Structural Diversity of Occluding junctions in the Low - resistance Chloride - secreting Opercular Epithelium of Seawater - adapted Killifish ( Fundulus heteroclitus ) STEPHEN A

The structural features of the chloride-secreting opercular epithelium of Seawateradapted killifish (Fundulus heteroclitus) were examined by thin-section and freeze-fracture electron microscopy, with particular emphasis on the morphological appearance of occluding junctions. This epithelium is a flat sheet consisting predominantly of groups of mitochondriarich chloride cells with their apices associated to form apical crypts . These multicellular groups are interspersed in an otherwise continuous pavement cell epithelial lining . The epithelium may be mounted in Ussing-type chambers, which allow ready access to mucosal and serosal solutions and measurement of electrical properties. The mean short-circuit current, potential difference (mucosal-side negative), and DC resistance for 19 opercular epithelia were, respectively, 120.0 ± 18.2 p,A/cm2, 12.3 ± 1 .7 mV, and 132.5 ± 26.4 UCm2. Short-circuit current, a direct measure of CI transport, was inhibited by ouabain (5 g,m) when introduced on the serosal side, but not when applied to the mucosal side alone. Autoradiographic analysis of [3H]ouabain-binding sites demonstrated that Na +,K+ -ATPase was localized exclusively to basolateral membranes of chloride cells; pavement cells were unlabeled . Occluding junctions between adjacent chloride cells were remarkably shallow (20-25 nm), consisting of two parallel and juxtaposed junctional strands. functional interactions between pavement cells or between pavement cells and chloride cells were considerably more elaborate, extending 0.3-0.5 ,um in depth and consisting of five or more interlocking junctional strands. Chloride cells at the lateral margins of crypts make simple junctional contacts with neighboring chloride cells and extensive junctions with contiguous pavement cells. Accordingly, in this heterogeneous epithelium, only junctions between Na + ,K+-ATPase-rich chloride cells are shallow. Apical crypts may serve, therefore, as focal areas of high cation conductivity across the junctional route. This view is consistent with the electrical data showing that transmural resistance across the opercular epithelium is low, and with recent studies demonstrating that transepithelial Na' fluxes are passive. The simplicity of these junctions parallels that described recently for secretory cells of avian salt gland (Riddle and Ernst, 1979, 1. Membr. Biol., 45:21-35) and elasmobranch rectal gland (Ernst et al ., 1979, 1. Cell Biol., 83:(2, Pt . 2) :83 a[Abstr]) and lends morphological support to the concept that paracellular ion permeation plays a central role in ouabain-sensitive transepithelial NaCl secretion . Although ouabain-sensitive Na' pumps appear to play a crucial role in transepithelial salt secretion, the unexpected but consistently observed subcellular localization of this enzyme to 488 the inward-facing tasolateral plasma membranes of secretory cells (13, 22, 23) is perplexing . Recently, several laboratories have rationalized this apparent contradiction by proposing a THE JOURNAL of CELL BIOLOGY VOLUME 87 NOVEMBER 1980 488-497 ©The Rockefeller University Press 0021-9525/80/11/0488/10$1 .00 on O cber 0, 2017 jcb.rress.org D ow nladed fom model for salt secretion in which intercellular space Na', generated by basolateral plasmalemmal Na+,K'-ATPase, enters the luminal compartment across the zonula occludens ("tight" junction) in response to a favorable electrical potential derived from secondary "active" transcellular transport of Cl(21, 25, 49, 50) . This model is predicated in part on the presence of a high conductance cation-selective shunt pathway across the zonula occludens . Accordingly, Riddle and Ernst (45) examined occluding junctions in avian salt gland secretory epithelium by thin-section and freeze-fracture electron microscopy and found them to exhibit a morphological simplicity resembling that described for several absorptive epithelia known to be leaky to ions (4, 5). Although the presence of structurally simple junctions in salt-gland epithelium is consistent with the proposed secretory transport model, this support remains somewhat circumstantial inasmuch as histological restraints preclude critical measurements of electrical parameters across this epithelium . Recently, Karnaky and collaborators (6, 36, 38) introduced the teleost opercular epithelium as a model system for studying salt secretion (reviewed by Karnaky [351) . This model system joins a small but growing list (cf. Frizzell et al . [271) of chloridesecreting epithelia that can be mounted in modified Ussing chambers for electrophysiological study. Like the teleost branchial epithelium with which it is structurally and functionally analogous, the opercular epitheliumconsists ofa heterogeneous cell population of ion-transporting chloride cells and nontransporting pavement cells and mucous cells . In the present study of opercular epithelia from seawater-adapted Fundulus heteroclitus, we describe the simplicity ofoccludingjunctions between transporting chloride cells, in contrast to the extensivejunctions between pavement cells, and suggest that the observed functional architecture between chloride cells is consistent with the observation that transmural resistances measured across in vitro preparations are quite low, as might be expected if these junctions exhibit a high conductivity for cations . In addition, we show the selective localization of Na' pumps to the basolateral plasmalemmal surfaces of chloride cells in this epithelium and establish the sensitivity of short-circuit current and transepithelial potential (seawater-side negative) to ouabain following serosal, but not to mucosal only, exposure to the inhibitor. A preliminary account of this work was presented