Differentiation-dependent expression of alpha-2,3-sialyltransferase in rabbit corneal epithelium

PURPOSE

Lectin studies have shown that in the rabbit corneal epithelium, alpha-2,3 sialylation of O-linked glycans differentiates limbal and corneal epithelial cell phenotypes. Because sialic acid can be regulated at the level of the expression of sialyltransferases (STs), the purpose of the present study was to analyze the expression of alpha-2,3STs in this tissue.

METHODS

Reverse transcription-polymerase chain reaction (RT-PCR) was used to generate ST cDNA from total rabbit corneal epithelium RNA using primers selected from the sequences of three previously cloned STs capable of catalyzing the transfer of sialic acid to O-linked oligosaccharides, human placental Galbeta-1,3GalNAc-Galbeta-1,4GluNAcalpha-2,3ST (STZ), and mouse brain Galbeta-1,3GalNAcalpha-2,3ST types I and II (ST3Gal I and ST3Gal II). Tissue distribution of mRNA was assayed by fluorescent in situ hybridization. A synthetic peptide whose sequence was deduced from a cloned cDNA fragment was synthesized and used to prepare an anti-ST goat antiserum. The molecular weights of immunodetectable polypeptides and their distribution in cryostat sections of the limbocorneal area were investigated by western blot analysis and indirect immunofluorescence, respectively.

RESULTS

RT-PCR yielded cDNA of expected basepair length for STZ and ST3(Gal II. The rabbit STZ cDNA was 86% identical with its human equivalent. Its mRNA was confined to the cornea, mainly in basal epithelial cells, and was not expressed in the limbus. Western blot analysis identified a band at 37 kDa whose binding was abolished by preincubation of the antiserum with the immunization peptide. Immunohistologic analysis revealed the presence of immunoreactive epitopes in all basal cells of the cornea but not in the limbus.

CONCLUSIONS

STZ mRNA and the enzyme itself are expressed in the basal layer of the corneal epithelium but are absent in the limbus. This enzyme's de novo expression seems thus responsible for the differential expression of alpha-2,3 sialylation along the limbocorneal differentiation axes. At least one more alpha-2,3ST is also present in the epithelium.

Connexin50, a gap junction protein of macrogliaP6n the mammalian retina and visual pathway

Reverse transcriptase-polymerase chain reaction (RT-PCR), Western blotting and immunocytochemistry were used to study the expression of gap junction proteins (connexins; Cx) in the rat and rabbit retina. RT-PCR of rabbit total retinal RNA using primers selected for the human Cx50 (alpha 8 Cx) DNA template yielded cDNA fragments of the predicted base pair size. Western blots of rat and rabbit retinal membrane preparations probed with a monoclonal antibody which recognizes Cx50 in the lens of several mammalian species revealed a single band (MW 50 kD), identical to that recognized in lens membrane extracts. In frozen retinal sections of both species, the same monoclonal antibody as well as two polyclonal antisera raised against a synthetic peptide from the C-terminal region of the human Cx50 polypeptide labeled Müller cells and astrocytes. In Müller cells, labeling was strongest in the endfeet and in the filamentous processes ensheathing the photoreceptors. Extending from the neural retina, Cx50-like immuno-reactivity was detected in astrocytes of the optic nerve and along retinal projections within the CNS. Our data indicate that Müller cells and astrocytes of mammalian retinas and throughout the visual pathway are coupled through gap junctions composed of connexin50.

Effect of heptanol on the short circuit currents of cornea and ciliary body demonstrates rate limiting role of heterocellular gap junctions in active ciliary body transport

Rabbit ciliary body and cornea were mounted in Ussing-type chambers in Tyrode's under voltage clamp and the effects of heptanol, a gap junction inhibitor, on the short circuit current generated by each of the respective epithelia were determined. Studies were carried out either in control conditions or following amphotericin B permeabilization of either the basolateral membrane of the nonpigmented epithelium of the ciliary body or the apical membrane of the corneal epithelium, respectively. Previous studies have shown that, following these permeabilizations, short circuit currents are established, reflecting aqueous (or tear)-to-serosa Na+ fluxes, and that Na+ translocation through gap junctions connecting the individual layers of these tissues constitutes the major rate limiting step. Heptanol inhibited most of the short circuit current of the amphotericin B-modified ciliary body and cornea and of the unmodified ciliary body epithelium (control). In all these cases, the apparent IC50 was about 0.8 M. In the unmodified corneal epithelium, where ion translocation across the apical membrane constitutes the main rate limiting step for active secretion, 0.4 or 0.8 mM heptanol induced short circuit current increases; partial inhibition was observed only at high concentrations known to cause maximal inhibition of junctional permeability. Heptanol also enhanced the volume regulatory decrease of cultured human NPE cells, a process dependent on cell swelling-induced stimulation of Cl- and K+ permeabilities. Combined with our previous results demonstrating the lack of heptanol effects on other epithelial functions, these data suggest that the effect of heptanol on the active ciliary body transepithelial transport is primarily due to inhibition of the nonpigmented-pigmented junctional path and that this path is a potential site of rate limitation for the secretory process.

Stem cells of the corneal epithelium lack connexins and metabolite transfer capacity

The stem cells of the corneal epithelial lineage are confined to the basal cell layer of the limbus, a vascularized outer corneal rim. These slow cycling cells of great proliferative potential maintain the corneal epithelial mass. Since cell-cell communication plays an important role in development and differentiation, we conducted a comparative examination of the expression of two corneal connexins, C x 43 and C x 50, and the tracer transfer capacity of the limbal and corneal epithelia using the scrape loading technique. C x 43 is abundantly expressed in the basal cell layer of the epithelium covering the cornea, but is essentially absent from the mouse, human, neonatal rabbit, and chicken limbal epithelium. In the adult rabbit the limbal epithelium displays an overall weak C x 43 immunoreactivity, but C x 43-free isolated basal cells can be distinguished. C x 50 is expressed throughout the corneal epithelium of the three mammalian corneas, but is not detectable in the limbus. Scrape loading experiments in the rabbit yielded results which were consistent with the immunohistological findings; limbal epithelium lacked tracer (lucifer yellow) transfer capacity, strongly suggesting the absence of functional gap junctions. Altogether, our results demonstrate the incompetence of stem cells for gap junction-mediated cell-to-cell communication. This property may reflect the need of these unique cells to maintain a distinct intracellular environment.

Alterations in connexin expression and cell communication in healing corneal epithelium

PURPOSE

The aim of this study was to examine cell-to-cell metabolite transfer and connexin distribution in the rabbit corneal epithelium, in the stationary state, and during wound healing.

METHODS

Rabbit corneas were wounded with a surgical tool, producing a 3-mm-wide elongated debridement. Corneas were allowed to heal in vivo for up to 45 hours. Monoclonal antibodies against connexins Cx 26, Cx 32, Cx 43, and Cx 50 were used to stain cryostat sections. Cell-to-cell metabolite transfer capacity was assessed by a modification of the scrape-loading technique using lucifer yellow as the organic ion tracer.

RESULTS

The rabbit corneal epithelium contains Cx 43 and Cx 50, localized in the cell's plasma membrane, as shown previously for other species. Cx 26 and Cx 32 are not detectable. Tracer transfer occurred in both basal and suprabasal cell layers. After wounding, the migrating epithelial monolayer lacked Cx 43 and Cx 50. This change was apparent 6 hours after injury and persisted until complete wound closure (approximately 24 hours). The Cx 50 membrane stain was increased elsewhere, in particular in the transition zone between monolayered and multilayered epithelium. Consistent with the expression changes, migrating cells displayed no or minimal cell-to-cell tracer transfer, whereas in the periphery of the wound, tracer transfer was enhanced in comparison to the control specimen.

CONCLUSIONS

Corneal epithelial healing involves biphasic changes in the expression of connexins and cell-to-cell communications. These alterations may be critical for the optimization of the healing response.

Connexin distribution in the rabbit and rat ciliary body. A case for heterotypic epithelial gap junctions

PURPOSE

To evaluate the distribution of different alpha- and beta-type connexins (Cx) present in the dual layered ciliary body epithelia (CBE) of both rabbit and rat.

METHODS

Immunocytochemical detection of Cx26, Cx32, Cx43, and Cx50 was performed on frozen sections of rabbit and rat ciliary body using indirect immunofluorescent methods. The identity of the antigens recognized by the monoclonal primary antibodies was further confirmed by Western immunoblots. Double labeling experiments based on either conventional or confocal microscopy were carried out to establish the exact spatial relationship between different connexins.

RESULTS

Connexin 50 was found only in the nonpigmented epithelium (NPE) at apical and basolateral membranes, whereas Cx43 was observed exclusively and at a very high concentration in the pigmented epithelium (PE), primarily in the apical cell membrane, with minimal extension to the proximal lateral zone. The correct antigenicity of the antibodies was confirmed by Western blots of rabbit ciliary body membranes. In rabbit, the Cx26 antibody detected an antigen that was abundant in the NPE and was weakly expressed in the PE. In rat, however, the Cx26 staining was confined to capillary wall endothelia. Western blots of ciliary body and liver membranes and liver immunohistology indicated that the Cx26 antibody used does not recognize rabbit Cx26. Cx32 did not yield any substantial epithelial labeling in either species.

CONCLUSIONS

The distribution of Cx50 around the entire NPE cell perimeter suggests its involvement in NPE-NPE cell homotypic gap junctions. The concentration of Cx43 and Cx50 at the apical membranes of the PE and NPE cells, respectively, and their complete absence from the opposite cell suggest that these connexins may participate in the formation of heterotypic gap junctions, either with each other or with other yet unidentified connexins.

Ca2+ mobilization and interlayer signal transfer in the heterocellular bilayered epithelium of the rabbit ciliary body

1. 'Ratiometric' fura-2 methodology in slice preparations and 'intensitometric' fluo-3 measurements of confocal images were used to simultaneously monitor Ca2+ mobilization in the two distinct, apically joined cell layers which constitute the ciliary body epithelium (CBE): the non-pigmented (NPE) and pigmented (PE) epithelia. 2. Both methods yielded comparable results regarding Ca2+ responses in the syncytium upon stimulation with adrenergic and cholinergic agonists. 3. The alpha 1-adrenoceptor agonist phenylephrine elicited a moderate [Ca2+]i increase in the PE, whereas NPE [Ca2+]i remained unchanged or exhibited a slight diminution. 4. In combination with carbachol, the alpha 2-adrenoceptor agonist brimonidine elicited large Ca2+ increases (> 10-fold) in both the NPE and PE cell layers, even though previous studies indicated the absence of an alpha 2-adrenergic effect on [Ca2+]i in the PE. The onset, as well as the peak of the Ca2+ responses in PE cells frequently exhibited a small delay with respect to adjacent NPE cells. No such time difference was observed between adjacent NPE cells. 5. Pre-incubation of the ciliary body in Ca(2+)-free solution under conditions known to elicit overt NPE-PE separation abolished the alpha 2-adrenocholinergic response in the PE. 6. Addition of heptanol to the perfusate, to block gap-junctional communication, caused a small [Ca2+]i decrease in the NPE and a slight increase in PE[Ca2+]i. Subsequently, the Ca2+ mobilization in the Pe in response to the brimonidine and carbachol combination was either blocked or showed a substantial delay. The Ca2+ mobilization in the NPE, in contrast, remained unchanged. 7. We conclude that the heterocellular syncytium exhibits rectificatory behaviour with respect to Ca2+ mobilization; responses originating within the NPE are easily transferred to the PE, while the reverse does not occur.

Adreno-cholinergic modulation of junctional communications between the pigmented and nonpigmented layers of the ciliary body epithelium

PURPOSE

Cell-to-cell communications between the epithelial layers of the ciliary body may be critical for aqueous humor production. The aim of this study was to identify pharmacologic agents that affect this path.

METHODS

Whole New Zealand rabbit ciliary bodies were mounted in Ussing-type chambers with Ca2+(-)free and Ca2+(-)rich Tyrode's in the nonpigmented (NPE; aqueous) and pigmented (PE; serosa) epithelial side hemichambers, respectively. The NPE of the PE were then permeabilized, either selectively to monovalent ions with amphotericin B or nonselectively to small solutes with digitonin. Resultant active transport activities were tracked as short circuit currents (ISCS).

RESULTS

Permeabilization of the NPE with either 10 microM amphotericin B or 10 micro M digitonin led to an aqueous-to-serosa-positive ISC. This ISC was inhibited by serosal-side ouabain and heptanol, indicating movement of Na+ from permeabilized NPE to the PE by the interlayer junctional path, followed by PE-to-serosa active Na+ transport. Permeabilization of the PE with amphotericin B elicited an ISC in the opposite direction, This ISC was abolished by aqueous-side ouabain and by heptanol, consistent with sequential PE to NPE Na+ translocation, followed by active, NPE-to-aqueous transport. Acetylcholine, epinephrine, norepinephrine, and the alpha 1-adrenergic agonist phenylephrine, but not brominidine, an alpha 2-adrenergic agonist, each caused an approximately 50% reduction of these currents. The inhibitions were fully dependent on serosal-side Ca2+ and were blocked by one calmodulin inhibitor, trifluoperazine, but not by another, calmidazolium.

CONCLUSIONS

The above observations provide evidence that cholinergic or alpha 1-adrenergic activation of the PE causes Ca2+(-)dependent inhibition of the NPE-PE junctional path. A triflouperazine-sensitive entity, which may be distinct from calmodulin, is involved in the inhibition.

Comparative adrenocholinergic control of intracellular Ca2+ in the layers of the ciliary body epithelium

PURPOSE

Both nonpigmented epithelia (NPE) and pigmented epithelia (PE) of the ciliary body are thought to participate in the formation of aqueous humor and its pharmacologic regulation. The aim of this study was to identify the similarities and differences in intracellular Ca2+ ([Ca2+]i) changes in each of the two layers in response to adrenergic and cholinergic inputs.

METHODS

Incubation of the Dutch belted rabbit ciliary body (CB) in low-[Ca2+] solution was used to induce the functional dissociation of the NPE and PE layers from each other. These layers or the intact undissociated CB were loaded with the fluorescent Ca2+ indicator fura-2 and mounted in superfusion chambers for fluorometric measurement of [Ca2+]i.

RESULTS

In the NPE of the intact CB epithelium or in the isolated NPE, 10 microM acetylcholine (ACh), 1 microM brimonidine (UK 14304), or 1 microM epinephrine each elicited minimal rises in [Ca2+]i. On the other hand, the combination of either adrenergic drug with ACh resulted in large mobilizations of this cation. The alpha 1-adrenergic agonist phenylephrine was unable to induce Ca2+ mobilization in the isolated NPE and failed to do so in 10 of 12 intact CB specimens. In the isolated PE, both epinephrine and phenylephrine elicited substantial similar [Ca2+]i increases, ACh induced a smaller and slower rise, and the response to its combination with either adrenergic drug was essentially additive. UK 14304 (+/- ACh) had no measurable effect in these cells.

CONCLUSIONS

Each layer of the CBE exhibits distinct alpha-adrenergic control mechanisms. The NPE contains an alpha 2-adrenergic mechanism highly dependent on the cholinergic tone. The PE, in contrast, contains an alpha 1-adrenergic pathway that operates independently of cholinergic input. This segregation of mechanisms provides a basis for highly complex regulatory responses in the intact, syncytially organized ciliary body epithelium.

Alpha-2,3 sialylation differentiate the limbal and corneal epithelial cell phenotypes

PURPOSE

The initial differentiation event for the corneal epithelial cell lineage occurs as the limbally localized stem cells yield, through mitosis, the highly proliferative, transiently amplifying corneal peripheral cells. This differentiation is characterized by the expression of tissue-specific cytokeratins, as well as the loss of alpha-enolase and pigmentation. All these are intracellular events. The aim of this study was to identify and characterize, through lectin analysis, changes in cell surface properties associated with differentiation.

METHODS

Cryostat sections of the limbo-corneal area from freshly dissected pigmented rabbit corneas were stained with fluorescent lectins.

RESULTS

Peanut lectin (PNA; binds to Ser/Threo-GalNAc-beta-1,3-Gal, if the Gal residue is not sialylated) stained the plasma membrane of all layers of the conjunctiva and limbus but was excluded from corneal cell membranes. Maakia amurensis agglutinin (MAA; binds to sialic acid attached to galactose through alpha-2,3 bonds in either N-glycans or O-glycans) stained exclusively corneal cell plasma membrane. After complete tissue desialylation, all corneal plasma membranes became PNA positive with equal stain intensity across both sides of the limbo-corneal margin. The binding of the agglutinins from Limax flavus (binds unselectively to sialic acid) and Sambucus nigra (binds to sialic acid attached through alpha-2,6 bonds) to the basement membrane displayed a large increase at the corneal side of limbo-corneal demarcation.

CONCLUSIONS

Limbal (stem) cells express on the cell surface unsialylated galactose residues that are recognized by PNA and that lack any sialic acid bound through alpha-2,3 bonds. The initial differentiation involves sialylation of these residues and the concurrent appearance of alpha-2,3 sialic acid residues, suggesting expression or activation of alpha-2-3 sialytransferase. Changes in basement membrane composition, charge, or both may underpin this expression.

Na(+-) Cl(-)- and HCO3(-)-dependent base uptake in the ciliary body pigment pigment epithelium

Segments of whole ciliary body dissected from Dutch belted rabbits were incubated for 60 min at 36 degrees C in a 30 microM Ca2+ Ringer's. The inner limiting membrane with its adherent non-pigmented epithelium then was mechanically removed from the surface. The newly-exposed viable layer of pigmented cells was then loaded with the fluorescent probe 2'-7'-bis (carboxymethyl)-5(6) carboxyfluorescein by incubating the segments for 45 min at RT with the cell permeable acetoxymethoxy form of the dye. These loaded tissues were perfused in a flow-through chamber, mounted on the heated stage of a microscope equipped for quantitative epifluorescence, and the intracellular pH (pHi) of small groups of cells was derived from the ratio of emission intensities generated by excitations at 490 and 440 nm, respectively. In N[2-hydroxyethyl] piperazine-N"-[2 ethane sulfonic acid](Hepes)-buffered Ringer's the intracellular pH was 7.23 +/- 0.21 (+/- S.D., n = 20). Replacement of 28 mM Hepes by 28 mM HCO3-/5% CO2 led to a 0.14 +/- 0.04 increase in pHi. This increase required the presence of Na+ and Cl- and was inhibited by 0.2 mM diisothiocyanatostilbene-2-2'-disulfonic acid. These observations as well as characteristic pHi, responses to the removal or introduction of Na+ or Cl- indicated the presence in the pigmented cells of a Na(+)- and Cl(-)-dependent HCO3- transporter responsible for base uptake.

Synthesis of rabbit corneal epithelial glycocalyx in vitro

A method to study the synthesis and cellular processing of epithelial apical membrane glycoproteins in the rabbit cornea was developed. Fluorescent derivatives of wheat germ agglutinin (WGA; alpha-N-acetylglucosamine and sialic acid hapten affinities), succinylated WGA (alpha-N-acetylglucosamine hapten affinity) and concavalin A (Con A; D-mannose and D-glucose hapten affinity) were reacted with the corneal surface and the extent of binding attained was measured by en face, microscope-aided fluorophotometry. Minimal binding of succinylated WGA and a large reduction in WGA binding following neuraminidase treatment demonstrated that the attachment of WGA to the corneal surface occurred via sialic acid residues, i.e. via structures associated with terminal glycosylation. Corneas were treated with digitonin to induce the exfoliation of the outer squamous-like cell layers. The time-dependent changes in lectin binding density at the apical surface of the newly exposed intrastratal cells were then determined. Binding densities for WGA and Con A at the time of exfoliation of the digitonin-devitalized squamous cell layers (< 2 hr post-devitalization) were similar to the densities measured at the surface of untreated corneas. Over the subsequent 18-20 hr, the WGA and Con A binding increased by 2.63 +/- 0.24 and 3.0 +/- 0.68 (+/- S.D., n = 4) fold, respectively. The effect of inhibitors of transcription (actinomycin D, alpha-amanitin), translation (cycloheximide), core glycosylation of polypeptides (tunicamycin), endoplasmic reticulum glucosidases (deoxinojirimycin) and Golgi mannosidase (swainsonine) indicated that the increases were underpinned by new glycoprotein synthesis driven by a stable, pre-existing mRNA pool. Retinoic acid (2 microM) inhibited the increase in WGA binding by 55 +/- 6% (n = 4) but did not affect the Con A density increase suggesting that this agent either, modifies the terminal glycosylation pattern of apical membrane proteins and/or inhibits the synthesis of proteins bearing sialic acid. Actinomycin D or alpha-aminitin reverted the retinoic acid action, indicating that the retinoid effect is mediated by induced gene expression.

ZO-1 in corneal epithelium; stratal distribution and synthesis induction by outer cell removal

The tight junctions (TJs) present exclusively in between the superficial cells (SC) are an important component for the barrier and ion secretory functions of the corneal epithelium. Electrophysiological studies have demonstrated that digitonin-induced devitalization of overlying cells induces the de novo generation of a paracellular barrier between the intrastratal cells becoming exposed to the outer surface. It was also shown that the closer the exposed cells to the SC position the higher their competency for this activity. We have now examined the spatial and quantitative distribution of ZO-1, a protein closely associated with the cytosolic face of TJs in the rabbit. Immunohistology showed: (a) no detectable ZO-1 in basal cells; (b) incipient punctuate accumulations in the wing cells; (c) numerous foci and a diffuse cytosolic staining in the squamous (SQ) cells; and (d) strong staining at the apical TJ locations in the superficial SQ layer. Immunoblot analysis of separated layers showed a SQ:basal cell ZO-1 concentration ratio in excess of 100. Devitalization of the SQ layers induced strong ZO-1 synthesis in the newly exposed wing cells where electron microscopy showed time-dependent development of TJs. A similar ZO-1 increase was observed in the basal cells after removal of all suprabasal cells by a low [Ca2+] incubation method. These results provide a biochemical correlate to the previous electrophysiological findings and show that expression of ZO-1 in the corneal epithelium is intimately related to the development of the superficial cell phenotype.

Ontogeny of corneal epithelial tight junctions: stratal locale of biosynthetic activities

PURPOSE

On reaching the surface position, corneal epithelial cells assemble a tight paracellular barrier rapidly. The purpose of this study was to identify at what maturation stage the cells acquire biochemical elements underpinning this ability.

METHODS

Rabbit corneas were subjected to a digitonin-exposure protocol that results in the devitalization and desquamation of the three outermost cell layers of the 5-6-layer thick epithelium and thus, exposes winglike cells to the tear surface. Corneas were then mounted in Ussing-type chambers to measure transepithelial resistance and solute fluxes.

RESULTS

After the devitalization treatment transepithelial resistance remained near baseline for 2 hours and then rose to 8.24 +/- 3.12 K omega.cm2 (+/- SD, n = 15) during the next 8 hours. This increase was matched by increases in the resistance to the paracellular flow of mannitol. Transmission electron microscopy confirmed de novo formation of tight junctions over this time span in the new surface. The addition of transcription (actinomycin D; alpha-amanitin), translation (cycloheximide), core glycosylation (tunicamycin), and endoplasmic reticulum-to-Golgi traffic (brefeldin A) inhibitors immediately after the devitalization protocol prevented the transepithelial resistance raise. Introducing time delays between the cell devitalization and the addition of these biosynthesis inhibitors reduced their effect; a time delay of 90 minutes abolished the inhibition by alpha-amanitin.

CONCLUSION

Devitalization of overlying cells induces the assembly of a tight paracellular barrier in the wing cells. The process requires and includes the synthesis of essential transcripts and proteins that were initially absent in these cells. In vivo, the synthesis of these elements must occur as cells reach the squamous stage of maturation.

Separation of the rabbit ciliary body epithelial layers in viable form: identification of differences in bicarbonate transport

Sections of whole ciliary (CB) dissected from Dutch belted rabbits were incubated for 2 hr at 36 degrees C in a 30 microM Ca2+ Ringer's. This incubation resulted in the spontaneous dissociation of the two cell layers comprising this epithelium, each remaining firmly cohesive with its own basement membrane. The inner limiting membrane with its adherent non-pigmented epithelium (NPE) was then mechanically removed from the surface exposing the apical surface of the pigmented epithelium (PE). Ultrastructural examination revealed no noxious effects in most cells although gross morphological changes in the NPE cells were noted. The newly separated layers were loaded with the cell-permeable acetoxymethyl ester form of the fluorescent probe BCECF. Most cells of both layers acquired stable BCECF fluorescence indicating viability. To achieve a preliminary evaluation of differences in PE and NPE bicarbonate transport, dye-loaded tissues were perfused in a flow-through chamber which was mounted on a microscope equipped for quantitative epifluorescence. The intracellular pH (pHi) of groups of cells (5-10) was derived from the ratio of emission intensities generated by excitations at 490 and 440 nm. In Hepes-buffered Ringer's the pHis for the PE and NPE were 7.20 +/- 0.10 and 7.33 +/- 0.14 (+/- S.D., n = 6), respectively. Replacement of 28 mM Hepes by 28 mM HCO3-/5% CO2 led to a 0.13 pHi increase in the PE and a decrease of 0.27 U in the NPE. The pHi responses of the two cell layers to removal and/or reintroduction of Na+ and/or Cl- were also highly dissimilar.(ABSTRACT TRUNCATED AT 250 WORDS)

Evidence for parallel Na(+)-H+ and Na(+)-dependent Cl(-)-HCO3- exchangers in cultured bovine lens cells

BCECF, a cell-entrapable dye with a pH-sensitive fluorescence spectrum, was used to identify transport mechanisms contributing to pH homeostasis of cultured bovine lens epithelial cells. Cells from a spontaneously established lineage were grown on glass coverslips that fit diagonally in a standard curvette and intracellular pH (pHi) was measured. Under perfusion with a CO2-HCO3(-)-free medium (pH 7.45), pHi was 7.19 +/- 0.21 (mean +/- S.D., n = 94 cell preparations). Cell acidifications (pHi to 6.65, n = 8) induced by the 'NH(4+)-loading' method were rapidly followed by a Na(+)-dependent, amiloride-inhibitable pHi recovery. Introduction of a CO2-HCO3(-)-rich medium (pH 7.45) resulted in a small acidification (0.18 +/- 0.04 U, n = 16; P < 0.002) due to rapid CO2 entry and an ensuing slow alkalinization to a pHi near the control CO2-HCO3(-)-free value. Subsequent removal of Cl- resulted in a further alkalinization of 0.18 +/- 0.02 U (n = 13; P < 0.001). This Cl- effect was completely inhibited by the absence of Na+, but was insensitive to amiloride, suggesting the presence of a Na(+)-dependent Cl(-)-HCO3- exchanger. Consistent with this posit, the reintroduction of Na+ to cells perfused in the absence of the cation with a HCO3(-)-containing, amiloride-complemented solution resulted in a gradual recovery from the acidic pHi induced by the baseline conditions (n = 6). The amiloride-insensitive, Na(+)- and HCO3(-)-dependent recovery was completely inhibited in cells pre-incubated with DIDS.(ABSTRACT TRUNCATED AT 250 WORDS)

Gap junctions in rabbit corneal epithelium: limited permeability and inhibition by cAMP

Rabbit corneas were mounted in Ussing chambers, and the apical membrane of the superficial cells (SCs) was permeabilized by exposure to digitonin in a Ca(2+)-free Ringer solution. This treatment resulted in the generation of large (60.7 +/- 13.2 microA/cm2, n = 25) Na(+)-dependent tear (T)-to-stroma (S) short-circuit currents (Isc). The Isc was abolished by ouabain and by 1.4 mM Ca2+ and was inhibited by heptanol, 18 alpha-glycyrrhetinic acid, and dieldrin, effects consistent with the notion that corneal transepithelial fluxes include translocations through gap junctions (GJs) before basolateral membrane transport. T-to-S Isc were also generated when T-side Na+ was replaced by K+, eliciting a T-to-S K+ flux via basolateral K+ channels and when, with either Na+ or K+ on the T side, channels were introduced at the apical membrane with amphotericin B. The Isc in all four conditions exhibited similar sensitivity to GJ inhibitors and were inhibited by adenosine 3',5'-cyclic monophosphate (cAMP) elevation. Fluorophotometry combined with SC permeabilization with digitonin demonstrated that the half-time for the SC to sub-SC movement of 2',7'-bis(carboxyethyl)-5,6-carboxyfluorescein (mol wt 540) exceeded 3 h. These results indicate that junctional communications along the epithelial stratification axis are highly restricted and modulated by cAMP concentration.

A vesicular preparation enriched in basolateral membranes from cow ciliary body epithelium

A procedure is described for the preparation of a tissue fraction enriched in basal-lateral membranes of the epithelium of ciliary body dissected from abattoir-derived cow eyes. Microscopical methods established the viability of the epithelial cells in the dissected tissue. A two-step homogenization procedure was employed. First, mild trituration with a Potter-Elvehjem tissue grinder was used to release the epithelial cells with minimal disintegration of the stroma and muscle components. Then, complete vesiculation of the released material was induced with a Dounce homogenizer. This material was fractionated through sequential differential and (Ficoll 400) equilibrium density gradient centrifugation steps. The procedure led to the isolation of a fraction enriched 16.8 +/- 5.5 (+/- SE, n = 5) fold in Na(+)+K(+)-ATPase activity with respect to the initial homogenate. Using a method to artificially generate pH gradients it was demonstrated that the preparation contained sealed vesicles and that the membranes of these vesicles exhibited a K+ selectivity consistent with a plasma membrane origin.

Bicarbonate transport mechanisms in rabbit ciliary body epithelium

Sections of whole ciliary body dissected from Dutch belted rabbits were incubated with the cell entrappable pH probe BCECEF-AM. This led to a highly specific localization of epifluorescence emission at the exposed, non-pigmented cell layer (npe) of the dual layered epithelium that covers this organ. The BCECF-loaded tissue sections were superfused in a flow-through chamber and the intracellular pH (pHi) of small groups (10-20) of cells was derived from the ratio of the emission intensities derived from excitations at 490 and 440 nm. In CO2/HCO3- Ringer's, npe pHi = 7.09 +/- 0.11. Replacement of CO2/HCO3- by Hepes increased pHi by 0.22 +/- 0.02, indicating alkali secretory activity under the bicarbonate-rich conditions. Replacement of Cl- by gluconate elicited a rapid, 0.6-U increase in pHi. This effect exhibited little dependence on Na+ and was inhibited by 0.5 mM dihydro-4,4'-diisothiocyanatostilbene -2,2'-disulfonate (H2DIDS). These results indicate the presence of an electroneutral Cl-/base exchange activity. Elevation of [K-] (by partial replacement of Na+) also elicited increases in pHi. In Cl(-)-free media pHi reached 7.8-8.0, a condition under which intracellular [HCO3-] is at least twice as high as its extracellular value. This effect did not occur in the absence of Na+. The Na(+)-dependent high [K+]-induced pHi increase was inhibited by H2DIDS. The effects of Ba2+ on pHi, alone and in combination with high [K+], as well as that of full K+ removal, suggested that the link between high [K+] and pHi increase was mainly due to the effect of cell depolarization on an electronegative Na+ dependent HCO3- transporter. Under normal physiological conditions, the two acid/base transport systems are the main determinants of npe pHi.

Contribution from a pH- and tonicity-sensitive K+ conductance to toad translens short-circuit current

Studies of toad (Bufo marinus) lenses mounted in Ussing-type chambers revealed that: (1) the translens short-circuit current (Isc) across the posterior surface is primarily carried by the movement of Na+ from the posterior bathing solution to the lens; (2) across the anterior face the majority of the Isc is mediated by Ba(2+)-sensitive channels and the remaining current is rapidly reduced by ouabain; (3) most of the anterior K+ conductance is of the tonicity-sensitive, quinidine-inhibitable type (i.e. hypotonic shifts increase Isc and hypertonic shifts decrease Isc; quinidine pretreatment eliminates such responses); (4) 86Rb+ uptake is stimulated by alkaline pH and occurs primarily across the anterior surface with quinidine the most potent inhibitor of this process; and (5) the Na(+)-K+ pump can maintain lens [Na+] and [K+] for at least 20 hr in a Ringer's solution near neutral pH but not at pH 8.7 (a pH used in some studies with this lens). It is concluded that the Isc can be viewed as a representation of the current across the epithelial basolateral membrane, a surface dominated by pH- and tonicity-sensitive K+ channels. The direction of the Isc response to tonicity changes suggests a role for these channels in epithelial volume regulation.

HCO3- transport in the toad lens epithelium is mediated by an electronegative Na(+)-dependent symport

The pH-sensitive cell-entrapable dye 2',7'-bis(carboxyethyl)-5,6-carboxyfluorescein (BCECF) was used to continuously monitor epithelial intracellular pH (pHi) of intact toad lenses, enabling a description of a HCO3- transport mechanism that contributes to pHi homeostasis of this organ. In physiological medium, pH 7.40, the steady-state pHi was 7.48 +/- 0.03 (SE; n = 93). Induction of cell depolarization by either elevation of [K+] to 50 mM, addition of 0.2 mM quinidine, a K(+)-channel blocker, or addition of 0.1 mM Li+ ionophore that equalizes Na+ and K+ permeabilities elicited pHi increases (delta pHi = 0.18 +/- 0.02; P less than 0.0005; n = 13, for K+). These increases could be blocked or reverted by DIDS and were not affected by amiloride. Removal of Na+ induced an amiloride-insensitive acidification. pHi recovery seen upon Na+ reintroduction in the presence of amiloride was inhibited by DIDS. Despite the effects of DIDS on induced pHi changes, the agent did not affect control pHi. Elevation of medium HCO3- (pH to 7.7) produced a pHi increase followed by a spontaneous reversal. This increase was both DIDS and Na+ sensitive. pHi was not affected in any condition by removal (or addition) of Cl-, unless the lens was pretreated with the artificial Cl(-)-HCO3- exchanger tributyltin. Collectively, these results suggest that the primary mechanism for HCO3- movement across the lens epithelial membrane is an electronegative Na+ cotransporter and that this system is near equilibrium under normal physiological conditions.

Layer-by-layer desquamation of corneal epithelium and maturation of tear-facing membranes

A method to devitalize single layers of apically exposed rabbit corneal epithelial cells through the use of digitonin is described. Devitalized cells exfoliate spontaneously as loosely cohesive, trypan-blue-stained layers, exposing underlying viable cells. Repeated application of this devitalization-exfoliation methodology results in the gradual elimination of each of the epithelial cells. The generation of corneal surfaces composed of the tear-facing membranes of all intraepithelial cell types--subsurface, wing, and basal--is thus attainable. Exposed surfaces were studied with respect to microanatomy, the binding of lectins, and the adherence of Pseudomonas aeruginosa. Microprojections (microvilli or microplicae) were absent in the basal cells but were present in all suprabasal layers, and increased gradually in density as cells approached the surface position. Wheat germ agglutinin and concanavalin A were found to bind to the tear-facing membranes of all suprabasal cell layers. The tear-facing membrane of the basal cells, in contrast, was not labeled. Within each labeled layer, the magnitude of lectin binding differed markedly from cell to cell; lectin binding decreased as the cellular area exposed to the tear surface increased. Pseudomonas were found exclusively at microprojection-free cellular areas, suggesting that inhibition of attachment is linked to the ontogeny of these microprojections.

Pumps and pathways for gastric HCl secretion

Data reviewed herein show that the HCl-secreting parietal cell is an exaggerated example of dynamic membrane transformation. Recruitment and recycling of membrane provide the means for the massive redistribution of the gastric proton pump, the H,K-ATPase, from one membrane domain (cytoplasmic tubulovesicles) to another (apical plasma membrane) as a function of parietal cell activation and inactivation. Functional activation of HCl secretion requires not only the redistribution of pump protein, but also the participation of pathways for the rapid flux of K+ and Cl- across the apical membrane. In apical plasma membrane vesicles from stimulated cells these pathways appear to be conductive and can operate independently. Thus, our model for the parietal cell proposes that K+ and Cl- flux from cell to lumen, operating in parallel and in concert with ATP-driven H+/K+ exchange, provides the concentration and osmotic forces required for net HCl secretion. Whether and how the K+ and Cl- pathways are activated by stimulation and/or how they get to the apical membrane domain remain important questions. With respect to mechanisms of parietal cell activation, secretagogue-coupled elevation of cAMP and activation of protein kinase A form the basis of a well-established second messenger pathway. Several laboratories have identified various proteins that are phosphorylated concomitant with parietal cell stimulation, representing numerous candidates for effectors in stimulus-secretion coupling. Here, we emphasized the possible involvement of an 80-kDa protein whose phosphorylation was correlated with the cAMP pathway of HCl secretion. Immunocytolocalization of the 80-kDa phosphoprotein to the apical membrane and associated actin microfilaments prompted our suggestion that this protein might serve as a linkage between plasma membrane and cytoskeleton. Search for a possible role for the 80-kDa phosphoprotein in apical surface organization, stability, and turnover should represent an important thrust of research. Further understanding of the mechanism of cell activation will require a more complete elaboration of the functional role of many activation-related proteins.

Stimulation of toad lens epithelial Na+/H+ exchange activity by hypertonicity

Incubation of toad lenses with the acetoxymethyl ester of 2',7'-biscarboxyethyl-5(6)-carboxy-fluorescein led to a highly selective accumulation of the de-esterified, pH-sensitive form of the dye in the epithelial cells, enabling the continuous fluorometric monitoring of epithelial intracellular pH (pHi) in intact lenses. The effects of changes in extralenticular [Na+] and of amiloride-addition indicated that the epithelium contains an amiloride-sensitive Na+/H+ antiport. Exposure of lenses to hypertonic conditions (by the addition of sucrose to the medium) resulted in a biphasic change in pHi; a rapid initial, 'spike-like' decrease was immediately followed by a persistent reversal that raised pHi in CO2/HCO3- -rich and -free media by 0.13 and 0.18 units, respectively. Under CO2/HCO3- -free conditions, the hypertonic exposure raised pHi to a value near the calculated equilibrium position for a lens Na+/H+ exchanger. At this point, monensin addition did not affect pHi, suggesting that the tonicity shift had induced a rapid endogenous Na+/H+ exchange activity. In contrast, in the presence of 1 mM amiloride or in the absence of extralenticular Na+, sucrose addition induced only a persistent pHi decrease, which could be reversed (in the 'amiloride' case) by monensin addition. These results demonstrate that the hypertonic exposure induced an epithelial cell acidification as well as a stimulation of the Na+/H+ exchange activity which reverted the acidification. The hypertonic exposure also elicited pHi increases in lenses that had been preacidified by the 'NH4+ loading' or 'pCO2 raise' methods, indicating that the onset of the stimulation could not be attributed to a pHi decrease.(ABSTRACT TRUNCATED AT 250 WORDS)

Cellular pH and Na+-H+ exchange activity in lens epithelium of Bufo marinus toad

Incubation of lenses with the acetoxymethyl ester of 2,7-biscarboxyethyl-5(6)-carboxyfluorescein (BCECF-AM) led to selective, prolonged entrapment of the deesterified, pH-sensitive, membrane-impermeable form of the dye (BCECF) in the lens epithelial cells allowing, thereby, for the continuous fluorometric monitoring of epithelial intracellular pH (pHi) in intact lenses. Under anaerobic conditions in physiological solution, the epithelium maintained a pHi of 7.36 +/- 0.14 (mean +/- SD, n = 6 lenses). In the absence of CO2 and HCO3-, the pHi increased to 7.83 +/- 0.38 (n = 55). In this condition, either the removal of Na+ or the addition of amiloride to the medium resulted in a decrease of pHi. This acidification was increased in amiloride-inhibitable fashion when lenses were loaded with Na+ before extracellular Na+ removal. The extent of artificially generated cellular acidosis (induced by the "NH4+ loading" technique) was markedly decreased by the presence of extracellular Na+ or reverted on its addition when the ion was initially absent; the latter reversal was inhibited by amiloride. This inhibition was by-passed by monensin addition. The existence of a Na+-H+ antiport in the epithelial cells is concluded.

Regeneration of resistance and ion transport in rabbit corneal epithelium after induced surface cell exfoliation

Exposure of the rabbit corneal surface to a 20-microM digitonin-0.9% NaCl solution leads to permeabilization of the most superficial cells of the stratified epithelium. The devitalized cells exfoliate spontaneously from the corneal surface. Detergent exposure limited to 4-8 min leads to permeabilization and rapid exfoliation of a monolayer of surface cells. Consistent with the presence of the epithelial paracellular permeability barrier in this cell layer, their permeabilization results in complete loss of transepithelial resistance (Rt). Within minutes after detergent removal an initial recovery of Rt can be noticed indicating generation of a new paracellular permeability barrier by the viable sub-surface cells. This recovery proceeds rapidly and Rt reaches within 70 min a maximum equal to greater than 90% of the preexfoliation values (= 2.43 k omega.cm2, n = 22). The Rt recovery is fully blocked in a reversible manner by 10 microM dihydrocytochalasin B. The recovery is not affected by inhibition of protein synthesis with 5 microM cycloheximide. When the ocular surface is treated again with digitonin the permeabilization and exfoliation of a monolayer of cells and loss of Rt are repeated. After the second detergent exposure an initial recovery of Rt occurs as before within minutes. However, the pace of Rt recovery is much slower: 4-5 hr are required to reach a stable maximal Rt values amounting to about 73% of initial control. This recovery can be fully blocked by 5 microM cycloheximide indicating that protein synthesis is required for generation of tight junctions by the second subcellular layer. With only a fraction of Rt recovered, short-circuit currents amounting to, at least, 50% of control values and attributable in part to cell-to-tear movement of Cl- through the apical surface can be measured. This suggests that apical-type Cl- channels are either present in the apically facing membrane of subsurface cells or that they are rapidly inserted in it from preexisting intracellular pools immediately following the devitalization of the surface cells by digitonin.

Cl- secretagogues increase basolateral K+ conductance of frog corneal epithelium

The stromal-to-tear transport of Cl- by the corneal epithelium of the frog is increased by pharmacological effectors (secretagogues) that are known to raise the intracellular levels of cyclic AMP or Ca2+. It has been shown in the past that the Cl- secretagogues increase the apical membrane permeability to Cl- and thus facilitate the cell-to-tear flux of the anion. In this report, we combine transepithelial and microelectrode studies to show that three of these secretagogues, epinephrine, the Ca2+ ionophore A23187, and forskolin, also increase the K+ conductance of the basolateral membrane by two- to threefold. The increase in the K+ conductance is not dependent on membrane potential, since this increase occurred equally when the basolateral membrane potential either exceeded 60 mV, as measured with microelectrodes, or was depolarized by voltage clamping after apical permeabilization with amphotericin B. It is proposed that both Cl- and K+ conductances are under the control of intracellular mediators that act independently on each pathway. The increase in basolateral K+ conductance favors the Cl- secretory process.

K+ and Cl- conductances in the apical membrane from secreting oxyntic cells are concurrently inhibited by divalent cations

This study concerns the properties of rapid K+ and Cl transport pathways that are present in the (H+ + K+)-ATPase membrane from stimulated, and secreting, gastric oxyntic cells. Ion permeabilities in the isolated stimulation-associated vesicles were monitored via the rates of H+ efflux under conditions of exclusive H+/K+ counterflux or H+ - Cl co-efflux, as well as by comparison of equilibration rates for 86Rb and 36Cl under conditions of equilibrium exchange and unidirectional salt flux. These latter studies suggest that Rb+ and Cl pathways are conductive and independent. In spite of the functional independence of the ion pathways, several divalent cations inhibit Rb+ and Cl isotopic exchange as well as the H+ efflux that is dependent on either K+ or anion (Cl, SCN, NO2) fluxes. Zn2+ is the more potent inhibitor, reducing by 50% the sensitive component of K+, Cl, and NO2 fluxes at about 20 microM; Mn2+ has a similar effect at 200 microM. Ni2+ and Co2+ were roughly equipotent to Mn2+ while Mg2+ and Ca2+ had no inhibitory effect. These results suggest that the stimulation-induced permeabilities, while functioning independently, may be physically linked, i.e., residing within a single entity. In similar studies carried out in (H+ + K+)-ATPase vesicles obtained from nonstimulated cells, no vestiges of sensitivity to the inhibitory divalent cations could be detected. The implications of these findings for the physiology of the oxyntic cell in the context of a model for membrane fusion are discussed.

Ion transport studies with H+-K+-ATPase-rich vesicles: implications for HCl secretion and parietal cell physiology

A summary of recent studies on relations between the properties of the membrane incorporating the H+-K+-ATPase, the H+ motive force in gastric acid secretion, and the secretory state of the parietal cell is presented. Depending on tissue secretory state, two distinct H+-K+-ATPase-rich membranes predominate in tissue homogenates, the gastric microsomes derived from the intracellular tubulovesicles of the resting cell and the stimulation-associated (SA) vesicle derived from the apical membrane of the acid-secreting cell. Structural and chemical differences between both vesicular types lend support to the notion that the formation of an expanded, elaborated apical membrane in the secreting parietal cell results from fusion of tubulovesicles containing the H+-K+-ATPase to an apical membrane of different chemical composition. Comparison of polypeptide composition of microsomes and SA membranes provides a way to identify and isolate membrane and cytoskeletal components putatively involved in the membrane interconversion process. Comparison of transport properties between gastric microsomes and SA vesicles demonstrates that stimulation triggers the appearance of rapid K+ and Cl- permeabilities in the H+-K+-ATPase membrane, allowing efficient acid accumulation in SA vesicles by the combination of rapid KCl influx followed by ATPase-driven H+ for K+ exchange, i.e., by K+ recycling. These stimulation-triggered conductances are functionally independent. Nevertheless, their concurrent inhibition by certain divalent cations (Mn2+,Zn2+) suggests their location within a single physical domain. The compatibility of the K+-recycling model for HCl accumulation in SA vesicles with gastric HCl secretion and selected electrophysiological observations and certain implications of the findings for cellular mechanisms of transport regulation in the context of a membrane fusion and recycling model are discussed.

Na+/H+ exchange in the cyanobacterium Synechococcus 6311

The cyanobacterium Synechococcus 6311 adapts to grow in 0.6 M NaCl by developing an efficient system for sodium extrusion. In the present investigation cells loaded with NaC1 were subjected to a large dilution. Changes in fluorescence quenching of acridine orange as a function of transmembrane Na+ gradients provide evidence that Na+/H+ exchange activity greatly enhanced in salt-adapted cells.

Stimulation of oxyntic cell triggers K+ and Cl- conductances in apical H+-K+-ATPase membrane

Vesicles isolated from the apical membrane of stimulated oxyntic cells [stimulation-associated (SA) vesicles] are highly permeable to KCl. The KCl flux is coupled to an electroneutral ATP-driven H+-K+ exchange (the H+-K+-ATPase) to produce net intravesicular HCl accumulation. In the past, we observed that rates of KCl transport were not accelerated by valinomycin and that dissipation of preformed H+ gradients in the presence of a protonophore (carbonyl cyanide, m-chlorophenylhydrazone, 10 microM) required the simultaneous presence of valinomycin. Consequently the fast KCl transport was attributed to an electroneutral cotransport system. Now we have been able to elicit fast H+ gradient dissipation in the absence of valinomycin by using the protonophore tetrachlorosalicylanilide. Experiments carried out in the absence of Cl- demonstrated the existence of a specific high-conductance pathway for K+. Experiments in K+-free medium demonstrated the existence of a high Cl- conductance. Parallel experiments in the equivalent H+-K+-ATPase-rich vesicles from nonsecreting oxyntic cells showed very little K+ and Cl- conductivity, suggesting that the appearance of large ionic conductance in the membrane is associated with the stimulation of the cell.

Actin and associated proteins in gastric epithelial cells

A quantitative assessment of the distribution and state of microfilament-related proteins in the heterocellular fundic gastric epithelium was carried out. Actin content, as determined by the DNAase inhibition assay, ranged from 29 to 42 micrograms/mg of tissue protein, depending upon the tissue source. About 60% of the total actin existed in fresh tissue in the polymeric form (F-actin). The distribution of fluorescent-labelled phallicidin demonstrated that F-actin was concentrated predominantly in the acid-secreting oxyntic cells. The patterns of distribution corresponded to the location of the numerous elongated apical surface microvilli seen within oxyntic cell canaliculi. In the isolated apical membrane, actin represented about 10% of the total protein and was present entirely as F-actin. After mild treatment of membranes with Triton X-100, filaments could be readily visualized by negative staining. More extensive Triton X-100 extraction solubilized intrinsic membrane protein and yielded an insoluble residue highly enriched in actin and containing several additional polypeptides. Homogenization and fractionation of the gastric epithelium in low ionic strength media led to the depolymerization of a significant proportion of the tissue actin which was recovered in the homogenate supernatant. When purified by DNAase affinity chromatography, this gastric actin displayed structural and functional properties similar to muscle actin. Incubation of the homogenate supernatant in KCl-Mg2+ induced the formation of actin-rich gels. The gels contained myosin as well as several other peptides that may be actin-binding proteins.

Muscarinic receptors and guanylate cyclase in mammalian gastric glandular cells

To investigate the involvement of guanosine 3',5'-cyclic monophosphate (cGMP) in the cholinergic activation of gastric acid and pepsinogen secretion, we studied the subcellular and cellular relation between particulate guanylate cyclase and muscarinic cholinergic receptor sites. Subcellular fractionation of homogenates from rabbit gastric glands showed that particulate guanylate cyclase and muscarinic receptors were distributed in similar patterns, which differed from the pattern found for Na+-K+-ATPase, a marker for basal-lateral plasma membranes. Assuming a basal-lateral membrane localization for particulate guanylate cyclase and cholinergic receptors, these results suggested a heterogeneity of glandular basal-lateral membranes. The distributions of these markers among fractions enriched in isolated canine parietal or chief cells were also followed. Na+-K+-ATPase correlated with parietal cell distribution (r = 0.86) and guanylate cyclase with chief cell distribution (r = 0.76). The distribution of quinuclidinyl benzilate (QNB) binding sites indicated association of muscarinic receptors with both cell types. The similar subcellular and cellular distributions of guanylate cyclase and QNB binding sites may reflect a functional relationship of these markers in muscarinic-activated pepsinogen secretion. As seen in most other tissues, gastric glandular guanylate cyclase was not stimulated by various gastric secretagogues. We found that small changes in Ca2+ concentration, within the micromolar range, can regulate glandular guanylate cyclase activity. These results are discussed in terms of the cholinergic activation of parietal and chief cell function.

Correlation of parietal cell structure and function

The apical surface of the gastric parietal cell is greatly expanded (5-10-fold) during maximal HCl secretion, as compared to the resting cell. The membrane recycling hypothesis has been proposed to account for the extensive, functionally related, rearrangement of cell membranes. Cytoplasmic membranes within the resting cell, the tubulovesicles, contain the H+/K+-ATPase. Fusion of tubulovesicles with the apical plasma membrane occurs when the cells are stimulated, thus providing the increased surface area and proper disposition of the H+ pump enzyme. Microfilaments, composed of actin and other regulatory proteins, serve to direct the reordering of the apical surface during stages of the secretory cycle. Cell fractionation of resting oxyntic mucosa reveals that virtually all of the H+/K+-ATPase activity is associated with light microsomal membrane vesicles, presumably derived from tubulovesicles. Although the enzyme from resting tissue is fully competent (e.g. ATP-driving pump, H+-K+ exchange), the microsomal vesicles lack an endogenous pathway to provide rapid access for K+ to its intravesicular activity site. In stimulated stomach, there is a redistribution of H+/K+-ATPase to a larger, denser membrane fraction, the so-called stimulation-associated vesicles. Morphological features and chemical content (e.g. microfilament proteins) suggest that the stimulation-associated vesicles are derived from the expanded apical surface of the stimulated oxyntic cell. A KCl cotransport system has been identified in the stimulation-associated membranes, which operates in parallel with the ATP-driven H+-K+ exchange pump. These two transport systems operate in concert within the apical membrane to provide the machinery for net HCl transport by the parietal cell.

Anion exchange in oxyntic cell apical membrane: relationship to thiocyanate inhibition of acid secretion

The effects of SCN- on H+-accumulation by inside-out gastric vesicles derived from the apical membrane of secreting oxyntic cells are reported. SCN- inhibited the formation of pH gradients in Cl- and isethionate media. In Cl-, the concentration of SCN- required to achieve a certain degree of inhibition of H+ uptake (or dissipation of performed gradients) was increased with the increase in Cl- concentration, indicating some competitive phenomena between these anions. Comparison of the rates of dissipation of similar pH gradients achieved in Cl- vs. isethionate suggested the existence of a fast Cl-/SCN- exchange. In addition, direct isotopic fluxes confirmed the existence of rapid anion exchange and K-salt transport for both Cl- and SCN-. The rates of anion-exchange and K-salt transport were of similar magnitude, and the rates for SCN- in either countertransport against Cl- or cotransport with K+ were twice as fast as the equivalent values for Cl-. These mediated pathways in the apical membrane provide the possible means for rapid access of SCN- to the acidic canalicular spaces of the oxyntic cell that is implicit in recent proposals to explain SCN- inhibition of gastric HCl secretion.

Ultrastructural changes related to functional activity in gastric oxyntic cells

When stimulated to secrete HCl the gastric oxyntic cell undergoes profound morphological change. The identifiable apical cell surface is greatly expanded in the stimulated oxyntic cell as compared with nonsecreting ones. To account for this change, one hypothesis proposes that the expanded surface is derived from the fusion of cytoplasmic tubulovesicular membranes with the existing limited apical membrane surface. An alternative hypothesis suggests that the tubulovesicular compartment is actually confluent with the apical surface at all times and that the morphological appearance follows the expansion of this supercollapsed compartment as HCl secretion commences. A variety of morphological evidence is reviewed here including transmission electron microscopy during various stages of secretion and inhibition, analysis of freeze-fracture replicas, penetration of macromolecular tracers, and membrane surface-staining characteristics. It is concluded that the weight of evidence favors a membrane fusion process. Moreover, recent comparative studies of membrane fractions from resting and secreting stomachs show different morphological and functional properties that are also consistent with a fusion hypothesis as a fundamental event in the membrane transformation of the oxyntic cell.

Changes in the membrane environment of the (K+ + H+)-ATPase following stimulation of the gastric oxyntic cell

Biochemical evidence is presented for changes in the membrane environment of the (K+ + H+)-dependent ATPase enzyme of the oxyntic cell following in vivo gastric stimulation of young New Zealand rabbits. The changes are inferred from the marked differences in the sedimentation properties of the (K+ + H+)-ATPase when obtained from homogenates of either stimulated or nonstimulated (resting) fundic gastric epithelium. Stimulation resulted in a redistribution of K+-ATPase activity that was reduced to less than half in the microsomal pellet and concomitantly increased in the membrane fractions normally associated with nuclei and mitochondria. Density gradient fractionation of the mitochondrial pellet yield a preparation rich in (K+ + H+)-ATPase. Our studies indicated that the membranes in this preparation are far larger and apparently denser than the microsomal vesicles associated with the nonstimulated state of the cell. The specific nature of the relationship between stimulation and the observed changes is suggested by the lack of change in the distribution of enzymatic activities unrelated to the apical pole of the oxyntic cell. Preliminary, tentative information aimed at identifying the processes responsible for the observed changes is presented.

Stoichiometry of the ligand-binding sites in the acetylcholine-receptor oligomer from muscle and from electric organ. Measurement by affinity alkylation with bromoacetylcholine

1. The affinity alkylation reaction of the cholinergic, depolarising ligand, bromoacetylcholine with reduced acetylcholine receptor in the membrane fragments of Torpedo marmorata and in Triton-solubilised receptor from cat denervated muscle has been studied. 2. Brief pretreatment with 100 microM bromoacetylcholine abolishes all [3H]alpha-neurotoxin binding in both cases. 3. In the receptor from each of these sources, the number of sites of specific alpha-neurotoxin binding is exactly equal to the number of sites that can be specifically alkylated by bromol[3H]acetytlcholine, at saturation of either ligand. 4. The concentration-dependence of specific bromo[3H]acetylcholine binding is found to be biphasic. A first phase can be clearly discerned in which one-half of the total specific ligand-binding sites are alkylated readily, and a second phase in which the remainder react at higher reagent concentrations. The same discrimination of two equal sets of ligand sites can be obtained by pre-blockade using low concentrations of unlabelled bromoacetylcholine followed by reaction with [3H]alpha-neurotoxin or bromo[3H]acetylcholine. 5. In both phases, a single subunit of Mr about 43000 is the sole site of specific alkylation in both Torpedo and muscle. The reasons for the appearance of two equal but distinct populations in the ligand binding sites in the receptors are discussed.

A procedure for membrane-protein reconstitution and the functional reconstitution of the anion transport system of the human-erythrocyte membrane

A short procedure for the isolation of band-3 protein, the protein responsible for anion exchange in erythrocytes, in a reasonable degree of purity was developed. Using this protein preparation and a novel procedure for membrane-protein reconstitution, vesicles displaying the basic features of the anion-exchange system of the erythrocyte were obtained. The reconstitution procedure is based on slow direct removal of Triton X-100 from aqueous lipid/detergent solutions. According to the composition of the reconstitution medium, either small single-walled or large multi-walled vesicles are obtained. The procedure conserves protein properties well, as is revealed by the similarity of the rates of SO4(2-) exchange in erythrocytes and reconstituted vesicles when corrected for the relevant volumes. A number of functional features of the exchange system were studied and compared with those of the native membrane.

Diffusion within egg lecithin bilayers resembles that within soft polymers

An analysis is presented of how the permeability coefficient/octanol:water partition coefficient ratio for 33 different chemical substances crossing egg lecithin bilayers depends on the molecular volume of the substances. From this analysis we conclude that bilayers made from egg lecithin behave as soft polymers in their discrimination between permeants of different sizes and shapes.

Functional characterization of anion transport system isolated from human erythrocyte membranes

The anion transport system of human red blood cells was isolated in vesicles containing the original lipids of the membrane and predominantly the 95,000-dalton polypeptides (Band 3) associated with intralipid particles. The vesicles display various characteristic properties of anion permeation closely resembling those of the native system. The properties include energy of activation, pH dependence, anion sleectivity, sensitivity to specific inhibitors, and exchange and net rates of sulfate transport. Based on these and other criteria, the functional properties of isolated vesicles could be equated with those of the intact cell system. Direct support for the involvement of 95,000-dalton polypeptides in permeation functions is provided.

The permeation of organic acids through lecithin bilayers. Resemblance to diffusion in polymers

1. The fluxes of aliphatic acids and their derivatives through black lipid membranes made of egg lecithin in decane were measured by means of a proton titration method. 2. Permeability coefficients were calculated and these were divided by the partition coefficient of the diffusing solute in different solvent systems: n-decane, olive oil, ether and octanol. The logarithms of the diffusion coefficients thus obtained were plotted against the logarithm of the molecular weight. The data could not be fitted to a single regression line in any solvent system. 3. When the logarithm of the diffusion coefficients were correlated to the logarithm of the molecular volume (equals molecular weight/ specific gravity) all the diffusants could be fitted to the same regression line, indicating that the molecular volume is a better index of molecular size and shape than the molecular weight. 4. Analysis of the experimental results assuming a model of diffusion through soft polymers (Lieb, W.R. and Stein, W. D. (1971) Current Topics in Membranes and Transport, vol. 2, pp. 1-39, Academic Press, New York) showed that decane and olive oil are not adequate model solvents for planar lecithin membranes but ether and octanol are good models. 5. The differential mass selectivity coefficient was found to be similar to that for soft polymers and biological membranes, i.e. greater than 3.0. 6. Water could be fitted by the same regression line, thus emphasizing the generality of passive transfer and implying that water crosses lipid membranes as single molecules.