Epithelial cell junctions. Left: two adjacent epithelial cells viewed in a section normal to the apical surface. Right: lateral view of one cell. Abbreviations denote features shown in both diagrams: MV = microvilli; zo = zonula occludens; za = zonula adherens (or belt desmosome); sd = spot desmosome; gj = gap junction; hd = hemidesmosome; bl = basal lamina. Tight junction (junctional complex) includes zo and za.

Main features of epithelial cell polarity. A. Diagram depicting structural polarity: MV = microvilli; G = Golgi apparatus; N = nucleus; M = mitochondria; BLI = basolateral‐membrane infoldings; zo = zonula occludens; za = zonula adherens; sd = spot desmosome; gj = gap junction; hd = hemidesmosome; (bl) = basal lamina. Positions of organelles are typical of most epithelial cells. B. Specific transport proteins are confined to different domains (apical or basolateral) of the plasma membrane. In example depicted, organic solute (OS, e.g., glucose) is transported into cell (across apical membrane) by secondary‐active transport via Na⁺‐OS cotransporters and out of cell (across basolateral membrane) by uncoupled passive transport via OS carriers. Na⁺ transport is downhill at apical membrane (via the cotransporters) and uphill at basolateral membrane (via the Na⁺, K⁺‐ATPase). The K⁺ channels in basolateral membrane mediate efflux of the K⁺ that enter the cell via the Na⁺, K⁺‐ATPase.

Rapid and reversible changes in intracellular ion activities following removal of Na⁺ (replaced with tetramethylammonium, TMA⁺) or Cl⁻ (replaced with cyclamate) from apical bathing solution. Numbers on the left denote voltages (in mV) at the beginning of each record. A. Top, voltage measured by intracellular Na⁺‐selective electrode, referenced to the membrane voltage (V_Na – V_cs), intracellular Na⁺ activity (aNa_i) scale on the right; bottom, transepithelial voltage (V_ms). Mucosa‐positive change in V_ms was caused by Na⁺‐TMA⁺ paracellular bi‐ionic potential. Na⁺ removal from apical bathing solution causes a rapid, reversible fall in aNa_i. B. Top, basolateral‐membrane voltage (V_cs); bottom, voltage measured by intracellular Cl⁻‐selective electrode, referenced to the membrane voltage (V_Cl–V_cs), intracellular Cl⁻ activity (aCl_i) scale on the right. During exposure to Cl⁻‐free mucosal medium, cell hyperpolarizes and aCl_i falls rapidly and reversibly. These results indicate that the Na⁺ and Cl⁻ transport pools are accessible to the ion‐sensitive microeletrode and hence involved the entire cytoplasm.

A. Steady‐state equivalent circuit of an epithelium with one cell type and a paracellular pathway of finite conductance. Each element in the circuit (a: apical membrane, b: basolateral membrane, s: paracellular pathway) is represented by a Thévenin electrical equivalent, i.e., an electromotive force (EMF), E, in series with a resistance, R. The Thévenin equivalent is an acceptable representation of any combination of linear electrical elements at the steady state. B. Membrane voltages for an epithelium with the indicated values of EMFs (in mV) and resistances (in Ω·cm²). Polarities referred to basolateral solution (transepithelial values) or to adjacent solution (cell membrane values). Note that all voltages differ from the respective EMF values.

Ion pumps in epithelial cells. A. Predicted general structure of α‐subunit of P‐type pump (representative of Na⁺, K⁺‐, Ca²⁺‐and H⁺,K⁺‐ATPases). The α‐subunit contains binding site(s) for transported substrate(s), ATPase activity, regulatory domains, and inhibitor binding sites. It is a single peptide chain, with amino and carboxyl termini in the cytoplasmic side, probably ten transmembrane α‐helices and ATP binding site in the second cytoplasmic loop. Based on Lingrel and Kunzweiler 420. B. Generalized catalytic cycle for P‐type ATPases modeled as countertransport pumps exchanging Na⁺ (A) for K⁺ (B), Ca²⁺ (A) for H⁺ (B), and H⁺ (B) for K⁺ (A) (Na⁺, K⁺‐, Ca²⁺‐ and H⁺, K⁺‐ATPase, respectively. E1 = conformation with ion‐binding sites accessible from cytoplasm, E2 = conformation with ion‐binding sites accessible from the extracellular face, E(A) or E(B) = ion “occluded” in the protein.

Transport mechanism in Na⁺‐absorptive epithelia. The two‐membrane hypothesis for Na⁺‐absorbing “tight” epithelia 366. At the apical‐cell membrane, Na⁺ entry is via an Na⁺‐channel blockable by amiloride. At the basolateral cell membrane, extrusion of Na⁺ is mediated by an Na⁺, K⁺‐ATPase inhibitable by ouabain. K⁺ “recycles” across the basolateral cell membrane via a K⁺ channel.

Short‐circuit current technique. Epithelium (vertical line between ve's) is tightly mounted in the middle partition of an Ussing chamber, separating solutions of identical composition, at the same pressure. Current is applied across the epithelium with an external circuit (ce = current electrode) and the transepithelial voltage (V_t) is measured (ve = voltage electrode). The short‐circuit current (I_sc) is the current needed to make V_t = 0. This current, in the conditions defined, is equivalent to the sum of active ion fluxes across the epithelium. Pitfalls in the use of this technique include: (1) improper short‐circuiting: series resistances in preparation produce voltage drops and the true transepithelial voltage is not zero; (2) the compositions of the bathing solution are different, i.e., there are passive (electrodiffusive or osmotic) driving forces that can account for net ion fluxes.

Demonstration that some epithelia have “leaky” (high‐permeability) junctions. A. Voltage scanning in Necturus gallbladder epithelium. High‐conductance pathways for transepithelial current flow were detected with microelectrode (S) exploring electric field with respect to distant electrode in apical bathing solution (I). S was moved along apical surface while large current was passed across the tissue (electrodes A in the inset). Microelectrode path is shown with respect to cell borders; × = tip position during voltage measurement, over junction (J) or over cell (C). In voltage scan, downward deflections mean current sinks (high‐conductance pathway). This experiment demonstrates junctional location of a high‐conductance transepithelial pathway. B. Junctional region of rabbit gallbladder epithelial cells after exposure to La³⁺ on the apical side. Lanthanum is clearly seen in intercellular pathway, indicating junctional permeability. Specimen fixed in glutaraldehyde, postfixed in OsO₄, and stained on grid with uranyl acetate and lead citrate. X 250,000.

Expected changes in epithelial‐cell solute and/or water content by lack of adjustment of solute transport rates at the two membranes. Center: salt‐absorbing epithelial cell at steady state, i.e., apical solute entry equals basolateral solute exit and cell solute content remains constant. Left: increases in solute content, by primary increase in entry (top) or primary decrease in exit (bottom). Right: decreases in solute content, by primary decrease in entry (top) or primary increase in exit (bottom). In all four instances, cell water volume would change in same direction as cell solute content, because of osmotic water flow (at least one of the cell membranes has a high water permeability).

Mechanisms of maintenance of intracellular electroneutrality and solution electroneutrality during steady‐state transepithelial ion transport. In all diagrams, Na⁺ or K⁺ transport is the primary event, indirectly or directly coupled to anion flux in same direction or to cation flux in the opposite direction. Examples of epithelia in which the scheme depicted has been proven or is suspected are denoted in parentheses. A. Na⁺ and Cl⁻ fluxes are directly coupled by cotransport (mammalian renal distal tubule). B. K⁺ and Cl⁻ fluxes occur via independent channels; the fluxes could be “coupled” by the membrane voltage, both as driving force for electrodiffusion and as gating mechanism (amphibian gallbladder stimulated by cAMP). C. Na⁺ and K⁺ fluxes occur in opposite directions, via separate channels; they may be “coupled” by the membrane voltage (mammalian renal cortical collecting data). D. Absorptive Na⁺ and Cl⁻ transepithelial fluxes occur via different pathways, transcellular and paracellular, respectively (frog skin epithelium). E. Secretory Cl⁻ and Na⁺ fluxes occur via different pathways, transcellular and paracellular, respectively (airway epithelium).

Transport mechanisms in “leaky” electroneutral NaCl‐absorbing epithelia. A. Na⁺ and Cl⁻ influxes across apical cell membrane via exchanges with H⁺ and , respectively. Cl⁻ exit across basolateral cell membrane is via Cl⁻ channels and/or via K⁺‐Cl⁻ cotransport. Na⁺ extrusion is mediated by Na⁺,K⁺‐ATPase. K⁺ recycles via channels (basolateral or both cell membranes). This model accounts for NaCl transport in most “leaky” NaCl‐absorptive epithelia. B. The mechanism of salt entry is Na⁺‐Cl⁻ cotransport (demonstrated for flounder urinary bladder and mammalian renal distal tubule and proposed for rabbit and Necturus gallbladder under certain conditions). Basolateral transport mechanisms as in top diagram. C. Mechanism of salt entry is Na⁺‐K⁺–2Cl⁻ cotransport (demonstrated for flounder intestine). Basolateral transport mechanisms as in top diagram.

Transport mechanisms in a “leaky” rheogenic Na⁺‐absorbing epithelium. Na⁺ influx is by cotransport with organic solutes (OS, e.g., sugar or amino acid). The transporters at the basolateral cell membrane are as shown in Figure 11.

Predicted secondary structure of human SGLT1. The protein consists of single chain of 664 amino acids, 12 putative transmembrane α‐helices, N‐glycosylation at Asn248, and cytoplasmic location of both amino and carboxyl termini. The carboxyl terminus is closely associated to the plasma membrane. Model for low‐affinity Na⁺‐glucose cotransporter SGLT2 is almost identical.

Predicted secondary structure of ROMK1. In contrast with other cloned ion channels, the protein has only two transmembrane α‐helices. The extracellular loop has an N‐glycosylation site and appears to be responsible for formation of the conductive pathway. Amino and carboxyl termini are on the cytoplasmic side. The carboxyl‐terminus region is long and contains the nucleotide‐binding domain (NBD).

Mechanisms of ion transport in Cl⁻‐transporting epithelia. A. Transport model for exocrine‐gland Cl⁻‐secreting epithelium. Intracellular [Cl⁻] is maintained above electrochemical equilibrium by electroneutral Na⁺‐K⁺–2Cl⁻ cotransport across the basolateral cell membrane. Na⁺ influx is balanced by efflux via Na⁺, K⁺ ‐ATPase; K⁺ influxes are balanced by efflux via K⁺ channels. Transepithelial secretion is induced by activation of apical‐membrane Cl⁻ channel(s). B. Model of Cl⁻ transport in thick ascending limb of Henle's loop, a Cl⁻‐absorptive epithelium. Cl⁻ entry across the apical membrane is mediated by Na⁺ ‐K⁺–2Cl⁻ cotransporter. Most basolateral Cl⁻ efflux proceeds through Cl⁻ channels.

Predicted secondary structure of CFTR. The protein consists of single peptide chain with two halves, each compromising six putative transmembrane α‐helices and a nucleotide‐binding domain (NBD). The two halves are linked by the R domain, a highly charged regulatory region. R domain possesses consensus phosphorylation sites for protein kinases A and C and Ca²⁺‐calmodulin kinase. ΔF508 deletion found in 70% of all CF cases is located in first NBD. Two consensus sites for N‐linked glycosylation are present between transmembrane domains 7 and 8.

Predicted secondary structure for shark rectal gland Na⁺‐K⁺–2Cl⁻ cotransporter (NKCC1). The protein comprises 1,191 amino acid residues. Amino and carboxyl termini are on the cytoplasmic side. There are two domains of seven and five putative transmembrane α‐helices, respectively, joined by an extracellular loop containing several N‐glycosylation sites (branched lines). This predicted secondary structure is shared by other members of putative caution‐Cl⁻ cotransporter family (see text).

Basic mechanisms of ion transport in H⁺ and secretory epithelia. Carbonic anhydrase catalyzes formation of H⁺ and OH⁻ from water; is then formed by reaction of OH⁻ and CO₂. Net result, shown in figure, is production of H⁺ and . In H⁺‐secreting epithelium there is apical‐membrane H⁺ extrusion and basolateral‐membrane extrusion. In secreting epithelium, polar expression of these transporters is reversed. Cl⁻ channel in parallel with exchanger maintains intracellular [Cl⁻]. Precise basolateral‐membrane transporters vary among H⁺‐secreting epithelia (see text).

Specific transport mechanisms in H⁺‐secreting epithelia: A. Renal proximal tubule. B. α‐intercalated cell of the renal collecting tubule. C. Oxyntic cell of the gastric mucosa. See text for details.

Transport mechanisms in ‐secreting epithelia. Net operation is secretion and Cl⁻ absorption. Note that polarized expression of H⁺ and transporters is opposite to that in H⁺‐secreting epithelia.

Mechanisms of water transport in leaky epithelia. Panel A. Three‐compartment model of Curran and MacIntosh 146. Because of solute entry (solid arrow) C_s(M) > C_s(A), where C_s is total solute concentration or osmolality. This causes osmotic water flow from compartment A to compartment M (open arrow). The elevation of the hydrostatic pressure in compartment M causes solution flow from compartment M into compartment B (open/solid arrow), regardless of the solute concentration (and hence osmotic pressure) in the latter. Modified from Whittembury and Reuss 753, with permission. Panel B. Standing‐gradient hypothesis of Diamond and Bossert 163. Solute transport (solid arrows) into channel (lateral intercellular space) causes a local increase in osmolality; water flows osmotically across bounding membranes (open arrows), “diluting” the solution in the channel. Transport toward open end is by bulk flow and diffusion. Modified from Reuss and Cotton 571, with permission. Panel C. Near‐isosmotic transport model. Because of high osmotic water permeability of cell membranes, differences in solution osmolality (C_s) needed to account for fluid transport are small, probably localized at epithelium‐solution interfaces. Salt transport causes dilution of solution on cis side and concentration of solution on trans side; these differences cause osmotic water flow from cis to trans. Magnitude of paracellular water flow is uncertain. Because of high surface area of lateral membranes and small volume of lateral intercellular spaces, space osmolality is “clamped” by cell osmolality, making longitudinal gradients small at most.

Membrane topology and putative structure of aquaporin 1 (CHIP28). A. Predicted secondary structure: CHIP comprises two repeats of three putative transmembrane domains each. Amino and carboxyl termini are on the cytoplasmic side. There is a single N‐glycosylation site in loop A (not shown). Loops B and E contain highly conserved Asn‐Pro‐Ala motifs thought to contribute to the formation of the water pore. Loop E contains Cys189, the residue conferring Hg‐sensitivity to water permeation. B. Proposed structure of the water pore of a CHIP monomer. α‐helical transmembrane domains form the peripheral wall, whereas loops B and E assemble in the membrane to form the thin barrier accounting for the selectivity for water. The diagram shows a monomer that is functional per se, but CHIP monomers assemble into tetrameric complexes.