Proc. Nat. Acad. Sci. USA Vol. 73, No. 4, pp. 1348-1350, April 1976 Physiology Eel esophagus as an osmoregulatory organ (ion and water permeabilities/dilution of sea water/seawater adaptation) TETSUYA HIRANO*t AND NICOLE MAYER-GOSTANt * Ocean Research Institute, University of Tokyo, Nakano, Tokyo 164, Japan, and * Groupe de Biologie Marine, DMpartement de Biologie, Cemmissariat a 1'Energie Atomique, 06230 Villefranche-sur-Mer, France Communicated by Howard A. Bern, January 19,1976

ABSTRACT Ion and water permeabilities were mea- After decapitation, about two-thirds of the esophagus (2-3 sured in the isolated esophagus of the eel (Anguilla anguilla cm in length) and the entire stomach and the intestine were and A. japonica), and compared with those in the stomach and the intestine. The freshwater eel esophagus was imper- isolated and washed by flushing with modified Krebs-Ringer meable both to Na+ and Cl- ions and to water, whereas per- bicarbonate (Ringer) solution with the following concentra- meabilities to the ions increased selectively after seawater tions: 127 mM NaCl, 3.0 mM KC1, 2.5 mM CaCl2, 3 mM adaptation. The ion permeabilities of both the freshwater MgSO4, 1.25 mM KH2PO4, 25 mM NaHCO3, gassed with and the seawater eel stomach were lower than in the seawa- 95% 02-5% C02. The intestine was not used in A. anguilla. ter eel esophagus, although water permeability was greater After equilibration for 30 min with Ringer solution, the gut than in the esophagus. Sea water enclosed in the lumen was filled with diluted three times more efficiently in the seawater eel regions were tied at one end with a cotton thread, esophagus than in the stomach. The intestinal permeabilities Ringer solution, and tied at the other end. Care was taken to were greater than those of the esophagus and the stomach, avoid excess distension of the sac. The filled sac was lightly and increased after seawater adaptation. In the eel, ingested blotted on filter paper, weighed, and transferred to an incu- sea water seems to be diluted mainly in the esophagus by bation vessel containing 50 ml of Ringer solution. The vessel passive diffusion of the ions into the blood without addition was gassed with 95% 02-5% CO2 and incubated with shak- of water. After further but less important dilution in the weighing the stomach with salt removal and with water addition, the ing at 20°. The final volume was obtained by water is absorbed by the intestine, following active absorp- sac before and after emptying. Each sac was then washed tion of the ions. Thus the eel in sea water is able to replace and filled with sea water (artificial sea water, Jamarin: 440 water lost osmotically by drinking hypertonic sea water. meq/liter of Na+, 516 meq/liter of Cl- or Villefranche Bay sea water: 540 meq/liter of Na+, 600 meq/liter of Cl-) and It is well accepted that marine teleosts drink sea water to re- incubated in Ringer solution. The incubation time was 60 place water lost osmotically across the body surface. The min for the esophagus and the stomach, while the intestine common view is that swallowed sea water, which is hyperos- was incubated for 30 min to keep the change in concentra- motic to the body fluid, is diluted in the stomach and later in tion gradient minimal. the intestine by osmotic influx of water from blood to The surface area of the sac was measured after cutting the lumen, and that the water is then absorbed from the intes- segment along its long axis and spreading on graph paper. tine by some mechanism dependent on the active uptake of Net movements of Na+ and Cl- ions were calculated for monovalent ions into the blood (1-3). each incubation from the difference between the initial and Unlike most of the teleost fishes, the eel has a gastrointes- the final concentrations of the ions. Sodium and chloride tinal tract that is anatomically divided into three distinct re- concentrations were estimated using a Hitachi 203 atomic gions: esophagus, stomach, and intestine (4). When the absorption spectrophotometer or an Eppendorf flame pho- drinking rate of the seawater eel was measured by cannulat- tometer and Buchler-Cotlove chloridometer. ing the esophagus, we noticed that the eel does not gulp water intermittently, but ingests it continually; the ingested RESULTS water seems to move along the esophagus rather slowly (5). Recently, Kirsch and Laurent (6) found that swallowed sea Since essentially the same results were obtained in A. anguil- water is already diluted to one-half by the time it reaches la and in A. japonica, the data from two species of the eels the posterior end of the eel esophagus. The present investi- were combined in the following figures and table. When the gation was carried out in order to elucidate the role of the organ was incubated on both sides with isotonic Ringer solu- esophagus in the dilution of ingested sea water. tion, no movement of either water or Na+ and Cl- ions was seen in the esophagus (Fig. 1). There was also no water and MATERIALS AND METHODS Na+ movement in the stomach, whereas significant (P < 0.01) secretion of Cl- ion into the lumen was observed. Ten Japanese cultured eels, Anguilla japonica, weighing There was no significant difference in the C1- secretion rate about 200 g each, were purchased from a commercial source between the seawater and the freshwater eel stomach. In and kept in a freshwater tank at 200 for at least 2 weeks be- contrast to the situation in the esophagus and the stomach, fore use. Five eels were then transferred to a seawater tank net absorption of water and the ions was seen in the intes- and kept there for 2 more weeks. Experiments were also tine, and the rates of absorption were significantly greater in done on the European eel, A. anguilla, weighing about 250 the seawater eel than in the freshwater eel. g. These eels were collected from the estuary of the river Fig. 2 shows net movements of water and ions in the gut, Rhone and kept in running fresh water (four eels) or sea filled with sea water and incubated in Ringer solution. In water (four eels) at 170 for 3 weeks before use. the freshwater eel esophagus, a small amount of water moved into the lumen and Na+ and Cl- ions moved out of t To whom requests for reprints should be addressed. the lumen, probably following the concentration gradient. 1348 Downloaded by guest on September 26, 2021 Physiology: Hirano and Mayer-Gostan Proc. Nat. Acad. Sci. USA 73 (1976) 1349

H6 Freshwater Eel Seawater Eel 4 Freshwater Eel Eel I Seawater 46 r -1 E 0 fI'l 36. a) I- 26. 16-

E S E $ I FIG. 1. Net movements of water and Na+ and Cl- ions in the isolated esophagus (E), stomach (S), and intestine (I) of the seawa- FIG. 2. Net movements of water and Na+ and Cl- ions in the ter and the freshwater eels, bathed on -both sides with isotonic isolated gut of the eel, filled with sea water and incubated in Ring- Ringer solution. Empty bars represent water movement, black er solution. Legend as in Fig.1. bars Na+ movement, and shaded bars Cl- movement. Positive values indicate mucosa to serosa movement. Standard errors of the means (n = 9, 5 A. japonica and 4 A. anguilla for the esophagus mmn after incubation in the intestine and after 60 mmn in the and the stomach, and n = 5 A. japonica for the intestine) are indi- esophagus. cated by vertical lines. * Significantly different from the freshwater value at 5% level. DISCUSSION Smith (1) proposed that marine teleosts drink sea water to The ion movements increased greatly after seawater adapta- replace water lost osmotically across the body surface. Based tion, without any change in water permeability. In the stom- on changes in ion composition of gastrointestinal fluid, he ach, water movement was greater than in the esophagus, suggested that swallowed sea water is diluted first in the whereas the ion movements were greater than those in the stomach and later in the intestine with an osmotic influx of freshwater eel esophagus but smaller than in the seawater water from the blood; subsequent absorption of water takes eel esophagus. There was no difference in water and ion place in the intestine following active absorption of monova- movements in the stomach between freshwater and seawater lent ions. Various authors tested this hypothesis on in vitro eels. The rates of ion movements in the freshwater eel intes- preparations from several teleost species by introducing sea tine were as great as those in the seawater eel esophagus, and water into the gut lumen. They observed that intestinal sacs increased further after seawater adaptation. The water filled with sea water and incubated in Ringer solution rapid- movement in the intestine was greater than in the esophagus ly gained weight, followed by a weight loss (absorption) and the stomach. However, it was highly variable, absorp- after some time (7-9). In Anguilla japonica, the reversal of tion of water being seen in some preparations, and no signif- water flow occurred earlier in the seawater eel intestine than icant difference was detected between freshwater and sea- in the freshwater eel (9). A similar observation was also water eels. made in vivo by Skadhauge (10) in the perfused intestine of Efficiency of seawater dilution was compared in the sea- A. anguilla, and the osmolality of the perfusion fluid which water eel gut filled with sea water by expressing water and corresponds to zero net flow of water (the turning point os- ion movements in water gain or Na+ loss per ml of fluid molality) was higher than plasma osmolality by 126 millios- (Table 1). The seawater eel esophagus gained little water but moles/kg in seawater eels and by 73 milliosmoles/kg in lost a large amount of Na+ (and also Cl-) ions; sea water is freshwater eels. This indicates that the seawater eel intestine diluted in the esophagus to two-thirds in 1 hr, mainly by re- is able to absorb water from hypertonic salt solution (about moval of the ions. Sea water seems to be diluted in the stom- half-strength sea water). ach by passive diffusion of the ions (less than in the esopha- However, our results show that ingested sea water is dilut- gus) and also by an osmotic influx of water (more than in the ed before it reaches the intestine and even before it reaches esophagus). However, 1 ml of sea water enclosed in the the stomach. The esophagus thus plays an important role in stomach was diluted only by 10% in an hour; dilution of sea dilution of ingested sea water. We have shown that the water in the esophagus was about three times more efficient esophagus of the seawater eel is highly permeable to Na+ than in the stomach. On the other hand, dilution of sea and Cl- ions but not to water: sea water is diluted in the water enclosed in the intestine was twice as fast as in the esophagus mainly by salt removal without water addition. esophagus; Na+ concentration decreased to two-thirds 30 Dilution is therefore a misleading term; desalting is more

Table 1. Movements of water and Na+ ion in the isolated gut of the seawater eel (A. japonica), filled with sea water and incubated in Ringer solution

Na+ concentration (meq/liter) Incubation Water gain/ml Na+ loss/ml time (min) initial fluid (pal) initial fluid (,geq) Initial Final

Esophagus 60 7.3 ± 4.2 135.9 ± 13.9 443 ± 3.1 310 ± 8.7 Stomach 60 26.4 ± 7.4 31.4 ± 8.4 443 ± 3.1 401 ± 8.5 Intestine 30 47.9 ± 13.5 105.0 ± 15.4 442 ± 3.7 323 ± 17.2

* Mean + SEM (n = 5). Downloaded by guest on September 26, 2021 1350 Physiology: Hirano and Mayer-Gostan Proc. Nat. Acad. Sci. USA 73 (1976) accurate. A similar conclusion was obtained in vivo by per- In contrast to the esophagus of the seawater eel, the fresh- fusing the esophagus of A. anguilla with sea water (6). water eel esophagus was almost impermeable to both water The sequence of events in replacing water lost osmotically and the ions. The physiological importance of the imper- in the seawater eel may be as follows. In the steady state, meability of esophagus to water and ions is not clear, since eels in sea water drink water continually, the drinking rate the eel in fresh water is known to drink little water (1, 5, 11). being modified probably by sphincter action at the begin- Since the water permeability of the intestine-was still greater ning of the esophagus. During the slow passage along the than that of the esophagus and the stomach, the small esophagus, Na+ and Cl- ions diffuse passively into the amount of fresh water the eels ingest will be absorbed partly blood; these ions are quickly excreted from the gills, thus in the stomach and completely in the intestine. At any rate, keeping the elevation of plasma ion concentration minimal. the change in ion permeability observed in the eel esophagus The seawater eel in the steady state drinks water at the rate seems to be part of the adaptation to different environmen- of about 0.35-0.4 ml/100 g-hr (11). Since the eels used in the tal salinities. In the present study, an increase in ion and present study weighed about 200 g, 0.7-0.8 ml of sea water water permeability was also seen in the intestine following would move along the esophagus in an hour. When 1 ml of seawater adaptation. This increase in absorptive capacity of sea water was enclosed in the seawater eel esophagus, 135 the intestine has been repeatedly observed in the eel, and is ,ueq of Na+ ion were removed from the lumen in an hour, known to be inhibited by prolactin in the freshwater eel and resulting in two-thirds dilution. Considering that ion move- facilitated by cortisol in the seawater eel (3, 14). It seems ment in the isolated preparation would be less than the highly probable that the ion permeability of the eel esopha- movement in vivo with intact blood circulation, and also gus is also under hormonal control. The involvement of hor- that about two-thirds of the esophagus was used in the mones, permeability characteristics, and morphological present study, dilution of ingested sea water in the esopha- changes associated with the functional changes in the eel gus would be even more efficient in vivo. esophagus invite further study. Thus, stomach receives partly "desalted" sea water, and further dilution will take place by an osmotic influx of water We are grateful to Prof. Howard A. Bern and to Dr. Jean Maetz and passive loss of ions following concentration gradient. In for comments on the manuscript. This work was supported in part the goosefish, Lophius piscatorius, Smith (1) found that con- by grants from the Japanese Ministry of Education and the French centrations of Mg2+ ions in the gastric fluid are lower than Atomic Energy Commission. in sea water, and ascribed this to dilution with gastric juice and to an osmotic influx of water from the blood. In the eel, 1. Smith, H. W. (1930) Am. J. Physiol. 93,480-505. however, dilution with gastric juice may not be so impor- 2. Maetz, J. (1974) in Biochemical and Biophysical Perspectives tant, since no appreciable water movement was seen when in Marine Biology, eds. Malins, D. C. & Sergent, J. R. (Aca- the stomach was incubated on both sides with identical demic Press, London, New York, San Francisco), pp. 1-167. Ringer solution. Moreover, the stomach was less permeable 3. Hirano, T., Morisawa, M., Ando, M. & Utida, S. (1975) in In- to ions' than the seawater eel esophagus, and dilution of the testinal Ion Transport, ed. Robinson, J. W. L. (Medical & Technical Pub. Co., Lancaster, England), pp. 300-317. enclosed sea water in the esophagus was about three times 4. Suehiro, Y. (1942) Jpn. J. Zool. 10, 1-303. more efficient than in the stomach. Considering the fact that 5. Hirano, T. (1974) J. Exp. Biol. 61, 737-747. the stomach receives already "desalted" sea water, dilution 6. Kirsch, R. & Laurent, P. (1975) C. R. Acad. Sci. Paris 280, in the stomach may not be important, at least in the eel. In 2013-2015. this respect, it is interesting to find that Na+ concentration 7. Sharratt, B. M., Bellamy, D. & Chester Jones, I. (1964) Comp. and osmolality of the gastric fluid are always greater than Biochem. Physiol. 11, 19-30. those of the body fluid in the goosefish (1), the eel (7), and 8. House, C. R. & Green, K. (1965) J. Exp. Biol. 42, 177-186. the flounder (12). According to Skadhauge (13), the osmolal- 9. Utida, S., Isono, N. & Hirano, T. (1967) Zool. Mag. 76, 203- ity of the luminal fluid taken from the upper end of the an- 204. 10. Skadhauge, E. (1969) J. Physiol. (London) 204, 135-158. terior intestine of A. anguilla is about 460 milliosmoles/kg, 11. Oide, H. & Utida, S. (1968) Mar. Biol. 1, 172-177. corresponding to about half-strength sea water. Considering 12. Hickman, C. P., Jr. (1968) Can. J. Zool. 46,457-466. the "turning-point osmolality" or the capacity of the seawa- 13. Skadhauge, E. (1974) J. Exp. Biol. 61, 737-747. ter eel intestine to absorb water from hypertonic salt solution 14. Utida, S., Hirano, T., Oide, H., Ando, M., Johnson, D. W. & (10), the seawater eel seems to be able to absorb water from Bern, H. A. (1972) Gen. Comp. Endocrinol. Suppl. 3, 317- the intestine without losing water from the body. 327. Downloaded by guest on September 26, 2021