course data and by paired t-test for inhibitor experiments using KyPlot (Ver. 5.0, Kyens Laboratory, Tokyo, Japan). The data of qPCR were analyzed between FW- and SW-acclimated fish by Student's t-test, and Mann-Whitney U-test was applied when normal distribution was not demonstrated (GraphPad Prism, ver. 5 for Windows, San Diego, CA). All results are given as means ± SE and are considered significant at P < 0.05.
Basic Characteristics of Esophageal Desalination
In the esophageal sac preparation, Na+ and Cl− uptake from luminal SW were low in FW eels (Fig. 1A), but the uptake of these ions was enhanced >10-fold in SW eels (Fig. 1B). A slight NaCl uptake occurred in SW eel esophagus when isotonic Ringer was on both mucosal and serosal sides, suggesting active uptake (Fig. 1C), and it increased equally when luminal NaCl concentration was increased to the SW level (Fig. 1D). However, this uptake was lower when only Na+ or Cl− concentration was increased to the SW level (Fig. 1, E and F) using Na gluconate or choline Cl (Table 1).
Effects of Blockers on NaCl Uptake
Among the transporter blockers examined, mucosal application of HCTZ, DMA, or DIDS profoundly decreased both Na+ and Cl− uptake equally, suggesting an involvement of NCC, NHE, and AE on the apical membrane in NaCl uptake (Fig. 2, A–C). The NaCl uptake recovered after removal of the blockers from luminal SW. Mucosal DPC (Fig. 2D) and bumetanide at 10−5 M (Fig. 2E) were without effect, but 10−4 M of bumetanide inhibited Na+ uptake significantly (Fig. 2F).
Among serosal blockers, DPC and ouabain decreased NaCl uptake significantly, but the pattern of inhibition was different (Fig. 2, G and H); the uptake recovered after removal of DPC from the serosal fluid, but the effect of ouabain was greater after its removal.
Identification of Transporter Genes by RNA-seq
Among approximately 165 million reads sequenced from cDNAs of FW and SW eel esophagus, 78% were mapped to the 9,308 genes annotated on the basis of the medaka genome database (27, 28), from which candidate transporter proteins for esophageal desalination were sought. As possible targets of blockers, several gene families were identified; DIDS-sensitive SLC4 and SLC26 families, DMA-sensitive SLC9 family, HTCZ and/or bumetanide-sensitive SLC12 family, and DPC-sensitive CLCN family. Some of the genes were expressed in substantial amounts and upregulated in SW eel esophagus (Table 3), which were subjected to further analyses by qPCR. Concerning AE and NHE, their scaffolding/regulatory proteins (NHE regulatory protein, NHERF) were also expressed in the esophagus (Table 3). Carbonic anhydrases (CAs) that activate AE and NHE were also expressed in the esophagus. In addition, aquaporins (AQPs) and cell adhesion molecules (claudins, CLDNs), which may be involved in water permeability, were also found in the RNA-seq data of eel esophagus (Table 3).
Comparison of expression levels between FW and SW eels by qPCR.
The candidate genes obtained by RNA-seq were blasted to obtain all paralogs, and paralog cDNAs were cloned to use gene-specific primers when necessary. Among candidates of HCTZ-sensitive transporters, expression of the NCC genes (slc12a3 and 10) was low, but slc12a8 (cation-chloride cotransporter 9, CCC9) was expressed substantially (Fig. 3, A–D). Among DIDS-sensitive AEs, expression of SLC26 family genes was low, and some decreased after SW acclimation (Fig. 3, E–G). In the SLC4 family, slc4a1s were also low or downregulated in SW eel esophagus despite coexpression of its adaptor protein (SLC4A1AP) (Fig. 3, H, I, L), but slc4a2 mRNAs were in significant amounts in SW eel esophagus (Fig. 3, J and K). Concerning DMA-sensitive NHEs, nhe1 and nhe2 were expressed in the esophagus, but the expression was not upregulated in SW eels (Fig. 3, M and N). However, expression of nhe3 was profoundly upregulated in SW eels, and its scaffolding protein NHERF1, which combines NHE3 with AE, was also found (Fig. 3, O and P).
All five ouabain-sensitive NKA genes were expressed in eel esophagus, of which nkaα1c1 and nkaα3 were strongly upregulated in SW eels (Fig. 4, A–D). Another gene, nkaα2, was downregulated in SW eel esophagus (data not shown). Among DPC-sensitive CLCs, clcn1 and clcn2 were expressed in the esophagus, and clcn2 was greatly upregulated in SW eels (Fig. 4E). In addition, the K+-Cl− cotransporter 1 gene (slc12a4, kcc1) was expressed substantially (Fig. 4G). Carbonic anhydrase provides H+ and for activation of coupled NHE and AE for NaCl uptake. Two intracellular CA2 paralogs and an extracellular CA4 gene were expressed in the esophagus, of which ca2a was most abundant and apparently upregulated in SW eels (Fig. 4, H–J). Both apical aqp1a and basolateral aqp3 were identified in the esophagus and both genes were apparently downregulated in SW eels (Fig. 4, K and L). Among CLDN genes detected in the esophagus (Table 3), cldn15a was profoundly upregulated in SW eels (Fig. 4M).
Cellular localization of transporter mRNA.
The SW eel esophagus was regionally differentiated; the anterior region had abundant mucus cells stained by Periodic acid-Schiff staining, while the middle and posterior regions were rich in columnar cells with apical microvilli (Fig. 5). In situ hybridization signals of nkas, nhe3, ca2a, clcn2, and ae2, which were expressed substantially in the eel esophagus and/or upregulated after SW acclimation, were observed in the epithelial cells but not in other cells in the lamina propria, submucosa, muscle layers, and subserosa of the esophagus (Fig. 6). In the epithelial cells, the genes were expressed in the columnar cells but not in the mucus cells (Fig. 6). The intensities of the signals were generally strongest in the middle region, particularly that of nhe3, whose signal was faint in other regions. On the other hand, ae2 was expressed in both FW and SW eels, and the signals were found more ubiquitously in different esophageal segments. The sense probes did not show any signals (data not shown).
The blood of marine teleosts contacts environmental SW only across thin respiratory epithelia, and thus, these animals are constantly exposed to the threat of osmotic dehydration (33). They can gain water from food (preformed water) and by oxidation of nutrients (metabolic water), but marine teleosts obtain water mainly by oral drinking as is the case of terrestrial animals (42, 43). In contrast to terrestrial animals that lose water by drinking SW, marine teleosts can gain water by absorbing more than 80% of ingested SW by the intestine. This exceptional ability is achieved by removal of monovalent ions (Na+ and Cl−) by the esophagus, removal of divalent cations (Ca2+ and Mg2+) as carbonate precipitates by active intestinal secretion of , and facilitated absorption of water in parallel with monovalent ions by the intestine (17). The excessive monovalent ions that entered the body are excreted by the gill ionocytes (22, 23) and divalent ions by the kidney (7). Among these processes, esophageal desalination is a unique feature that enables effective water absorption by the intestine of marine teleost fishes (21, 27, 38). In the present study, we have delineated a complete picture of the molecular mechanisms of esophageal desalination in SW eels using physiological, molecular biological, and histological studies. Molecular mechanisms for NaCl absorption have been investigated intensively in marine teleost intestine with apical NKCC2 and K+ channel and basolateral NKA and CLC playing a major role (17). However, only NKCC2 was molecularly identified (3, 10), while molecular subtypes of K+ channel, CLC, and NKA have not been identified yet even in the intestine.
The sac experiment showed that NaCl uptake by eel esophagus is enhanced after SW acclimation and that the two ions are strongly coupled, as removal of either Na+ or Cl− from luminal fluid abolished the enhanced absorption. Thus a coupled Na+ and Cl− transport system has developed in the epithelial cells of SW eel esophagus. The NaCl transport appears to be mostly transcellular, as various transporter blockers profoundly decreased the transport when applied to mucosal or serosal side. However, a small portion (<20%) may be paracellular, as NKA blockade by ouabain, which abolishes the energy supply for transcellular NaCl transport, did not abolish the transport completely. Similar results have been reported in SW-acclimated eels (21, 27, 37) and marine flounder (38). These studies also found that esophageal epithelia have high transepithelial resistance and low water permeability.
To pursue candidate transporters for desalination, various blockers were applied to either mucosal or serosal side of the epithelia. Among those examined, mucosal HCTZ (NCC blocker), DMA (NHE blocker), DIDS (AE blocker), serosal ouabain (NKA blocker), and DPC (CLC blocker) were effective to decrease both Na+ and Cl− uptake from the luminal SW. However, mucosal DPC and bumetanide (NKCC blocker) at 10−5 M had no significant effects. This dose of bumetanide completely inhibited NKCC2 activity in SW eel intestine (3). The inhibition by mucosal bumetanide at a higher dose may be due to the nonspecific inhibition such as NCC (37). These results strongly suggest that mucosal NCC and coupled NHE and AE, and serosal NKA and CLC are responsible for the coupled Na+ and Cl– transport through the epithelial cells of SW eel esophagus. The inhibition of NaCl uptake by mucosal HCTZ, amiloride (ENaC blocker), DIDS, and bumetanide was reported previously in SW eel esophagus (37), by mucosal amiloride in the flounder (38), and by mucosal ethylisopropyl amiloride (EIPA, NHE blocker) in the toadfish (15). ENaC ortholog appears to be absent in teleosts but acid-sensing ion channel (ASIC) that belongs to the ENaC/degenerin family is involved in Na+ uptake by the gill of rainbow trout (14). However, amiloride does not seem to be an effective blocker for ASIC, and ASIC does not seem to be expressed in the eel esophagus, as shown by the RNA-seq of this study. In addition, mucosal furosemide (NKCC blocker) was ineffective in the flounder (38) and mucosal amiloride, furosemide, and thiazide (NCC blocker) were ineffective in the toadfish (15). Thus, transporters involved in esophageal desalination could be diverse among teleost species.
Molecular Biological Studies
To identify all candidate transporter genes expressed in the eel esophagus, transcriptomic analyses (RNA-seq) were performed. The candidates were narrowed down by the following criteria: target of blockers in physiological studies, sufficient amounts of expression as functional transporters, and upregulation after SW acclimation. The expression of the selected genes was compared between FW and SW eels by qPCR. Detailed analyses indicated that RNA-seq could not distinguish the paralogous genes with high-sequence identity, such as NKA α-subunit subtypes (50). Thus, we cloned paralogous cDNAs in the eel for accurate qPCR and in situ hybridization, and only unique sequences were used for quantification. Annotation of the paralogous genes was determined by comparison with those of medaka (24, 51). The selected genes were finally subjected to in situ hybridization to confirm their expression in the absorptive epithelial cells. In addition to the NaCl transporters, AQPs and cell adhesion molecules, such as CLDN, were sought to assess their involvement in low water permeability of SW eel esophagus using the same criteria used for NaCl transport (except for downregulation of AQPs). The candidate molecules involved in NaCl and water transport are shown in Fig. 7 with their location in the epithelial cells, which is discussed in detail below.
NaCl transporters on the apical membrane.
As target of DMA, nhe3 that was upregulated profoundly after SW acclimation is the best candidate. The nherf1 may be a scaffold for NHE3, as it is expressed significantly in eel esophagus and is necessary for the regulation of NHE3 activity (5, 13). NHERF1 has two PDZ binding sites, and NHE3 binds one of them with its PDZ domain (30). As nhe2 was also expressed considerably in both FW and SW eels, NHE2 may have a housekeeping role in Na+ uptake, while NHE3 is involved in facilitated desalination in SW. Such differential roles of NHE2 and NHE3 have been reported for NaCl absorption in the murine intestine (6).
The DIDS-sensitive AE coupled with NHE3 for NaCl uptake may be SLC26A3 (downregulated in adenoma, DRA), although its gene expression was low in SW eel esophagus. In the murine small intestine, DRA is a preferred partner of NHE3 (46), and DRA is advantageous for Cl− absorption as it absorbs two Cl− in exchange of one (25). Another possible partner, SLC26A6, takes up one Cl− in exchange of two , and its gene expression was low and downregulated in SW eel esophagus. In addition, NHE2 and AE2 are another possible combination for persistent NaCl absorption on the mucosal membrane because of their high levels of expression in both FW and SW eels. AE2 is preferentially localized on the basolateral membrane of epithelial cells, but plasma membrane targeting is flexible and dependent on the binding to cytoskeleton-related proteins (9). The activity of these coupled transporters may be enhanced by the supply of H+ and by cytosolic CA2, as shown in the teleost intestine (17). In this study, ca2a was apparently upregulated in SW eel esophagus.
As a target of HCTZ, NCC1 (SLC12A3) and teleost-specific paralog named NCC2 (SLC12A10) are likely candidates because NCC incorporates Cl− and Na+ together into the cell at the apical membrane (1). However, the expression of these genes was low and rather suppressed in SW eels. Despite the low expression, eel NCC proteins might be sufficient for significant NaCl uptake. Another candidate is CCC9 (SLC12A8), as the gene was expressed significantly in SW eel esophagus. CCC9 exists widely in vertebrates, but its phylogenetic position is most diverged from other SLC12 family members (18). Furthermore, a splice variant of human CCC9 transported polyamines and amino acids (12). However, it is yet possible that eel CCC9 is a target of HCTZ and transports Na+ and Cl− together in teleosts.
NaCl transporters on the serosal side.
Na+ and Cl− taken up into the cells seem to be transported into the serosal side by the combination of ouabain-sensitive NKA and DPC-sensitive CLC in SW eel esophagus. In fact, two NKA genes (nkaα1c1 and nkaα3) and a CLCN gene (clcn2) were profoundly upregulated after SW acclimation (Fig. 7). This is the first identification of functional NKA and CLC subtypes in the gastrointestinal tracts of teleosts. NKA subtypes are highly divergent in teleosts, and eels possess three NKAα1c subtypes and one NKAα2 and NKAα3 but lack NKAα1a and NKAα1b (50). Other nkas and clcn3 expressed in the esophagus may have a housekeeping role in NaCl absorption in both FW and SW eels. In zebrafish gill ionocyte, apical NCC and basolateral CLCN2 and NKA have been implicated in NaCl absorption from environmental FW (47). Three CLCN2 subtypes have been identified in zebrafish, but we identified only one CLCN2 in the eel genome database.
For proper functioning of NKA, the KCC1 gene expressed significantly in eel esophagus may be responsible for K+ recycling in addition to Cl− absorption to the serosal side (Fig. 7). It is possible that the inhibitory effect of serosal DPC on NaCl absorption is not only on CLCN2 but also on KCC1 because NCC1 that belong to the same SLC12 family was sensitive to 10−3 M DPC in the SW eel intestine (3). In addition, as the Na+- cotransporter gene (nbce1a) was expressed appreciably in the eel esophagus (data not shown), NBCe1a (SLC4A4a) might be coupled with AE1 (SLC4A1) and transport Na+ and Cl− into the circulation with recycling of at the basolateral membrane. The adaptor protein gene for AE1 (slc4a4ap) was also expressed in the eel esophagus.
Proteins related to water permeability.
Another unique feature of SW eel esophagus is low water permeability to protect from osmotic water loss when ingested SW enters the esophagus (21). This is another important cue for marine teleosts to survive in hyperosmotic SW. In mammals, water is osmotically lost by the anterior digestive tracts if they drink SW (42). In this study, we found that aqp1a and aqp3 were expressed in the eel esophagus, and both were downregulated after SW acclimation (Fig. 7). AQP1 was localized in the mucosal membrane in the eel intestine and upregulated after SW acclimation, although it is not known whether it was AQP1a or AQP1b (4). Unlike current data, aqp1a expression was upregulated after administration of cortisol, a SW-adapting hormone, in the eel esophagus (34). The apparent difference may be due to the different condition of the eels used, silver European eels ready for downstream migration (34) and cultured Japanese eels (current study), since cortisol was ineffective in yellow European eels in FW. The aqp3 expression was detected in the digestive tracts of European eels, but the signal was not detected in the epithelial cells involved in ion and water transport (31). In mammals, AQP3 was localized on the basolateral membrane of epithelial cells in the digestive tracts (53).
In addition to the transcellular pathway, water movement across epithelia is regulated intercellularly (between cells) by tight junction proteins, such as CLDN in teleosts (29). Teleosts have more than 50 CLDN genes (41), of which cldn15a,b and cldn25 are predominantly expressed in the intestine of medaka (8) and Atlantic salmon (45). Nine CLDN genes were expressed significantly in the eel esophagus, as detected by RNA-seq, of which cldn15a was most abundant, and its expression was profoundly upregulated after SW acclimation (Fig. 4M). The cldn15b was also expressed in the eel esophagus, and its expression was unaltered after SW acclimation (data not shown). Thus, it is likely that CLDN15A is involved in low paracellular water permeability in SW eel esophagus (Fig. 7). In medaka intestine, cldn15a expression did not differ between FW and SW fish, but cldn15b expression was downregulated in SW fish (41).
The luminal surface of FW eel esophagus has longitudinal, straight folds, which underwent marked morphological changes to higher and meandering folds after SW acclimation (52). These morphological changes certainly increase the surface area to absorb ions when distended by active drinking of SW. Esophageal epithelia of FW eels are occupied by mucous cells throughout the whole length, but regional differences in cell types become apparent after SW acclimation; the anterior part, named mucous esophagus, is a thick and stratified epithelium consisting of mucus and microridge cells just like FW eel esophagus, and the mucous cells are gradually replaced by a layer of mitochondrion-rich columnar cells with microvilli on the apical surface along the anterior-posterior axis (35). In this study, columnar cells are prominent in the middle to posterior regions of the SW eel esophagus. Most of the NaCl transporter genes upregulated in SW eels were found in the columnar cells by in situ hybridization. As a sphincter exists between the esophagus and stomach, ingested SW may be stored for some time in the esophagus for active desalination in the middle/posterior part. The lateral membrane of adjacent columnar cells in the posterior esophagus was interdigitated to form intricate contacts in SW eel esophagus (35). This may be morphological evidence for reduced osmotic water loss between columnar cells, as evidenced by the upregulated CLDN15A.
Perspectives and Significance
The significance of this study is that we could depict a whole scheme for the molecular basis of active desalination (NaCl absorption) and low water permeability of SW-acclimated eel esophagus, which are characteristic in the marine teleosts and are essential for survival in the hyperosmotic SW environment. Without this unique mechanism in the esophagus, marine teleosts would undoubtedly lose water after SW ingestion and would not be able to absorb water in the intestine. Even in the well-studied intestine, where water is absorbed to counter dehydration, transporter/cell adhesion molecules have not been identified yet at the paralog level (17). As we have transcriptome (RNA-seq) data of the eel intestine during the course of SW acclimation, we may be able to identify such molecules in the intestine, which are upregulated after SW acclimation and are important for ion and water absorption for survival in SW, as reported previously, in the medaka intestine (51). As functional molecules responsible for osmoregulation are highly diversified among teleost species (44), molecular mechanisms for esophageal desalination need to be investigated in different teleost species to identify the essential mechanism for euryhalinity and/or SW acclimation.
This work was supported by the Grant-in-Aid for Scientific Research (A) from the Japan Society for the Promotion of Science to Y. Takei (23247010).
No conflicts of interest, financial or otherwise, are declared by the authors.
Y.T., W.I., and M.K. conceived and designed research; Y.T., M.K.-S.W., H.O., Y.S., and M.K. analyzed data; Y.T., M.K.-S.W., H.O., and W.I. interpreted results of experiments; Y.T., M.K.-S.W., and S.P. prepared figures; Y.T. drafted manuscript; Y.T. and W.I. edited and revised manuscript; Y.T., M.K.-S.W., S.P., H.O., Y.S., W.I., and M.K. approved final version of manuscript; M.K.-S.W., S.P., H.O., Y.S., and M.K. performed experiments.
We thank Kanako Taguchi for help in the physiological studies and Dr. Masaaki Ando of Atmosphere and Ocean Research Institute for guidance. We also thank Dr. Christopher A. Loretz of State University of New York at Buffalo for valuable comments on the manuscript and polishing the English.
- Copyright © 2017 the American Physiological Society