Search

Clathrin-coated vesicles execute receptor-mediated endocytosis at the plasma membrane. However, a role for clathrin in later endocytic trafficking processes, such as receptor sorting and recycling or maintaining the organization of the endocytic pathway, has not been thoroughly characterized. The existence of clathrin-coated buds on endosomes suggests that clathrin might mediate later endocytic trafficking events. To investigate the function of clathrin-coated buds on endosomal membranes, endosome function and distribution were analyzed in a HeLa cell line that expresses the dominant-negative clathrin inhibitor Hub in an inducible manner. As expected, Hub expression reduced receptor-mediated endocytosis at the plasma membrane. Hub expression also induced a perinuclear aggregation of early endosome antigen 1-positive early endosomes, such that sorting and recycling endosomes were found tightly concentrated in the perinuclear region. Despite the dramatic redistribution of endosomes, Hub expression did not affect the overall kinetics of receptor sorting or recycling. These data show that clathrin function is necessary to maintain proper cellular distribution of early endosomes but does not play a prominent role in sorting and recycling events. Thus, clathrin's role on endosomal membranes is to influence organelle localization and is distinct from its role in trafficking pathways at the plasma membrane and trans-Golgi network.

The role of clathrin in the sorting of proinsulin to secretory granules, the formation of immature granules and their subsequent maturation is not known. To this end, primary rat pancreatic beta-cells were infected with a recombinant adenovirus co-expressing the Hub fragment, a dominant-negative peptide of the clathrin heavy chain and enhanced green fluorescent protein (EGFP as a marker of infected cells). A population of cells expressing the highest levels of EGFP (and thus Hub) was obtained using a fluorescence-activated cell sorter (FACS). Control cells were infected with an adenovirus expressing EGFP alone. By immunofluorescence, control cells showed intense staining for both clathrin light chain and proinsulin in a perinuclear region. In cells expressing high levels of Hub, the clathrin light-chain signal was faint and diffuse in keeping with its displacement from membranes. There was, however, no detectable effect of Hub expression on proinsulin staining or disposition within the cell. Proinsulin sorting and conversion, and the fate (release and/or degradation) of insulin and C-peptide, was studied by pulse-chase and quantitative reverse phase HPLC. In both Hub-expressing and control cells, >99% of all newly synthesized proinsulin was sorted to the regulated pathway and there was no effect of Hub on proinsulin conversion to insulin. In presence of Hub there was, however, a significant increase in the percentage of C-peptide truncated to des-(27-31)-C-peptide at early times of chase as well as more extensive degradation of C-peptide thereafter. It is concluded that clathrin is not implicated in the sorting or processing of proinsulin or in regulated exocytosis of secretory granules. These results confirm a role for clathrin in the removal of proteases from maturing granules, thus explaining the increased truncation and degradation of C-peptide in cells expressing Hub.

Adaptors are the membrane-binding components of clathrin-coated vesicles. The interaction of the trans-Golgi coat adaptor AP1 with membrane-associated proteins was analyzed by affinity chromatography. Proteins of 83 and 52 kDa bound specifically to the core domain of AP1 and showed no interaction with AP2 or other clathrin-coated vesicle proteins. The AP1-binding proteins were tightly membrane-associated, though behaved as peripheral membrane proteins. They were detected in membranes depleted of clathrin-coated vesicles and not in coated vesicles, suggesting that the interaction of these proteins with AP1 may precede coated vesicle budding. Co-fractionation of the AP1-binding proteins with trans-Golgi network membrane was also observed. Upon gel filtration, both AP1-binding proteins eluted in a high molecular mass complex which was labile at high concentrations of Tris. The 83 kDa protein bound to AP1 affinity resin in the absence of the 52 kDa protein. In contrast, the separated 52 kDa protein did not bind AP1, suggesting that the 83 kDa protein is the AP1-binding component of the complex. Characterization of this protein complex defines a novel membrane-associated component that specifically interacts with AP1 and may contribute to its function in forming clathrin-coated vesicles.

The role of the 100-kDa polypeptide components of clathrin-coated vesicles in endocytosis was investigated by microinjection of specific monoclonal antibodies. Receptor-mediated uptake of transferrin and liposomes was quantitatively inhibited. These results show that the 100-kDa polypeptides are directly involved in localized clathrin assembly at the cell periphery and are markers for the endocytic pathway. This demonstrates an in situ function of these polypeptides and the protein complexes in which they are found.

The light chains (LCa and LCb) of bovine brain clathrin are resistant to heat denaturation by boiling, a property shared by tropomyosin (Bailey, K., 1948, Biochem. J., 43:271-281). Light chains were partially purified by boiling and centrifugation of a Tris-extract of crude membranes prepared from bovine brains (Keen, J. H., M. C. Willingham, and I. H. Pastan, 1979, Cell., 16:303-312). Contaminant polypeptides were then removed by size-exclusion high-pressure liquid chromatography. The purified light chains were separated from each other by using an immunoaffinity column prepared from a monoclonal antibody CVC.7 specific for LCa and not LCb.

Clathrin was isolated from detergent-solubilized, biosynthetically radiolabeled cells by immunoprecipitation with anti-clathrin monoclonal antibodies. Immunoprecipitates obtained after pulse-chase labeling demonstrated that after biosynthesis the LCa light chain of clathrin could be found either complexed to heavy chain or in a free pool (not associated with heavy chain) which decreased steadily over time. More than half of the free LCa disappeared within the first hour after biosynthesis, but some was still detectable after several hours. Incorporation of clathrin LCa light chain and heavy chain into coated vesicles was coordinate and increased up to 4 h after biosynthesis. Comparison of these kinetics suggested that once incorporated into coated vesicles, LCa and heavy chain did not dissociate, even during depolymerization of the vesicle. There was also little apparent degradation of clathrin found in coated vesicles for up to 22 h after biosynthesis. Immunoprecipitation with anti-clathrin monoclonal antibodies was carried out after fractionation of continuously radiolabeled cell lysates using two different sizing columns. These experiments indicated that the triskelion form of clathrin that has been isolated from coated vesicles in vitro also exists in vivo. They also confirmed the existence of a transient but detectable pool of newly synthesized free LCa light chain.

Three monoclonal antibodies that react with previously undefined antigenic determinants on the clathrin molecule have been produced and characterized. They were isolated from a fusion between myeloma cells and popliteal lymphocytes from SJL mice that had received footpad injections of human brain clathrin. This protocol was chosen to favor the production of antibodies to poorly immunogenic proteins and thereby increase the repertoire of anti-clathrin monoclonal antibodies. One antibody (X16) reacts preferentially with the heavier of the two clathrin light chains (LCa) when it is not associated with heavy chain. This specificity is different from that of the anti-LCa antibody, CVC.6, which has preferential reactivity with heavy chain-associated LCa. In addition, X16 and CVC.6 bound simultaneously to LCa, confirming that they react with different sites. The other two antibodies produced, X19 and X22, react with two different determinants on the clathrin heavy chain, based on immunoprecipitation, Western blot, and binding studies. Competitive binding studies with anti-clathrin monoclonal antibodies showed that they define a total of five distinct antigenic determinants on bovine clathrin.

The binding of clathrin-coated vesicles, clathrin triskelions, and free clathrin light chains to calmodulin-Sepharose was compared. When isolated from bovine brain, all three components bound to calmodulin-Sepharose in the presence of calcium and could be eluted by its removal. In contrast, coated vesicles and triskelions isolated from bovine adrenal gland did not bind to calmodulin-Sepharose, although the free light chains from adrenal gland bound as effectively as those from brain. As distinct isoforms of the clathrin light chains are expressed by brain and adrenal gland, these results implicate the clathrin light chains as the calmodulin-binding component of coated vesicles and triskelions. Furthermore, the insertion sequences found in the neuron-specific isoforms, although not necessary for the binding of free clathrin light chains to calmodulin, must facilitate the interaction of heavy chain-associated light chains with calmodulin. Recombinant mutants of LCa, with deletions spanning the entire sequence, were tested for binding to calmodulin-Sepharose. Those mutants retaining structural integrity, as assessed by the binding of a panel of monoclonal antibodies, exhibited varying amounts of calmodulin binding activity. However, deletion of the carboxyl-terminal 20 residues abolished calmodulin interaction. Thus, the carboxyl terminus of LCa appears to constitute a calmodulin-binding site. Peptides corresponding to the carboxyl terminus of LCa or LCb inhibited the interaction of the light chains with calmodulin, suggesting that this region forms the calmodulin-binding site of both LCa and LCb. The carboxyl-terminal peptides of LCa and LCb inhibited the interaction of light chains with calmodulin approximately 10-fold less effectively than a calmodulin-binding peptide derived from smooth muscle myosin light chain kinase, but much more effectively than a calmodulin-binding peptide derived from adenylate cyclase. This comparison places the clathrin light chain-calmodulin interaction within the physiological range seen for other calmodulin-binding proteins.

Three forms of clathrin light chain contain two cysteine residues. These are the predominant brain-specific forms of LCa and LCb and the non-brain form of LCb. After purification in the absence of thiols they contain intramolecular disulphide bonds. The reduced and the oxidized forms show differences in electrophoretic mobility, explaining the variable and heterogeneous patterns observed on electrophoresis. Accessibility of the thiol groups in the free light chains is greater than when they are associated with the heavy chain. In contrast the cysteine residues of the clathrin heavy chain are completely inaccessible in the absence of denaturants and are not found in disulphide bonds. The antigenic properties of the oxidized and the reduced forms of the clathrin light chains are similar, as is their capacity to bind to the clathrin heavy chain. After isolation in the presence of 10 mM-iodoacetamide, the light-chain cysteine residues are fully alkylated. The results are consistent with the reduced form being the native state and the light-chain disulphide bonds an artifact of isolation.

Two regions on the clathrin heavy chain that are involved in triskelion interactions during assembly have been localized on the triskelion structure. These regions were previously identified with anti-heavy chain monoclonal antibodies X19 and X35, which disrupt clathrin assembly (Blank, G. S., and F. M. Brodsky, 1986, EMBO (Eur. Mol. Biol. Organ.) J., 5:2087-2095). Antibody-binding sites were determined based on their reactivity with truncated triskelions, and were mapped to an 8-kD region in the middle of the proximal portion of the triskelion arm (X19) and a 6-kD region at the triskelion elbow (X35). The elbow site implicated in triskelion assembly was also shown to be included within a heavy chain region involved in binding the light chains and to constitute part of the light chain-binding site. We postulate that this region of the heavy chain binds to the interaction site identified on the light chains that has homology to intermediate filament proteins (Brodsky, F. M., C. J. Galloway, G. S. Blank, A. P. Jackson, H.-F. Seow, K. Drickamer, and P. Parham, 1987, Nature (Lond.), 326:203-205). These findings suggest the existence of a heavy chain site, near the triskelion elbow, which is involved in both intramolecular and intermolecular interactions during clathrin assembly.

Isolated human fibroblast plasma membranes that were attached by their extracellular surface to a solid substratum contained numerous clathrin coated pits that could be removed with a high pH buffer (Moore, M.S., D.T. Mahaffey, F.M. Brodsky, and R.G.W. Anderson. 1987. Science [Wash. DC]. 236:558-563). When these membranes were incubated with coat proteins extracted from purified bovine coated vesicles, new coated pits formed that were indistinguishable from native coated pits. Assembly was dependent on the concentration of coat protein with half maximal assembly occurring at 7 micrograms/ml. Assembly was only slightly affected by the presence of divalent cations. Whereas normal appearing lattices formed in a low ionic strength buffer, when assembly was carried out in a low pH buffer, few coated pits were evident but numerous small clathrin cages decorated the membrane. Coated pits did not form randomly on the surface; instead, they assembled at differentiated regions of membrane that could be distinguished in carbon/platinum replicas of frozen and etched membranes by the presence of numerous particles clustered into patches the size and shape of a coated pit.