Differences between liver gap junction protein and lens MIP 26 from rat: implications for tissue specificity of gap junctions

Experimental and Simulation Studies of Aquaporin 0 Water Permeability and Regulation

We begin with the history of aquaporin zero (AQP0), the most prevalent membrane protein in the eye lens, from the early days when AQP0 was a protein of unknown function known as Major Intrinsic Protein 26. We progress through its joining the aquaporin family as a water channel in its own right and discuss how regulation of its water permeability by pH and calcium came to be discovered experimentally and linked to lens homeostasis and development. We review the development of molecular dynamics (MD) simulations of lipid bilayers and membrane proteins, including aquaporins, with an emphasis on simulation studies that have elucidated the mechanisms of water conduction, selectivity, and proton exclusion by aquaporins in general. We also review experimental and theoretical progress toward understanding why mammalian AQP0 has a lower water permeability than other aquaporins and the evolution of our present understanding of how its water permeability is regulated by pH and calcium. Finally, we discuss how MD simulations have elucidated the nature of lipid interactions with AQP0.

The topological structure of connexin 26 and its distribution compared to connexin 32 in hepatic gap junctions

Of the gap junction proteins characterized to date, Cx26 is unique in that it is usually expressed in conjunction with other members of the family, typically Cx32 (liver [Nicholson et al., Nature 329:732-734, 1987], pancreas, kidney, and stomach [J.-T. Zhang, B.J. Nicholson, J. Cell Biol. 109:3391-3410, 1989]), or Cx43 (leptomeninges [D.C. Spray et al., Brain Res. 568:1-14, 1991] and pineal gland [J.C. Sáez et al., Brain Res. 568:265-275, 1991]). We have used specific antisera both to investigate the distribution of Cx32 and Cx26 in isolated liver gap junctions, and empirically establish the topological model of Cx26 suggested by its sequence and analogy to other connexins. Antipeptide antisera were prepared to four of the five hydrophilic domains which flank the four putative transmembrane spanning regions of Cx26. Antibodies to N-terminal residues 1-17 (alpha Cx26-N), to residues 101-119 in the putative cytoplasmic loop (alpha Cx26-CL), and to C-terminal residues 210-226 (alpha Cx26-C) were all specific for Cx26. An antibody to residues 166-185 between hydrophobic domains 3 and 4 of Cx32 had affinity for both Cx26 and Cx32 (alpha Cx32/26-E2). The antigenic sites Cx26-N, -CL and -C were each demonstrated to be cytoplasmically disposed, although the latter was conformationally hidden prior to partial proteolysis. The antigenic site for alpha Cx32/26-E2 was only accessible after exposure of the extracellular face by separation of the junctional membranes in 8 M urea, pH 12.3. This treatment also served to reveal the region between residues 45 and 66 to Asp-N protease. The topology thus demonstrated for Cx26 is consistent with that deduced for other connexins (i.e., Cx32 and Cx43). Comparison of immunogold decorated gap junctions reacted with antibodies specific to Cx26 (alpha Cx26-N and -CL), or to Cx32 [alpha Cx32-CL], indicates that these connexins do not aggregate in subdomains within a junction, at least within the resolution provided by the labeling density (one antibody per 15-22 connexons). Although the presence of both connexins within a single channel could not be distinguished, possible interactions between channels is discussed.

Multiple structural types of gap junctions in mouse lens

Gap junctions in the epithelium and superficial fiber cells from young mice were examined in lenses prepared by rapid-freezing, and processed for freeze-substitution and freeze-fracture electron microscopy. There appeared to be three structural types of gap junction: one type between epithelial cells and two types between fiber cells. Epithelial gap junctions seen by freeze-substitution were approximately 20 nm thick and consistently associated with layers of dense material lying along both cytoplasmic surfaces. Fiber gap junctions, in contrast, were 15–16 nm (type 1) or 17–18 nm thick (type 2), and had little associated cytoplasmic material. Type 1 fiber gap junctions were extensive in flat expanses of cell membrane and had a thin, discontinuous central lamina, whereas type 2 fiber gap junctions were associated with the ball-and-socket domains and exhibited a dense, continuous central lamina. Both types of fiber gap junction had a diffuse arrangement of junctional intramembrane particles, whereas particles and pits of epithelial gap junctions were in a tight, hexagonal configuration. The type 2 fiber gap junctions, however, had a larger particle size (approximately 9 nm) than the type 1 (approximately 7.5 nm). In addition, a large number of junctional particles typified the E-faces of both fiber types but not the epithelial type of gap junction. Gap junctions between fiber and epithelial cells had structural features of type 1 fiber gap junctions.(ABSTRACT TRUNCATED AT 250 WORDS)

The crystalline lens. A system networked by gap junctional intercellular communication

The vertebrate eye lens is a solid cyst of cells which grows throughout life by addition of new cells at the surface. The older cells, buried by the newer generations, differentiate into long, prismatic fibers, losing their cellular organelles and filling their cytoplasms with high concentrations of soluble proteins, the crystallins. The long-lived lens fibers are interconnected by gap junctions, both with themselves and with an anterior layer of simple cuboidal epithelial cells at the lens surface. This network of gap junctions joins the lens cells into a syncytium with respect to small molecules, permitting metabolic co-operation: intercellular diffusion of ions, metabolites, and water. In contact with nutrients at the lens surface, the epithelial cells retain their cellular organelles, and are able to provide the metabolic energy to maintain correct ion and metabolite concentrations within the lens fiber cytoplasms, such that the crystallins remain in solution and do not aggregate (cataract). Gap junctions are formed by a family of integral membrane channel-forming proteins called connexins. Gap junctions between lens epithelial cells are composed of a connexin which is common between many different cell types, notably myocardial cells and connective tissue fibroblasts. The gap junctions between epithelial cells and lens fibers have not yet been biochemically characterized. The gap junctions formed between lens fibers are composed of at least two different connexins, one of which has not been detected between other cell types. The unusual physiology and longevity of the lens fibers may require the special set of connexins which are found joining these cells.

MP26, a protein of intercellular junctions in the bovine lens: electrophoretic and chromatographic characterization

We have characterized the membrane protein of apparent molecular weight 26 kD from bovine lenses (MP26 or MIP) with respect to six different electrophoretic and chromatographic procedures. These include one- and two-dimensional gel electrophoretic procedures, as well as SDS-hydroxylapatite chromatography. The two-dimensional gels include isoelectric focusing with both conventional ampholytes and buffer focusing methods. With buffer focusing, the membranes are solubilized without the use of SDS and the isoelectric focusing is performed in the absence of SDS. As specific probes for MP26, a monoclonal antibody and an anti-MP26 rabbit serum were used, the latter prepared against electrophoretically purified MP26. These separation techniques were adapted to MP26 in order to permit a more detailed characterization of this protein and to search for any heterogeneity in this size range, specifically other junctional proteins or protein fragments. We have found evidence for charge heterogeneity in MP26, but no evidence for multiple membrane proteins of Mr 26,000 in urea-treated membranes. The charge heterogeneity appears to be related to a phosphorylation of MP26. The results reported here aid the interpretation of a variety of data, especially findings on the reconstitution of MP26 in artificial membranes and results from work with polyclonal MP26 antibodies. These investigations are all designed to evaluate the proposed role of MP26 as a protein of cell-to-cell channels in the lens fiber cell.

Evaluation of lens epithelial cell differentiation by quantitation of MP26 mRNA relative to gamma-crystallin mRNA in initiation of galactose cataracts in the rat

It is well established that in response to feeding of galactose to four-week-old rats cataracts develop (Unakar, Genyea, Reddan & Reddy, Exp. Eye Res. 26, 123-33, 1978). Initiation of cataracts leads to epithelial cell proliferation. In the lens, the main intrinsic membrane protein, MP26, is recognized as a fiber cell-specific gene product, and gamma-crystallin is found in terminally differentiated fiber cells. By Northern blot analysis we quantitated the MP26 and gamma-crystallin mRNAs found in the lens at various times spanning from 0 to 96 hr on galactose. We find that by 24 hr, MP26 mRNA, as well as gamma-crystallin mRNA, increased significantly above the zero time levels. The data was also confirmed by in situ hybridization of various lens sections to both gamma-crystallin and MP26 35S-labeled antisense RNA probes. As expected, at later periods beyond 24 hr, the levels of gamma-crystallin and MP26 mRNAs dropped to below the control levels. The initial increase in MP26 and gamma-crystallin mRNAs lead us to conclude that initiation of galactose cataracts appears to support, although for a short time period, both epithelial cell elongation and fiber cell differentiation.

Amino acid sequence of in vivo phosphorylation sites in the main intrinsic protein (MIP) of lens membranes

The main intrinsic membrane protein of the lens fiber cell, MIP, has been previously shown to be phosphorylated in preparations of lens fragments. Phosphorylation occurred on serine residues near the cytoplasmic C-terminus of the molecule. Since MIP is thought to function as a channel protein in lens plasma membranes, possibly as a cell-to-cell channel protein, phosphorylation could regulate the assembly or gating of these channels. We sought to identify the specific serines which are phosphorylated in order to help identify the kinases involved in regulating MIP function. To this end we purified a peptide fragment from native membranes that had not been subjected to any exogenous kinases or kinase activators. Any phosphorylation detected in these fragments must be due to cellular phosphorylation and thus is termed in vivo phosphorylation. Purified membranes were also phosphorylated with cAMP-dependent protein kinase to determine the mobility of phosphorylated and unphosphorylated MIP-derived peptides on different HPLC columns and to determine possible cAMP-dependent protein kinase phosphorylation sites. Lens membranes, which contain 50-60% of the protein as MIP, were digested with lysylendopeptidase C. Peptides were released from the C-terminal region of MIP and a major product of 21-22 kDa remained membrane-associated. Separation of the lysylendopeptidase-C-released peptides on C8 reversed-phase HPLC demonstrated that one of these fragments, corresponding to residues 239-259 in MIP, was partially phosphorylated. The phosphorylated and nonphosphorylated forms of this peptide were separated on QAE HPLC. In vivo phosphorylation sites were found at residues 243 and 245 through phosphoserine modification via ethanethiol and sequence analysis. Phosphorylation was never detected on serine 240. The phosphorylation level of serine 243 could be increased by incubation of membranes with cAMP-dependent protein kinase under standard assay conditions. Other kinases that phosphorylate serines found near acidic amino acids must be responsible for the in vivo phosphorylation demonstrated at serine 245.

Properties of channels reconstituted from the major intrinsic protein of lens fiber membranes

Detergent-solubilized plasma membrane protein of either adult bovine or calf lens and high-performance liquid chromatography-purified major intrinsic protein (MIP) of the lens were reconstituted into unilamellar vesicles and planar lipid bilayers. Freeze-fracture studies showed that the density of intramembrane particles in the vesicles was proportional to the protein/lipid ratio. At high ratios, these particles crystallized into tetragonal arrays as does MIP in lens fibers. Channels induced by either purified MIP or detergent-solubilized protein had essentially identical properties. The conductance of multichannel membranes was maximal near 0 mV and decreased to 0.49 +/- 0.08 of the maximum value at voltages greater than 80 mV. The dependence of the conductance on voltage was well fit by a two-state Boltzmann distribution. Voltage steps greater than 30 mV elicited an ohmic current step followed by a slow (seconds) biexponential decrease. The amplitudes and time constants depended on the magnitude but not the sign of the voltage. Steps from 100 mV to voltages less than 30 mV caused the channels to open exponentially with a millisecond time constant. Analysis of latency to first closure after a voltage step gave nearly the same time constants as multichannel kinetics. Single-channel conductance is proportional to salt concentration from 0.1 to 1.0 M in KCl. In 0.1M KCl, the channel had two preferred conductance states with amplitudes of 380 and 160 pS, as well as three additional substates. Multi- and single-channel data suggest that the channel has two kinetically important open states. The channel is slightly anion selective. The properties of the channel do not vary appreciably from pH 7.4 to 5.8 or from pCa 7 to 2. We propose that a channel with these properties could contribute to maintenance of lens transparency and fluid balance.

Membrane and junctional properties of dissociated frog lens epithelial cells

Individual cells and cell pairs were isolated from frog lens epithelium. Individual cells were whole cell voltage clamped and the current-voltage relationship was determined. The cells had a mean resting voltage of -54.3 mV and a mean input resistance of 1.4 G omega. The current-voltage relationship was linear near the cell resting voltage, but showed decreased resistance with large depolarization or hyperpolarization. Junctional currents between pairs of cells were recorded using the dual whole cell voltage-clamp technique. The corrected junctional resistance was 15.5 M omega (64.5 nS). The junctional current-voltage relationship was linear. A combination of ATP and cAMP, in the electrodes, stabilized junctional resistance. Currents recorded when uncoupling was nearly complete, showed evidence of single connexion gating events. A single-channel conductance of about 100 pS was prominent. Dye spread between isolated cell pairs was demonstrated using Lucifer Yellow CH in a whole cell configuration. Photodamage to the cells due to the dye was apparent. Dye loaded cells, in the presence of exciting light, showed decreased resting voltages, decreased input resistances and morphological changes. Glutathione (20 mM) delayed this damage.

The structural organization and protein composition of lens fiber junctions

The structural organization and protein composition of lens fiber junctions isolated from adult bovine and calf lenses were studied using combined electron microscopy, immunolocalization with monoclonal and polyclonal anti-MIP and anti-MP70 (two putative gap junction-forming proteins), and freeze-fracture and label-fracture methods. The major intrinsic protein of lens plasma membranes (MIP) was localized in single membranes and in an extensive network of junctions having flat and undulating surface topologies. In wavy junctions, polyclonal and monoclonal anti-MIPs labeled only the cytoplasmic surface of the convex membrane of the junction. Label-fracture experiments demonstrated that the convex membrane contained MIP arranged in tetragonal arrays 6-7 nm in unit cell dimension. The apposing concave membrane of the junction displayed fracture faces without intramembrane particles or pits. Therefore, wavy junctions are asymmetric structures composed of MIP crystals abutted against particle-free membranes. In thin junctions, anti-MIP labeled the cytoplasmic surfaces of both apposing membranes with varying degrees of asymmetry. In thin junctions, MIP was found organized in both small clusters and single membranes. These small clusters also abut against particle-free apposing membranes, probably in a staggered or checkerboard pattern. Thus, the structure of thin and wavy junctions differed only in the extent of crystallization of MIP, a property that can explain why this protein can produce two different antibody-labeling patterns. A conclusion of this study is that wavy and thin junctions do not contain coaxially aligned channels, and, in these junctions, MIP is unlikely to form gap junction-like channels. We suggest MIP may behave as an intercellular adhesion protein which can also act as a volume-regulating channel to collapse the lens extracellular space. Junctions constructed of MP70 have a wider overall thickness (18-20 nm) and are abundant in the cortical regions of the lens. A monoclonal antibody raised against this protein labeled these thicker junctions on the cytoplasmic surfaces of both apposing membranes. Thick junctions also contained isolated clusters of MIP inside the plaques of MP70. The role of thick junctions in lens physiology remains to be determined.

Tissue-specific granularity of gap junction cytoplasmic surfaces revealed by rapid-freeze, deep-etch replicas

Previous rapid-freeze, deep-etch replica studies have revealed the differences between heart and liver gap junctions; cytoplasmic surfaces of in situ and phenylmethylsulfonyl fluoride (PMSF)-unproteolyzed isolated cardiac gap junctions (MW 47 kD) have a particulate substructure, which is absent both in the proteolyzed heart junctions (MW 29 kD) and in the liver junctions isolated with PMSF (MW 28 kD). The present deep-etch replica studies of gap junction cytoplasmic surface (CS) membranes in several tissues of rats and mice were performed to examine whether or not this difference between liver and heart is typical of variations in gap junction proteins from tissue to tissue. In surface mucous cells of the stomach, intestinal epithelial cells, and kidney tubule cells, these epithelial gap junctions always showed smooth cytoplasmic surfaces, similar to the liver gap junctions. In contrast, in the atrial myocardium, aortic endothelium, and the ciliary process, cytoplasmic surface membranes of the gap junctions consistently revealed particulate patterns. Close examinations disclosed that those granular structures were not merely attached to the memvrane surface, but they also protruded from the membrane interior as an integral component of gap junction particles. Furthermore, in the pregnant rat uterus at term, cytoplasmic surface membranes of myometrial smooth muscle gap junctions were particulate, but those of endometrial epithelium were smooth. The present observation suggest that tissue specificity exists in cytoplasmic surface structures of gap junctions between the "true" epithelial and the nonepithelial tissues: the nonepithelial gap junctions contain the additional cytoplasmic surface domain that is absent in the gap junctions of "true" epithelial origin.

Inhibition of gap-junctional intercellular communication between epithelial cells transformed by the activated H-ras-1 oncogene

In order to study the effects of an activated H-ras-1 oncogene on gap-junctional intercellular communication, we introduced the EJ/T24 H-ras-1 oncogene into cells of the epithelial Clone 9-3 cell line. Gap-junctional intercellular communication was significantly reduced in H-ras-1-transformed Clone 9-3 derivatives; this result shows that transformation by the activated H-ras-1 oncogene can inhibit gap-junctional intercellular communication. We postulate that the activated H-ras-1 oncogene product could mediate this effect through a change in the phosphorylation of the major gap-junction protein.

Structural and molecular biology of the eye lens membranes

Lens transparency is associated with a unique design in tissue development and architecture. The fiber plasma membrane has domains which link with the cytoskeleton, thus maintaining cell shape. Other membrane regions form processes which interlock adjacent lens fibers, and intercellular junctions contain transmembrane pores which allow passage of metabolites between cells. Much interest has recently focused on the study of lens membrane structure and function, mainly because membrane dysfunction may be associated with cataract formation. This article reviews what is known about the structure of membrane domains, about the identification of domain-specific proteins, and describes current attempts to relate these results to function. Much of the presently available data is controversial, and an attempt will be made to reconcile them in revised models and testable hypotheses.

The arthropod gap junction and pseudo-gap junction: isolation and preliminary biochemical analysis

The hepatopancreas of the crayfish, Procambarus clarkii, contains an unusual abundance of gap junctions, suggesting that this tissue might provide an ideal source from which to isolate the arthropod-type of gap junction. A membrane fraction obtained by subcellular fractionation of this organ contained smooth septate junctions, zonulae adhaerentes, gap junctions and pentalaminar membrane structures (pseudo-gap junctions) as determined by electron microscopy. A further enrichment of plasma membranes and gap junctions was achieved by the use of linear sucrose gradients and extraction with 5 mM NaOH. The enrichment of gap junctions correlated with the enrichment of a 31 Kd protein band on polyacrylamide gels. Extraction with greater than or equal to 20 mM NaOH or greater than or equal to 0.5% (w/v) Sarkosyl NL97 resulted in the disruption and/or solubilization of gap junctions. Negative staining revealed a uniform population of 9.6 nm diameter subunits within the gap junctions with an apparent sixfold symmetry. Using antisera to the major gap junctional protein of rat liver (32 Kd) and to the lens membrane protein (MP 26), we failed to detect any homologous antigenic components in the arthropod material by immunoblotting-enriched gap junction fractions or by immunofluorescence on tissue sections. The enrichment of another membrane structure (pseudo-gap junctions), closely resembling a gap junction, correlated with the enrichment of two protein bands, 17 and 16 Kd, on polyacrylamide gels. These structures appeared to have originated from intracellular myelin-like figures in phagolysosomal structures. They could be distinguished from gap junctions on the basis of their thickness, detergent-alkali insolubility, and lack of association with other plasma membrane structures, such as the septate junction. Pseudo-gap junctions may be related to a class of pentalaminar contacts among membranes involved in intracellular fusion in many eukaryotic cell types. We conclude that pseudo-gap junctions and gap junctions are different cellular structures, and that gap junctions from this arthropod tissue are uniquely different from mammalian gap junctions of rat liver in their detergent-alkali solubility, equilibrium density on sucrose gradients, and protein content (antigenic properties).

MP17, a fiber-specific intrinsic membrane protein from mammalian eye lens

A major protein with a molecular weight of 17,000, designated as MP17, has been identified in mammalian eye lens plasma membranes. Hydrophobic photolabeling experiments revealed that MP17 is a genuine intrinsic membrane protein. By using monoclonal antibodies we demonstrated that MP17 is not detectable in liver, heart, muscle, spleen and kidney, and thus can be considered, like MP26, as a lens-specific membrane protein. Furthermore, we showed that MP17 is a substrate for cAMP-dependent protein kinase and that it is a calmodulin-binding protein.

Roles of cell junctions in gametogenesis and in early embryonic development

Recent reviews of the role of cell junctions in development have focused primarily upon functions related to the relatively subtle physiological modulation of their subunits in relation to fundamental developmental processes in a wide variety of organisms. There is, however, considerable support from numerous laboratories that the more radical modulation of the presence and number of junctional subunits in many diverse tissues may play a pivotal role in a wide spectrum of developmental phenomena ranging from gametogenesis to organogenesis. Since a great deal of recent interest in this latter subject has concentrated upon vertebrate systems including mammals, this review will examine the functional significance of the modulation of gap junctions, tight junctions and desmosomes in a developing idealized mammalian system from gamete formation to tissue and organ differentiation during embryo-genesis.

Dopaminergic regulation of horizontal cell gap junction particle density in goldfish retina

Light- or dark-adapted goldfish (Carassius auratus) retinas were treated with dopamine, which is believed to uncouple horizontal cells via D1 receptors, or with the dopamine antagonist haloperidol. Aldehyde-fixed retinas were freeze-fractured and the replicas examined by electron microscopy to identify horizontal gap junctions. The density (number per micron2) of intra-membrane particles of horizontal cell soma gap junctions was significantly lower in light-adapted and dopamine-treated retinas than in dark-adapted and haloperidol-treated retinas. There was no statistically significant difference between gap junction particles densities in (I) light-adapted (untreated) and in dopamine-treated (light- or dark-adapted) retinas, or between (II) dark-adapted (untreated) and haloperidol-treated (light- or dark-adapted). These results suggest that the uncoupling of horizontal cell somas by dopamine is accompanied by a decrease in gap junction particle density and that there is a greater release of dopamine during light-adaptation than dark-adaptation. Unlike horizontal cell somas, horizontal cell axon terminals did not show consistent changes in gap junction particle density with light- or dark-adaptation. Although the data suggests that there may be a reduction in axon terminal gap junction particle density with dopamine treatment, this effect is not reversible with haloperidol treatment. Our results suggest that the regulation of gap junctions may differ at two sites within the same cell.

Survival of fiber cells and fiber-cell messenger RNA in lens of rats maintained on a 50% galactose diet for 45 days

Previous work from this laboratory has shown that in cataractous lens from rats fed a 50% galactose diet up to 32 days, synthesis of crystallin mRNAs was reduced while synthesis of mRNAs for the non-crystallin proteins appeared to be maintained at equivalent or higher levels than found in the controls. In this study, we find that the population of mRNA from the 45-day cataractous lens included all of the crystallin mRNAs and substantial amounts of mRNAs for the non-crystallin proteins, in particular those proteins of molecular weights ranging from 45,000 to 92,000, and pIs from 5.5 to 7.0. At the 45-day cataractous stage, the observed mRNA products totaled between 100 and 120, while for the control they numbered between 40 and 50. In addition, hybrid-select measurements with a lambda gt11 MP26 clone established that MP26 mRNA persisted in the fiber cells of 11-, 20-, 32- and 45-day galactosemic lens. The data, therefore, suggest that continuous exposure of the lens to galactose apparently leads to significant fluctuations in mRNA synthesis and survival. Indirect immunofluorescence (with a monospecific polyclonal rabbit anti-MP26 antibody) and light-microscopy studies demonstrated that although the cortical fiber cells were swollen and structurally disoriented, they retained their nuclei, while the enucleated fiber cells were absent. The microscopy data also suggest that differentiation of the epithelial cells to fiber cells continued in the cataractous lens. Fiber-cell migration was backward as compared with the control, and the surviving cells were localized within the cortex and appeared to be surrounded by damaged cells or cell debris. It is suggested that persistent fiber-cell survival, and continued synthesis of various classes of mRNA by the nucleated cells in the cataractous lens could explain why reversal of cataracts results in recovery in both cell morphology and transparency in the non-nuclear portion of the lens.

The cardiac gap junction protein (Mr 47,000) has a tissue-specific cytoplasmic domain of Mr 17,000 at its carboxy-terminus

The molecular weight of the heart gap junctional protein subunit was, until recently, believed to be about Mr 28,000-30,000, similar to that of other previously characterized gap junctional proteins. A larger polypeptide of about Mr 44,000-47,000, which undergoes proteolysis during isolation, has recently been proposed as the form of the heart junction protein in vivo. We show here that this entity has the same amino-terminal sequence as the previously characterized Mr 29,000-30,000 component. Thus, the cardiac junctional protein has, at its carboxy-terminus, cytoplasmic domain of Mr 17,000; this domain is absent in the liver protein. These observations provide further evidence that gap junction proteins form a highly diversified family.

Phosphorylation of liver gap junction protein by protein kinase C

The 27 kDa protein, a major component of rat liver gap junctions, was shown to be phosphorylated in vitro by protein kinase C. The stoichiometry of the phosphorylation indicated that approx. 0.33 mol phosphate was incorporated per mol 27 kDa protein. Phosphorylation was entirely dependent on the presence of calcium and was virtually specific for serine residues. For comparison, the gap junction protein was also examined for its phosphorylation by cAMP-dependent protein kinase, the extent of phosphorylation being one-tenth that exerted by protein kinase C.

Cloning and characterization of human and rat liver cDNAs coding for a gap junction protein

An extended synthetic oligonucleotide (58-mer) has been used to identify and characterize a human liver gap junction cDNA. The cDNA is 1,574 bases long and contains the entire coding region for a gap junction protein. In vitro translation of the RNA products of this cDNA is consistent with it coding for a 32,022-D protein. Southern blot analysis indicates that the gap junction gene is present as a single copy, and that it can be detected in a variety of organisms using the human liver cDNA as a probe. The human cDNA has been used to screen a rat liver cDNA library, and a rat liver junction cDNA clone has been isolated. The rat liver clone is 1,127 bases in length, and it has strong sequence homology to the human cDNA in the protein-coding region, but less extensive homology in the 3'-untranslated region.

Molecular cloning of cDNA for rat liver gap junction protein

An affinity-purified antibody directed against the 27-kD protein associated with isolated rat liver gap junctions was produced. Light and electron microscopic immunocytochemistry showed that this antigen was localized specifically to the cytoplasmic surfaces of gap junctions. The antibody was used to select cDNA from a rat liver library in the expression vector lambda gt11. The largest cDNA selected contained 1,494 bp and coded for a protein with a calculated molecular mass of 32,007 daltons. Northern blot analysis indicated that brain, kidney, and stomach express an mRNA with similar size and homology to that expressed in liver, but that heart and lens express differently sized, less homologous mRNA.

Electrophysiological properties of gap junctions between dissociated pairs of rat hepatocytes

Physiological properties of isolated pairs of rat hepatocytes were examined within 5 h after dissociation. These cells become round when separated, but cell pairs still display membrane specializations. Most notably, canaliculi are often present at appositional membranes which are flanked by abundant gap and tight junctions. These cell pairs are strongly dye-coupled; Lucifer Yellow CH injected into one cell rapidly diffuses to the other. Pairs of hepatocytes are closely coupled electrically. Conductance of the junctional membrane is not voltage sensitive: voltage clamp studies demonstrate that gj is constant in response to long (5 s) transjunctional voltage steps of either polarity (to greater than +/- 40 mV from rest). Junctional conductance (gj) between hepatocyte pairs is reduced by exposure to octanol (0.1 mM) and by intracellular acidification. Normal intracellular pH (pHi), measured with a liquid ion exchange microelectrode, was generally 7.1-7.4, and superfusion with saline equilibrated with 100% CO2 reduced pHi to 6.0-6.5. In the pHi range 7.5-6.6, gj was constant. Below pH 6.6, gj steeply decreased and at 6.1 coupling was undetectable. pHi recovered when cells were rinsed with normal saline; in most cases gj recovered in parallel so that gj values were similar for pHs obtained during acidification or recovery. The low apparent pK and very steep pHi-gj relation of the liver gap junction contrast with higher pKs and more gradually rising curves in other tissues. If H+ ions act directly on the junctional molecules, the channels that are presumably homologous in different tissues must differ with respect to reactive sites or their environment.

Epithelial cell adhesion molecules

Recognition and binding between cells are of fundamental importance for a proper function of multicellular organisms, both during embryonic development and in the adult stage. Recently several cell surface proteins that are involved in these phenomena have been discovered. In the identification of these proteins, called cell adhesion molecules (CAMs), immunological methods have played a significant role. In a different approach to studies of cell-cell binding at the molecular level, the chemical composition of intercellular junctions is being studied. Intercellular junctions are specialized cell surface domains that have been identified by electron microscopy. They are particularly well developed in epithelia. Several proteins in the junctions have now been identified and characterized. This review deals with the biochemical properties of epithelial CAMs, and those proteins that are candidates for cell-to-cell binding in the junctions. In particular, the relationships between the various CAMs and junctional proteins are discussed. The tentative biological functions of these molecules are also considered.

Evidence for two physiologically distinct gap junctions expressed by the chick lens epithelial cell

Lens epithelial cells communicate with two different cell types. They communicate with other epithelial cells via gap junctions on their lateral membranes, and with fiber cells via junctions on their apices. We tested independently these two routes of cell-cell communication to determine if treatment with a 90% CO2-equilibrated medium caused a decrease in junctional permeability; the transfer of fluorescent dye was used as the assay. We found that the high-CO2 treatment blocked intraepithelial dye transfer but not fiber-to-epithelium dye transfer. The lens epithelial cell thus forms at least two physiologically distinct classes of gap junctions.

Is the C-terminal arm of lens gap junction channel protein the channel gate?

Lens gap junction channels are studied in a reconstituted system obtained by incorporating into liposomes, with or without calmodulin, the lens junction protein (MIP26) and its trypsin-cleaved product (MIP21) that lacks the C-terminal arm. Channel permeability is studied with an osmotic swelling assay. MIP26 and MIP21 liposomes swell in sucrose or polyethyleneglycol with or without Ca++ indicating the presence of large channels. Without Ca++, MIP26 and MIP21 liposomes swell in both permeants. With Ca++, MIP26-calmodulin liposomes do not swell in either permeant, indicating complete channel closure, while MIP21-calmodulin liposomes swell in sucrose but not in polyethyleneglycol. This suggests that the C-terminal arm participates in channel gating.

Purified lens junctional protein forms channels in planar lipid films

Junctions isolated from bovine lenses were solubilized with the detergent octyl glucoside, and their protein(s) was reconstituted in unilamellar vesicles. The protein(s) appears as annular-shaped intramembrane particles approximately equal to 10 nm in diameter on the vesicles' fracture faces. The addition of the vesicle-containing junctional protein(s) to both sides of preformed lipid films induced voltage-dependent channels. The channels have a conductance of 200 pS in 0.1 M salt solutions and are thus large enough to account for the electrical coupling observed between intact lens fibers; they turn off when the magnitude of the voltage is increased and in the presence of octanol. Although the identity of the reconstituted channels as the communicating pathway between lens fibers remains to be proven, it is most likely that the reconstituted channels are formed by MIP-26, the major protein component of the isolated lens junctions.

Lens junctional protein: analyzing MP26 with monoclonal antibodies

Specific antibodies are versatile tools for analyzing cell surface proteins. This study involves the characterization of monoclonal antibodies which are specific for the junctional protein found in the lens fiber cell. This protein can be expected to include regions on the external membrane surface for junction formation, others on the cytoplasmic surface for regulation of junctional properties and, if cell-cell channels are indeed involved, transmembrane domains forming the hydrophilic connection between adjacent cytoplasms. Antibodies to these various regions would provide for an experimental analysis of the junctional protein, e.g., the identification of "active sites" for junction formation. Three monoclonal antibodies specific for the lens junctional protein in the chicken are described here. The first, termed B2, also recognizes the bovine junctional protein, MP26 (5). We have characterized the submolecular specificity of B2 and have found that it binds approximately ten amino acid residues from the C-terminus of MP26. In isolated lens junction preparations, B2 binds to the cytoplasmic surfaces of the lens junctions (both 12 nm and 16 nm thick forms). Thus, we consider MP26 a component of the lens junction. Monoclonal A4, the second antibody considered in detail here, was produced by immunization with lens membranes after treatment with low pH. We have found that lens junctional membranes are separated, or "split," by treatment at pH 2.5-3.0. It appears that A4 binds to the external surface of the junctional membrane; EM studies to confirm this are in progress. In order to map the A4 binding site within the chicken junctional protein and to explore the arrangement of this protein within the membrane, a number of procedures were used to generate fragments of MP26. These included reactions with N-chlorosuccinimide and proteases after acid treatment. Antibody binding to fragments was evaluated with immunotransfer ("Western") procedures. These studies mapped the A4 binding site to the center of the molecule and suggested that MP26 projected externally from the membrane at two different points. These results are consistent with a recent model, based on sequence data (6), for the arrangement of MP26 within the bovine lens membrane.

The distribution of the main intrinsic membrane polypeptide in ocular lens

The Main Intrinsic Polypeptide (MIP) of the ocular lens fiber cell plasma membrane was immunocytochemically localized at the ultrastructural level on ultrathin frozen sections of rat lens, and on extracted, gradient-purified bovine lens membranes. The results indicate that both the junctional and non-junctional membrane domains of the cortical lens fiber cell are MIP immunoreactive. Frozen thin section immunocytochemistry of the lens epithelium and hepatocytes, also using anti-MIP antibodies, revealed that these cells, and their intercellular junctions, are not MIP-immunoreactive. From these findings we conclude that 1) MIP, a putative fiber cell junctional protein, is present throughout the plasma membrane of the lens fiber cell, and is not confined to the fiber cell junctional domain, 2) MIP is not a detectable component of the lens epithelial cell membrane, or its intercellular junctions, 3) MIP is not detectable in gap junctions of hepatocytes.

Lens junctions are communicating junctions

Lens fibers are electrically coupled with each other and directly exchange dyes and metabolites. In most cells, this form of communication is mediated by gap junctions. Lens fibers lack typical gap junctions. The lens junctions, although morphologically similar to gap junctions, differ from them structurally, chemically and immunologically. Nevertheless, recent evidence suggests that indeed lens junctions are communicating junctions. The lens junction protein, MIP26, displays structural characteristics similar to other channel proteins. Once incorporated into liposomes it forms channels permeable to molecules as heavy as 1.5 kDa. Like other communicating junctions, lens junctions assume crystalline arrays and uncouple with Ca++. The liposome incorporated channels close with Ca++ and H+ in the presence of calmodulin (CaM). Partial loss of gating competency occurs after proteolytic cleavage of the C-terminal arm of MIP26. The need for a unique type of communicating junction in lens is unclear. A possibility is that this tissue has some special cell-to-cell transport requirements, in terms of size and/or charge of permeants, not shared by coupled cells of other tissues.

Preparation of a gap junction fraction from uteri of pregnant rats: the 28-kD polypeptides of uterus, liver, and heart gap junctions are homologous

A procedure for the preparation of a gap junction fraction from the uteri of pregnant rats is described. The uterine gap junctions, when examined by electron microscopy of thin sections and in negatively stained preparations, were similar to gap junctions isolated from heart and liver. Major proteins of similar apparent molecular weight (Mr 28,000) were found in gap junction fractions isolated from the uterus, heart, and liver, and were shown to have highly homologous structures by two-dimensional mapping of their tryptic peptides. An Mr 10,000 polypeptide, previously deduced to be a proteolytic product of the Mr 28,000 polypeptide of rat liver (Nicholson, B. J., L. J. Takemoto, M. W. Hunkapiller, L. E. Hood, and J.-P. Revel, 1983, Cell, 32:967-978), was also studied and shown by chymotryptic mapping to be homologous in the uterine, heart, and liver gap junction fractions. An antibody raised in rabbits to a synthetic peptide corresponding to an amino-terminal sequence of the liver gap junction protein recognized Mr 28,000 proteins in the three tissues studied, showing that the proteins shared common antigenic determinants. These results indicate that gap junctions are biochemically conserved plasma membrane specializations. The view that gap junctions are tissue-specific plasma membrane organelles based on previous comparisons of Mr 26,000-30,000 polypeptides is not sustained by the present results.

Lens cell-to-cell channel protein: I. Self-assembly into liposomes and permeability regulation by calmodulin

Lens fibers are coupled by communicating junctions which contain a 28-kDalton protein (MIP26) believed to be the main component of the cell-to-cell channel. To study the permeability properties and regulation of these channels, an in vitro system has been developed in which MIP26 isolated from calf lens is incorporated into liposomes and the resulting channels are studied spectrophotometrically by a swelling assay. Liposome vesicles were prepared using a sonication/resuspension method. Incorporation efficiency was monitored by freeze-fracture. Vesicles were resuspended in 6% Dextran T-10. Assay buffer was identical, except for isotonic substitution of sucrose for T-10. MIP26-incorporated (but not control) vesicles swell under isotonic conditions indicating sucrose entry (via channels) followed by water to maintain osmotic balance. In the absence of calmodulin, calcium ion has no effect on channel permeability. On the contrary, vesicles prepared with equimolar amounts of MIP26 and CaM do not swell in the presence of calcium ion, indicating that the channels can be closed. Addition of EGTA to these vesicles reinitiates swelling--evidence that the channel gating mechanism is reversible. Magnesium ion has no effect on either type of vesicle.

Identification of the calmodulin-binding components in bovine lens plasma membranes

Lens membranes, purified from calf lenses, have been labeled by covalent cross-linking to membrane-bound 125I-calmodulin with dithiobis(succinimidyl propionate). Electrophoretic analysis in sodium dodecyl sulfate demonstrated two major 125I-containing products of Mr = 49 000 and 36 000. That the formation of these two components was specifically inhibited by unlabeled calmodulin, or calmodulin antagonists, would indicate that the formation of these components was calmodulin-specific. The size of these two 125I-labeled components was unchanged over a range of 125I-calmodulin or dithiobis(succinimidyl propionate) concentrations indicating that they represent 1:1 complexes between 125I-calmodulin (Mr = 17 000) and Mr-32 000 and Mr-19 000 lens membrane components respectively. Although formation of both cross-linked components exhibited an absolute dependence on Mg2+, the autoradiographic intensity of these components was enhanced when Ca2+ was included with Mg2+ during the cross-linking reaction. Labeling was maximal in 10 mM MgCl2 and approximately 1 microM Ca2+. Treatment of lens membranes with chymotrypsin resulted in the cleavage of MP26 (the major lens membrane protein), with the appearance of a major proteolytic fragment of Mr = 22 000. This proteolysis was not associated with any significant change in either the size or amount of the 125I-calmodulin-labeled membrane components. These results suggest that calmodulin interacts with two membrane proteins, but not significantly with MP26, in the intact lens cell membrane. Our results indicate the need to maintain caution in interpreting direct calcium plus calmodulin effects on MP26 and lens cell junctions.

Identification of a 70,000-D protein in lens membrane junctional domains

A 70,000-D membrane protein (MP70), which is restricted to the eye lens fibers and is present in immunologically homologous form in many vertebrate species, has been identified. By use of anti-MP70 monoclonal antibodies for immunofluorescence microscopy and electron microscopy, this polypeptide was localized in lens membrane junctional domains. Both immunofluorescence microscopy and SDS PAGE reveal an abundance of MP70 in the lens outer cortex that coincides with a high frequency of fiber gap junctions in the same region.

Proteolysis of cardiac gap junctions during their isolation from rat hearts

Gap junctions (GJ) isolated from rat hearts in presence of the protease inhibitor phenylmethylsulfonylfluoride (PMSF) contain a Mr 44,000 to 47,000 major polypeptide and have a urea-resistant layer of fuzz on their cytoplasmic surfaces, whereas junctions isolated without PMSF are proteolyzed to a Mr 29,500 polypeptide by a serine protease and have smooth cytoplasmic surfaces (C.K. Manjunath, G.E. Goings & E. Page Am. J. Physiol. 246:H865-H875, 1984). Rat liver GJ isolated with or without PMSF contain a Mr 28,000 polypeptide and have smooth cytoplasmic surfaces. Here we examine the origin, type and inhibitor sensitivity of the heart protease; why similar proteolysis is absent during isolation of rat liver gap junctions; and whether the Mr 44,000 to 47,000 cardiac GJ polypeptide is the precursor of the Mr 29,500 subunit. We show that the Mr 44,000 to 47,000 polypeptide corresponds to the unproteolyzed connexon subunit; that proteolysis of this polypeptide occurs predominantly during exposure to high ionic strength solution (0.6 M KI) which releases serine protease from mast cell granules; that this protease is inhibitable with PMSF and (less completely) soybean trypsin inhibitor and chymostatin; and that in vivo degranulation of mast cells by injecting rats with compound 48/80 fails to prevent breakdown of cardiac GJ during isolation. The results support the concept that GJ from rat heart and liver differ in protein composition.

Cell type-specific expression of nuclear lamina proteins during development of Xenopus laevis

The cell type-specific expression of the major nuclear lamina polypeptides ("lamins") during development of Xenopus was studied using two monoclonal antibodies (L(0)46F7: specific for LIII, the single lamin of oocytes; PKB8: specific for LI and LII of some somatic cells). In the oocyte, LIII localizes in the nuclear polymer, but upon nuclear envelope breakdown it is solubilized to a form sedimenting at 9 S. In early embryos, LIII contributes to nuclear lamina formation until its depletion. Correspondingly, LI and LII begin to be expressed at a specific point in embryogenesis and appear to be integrated with LIII into a common lamina structure. Later in development, LIII reappears as a prominent nuclear lamina protein but only in certain cells (neurons, muscle cells, and diplotene oocytes). We conclude that amphibian lamins represent a family of proteins expressed in relation to certain programs of cell differentiation.

Major intrinsic polypeptide (MIP26K) from lens membrane: reconstitution into vesicles and inhibition of channel forming activity by peptide antiserum

Bovine and human lens membrane, when reconstituted into lipid vesicles containing oxidized cytochrome C, will mediate the transmembrane passage of ascorbate into the vesicles, where the reduction of cytochrome C is measured spectrophotometrically. This channel forming activity is specifically inhibited by antiserum made against a synthetic octapeptide near the C-terminus of MIP26K. Together, these studies describe a direct and more sensitive assay system for measurement of channel-forming activity of MIP26K, and suggest that the C-terminus of this molecule may be particularly important in the regulation of channel formation.

Reduction of gap junctional conductance by microinjection of antibodies against the 27-kDa liver gap junction polypeptide

Antibody raised against isolated rat liver gap junctions was microinjected into coupled cells in culture to assess its influence on gap junctional conductance. A rapid inhibition of fluorescent dye transfer and electrical coupling was produced in pairs of freshly dissociated adult rat hepatocytes and myocardial cells as well as in pairs of superior cervical ganglion neurons from neonatal rats cultured under conditions in which electrotonic synapses form. The antibodies have been shown by indirect immunofluorescence to bind to punctate regions of the plasma membrane in liver. By immunoreplica analysis of rat liver homogenates, plasma membranes, and isolated gap junctions resolved on NaDodSO4/polyacrylamide gels, binding was shown to be specific for the 27-kDa major polypeptide of gap junctions. This and similar antibodies should provide a tool for further investigation of the role of cell-cell communication mediated by gap junctions and indicate that immunologically similar polypeptides comprise gap junctions in adult mammalian cells derived from all three germ layers.

Phosphorylation of lens membranes with a cyclic AMP-dependent protein kinase purified from the bovine lens

We report the phosphorylation of lens membranes with a cAMP-dependent protein kinase isolated from bovine lenses. The holoenzyme was eluted from DEAE agarose at less than 100 mM NaCl and from gel filtration columns with a relative molecular weight of 180 000. The regulatory subunit was identified with the affinity label 8-azido-[32P]cAMP. Four focusing variants with relative molecular weights of 49 000 were seen on two-dimensional gels. The catalytic subunit was purified approx. 5000-fold and migrated at 42 000 Mr on SDS gels. Based on these observations, the enzyme is classified as a Type I cAMP-dependent protein kinase. Purified lens plasma membranes were incubated with the holoenzyme or its catalytic subunit in the presence of 32P-labeled ATP. Several membrane proteins, including the major lens membrane polypeptide, MP26, were shown to be substrates for the kinase in this reaction. MP26 appears to be the major component of intercellular junctions in the lens. Studies with protease treatments on labeled membranes appeared to localize the phosphorylation sites to the cytoplasmic side of the membrane.

Sequence analysis of peptide fragments from the intrinsic membrane protein of calf lens fibers MP26 and its natural maturation product MP22

Calf lens fiber plasma membranes, containing only the intrinsic membrane protein MP26 and its maturation product MP22 were treated with proteolytic enzymes such as trypsin, protease V8 from S. aureus or with chemical agents as CNBr in formic acid. The cleavage products, purified by electrophoresis, were analysed for their amino acid composition and N-terminal sequences. Proteolysis gave rise to peptides which were mainly shortened at the C-terminal end of the molecules. While the V8 protease produced a fragment with a similar N-terminal sequence as the maturation product MP22, trypsin yielded another cleavage product. Chemical hydrolysis yielded large fragments (11-15 kDa) with hydrophobic N-terminal sequences. Our results suggest that MP26 is characterised by an N-terminal signal sequence and possesses other hydrophobic domains which could function as untranslocated insertion sequences.

Interaction of calmodulin and other calcium-modulated proteins with mammalian and arthropod junctional membrane proteins

Calmodulin and other calcium-modulated proteins bind in vitro to purified junctional polypeptides from rat liver gap junctions, bovine lens fiber junctions, a chymotryptic fragment from bovine lens junctions, and crayfish hepatopancreas gap junctions. The potential biological relevance of the interaction of calmodulin with junctional proteins is suggested by immunocytochemical localization of endogenous calmodulin in cortical regions of the cell where gap junctions exist. These observations provide a molecular basis for understanding the potential regulatory role of calmodulin on cell-cell communication channels in vivo. In addition, the calmodulin binding represents the first molecular homology that has been found for junctional channel proteins from mammalian and arthropod tissues.