The microvillus 110K cytoskeletal protein is an integral membrane protein

Calcium, nucleotide, and actin affect the interaction of mammalian Myo1c with its light chain calmodulin

To investigate the interaction of mammalian class I myosin, Myo1c, with its light chain calmodulin, we expressed (with calmodulin) truncation mutants consisting of the Myo1c motor domain followed by 0-4 presumed calmodulin-binding (IQ) domains (Myo1c (0IQ)-Myo1c (4IQ)). The amount of calmodulin associating with the Myo1c heavy chain increased with increasing number of IQ domains from Myo1c (0IQ) to Myo1c (3IQ). No calmodulin beyond that associated with Myo1c (3IQ) was found with Myo1c (4IQ) despite its availability, showing that Myo1c binds three molecules of calmodulin with no evidence of a fourth IQ domain. Unlike Myo1c (0IQ), the basal ATPase activity of Myo1c (1IQ) was >10-fold higher in Ca (2+) vs EGTA +/- exogenous calmodulin, showing that regulation is by Ca (2+) binding to calmodulin on the first IQ domain. The K m and V max of the actin-activated Mg (2+)-ATPase activity were largely independent of the number of IQ domains present and moderately affected by Ca (2+). In binding assays, some calmodulin pelleted with Myo1c heavy chain when actin was present, but a considerable fraction remained in the supernatant, suggesting that calmodulin is displaced most likely from the second IQ domain. The Myo1c heavy chain associated with actin in a nucleotide-dependent fashion. In ATP a smaller proportion of calmodulin pelleted with the heavy chain, suggesting that Myo1c undergoes nucleotide-dependent conformational changes that affect the affinity of calmodulin for the heavy chain. The studies support a model in which Myo1c in the inner ear is regulated by both Ca (2+) and nucleotide, which exert their effects on motor activity through the light-chain-binding region.

Calcium and the regulation of cytoskeletal assembly, structure and contractility

Calcium plays a central role in the regulation of cytoskeletal assembly, structure and contractility. In the case of actin there are a number of functional classes of actin-binding proteins which confer on a given actin filament its specific function in the cell. Among these various classes of actin-binding proteins are a subset of proteins whose activity is either regulated directly or indirectly (for example, through calmodulin) by Ca2+. This includes the regulation of actin-myosin interaction, actin assembly, actin filament interaction and the formation of supramolecular cytoskeletal networks, and the interaction of actin with membranes. Examples of these various modes of Ca2+-dependent regulation of cytoskeletal structure and contractility are discussed.

Identification of functional domains of adenylyl cyclase using in vivo chimeras

Adenylyl cyclase, the effector molecule of the cAMP signaling pathway, is composed of a family of isoforms that differ in their modes of regulation. Many of these modulatory interactions are dependent upon well characterized molecules from various second messenger pathways; however, very little is known about their mechanisms or sites of action on adenylyl cyclase. Chimeras were produced by a novel in vivo mechanism between two differentially modulated adenylyl cyclases to identify their regulatory domains. The basal activity of the type I adenylyl cyclase (AC1) is activated by calcium/calmodulin, inhibited by G protein beta gamma subunits, and insensitive to protein kinase C regulation. In contrast, type II adenylyl cyclase (AC2) is insensitive to calcium/calmodulin regulation and is activated by G protein beta gamma subunits as well as by activated protein kinase C. Expression and biochemical characterization of chimeras between AC1 and AC2 identified a single specific domain of AC1 responsible for calmodulin binding and a small, well defined region near the C terminus of AC2 required for protein kinase C activation.

Differential calmodulin binding to three myosin-1 isoforms from liver

We have previously purified and characterized two myosin-1 isoforms from rat liver (molecular masses 130 kDa and 110 kDa; L. M. Coluccio and C. Conaty (1993) Cell Motil. Cytoskel. 24, 189–199). Here, we describe the purification and characterization from liver of a third myosin-1 (molecular mass 105 kDa) and determine the number of calmodulin molecules associated with each of these three myosin-1 isoforms. The 105 kDa polypeptide, solubilized from liver homogenates with the addition of ATP, co-sediments with F-actin, co-purifies with calmodulin, and binds calmodulin in the presence of EGTA. Antibodies directed against chicken intestinal brush border myosin-1 cross-react with the 105 kDa polypeptide on immunoblots. Partial peptide sequence analysis indicates that the polypeptide corresponds with an MM1 gamma gene product that represents a myosin-1 isoform cloned from mouse brain (Sherr et al. (1993) J. Cell Biol. 120, 1405–1416). A comparison of calmodulin binding to the now three isolated forms of myosin-1 in liver shows that in solution the 105 kDa and 110 kDa polypeptides bind two molecules of calmodulin each whereas the 130 kDa binds six molecules of calmodulin.

Phase separation of biomolecules in polyoxyethylene glycol nonionic detergents

The advantage of aqueous two-phase systems based on polyoxyethylene detergents over other liquid-liquid two-phase systems lies in their capacity to fractionate membrane proteins simply by heating the solution over a biocompatible range of temperatures (20 to 37 degrees C). This permits the peripheral membrane proteins to be effectively separated from the integral membrane proteins, which remain in the detergent-rich phase due to the interaction of their hydrophobic domains with detergent micelles. Since the first reports of this special characteristic of polyoxyethylene glycol detergents in 1981, numerous reports have consolidated this procedure as a fundamental technique in membrane biochemistry and molecular biology. As examples of their use in these two fields, this review summarizes the studies carried out on the topology, diversity, and anomalous behavior of transmembrane proteins on the distribution of glycosyl-phosphatidylinositol-anchored membrane proteins, and on a mechanism to describe the pH-induced translocation of viruses, bacterial endotoxins, and soluble cytoplasmic proteins related to membrane fusion. In addition, the phase separation capacity of these polyoxyethylene glycol detergents has been used to develop quick fractionation methods with high recoveries, on both a micro- and macroscale, and to speed up or increase the efficiency of bioanalytical assays.

Dependence of calmodulin localization in the retina on the NINAC unconventional myosin

Calmodulin is a highly conserved regulatory protein found in all eukaryotic organisms which mediates a variety of calcium ion-dependent signalling pathways. In the Drosophila retina, calmodulin was concentrated in the photoreceptor cell microvillar structure, the rhabdomere, and was found in lower amounts in the sub-rhabdomeral cytoplasm. This calmodulin localization was dependent on the NINAC (neither inactivation nor afterpotential C) unconventional myosins. Mutant flies lacking the rhabdomere-specific p174 NINAC protein did not concentrate calmodulin in the rhabdomere, whereas flies lacking the sub-rhabdomeral p132 isoform had no detectable cytoplasmic calmodulin. Furthermore, a defect in vision resulted when calmodulin was not concentrated in the rhabdomeres, suggesting a role for calmodulin in the regulation of fly phototransduction. A general function of unconventional myosins may be to control the subcellular distribution of calmodulin.

Myosin-I in mammalian liver

Myosin-I refers to a class of proteins with a molecular weight of approximately 110-kDa, which have characteristics of conventional myosin but are unable to form filaments. Previous studies have implicated myosin-I in motile cellular processes including cell migration and phagocytosis. Although the first example of myosin-I in higher eukaryotes was the intestinal 110K-calmodulin complex, which forms in microvilli the lateral links connecting the core bundle of actin filaments to the membrane, myosin-I has now been shown to be a component of rat kidney and to be present in bovine adrenal gland and brain. We have now purified and characterized two polypeptides from rat liver which have several characteristics of the intestinal 110K-calmodulin complex. Both liver polypeptides are solubilized with ATP and co-elute on gel filtration with calmodulin. The polypeptides, of 110-kDa and 130-kDa, bind calmodulin in 1 mM EGTA. Both polypeptides bind to F-actin in an ATP reversible fashion, and crosslink actin filaments. The purified polypeptides possess an actin-activated Mg(2+)-ATPase activity typical of brush border myosin-I. A polyclonal antiserum directed against the chicken intestinal 110-kDa polypeptide recognizes both rat liver polypeptides, whereas another serum recognizes the 130-kDa but not the 110-kDa rat liver polypeptide. Controlled proteolysis of the purified polypeptides with alpha-chymotrypsin indicates that the two polypeptides are distinct but related. Immunofluorescence microscopy on isolated hepatocytes shows distribution of myosin-I to be vesicular, distributed throughout the cytoplasm, but more concentrated near the nucleus. These data contribute new evidence by several functional criteria that multiple myosin-I molecules are present in higher organisms and may coexist in a single cell type.

Acetonitrile as a substitute for ethanol/propylene oxide in tissue processing for transmission electron microscopy: comparison of fine structure and lipid solubility in mouse liver, kidney, and intestine

Tissue processing for transmission electron microscopy (TEM) is commonly accomplished using ethanol (EtOH) as a dehydrating solvent and propylene oxide (PO) as a transition fluid. Both solvents have some undesirable properties: EtOH solubilizes lipids; PO is highly flammable, volatile, toxic, and potentially carcinogenic. Their replacement by a compound devoid of these characteristics is therefore desirable. Acetonitrile (AN) appears to be such a solvent. It is freely miscible with water, alcohols, acetone, and epoxy resins; it does not interfere with epoxy polymerization; and the resulting cured resins have excellent cutting quality and beam stability. AN is also an excellent dehydrating agent whose use does not necessitate modification of current techniques. Most importantly, the low solubility of phospholipids (PL) in AN limits the loss of membrane lipids and, hence, leads to a better preservation of tissue features.

Identification and molecular characterization of the calmodulin-binding subunit gene (CMP1) of protein phosphatase 2B from Saccharomyces cerevisiae. An alpha-factor inducible gene

A method has been developed for the rapid purification of yeast calmodulin in high yield. Using a 125I-labeled calmodulin SDS/PAGE gel overlay procedure with either yeast or bovine calmodulin, we show that the bovine and yeast proteins recognize the same proteins in total yeast extracts. However, yeast calmodulin does not bind to many of the proteins in vertebrate cells identified using bovine calmodulin. A lambda gt11 yeast genomic expression library was screened with yeast or bovine brain 125I-calmodulin to identify sequences derived from calmodulin binding proteins. Twelve clones were recovered, all containing a common DNA insert; all bound calmodulin in a Ca(2+)-dependent manner. The complete coding sequence was recovered and sequenced. The predicted protein sequence show greater than 50% identity to the A subunit of vertebrate protein phosphatase 2B. The gene was designated CMP1 and shown to reside on chromosome IV. Disruption or over-expression of CMP1 have no obvious phenotype; yeast appears to contain one or more CMP1-related genes. The protein product of the CMP1 gene is elevated by alpha-factor treatment, suggesting an involvement of protein phosphatase 2B in the mating response.

Mutations in the regions of the Rous sarcoma virus 3' splice sites: implications for regulation of alternative splicing

Retrovirus RNA is synthesized as a single primary transcript that is differentially processed by RNA splicing. Three species of viral RNA (spliced env, spliced src, and unspliced full-length RNA) are produced in chicken embryo fibroblasts infected with Rous sarcoma virus, an avian retrovirus. The env and src mRNAs are synthesized by the alternative use of two 3' splice sites. The mechanism by which balanced splicing at the two sites is maintained was investigated in this report. Mutants that increase or decrease splicing at one of the two 3' splice sites were analyzed for the effect on splicing at the other site. The two splice sites differed in their response. Mutations that caused a specific increase in the level of spliced env mRNA were associated with reciprocal changes in the levels of src mRNA but with no change in overall splicing. In contrast, mutants in which the src 3' splice site was inactivated demonstrated a decreased overall level of spliced RNA but had little or no effect on the level of spliced env mRNA. Mutations that caused specific increases in the level of spliced src mRNA had variable effects on env mRNA levels. Deletion of regions in gag, which was previously shown to contain a cis-acting negative regulator of splicing, resulted in a corresponding increase of both spliced viral mRNAs and a decrease in unspliced RNA, suggesting that this element suppressed both env and src splicing. Several models are discussed which are possible mechanisms for regulation of alternative splicing of Rous sarcoma virus RNA, but none of these models appear to be consistent with all of the data.

Limited tissue distribution of the intestinal brush border myosin I protein

A myosinlike 105-110-kilodalton calmodulin-binding protein, brush border myosin I, found in the intestinal brush border has been linked to two seemingly disparate but possibly interacting functions of the brush border, namely, microvillar motility and vitamin D regulated calcium transport. If brush border myosin I were to function primarily as a myosinlike molecule powering cellular or microvillar motility, one might expect it to be found in a variety of tissues with microvilli such as the renal brush border and bile canaliculus. On the other hand, a more specialized function such as participation in vitamin D regulated calcium transport might dictate a more restricted tissue distribution for brush border myosin I. To determine the tissue distribution of brush border myosin I, we purified this protein to apparent homogeneity, generated antisera to it, and used the antisera to localize the protein within the intestinal epithelial cell by immunocytochemistry. We then screened a variety of other tissues (brain, lung, heart, liver, spleen, pancreas, kidney, and skeletal muscle) both for calmodulin-binding proteins as well as for brush border myosin I using Western blots and immunofluorescence. Our results indicate that the intestinal brush border myosin I is limited in its distribution to the intestinal brush border.

Correlation between the presence of microvilli and the growth or metastatic potential of tumor cells

We used an electron microscope to examine microvilli which appear on the surfaces of various tumor cells with high or low growth potential and/or metastatic ability. The results show that a greater number of microvilli appeared on the surfaces of tumor cells (QRpP and ERpP) which possess high growth potential than on tumor cells (QR and ER) with low growth potential. We also observed that microvilli were more abundant on the surface of highly metastatic clone cells, i.e. c-SST-2 (cl-2), mouse B16 melanoma (F-10) and human colon carcinoma (KM12SM) than on weakly metastatic clone cells, c-SST-2 (cl-4-2), B16 (F-1) and (KM12C). At the same time, more microvilli were observed on the surface of B16 BL6 cells, which were obtained from the metastatic site of the B16 F10 cells, than on the surface of the parent B16 F10 cells. Immunoelectron microscopy revealed that the c-neu oncogene product, which is closely related to an epidermal growth factor receptor, was positively stained in the microvilli of tumor cells (ERpP) with high growth potential and high metastatic ability, whereas the tumor cells (ER) with low growth potential and weak metastatic ability were not stained. These findings suggest that the increased presence of microvilli correlates closely with the growth potential and metastatic ability of tumor cells.

Partial deduced sequence of the 110-kD-calmodulin complex of the avian intestinal microvillus shows that this mechanoenzyme is a member of the myosin I family

The actin bundle within each microvillus of the intestinal brush border is laterally tethered to the membrane by bridges composed of the protein complex, 110-kD-calmodulin. Previous studies have shown that avian 110-kD-calmodulin shares many properties with myosins including mechanochemical activity. In the present study, a cDNA molecule encoding 1,000 amino acids of the 110-kD protein has been sequenced, providing direct evidence that this protein is a vertebrate homologue of the tail-less, single-headed myosin I first described in amoeboid cells. The primary structure of the 110-kD protein (or brush border myosin I heavy chain) consists of two domains, an amino-terminal "head" domain and a 35-kD carboxy-terminal "tail" domain. The head domain is homologous to the S1 domain of other known myosins, with highest homology observed between that of Acanthamoeba myosin IB and the S1 domain of the protein encoded by bovine myosin I heavy chain gene (MIHC; Hoshimaru, M., and S. Nakanishi. 1987. J. Biol. Chem. 262:14625-14632). The carboxy-terminal domain shows no significant homology with any other known myosins except that of the bovine MIHC. This demonstrates that the bovine MIHC gene most probably encodes the heavy chain of bovine brush border myosin I (BBMI). A bacterially expressed fusion protein encoded by the brush border 110-kD cDNA binds calmodulin. Proteolytic removal of the carboxy-terminal domain of the fusion protein results in loss of calmodulin binding activity, a result consistent with previous studies on the domain structure of the 110-kD protein. No hydrophobic sequence is present in the molecule indicating that chicken BBMI heavy chain is probably not an integral membrane protein. Northern blot analysis of various chicken tissue indicates that BBMI heavy chain is preferentially expressed in the intestine.

Characterization of intestinal microvillar membrane disks: detergent-resistant membrane sheets enriched in associated brush border myosin I (110K-calmodulin)

The actin bundle within each microvillus of the intestinal brush border (BB) is tethered laterally to the membrane by bridges composed of BB myosin I. Avian BB myosin I, formerly termed 110K-calmodulin, consists of a heavy chain with an apparent Mr of 110 kD and three to four molecules of calmodulin "light chains." Recent studies have shown that this complex shares many properties with myosin including mechanochemical activity. In this report, the isolation and characterization of a membrane fraction enriched in bound BB myosin I is described. This membrane fraction, termed microvillar membrane disks, was purified from ATP extracts of nonionic detergent-treated microvilli prepared from avian intestinal BBs. Ultrastructural analysis revealed that these membranes are flat, disk-shaped sheets with protrusions which are identical in morphology to purified BB myosin I. The disks exhibit actin-activated Mg-ATPase activity and bind and cross-link actin filaments in an ATP-dependent fashion. The mechanochemical activity of the membrane disks was assessed using the Nitella bead movement assay (Sheetz, M. P., and J. A. Spudich. 1983. Nature [Lond.]. 303:31-35). These preparations were shown to be free of significant contamination by conventional BB myosin. Latex beads coated with microvillar membrane disks move in a myosin-like fashion along Nitella actin cables at rates of 12-60 nm/s (average rate of 33 nm/s); unlike purified BB myosin I, the movement of membrane disk-coated beads was most reproducibly observed in buffers containing low Ca2+.

Characterization and partial purification of a high-molecular-mass, calmodulin-binding, tyrosyl-phosphorylated protein from lymphocyte plasma membranes

A plasma membrane phosphoprotein with a species-dependent molecular mass of 108 or 112 kDa (P108/112) was analyzed in lymphoid cells from rats and humans. After 24 h lectin stimulation its in vitro phosphorylation was raised to an important extent. Phosphotyrosine was analyzed by 2-D electrophoresis. Calmodulin was bound by P108/112 in a Ca2+-dependent manner. P108/112 remained insoluble after extraction with detergent, high salt, EDTA, or high pH. After chlorethanol extraction it was partially purified by gel filtration. P108/112 shows conspicuous similarities with the 110-kDa protein from chicken intestine microvilli.

Calmodulin binding to the intestinal brush-border membrane: comparison to other calcium-binding proteins

The intestinal brush-border membrane contains a high concentration of calmodulin bound to a 105,000 dalton (105 kDa) protein. Binding of radioiodinated calmodulin to this protein does not require calcium but is inhibited by trifluoperazine and excess unlabelled calmodulin. Recent evidence suggests that the 105 kDa protein in conjunction with calmodulin may be involved in the regulation of calcium transport across the brush-border membrane. In this report, we evaluated the binding of the 105 kDa protein to other radioiodinated calcium-binding proteins including the vitamin D-dependent intestinal calcium-binding protein. We observed that troponin C and S100 beta protein both bound strongly to the 105 kDa protein. The binding of S100 beta was inhibited by EGTA, but was little affected by trifluoperazine and excess unlabelled S100 beta, whereas that of troponin C was inhibited by trifluoperazine and excess unlabelled troponin C, but was little affected by EGTA. Both troponin C and S100 beta bound to a large number of proteins to which calmodulin did not bind. The vitamin D-dependent calcium-binding protein (calbindin) from chick intestine and rat kidney also bound to the 105 kDa protein, albeit more weakly than troponin C, S100 beta and calmodulin. The binding of the calbindins was increased by EGTA and was little affected by trifluoperazine and excess unlabelled calbindin. Parvalbumin, rat osteocalcin, and alpha-lactalbumin showed little binding to any brush-border membrane protein. Our results indicate that the 105 kDa calmodulin-binding protein of the intestinal brush border can bind to a variety of calcium-binding proteins all of which contain homologous regions thought to be the calcium-binding sites. Only the binding of troponin C resembles the binding of calmodulin, however, in being inhibited by trifluoperazine and excess unlabelled ligand. The functional significance of these observations in terms of regulating calcium transport across the brush-border membrane remains to be established.

Structural and immunological characterization of the myosin-like 110-kD subunit of the intestinal microvillar 110K-calmodulin complex: evidence for discrete myosin head and calmodulin-binding domains

The actin bundle within each microvillus of the intestinal brush border is tethered laterally to the membrane by spirally arranged bridges. These bridges are thought to be composed of a protein complex consisting of a 110-kD subunit and multiple molecules of bound calmodulin (CM). Recent studies indicate that this complex, termed 110K-CM, is myosin-like with respect to its actin binding and ATPase properties. In this study, possible structural similarity between the 110-kD subunit and myosin was examined using two sets of mAbs; one was generated against Acanthamoeba myosin II and the other against the 110-kD subunit of avian 110K-CM. The myosin II mAbs had been shown previously to be cross-reactive with skeletal muscle myosin, with the epitope(s) localized to the 50-kD tryptic fragment of the subfragment-1 (S1) domain. The 110K mAbs (CX 1-5) reacted with the 110-kD subunit as well as with the heavy chain of skeletal but not with that of smooth or brush border myosin. All five of these 110K mAbs reacted with the 25-kD, NH2-terminal tryptic fragment of chicken skeletal S1, which contains the ATP-binding site of myosin. Similar tryptic digestion of 110K-CM revealed that these five mAbs all reacted with a 36-kD fragment of 110K (as well as larger 90- and 54-kD fragments) which by photoaffinity labeling was shown to contain the ATP-binding site(s) of the 110K subunit. CM binding to these same tryptic digests of 110K-CM revealed that only the 90-kD fragment retained both ATP- and CM-binding domains. CM binding was observed to several tryptic fragments of 60, 40, 29, and 18 kD, none of which contain the myosin head epitopes. These results suggest structural similarity between the 110K and myosin S1, including those domains involved in ATP- and actin binding, and provide additional evidence that 110K-CM is a myosin. These studies also support the results of Coluccio and Bretscher (1988. J. Cell Biol. 106:367-373) that the calmodulin-binding site(s) and the myosin head region of the 110-kD subunit lie in discrete functional domains of the molecule.

Immunofluorescent and immunochemical evidence for the expression of cytovillin in the microvilli of a wide range of cultured human cells

We have previously purified a Mr 75,000 protein, cytovillin, from cultured human choriocarcinoma cells (JEG-3) and shown that this protein was specifically confined to the microvillus membrane of these cells. I have now studied the expression and the subcellular distribution of cytovillin in eighteen normal and transformed human cell lines and strains by using immunoblotting and indirect immunofluorescence microscopy. In all cell types, cytovillin was highly enriched in cell surface protrusions. When cell types were ranked according to their staining intensity, choriocarcinoma was highest, then amniotic epithelial cells, other choriocarcinoma cells and tumor cells, and finally fibroblastoid cells. The latter only gave faint diffuse fluorescence on the plasma membrane and, occasionally, on the microvilli. However, detergent extracts of all cell types could be shown to contain cytovillin by the use of immunoblotting techniques. Metabolic pulse-chase labelling experiments with JEG-3 cells demonstrated synthesis of cytovillin as a single-chain polypeptide. No precursor forms or specific proteolytic cleavage products could be seen either by immunoblotting or immunoprecipitation. The protein was found to be very stable with a biologic half-life of about 25 hours. The pI determined by isoelectric focusing was 6.1. These results were consistent with cytovillin being an integral component of the microvilli and other surface extensions of all human cell types examined.

Molecular organization of the intestinal brush border

The brush border of enterocytes represents one of the more specialized apical poles of epithelial cells. It is formed by particularly well-developed apical plasma membrane microvilli, whose shape is ensured by a highly organized cytoskeleton. The molecular organization of the cytoskeleton is described. Whereas several cytoskeleton proteins are ubiquitous, villin is highly specific for intestinal cells and can be used as a differentiation marker of these cells. The major glycoproteins, in particular hydrolases, of the brush border membrane have been characterized. They have many common structural features, in particular their mode of integration into the membrane by their N-terminal hydrophobic sequences that also plays the role of the 'signal peptide' responsible for their co-translational insertions into the endoplasmic reticulum. Studies on the biosynthesis and intracellular pathway of aminopeptidase N strongly suggest that sorting of apical and basolateral glycoproteins could occur after their integration into the basolateral domain.

Organization of the actin filament cytoskeleton in the intestinal brush border: a quantitative and qualitative immunoelectron microscope study

In the present study we have used immunogold labeling of ultrathin sections of the intact chicken and human intestinal epithelium to obtain further insight into the molecular structure of the brush-border cytoskeleton. Actin, villin, and fimbrin were found within the entire microvillus filament bundle, from the tip to the basal end of the rootlets, but were virtually absent from the space between the rootlets. This suggests that the bulk of actin in the brush border is kept in a polymerized and cross-linked state and that horizontally deployed actin filaments are virtually absent. About 70% of the label specific for the 110-kD protein that links the microvillus core bundle to the lipid bilayer was found overlying the microvilli. The remaining label was associated with rootlets and the interrootlet space, where some label was regularly observed in association with vesicles. Since the terminal web did not contain any significant amounts of tubulin and microtubules, the present findings would support a recently proposed hypothesis that the 110-kD protein (which displays properties of an actin-activated, myosin-like ATPase) might also be involved in the transport of vesicles through the terminal web. Label specific for myosin and alpha-actinin was confined to the interrootlet space and was absent from the rootlets. About 10-15% of the myosin label and 70-80% of the alpha-actinin label was observed within the circumferential band of actin filaments at the zonula adherens, where myosin and alpha-actinin displayed a clustered, interrupted pattern that resembles the spacing of these proteins observed in other contractile systems. This circular filament ring did not contain villin, fimbrin, or the 110-kD protein. Finally, actin-specific label was observed in close association with the cytoplasmic aspect of the zonula occludens, suggesting that tight junctions are structurally connected to the microfilament system.

Expression of brush border calmodulin-binding proteins during human small and large bowel differentiation

The expression and immunocytochemical localization of three brush border cytoskeletal calmodulin-binding proteins, caldesmon, fodrin, and the 110 kDa subunit of the 110 kDa calmodulin complex, have been studied in human intestinal epithelial cells as a function of their ontogenic differentiation. At immature stages (fetal week 8), caldesmon and fodrin were present in undifferentiated intestinal epithelial cells. However, no 110 kDa protein was detectable except a 135 kDa immunoreactive species. The 110 kDa form appeared at week 12, when microvilli differentiate, and became prominent at week 14 simultaneously with the disappearance of the 135 kDa species. Finally at week 14, the calmodulin-binding protein pattern was identical to that found in adults. Immunocytochemical experiments revealed that at week 8, antibodies to caldesmon and fodrin gave a fluorescence lining at the periphery of the cells, whereas the 110 kDa immunoreactive species was hardly detectable. Then, as early as week 12 of gestation, with the three antisera, a bright fluorescence lined the apex of the cells, as in adults. In the colon, the events were delayed. This study demonstrates that the developmental pattern of the three calmodulin-binding proteins investigated, caldesmon, fodrin and the 110 kDa subunit, parallels the temporal differentiation of human intestinal brush borders and the proximo-distal morphological intestinal maturation.

Studies on protein kinases involved in regulation of nucleocytoplasmic mRNA transport

The rate of energy-dependent nucleoside triphosphatase (NTPase)-mediated nucleocytoplasmic translocation of poly(A)-containing mRNA [poly(A)+mRNA] across the nuclear envelope is thought to be regulated by poly(A)-sensitive phosphorylation and dephosphorylation of nuclear-envelope protein. Studying the phosphorylation-related inhibition of the NTPase, we found that phosphorylation of one polypeptide of rat liver envelopes by endogenous NI- and NII-like protein kinase was particularly sensitive to poly(A). This polypeptide (106 kDa) was also phosphorylated by nuclear-envelope-bound Ca2+-activated and phospholipid-dependent protein kinase (protein kinase C). Activation of kinase C by tumour-promoting phorbol esters resulted in inhibition of nuclear-envelope NTPase activity and in a concomitant decrease of mRNA (actin) efflux rate from isolated rat liver nuclei. Protein kinase C, but not nuclear envelope NI-like or NII-like protein kinase, was found to be solubilized from the envelope by Triton X-100, whereas the presumable poly(A)-binding site [the 106 kDa polypeptide, representing the putative carrier for poly(A)+mRNA transport] remained bound to this structure. RNA efflux from detergent-treated nuclei lost its susceptibility to phorbol esters. Addition of purified protein kinase C to these nuclei restored the effect of the tumour promoters. Protein kinase C was found to bind also to isolated rat liver nuclear matrices in the absence but not in the presence of ATP. The NII-like nuclear-envelope protein kinase co-purified together with the 106 kDa polypeptide which specifically binds to poly(A) in an ATP-labile linkage.

Mapping of the microvillar 110K-calmodulin complex: calmodulin-associated or -free fragments of the 110-kD polypeptide bind F-actin and retain ATPase activity

The 110K-calmodulin complex isolated from intestinal microvilli is an ATPase consisting of one polypeptide chain of 110 kD in association with three to four calmodulin molecules. This complex is presumably the link between the actin filaments in the microvillar core and the surrounding cell membrane. To study its structural regions, we have partially cleaved the 110K-calmodulin complex with alpha-chymotrypsin; calmodulin remains essentially intact under the conditions used. As determined by 125I-calmodulin overlays, ion exchange chromatography, and actin-binding assays, a 90-kD digest fragment generated in EGTA remains associated with calmodulin. The 90K-calmodulin complex binds actin in an ATP-reversible manner and decorates actin filaments with an arrow-head appearance similar to that found after incubation of F-actin with the parent complex; binding occurs in either calcium- or EGTA-containing buffers. ATPase activity of the 90-kD digest closely resembles the parent complex. In calcium a digest mixture containing fragments of 78 kD, a group of three at approximately 40 kD, and a 32-kD fragment (78-kD digest mixture) is generated with alpha-chymotrypsin at a longer incubation time; no association of these fragments with calmodulin is observed. Time courses of digestions and cyanogen bromide cleavage indicate that the 78-kD fragment derives from the 90-kD peptide. The 78-kD mixture can also hydrolyze ATP. Furthermore, removal of the calmodulin by ion exchange chromatography from this 78-kD mixture had no effect on the ATPase activity of the digest, indicating that the ATPase activity resides on the 110-kD polypeptide. The 78 kD, two of the three fragments at approximately 40 kD, and the 32-kD fragments associate with F-actin in an ATP-reversible manner. Electron microscopy of actin filaments after incubation with the 78-kD digest mixture reveals coated filaments, although the prominent arrowhead appearance characteristic of the parent complex is not observed. These data indicate that calmodulin is not required either for the ATPase activity or the ATP-reversible binding of the 110K-calmodulin complex to F-actin. In addition, since all the fragments that bind F-actin do so in an ATP-reversible manner, the sites required for F-actin binding and ATP reversibility likely reside nearby.

Isolation of an immunoreactive analogue of brain fodrin that is associated with the cell cortex of Dictyostelium amoebae

We have used a polyclonal affinity-purified antibody made against chicken brain fodrin (both 240 and 235 Kd subunits) as a probe to determine if a fodrinlike protein exists in amoebae of Dictyostelium discoideum. In Western blots of whole cells and the isolated cell cortex, polypeptides measuring 220 and 70 Kd are recognized by the fodrin antibodies. In situ localization by indirect immunofluorescence with antifodrin indicates that the immunoreactive polypeptides are cortical. The immunoreactive analogues copatch and cocap with concanavalin A. At the level of resolution of the electron microscope, immunocytochemistry with antifodrin and colloidal gold confirms that the immunoreactive analogues are cortical proteins associated with microfilaments on the cytoplasmic side of the plasma membrane. We have isolated and characterized the 220 Kd protein to determine if it is similar to fodrin and to investigate its relationship to the 70 Kd polypeptide. The 220 Kd protein can be extracted from the cortex in the absence of detergent and isolated by gel filtration and sucrose density gradient sedimentation. The 220 Kd is a rod-shaped protein 118 +/- 17.8 nm (N = 37) in length. It has a sedimentation coefficient of 9.3 S and Stokes' radius of 13 nm and exists as a dimer of approximately 500,000 daltons (Mr). Isolated 220 Kd binds to actin filaments in vitro when assayed by rotary shadowing. Morphological criteria distinguish 220 Kd from Dictyostelium myosin II heavy chain (215 Kd) and the filaminlike protein at 240 Kd. The 70 Kd polypeptide appears to be a cleavage fragment of the 220 Kd, since it is found after prolonged storage when formerly only the 220 Kd was present. Furthermore, the 220 and 70 Kd polypeptides exhibit similar one-dimensional peptide maps when treated with TPCK trypsin. On the basis of its physical and immunoreactive characteristics, and location in the cell, the 220 Kd may be a fodrinlike protein.

Light and electron microscopical demonstration of c- erB-2 gene product-like immunoreactivity in human malignant tumors

C-erbB-2 is a human protooncogene homologous with the well-known c-erbB. Genes and gene products of the EGF receptor and c-erbB are known to be closely related and to be closely homologous in their intracellular domain. Inspection of the deduced amino acid sequence suggested that the c-erbB-2 gene encodes a receptor for a yet unidentified growth factor. An immunohistological study was performed by introducing an antibody raised in the rabbit by immunization with a synthetic peptide corresponding to a part of the intracytoplasmic domain of predicted gene product. Specimens from 13 normal human organs, fresh frozen tissue from 41 surgically excised human malignant tumors and eight cell lines maintained in nude mice were studied. Positive staining was found in 4 of the 41 (9.8%) malignant tumors. All of the positive tumors were adenocarcinomas and two adenocarcinoma cell lines were also positive. Amongst the normal human tissues, epithelial cells in stomach, small and large intestine were faintly stained. When the positively stained cell lines were studied by immunoelectronmicroscopy, the reaction was most prominent in the membrane of microvilli, but part of the nuclear membrane, the endoplasmic reticulum and the outer cell membrane were also stained. DNA and mRNA blot assays, as well as our immunoprecipitation test, revealed that immunohistologically positive cell lines bore amplified c-erbB-2 DNA, c-erbB-2 mRNA and 185 kD protein which is supposed to be the gene product, while negative cell lines did not.

Developmental pattern of calmodulin-binding proteins in rat jejunal epithelial cells

Calmodulin-binding proteins have been studied in presumptive rat jejunal epithelial cells and in purified rat brush borders during development. Incubation of nitrocellulose replicas with [125I] calmodulin revealed that, at immature stages (13-15 days of fetal life), only two calmodulin-binding bands were detectable with molecular masses of approximately 145,000 and 135,000 daltons. By fetal day 19, additional calmodulin-binding proteins of 240,000 and 110,000 daltons were observed. The 145,000- and 240,000-dalton calmodulin-binding bands contained polypeptides that were immunologically similar to caldesmon and to the alpha-subunit of the non-erythroid spectrin (fodrin) respectively. Antisera reactive with the 110K subunit of the microvillus 110K-calmodulin complex labelled a 135,000-dalton band which comigrated with one of the calmodulin-binding proteins. This 135,000-dalton immunoreactive polypeptide persisted until birth but was absent in brush borders isolated from adult intestine. In addition, the 110K antisera reacted with an approximately 110,000-dalton subunit by fetal day 19. At birth, numerous lower-molecular-mass 110K immunoreactive bands were also detectable. Immunocytochemical localization of the three calmodulin-binding proteins revealed that, at fetal day 14, caldesmon and fodrin displayed fluorescence lining the periphery of the epithelial cells, whereas staining with the 110K antisera was very weak. At fetal day 19, staining with the three antisera resulted in bright fluorescence localized in the apical part of the epithelial cells, in parallel to the differentiation of brush borders. At this stage, the apical staining of the calmodulin-binding proteins was similar to that of the adult.(ABSTRACT TRUNCATED AT 250 WORDS)

Assembly of the intestinal brush border: appearance and redistribution of microvillar core proteins in developing chick enterocytes

The assembly of the intestinal microvillus cytoskeleton during embryogenesis in the chick was examined by immunochemical and light microscopic immunolocalization techniques. For these studies, affinity-purified antibodies reactive with three major cytoskeletal proteins of the adult intestinal microvillus, fimbrin, villin, and the 110-kD subunit of the 110K-calmodulin protein complex were prepared. Immunocytochemical staining of frozen sections of embryonic duodena revealed that all three proteins were present at detectable levels at the earliest stages examined, day 7-8 of incubation (Hamilton/Hamburger stages 25-30). Although initially all three proteins were diffusely distributed throughout the cytoplasm, there was a marked asynchrony in the accumulation of these core proteins within the apical domain of the enterocyte. Villin displayed concentrated apical staining by embryonic day 8 (stage 28), while the apical concentration of fimbrin was first observed at embryonic day 10 (stage 37). Diffuse staining of the enterocyte cytoplasm with the anti-110K was observed throughout development until a few days before hatch. By embryonic day 19-21 110K staining was concentrated at the cell periphery (apical and basolateral). The restricted apical localization characteristic of 110K in the adult brush border was not observed until the day of hatching. Immunoblot analysis of whole, solubilized embryonic duodena confirmed the presence of 110K, villin, and fimbrin throughout development and indicated substantial increases in all three proteins, particularly late in development. Immunoblot staining with anti-110K also revealed the presence of a high molecular mass (200 kD) immunoreactive species in embryonic intestine. This 200-kD form was absent from isolated embryonic enterocytes and may be a component of intestinal smooth muscle.

Calcium-regulated cooperative binding of the microvillar 110K-calmodulin complex to F-actin: formation of decorated filaments

The 110K-calmodulin complex of intestinal microvilli is believed to be the link between the actin filaments comprising the core bundle and the surrounding cell membrane. Although not the first study describing a purification scheme for the 110K-calmodulin complex, a procedure for the isolation of stable 110K-calmodulin complex both pure and in high yield is presented; moreover, isolation is without loss of the associated calmodulin molecules since a previously determined ratio in isolated microvillar cytoskeletons of calmodulin to 110-kD polypeptide of 3.3:1 is preserved. We have found that removal of calmodulin from the complex by the calmodulin antagonists W7 or W13 results in precipitation of the 110-kD polypeptide with calmodulin remaining in solution. The interaction of 110K-calmodulin with beef skeletal muscle F-actin has been examined. Cosedimentation assays of 110K-calmodulin samples incubated with F-actin show the amount of 110K-calmodulin associating with F-actin to be ATP, calcium, and protein concentration dependent; however, relatively salt independent. In calcium, approximately 30% of the calmodulin remains in the supernatant rather than cosedimenting with the 110-kD polypeptide and actin. Electron microscopy of actin filaments after incubation with 110K-calmodulin in either calcium- or EGTA-containing buffers show polarized filaments often laterally associated. Each individual actin filament is seen to exhibit an arrowhead appearance characteristic of actin filaments after their incubation with myosin fragments, heavy meromyosin and subfragment 1. In some cases projections having a 33-nm periodicity are observed. This formation of periodically spaced projections on actin filaments provides further compelling evidence that the 110K-calmodulin complex is the bridge between actin and the microvillar membrane.

Calmodulin and calmodulin-binding proteins in the renal brush border

The calmodulin content of renal brush-border membrane vesicles, prepared by Mg2+-precipitation in EGTA-containing solutions, amounts to 1.8 micrograms per mg protein. The amount and the distribution of this EGTA-insensitive calmodulin was determined in membrane and cytoskeletal fractions prepared from the brush-border membrane vesicles by extraction with Triton X-100. The Triton X-100 insoluble pellet contains 21.2% of the protein and 52.2% of the EGTA-insensitive calmodulin, which amounts in this fraction to 4.4 micrograms per mg protein. Treatment of the Triton X-100 insoluble pellet, consisting of the microvillar core residue, with ATP and Mg2+ results in the solubilization of a relatively small number of proteins among which are actin, myosin, calmodulin and several calmodulin-binding proteins. The solubilization is partially reversible and a fraction of the proteins can be precipitated by centrifugation after the enzymatic hydrolysis of ATP. Readdition of ATP to the pellet results in the resolubilization of myosin, part of the actin, an 115-kDa calmodulin-binding protein and calmodulin. The calmodulin content of the final extract was 61.8 micrograms per mg protein. We have found roughly the same distribution pattern of calmodulin and ATP-solubilized, calmodulin-binding proteins in renal and intestinal brush-border preparations. The calmodulin content, however, as well as the relative amount of the calmodulin-binding proteins versus actin are about 4 to 5-times higher in intestinal than in renal microvillar core residues.

Further characterization of a fodrin-containing transmembrane complex from mouse T-lymphoma cells

A transmembrane complex containing fodrin (an actin-binding protein) and a major surface glycoprotein (GP 180) was previously isolated from mouse T-lymphoma cells by the complementary techniques of non-ionic detergent extraction and sucrose gradient centrifugation (Bourguignon et al. (1985) J. Cell Biol. 101, 477-487). The analysis of this complex has been extended to verify the structural association and further define the interaction between fodrin and GP 180. The association between fodrin and GP 180 has been confirmed by the following evidence: co-sedimentation of fodrin and GP 180 in a single peak on a sucrose gradient with a sedimentation coefficient of 20 S; a constant ratio of fodrin and GP 180 across the 20 S peak; the specific co-precipitation of GP 180 with fodrin from the 20 S peak using anti-fodrin antibody; and the colocalization of fodrin and GP 180 from the 20 S peak on actin filaments using an immuno-electron microscopic technique. Furthermore, this fodrin-GP 180 complex can be readily dissociated and reassembled in the presence and absence of 0.6 M NaCl, respectively. The fact that this fodrin-GP 180 complex displays actin-binding ability indicates that this transmembrane complex may play an important role in the linking event between receptors and the cytoskeleton during lymphocyte patching and capping.

Fractionation of detergent lysates of cells by ammonium sulphate-induced phase separation

A procedure is described for fractionating detergent lysates of cells based on the ability of (NH4)2SO4 to induce phase separation of detergents such as Triton X-100, sodium deoxycholate, and sodium cholate, into detergent-rich and detergent-depleted phases. An analysis of six murine lymphocyte cell surface molecules revealed that the partitioning in Triton X-100 of each molecule was highly dependent upon the (NH4)2SO4 concentration, each antigen partitioning into the detergent-rich phase at a defined salt concentration. In contrast, none of the six molecules appeared in the detergent-rich phase of a Triton X-114 phase separation, even though two of the molecules, namely Ly-2/3 and L3T4, are well-characterized integral membrane proteins. It was also observed that (NH4)2SO4 resulted in the partitioning of many nonmembrane proteins into the detergent-rich phase, indicating that the procedure can be used to fractionate all cellular proteins. By judicious choice of (NH4)2SO4 concentrations, precipitation of cellular proteins at two different (NH4)2SO4 concentrations, and combining the method with subcellular fractionation prior to detergent solubilization, substantial enrichment and concentration of particular cellular proteins could be achieved.

Phase separation of rat intestinal brush border membrane proteins using Triton X-114

Rat intestinal microvillus membrane contains at least 24 polypeptides, of which 18 can be solubilized using Triton X-114 at 4 degrees C. Upon phase separation at 32 degrees C, 11 proteins separated nearly completely into the detergent-rich phase, while 9 proteins were found exclusively in the aqueous phase. Enzymes which were uniquely included in the detergent phase were alkaline phosphatase, leucine aminopeptidase, gamma-glutamyl transpeptidase, and Ca2+-Mg2+ ATPase. The proteins which were excluded from the detergent phase and found exclusively in the aqueous phase included the disaccharidases (glucoamylase, sucrase-isomaltase, trehalase, lactase) and the ileal receptor for the intrinsic factor-cobalamin complex. Integral membrane proteins can thus be separated during solubilization into two groups prior to further purification or characterization.

Reevaluation of the hydrophobic nature of the 110-kD calmodulin-, actin-, and membrane-binding protein of the intestinal microvillus

A complex of calmodulin (CM) and the 110-kD (110K) subunit composes the helical array of cross-bridges linking the microvillus actin filament bundle with the membrane. The hydrophobic properties of the 110K protein, assessed by the detergent phase partitioning assay [Bordier C: J Biol Chem 256:1604, 1981], are highly dependent on the solution conditions used in its isolation. The ATP-dissociable 110K-CM complex [Howe and Mooseker: J Cell Biol 97:974, 1983] exhibits hydrophilic characteristics in this assay. In contrast, the 110K subunit extracted from brush borders by Triton X-100, sodium dodecyl sulfate, and sodium pyrophosphate (detergent-treated 110K) [Glenney JR, Glenney P: Cell 37:743, 1984] behaves as a hydrophobic protein. However, because the soluble hydrophilic 110K-CM can be rendered hydrophobic by treating the complex with the same detergent and salt conditions used in the preparation of detergent-treated 110K, the properties of detergent-treated 110K seem likely to be an effect of the solution conditions on its native conformation, sedimentability, or exposure of binding domains. In addition, the detergent-treated 110K is devoid of calmodulin and no longer exhibits the actin-binding activity characteristic of the ATP-dissociable 110K-CM and of the intact complex in situ. With two partially purified preparations of the 110K subunit exhibiting such dramatically distinct properties, it seems premature to define the nature of the 110K subunit's association with the membrane at this time.

Ciliary membrane tubulin and associated proteins: a complex stable to Triton X-114 dissociation

When either membranes from scallop gill cilia or reconstituted membranes from the same source are solubilized with Triton X-114 and the detergent is condensed by warming, no significant fraction of any major membrane protein partitions into the micellar detergent. Rather, most of the membrane lipids condense with the detergent phase, forming mixed micelles from which nearly pure lipid vesicles may be produced by adsorption of detergent with polystyrene beads. One minor membrane protein, with a molecular weight of about 20 000, is associated consistently with these vesicles. The aqueous phase contains a fairly homogeneous protein-Triton X-114 micelle sedimenting at 2.6 S in the analytical ultracentrifuge. Sucrose gradient velocity analysis in a detergent-free gradient indicates moderate size polydispersity but constant polypeptide composition throughout the sedimenting protein zone. Sucrose gradient equilibrium analysis (also in a detergent-free gradient) results in a protein-detergent complex banding at a density of 1.245 g/cm3. Sedimentation of the protein-detergent complex in the ultracentrifuge, followed by fixation and normal processing for electron microscopy, reveals a fine, reticular material consisting of 5-10-nm granules. These data are consistent with previous evidence that membrane tubulin and most other membrane proteins exist together as a discrete lipid-protein complex in molluscan gill ciliary membranes.

1,25-Dihydroxyvitamin D increases calmodulin binding to specific proteins in the chick duodenal brush border membrane

In previous studies we demonstrated that the biologically active vitamin D metabolite 1,25-dihydroxyvitamin D [1,25(OH)2D] increased the calmodulin (CaM) content of chick duodenal brush border membranes (BBM) without increasing the total cellular CaM content. Therefore, we evaluated the binding of CaM to discrete proteins in the BBM and determined whether 1,25(OH)2D could influence such binding. We observed one major and several minor CaM-binding bands on autoradiograms of sodium dodecyl sulfate polyacrylamide gels incubated with [125I]CaM. The major band had a molecular weight of 102,000-105,000. It bound CaM even in the presence of EGTA, but not in the presence of trifluoperazine or excess nonradioactive CaM. The administration of 1,25(OH)2D increased the apparent binding of CaM to this protein as assessed by densitometry of the autoradiogram. This increase in CaM binding coincided with the increased ability of the same BBM vesicles to accumulate calcium. Cycloheximide in doses that markedly reduced the incorporation of [35S]methionine into BBM proteins did not reduce the ability of 1,25-dihydroxyvitamin D3 to stimulate either calcium uptake by the BBM vesicles or CaM binding to the 102,000-105,000-mol-wt protein. These results suggest that 1,25(OH)2D administration increases the CaM content of duodenal BBM by increasing the ability of a 102,000-105,000-mol-wt protein to bind CaM. This mechanism may underlie the ability of 1,25(OH)2D to stimulate calcium movement across the intestinal BBM.

Interaction of alpha-crystallin with lens plasma membranes. Affinity for MP26

The binding of the major water-soluble lens protein alpha-crystallin to the lens plasma membrane has been investigated by reassociating purified alpha-crystallin with alpha-crystallin-depleted membranes and with phospholipid vesicles in which the lens membrane protein MP26 had been reconstituted. alpha-Crystallin reassociates at high affinity (Kd = 13 X 10(-8)M) with alkali-washed lens plasma membranes but not with lens plasma membranes treated with guanidine/HCl, nor with phospholipid vesicles or erythrocyte membranes. Binding to lens plasma membranes is dependent on salt, temperature and pH and occurs in a saturable manner. Reconstitution of MP26 into phospholipid vesicles and subsequent analysis of alpha-crystallin binding suggests the involvement of this transmembrane protein. Binding ist not influenced by pretreatment of membranes with proteases, suggesting that the 4-kDa cytoplasmic fragment of MP26 is not necessary for alpha-crystallin binding. Labeling experiments using (trifluoromethyl)-3-(m-[125I]iodophenyl)diazirine as a probe for intrinsic membrane proteins further showed that alpha-crystallin contains hydrophobic regions on its surface which might enable this protein to make contact with the lipid bilayer. Newly synthesized alpha-crystallin, in lens culture, is not associated with the plasma membrane, suggesting that the assembly of alpha-crystallin aggregates does not take place in a membrane-bound mode.

Localization of cell surface glycoproteins in membrane domains associated with the underlying filament network

To visualize the localization of cell surface constituents in relation to the plasma membrane-associated filament network, we developed a method based on a combination of immunogold labeling and dry-cleaving. For labeling we used trinitrophenyl-derivatized ligand, anti-TNP antibodies, and protein A-coated colloidal gold. Dry-cleaving (Mesland, D. A. M., H. Spiele, and E. Roos, 1981, Exp. Cell Res., 132: 169-184) involves cleavage of lightly fixed critical point-dried cells by means of adhesive tape. Since cells cleave close to the cell surface, the remaining layer is thin enough to be examined in transmission electron microscopy. Using this method, we studied concanavalin A-binding constituents on the medium-facing surface of H35 hepatoma cells. The distribution of the gold particles, which was partly dispersed and partly patchy, coincided strikingly with membrane-associated filaments, and label was virtually absent from areas overlying openings in the filament network. In stereo pairs we observed the label to be localized to areas of somewhat enhanced electron density at the plane of the membrane. These areas were interconnected in a pattern congruent with the filament network. Preliminary observations on wheat germ agglutinin receptors on the hepatoma cells as well as concanavalin A receptors on isolated hepatocytes yielded comparable results. It thus appears that surface glycoproteins, although seemingly randomly distributed as observed in thin sections, may actually be localized to particular membrane domains associated with underlying filaments.

The membrane skeleton of a unicellular organism consists of bridged, articulating strips

In this paper we show that a membrane skeleton associated with the plasma membrane of the unicellular organism Euglena consists of approximately 40 individual S-shaped strips that overlap along their lateral margins. The region of strip overlap is occupied by a set of microtubule-associated bridges and microtubule-independent bridges. Both cell form and plasma membrane organization are dependent on the integrity of this membrane skeleton. Removal of the membrane skeleton with a low-molar base results in loss of membrane form and randomization of the paracrystalline membrane interior characteristic of untreated cells. Conversely, removal of the plasma membrane and residual cytoplasm with lithium 3,5-diiodosalicylate/Nonidet P-40 yields cell ghosts that retain the form of the original cell but consist only of the membrane skeleton. Two major polypeptides of 86 and 80 KD persist in the skeleton and two other major proteins of 68 and 39 kD are associated with the plasma membrane fraction. None of these components appears to be the same as the major polypeptides (spectrins, band 3) of the erythrocyte ghost, the other cell system in which a well-defined peripheral membrane skeleton has been identified. We suggest that the articulating strips of euglenoids are not only the basic unit of cell and surface form, but that they are also positioned to mediate or accommodate surface movements by sliding, and to permit surface replication by intussusception.