Relationship between endosomes and lysosomes.

Delivery of endocytosed macromolecules to lysosomes occurs by means of direct fusion of late endosomes with lysosomes. This has been formally demonstrated in a cell-free content mixing assay using late endosomes and lysosomes from rat liver. There is evidence from electron microscopy studies that the same process occurs in intact cells. The fusion process results in the formation of hybrid organelles from which lysosomes are re-formed. The discovery of the hybrid organelle has opened up three areas of investigation: (i) the mechanism of direct fusion of late endosomes and lysosomes, (ii) the mechanism of re-formation of lysosomes from the hybrid organelle, and (iii) the function of the hybrid organelle. Fusion has analogies with homotypic vacuole fusion in yeast. It requires syntaxin 7 as part of the functional trans-SNARE [SNAP receptor, where SNAP is soluble N-ethylmaleimide-sensitive factor (NSF) attachment protein] complex and the release of lumenal calcium to achieve membrane fusion. Re-formation of lysosomes from the hybrid organelle occurs by a maturation process involving condensation of lumenal content and probably removal of some membrane proteins by vesicular traffic. Lysosomes may thus be regarded as a type of secretory granule, storing acid hydrolases in between fusion events with late endosomes. The hybrid organelle is predicted to function as a 'cell stomach', acting as a major site of hydrolysis of endocytosed macromolecules.

Syntaxin 7 is localized to late endosome compartments, associates with Vamp 8, and Is required for late endosome-lysosome fusion.

Protein traffic from the cell surface or the trans-Golgi network reaches the lysosome via a series of endosomal compartments. One of the last steps in the endocytic pathway is the fusion of late endosomes with lysosomes. This process has been reconstituted in vitro and has been shown to require NSF, alpha and gamma SNAP, and a Rab GTPase based on inhibition by Rab GDI. In Saccharomyces cerevisiae, fusion events to the lysosome-like vacuole are mediated by the syntaxin protein Vam3p, which is localized to the vacuolar membrane. In an effort to identify the molecular machinery that controls fusion events to the lysosome, we searched for mammalian homologues of Vam3p. One such candidate is syntaxin 7. Here we show that syntaxin 7 is concentrated in late endosomes and lysosomes. Coimmunoprecipitation experiments show that syntaxin 7 is associated with the endosomal v-SNARE Vamp 8, which partially colocalizes with syntaxin 7. Importantly, we show that syntaxin 7 is specifically required for the fusion of late endosomes with lysosomes in vitro, resulting in a hybrid organelle. Together, these data identify a SNARE complex that functions in the late endocytic system of animal cells.

The role of intraorganellar Ca(2+) in late endosome-lysosome heterotypic fusion and in the reformation of lysosomes from hybrid organelles.

We have investigated the requirement for Ca(2+) in the fusion and content mixing of rat hepatocyte late endosomes and lysosomes in a cell-free system. Fusion to form hybrid organelles was inhibited by 1,2-bis(2-aminophenoxy) ethane-N,N,N',N'-tetraacetic acid (BAPTA), but not by EGTA, and this inhibition was reversed by adding additional Ca(2+). Fusion was also inhibited by methyl ester of EGTA (EGTA-AM), a membrane permeable, hydrolyzable ester of EGTA, and pretreatment of organelles with EGTA-AM showed that the chelation of lumenal Ca(2+) reduced the amount of fusion. The requirement for Ca(2+) for fusion was a later event than the requirement for a rab protein since the system became resistant to inhibition by GDP dissociation inhibitor at earlier times than it became resistant to BAPTA. We have developed a cell-free assay to study the reformation of lysosomes from late endosome-lysosome hybrid organelles that were isolated from the rat liver. The recovery of electron dense lysosomes was shown to require ATP and was inhibited by bafilomycin and EGTA-AM. The data support a model in which endocytosed Ca(2+) plays a role in the fusion of late endosomes and lysosomes, the reformation of lysosomes, and the dynamic equilibrium of organelles in the late endocytic pathway.

Lysosome-endosome fusion and lysosome biogenesis.

Recent data both from cell-free experiments and from cultured cells have shown that lysosomes can fuse directly with late endosomes to form a hybrid organelle. This has a led to a hypothesis that dense core lysosomes are in essence storage granules for acid hydrolases and that, when the former fuse with late endosomes, a hybrid organelle for digestion of endocytosed macromolecules is created. Lysosomes are then re-formed from hybrid organelles by a process involving condensation of contents. In this Commentary we review the evidence for formation of the hybrid organelles and discuss the current status of our understanding of the mechanisms of fusion and lysosome re-formation. We also review lysosome biosynthesis, showing how recent studies of lysosome-like organelles including the yeast vacuole, Drosophila eye pigment granules and mammalian secretory lysosomes have identified novel proteins involved in this process.

Fusion of lysosomes with late endosomes produces a hybrid organelle of intermediate density and is NSF dependent.

Using a cell-free content mixing assay containing rat liver endosomes and lysosomes in the presence of pig brain cytosol, we demonstrated that after incubation at 37 degrees C, late endosome-lysosome hybrid organelles were formed, which could be isolated by density gradient centrifugation. ImmunoEM showed that the hybrids contained both an endocytosed marker and a lysosomal enzyme. Formation of the hybrid organelles appeared not to require vesicular transport between late endosomes and lysosomes but occurred as a result of direct fusion. Hybrid organelles with similar properties were isolated directly from rat liver homogenates and thus were not an artifact of cell-free incubations. Direct fusion between late endosomes and lysosomes was an N-ethylmaleimide-sensitive factor-dependent event and was inhibited by GDP-dissociation inhibitor, indicating a requirement for a rab protein. We suggest that in cells, delivery of endocytosed ligands to an organelle where proteolytic digestion occurs is mediated by direct fusion of late endosomes with lysosomes. The consequences of this fusion to the maintenance and function of lysosomes are discussed.

Dense core lysosomes can fuse with late endosomes and are re-formed from the resultant hybrid organelles.

Electron microscopy was used to evaluate the function and formation of dense core lysosomes. Lysosomes were preloaded with bovine serum albumin (BSA)-gold conjugates by fluid phase endocytosis using a pulse-chase protocol. The gold particles present in dense core lysosomes and late endosomes were flocculated, consistent with proteolytic degradation of the BSA. A second pulse of BSA-gold also accumulated in the pre-loaded dense core lysosomes at 37 degrees C, but accumulation was reversibly blocked by incubation at 20 degrees C. Time course experiments indicated that mixing of the two BSA-gold conjugates initially occurred upon fusion of mannose 6-phosphate receptor-positive/lysosomal glycoprotein-positive late endosomes with dense core lysosomes. Treatment for 5 hours with wortmannin, a phosphatidyl inositide 3-kinase inhibitor, caused a reduction in number of dense core lysosomes preloaded with BSA-gold and prevented a second pulse of BSA-gold accumulating in them. After wortmannin treatment the two BSA-gold conjugates were mixed in swollen late endosomal structures. Incubation of NRK cells with 0.03 M sucrose resulted in the formation of swollen sucrosomes which were morphologically distinct from preloaded dense core lysosomes and were identified as late endosomes and hybrid endosome-lysosome structures. Subsequent endocytosis of invertase resulted in digestion of the sucrose and re-formation of dense core lysosomes. These observations suggest that dense core lysosomes are biologically active storage granules of lysosomal proteases which can fuse with late endosomes and be re-formed from the resultant hybrid organelles prior to subsequent cycles of fusion and re-formation.

The effect of wortmannin on the localisation of lysosomal type I integral membrane glycoproteins suggests a role for phosphoinositide 3-kinase activity in regulating membrane traffic late in the endocytic pathway.

Addition of wortmannin to normal rat kidney cells caused a redistribution of the lysosomal type I integral membrane proteins Igp110 and Igp120 to a swollen vacuolar compartment. This compartment did not contain the cation independent mannose 6-phosphate receptor and was depleted in acid hydrolases. It was distinct from another swollen vacuolar compartment containing the cation independent mannose 6-phosphate receptor. The swollen Igp110-positive compartment was accessible to a monoclonal antibody against Igp120 added extracellularly, showing that it had the characteristics of an endosomal compartment. Wortmannin had no gross morphological effect on the trans-Golgi network or lysosomes nor any effect on the delivery to the trans-Golgi network of endocytosed antibodies against the type I membrane protein TGN38. We propose that the observed effects of wortmannin were due to inhibition of membrane traffic between cation independent mannose 6-phosphate receptor-positive late endosomes and the trans-Golgi network and to inhibition of membrane traffic between a novel Igp120-positive, cation independent mannose 6-phosphate receptor-negative late endosomal compartment and lysosomes. The effects of wortmannin suggest a function for a phosphatidylinositol 3-kinase(s) in regulating membrane traffic in the late endocytic pathway.

Identification of a lysosomal protein causing lipid transfer, using a fluorescence assay designed to monitor membrane fusion between rat liver endosomes and lysosomes.

In the present and previous studies [Mullock, Perez, Kuwana, Gray and Luzio (1994) J. Cell Biol. 126, 1173-1182], we have attempted to investigate endosome-lysosome fusion using an assay based on the dilution of the self-quenching fluorescent lipid probe octadecylrhodamine. Although some characteristics of fluorescence dequenching were consistent with those observed in other cell-free assays, we have now demonstrated that increased fluorescence was due to leakage of an intralysosomal lipid-transfer protein. This protein was purified and found to be a 22 kDa molecule with sequence, immunological and functional characteristics strongly suggesting that it is the rat homologue of human GM2-activator protein. Both the 22 kDa protein and recombinant human GM2-activator protein caused fluorescence dequenching either when mixed with octadecylrhodamine-loaded endosomes and lysosomal membranes or in a liposome system. The data were consistent with GM2-activator protein acting as an octadecylrhodamine-transfer protein. Antibodies to the 22 kDa protein added to cell-free endosome-lysosome content-mixing assays had no effect, although they could inhibit fluorescence dequenching caused by the protein. Thus this protein is not required in any fusion event involved in delivery of ligands from endosomes to lysosomes. The existence within an intracellular organelle of a protein capable of acting as an octadecylrhodamine-transfer protein suggests the need for caution in the interpretation of fluorescence-dequenching assays using mammalian subcellular fractions.

Lysosomes can fuse with a late endosomal compartment in a cell-free system from rat liver.

The passage of pulse doses of asialoglycoproteins through the endosomal compartments of rat liver hepatocytes was studied by subcellular fractionation and EM. The kinetics of disappearance of radiolabeled asialofetuin from light endosomes prepared on Ficoll gradients were the same as the kinetics of disappearance of asialoorosomucoid-horse radish peroxidase reaction products from intracellular membrane-bound structures in the blood sinusoidal regions of hepatocytes. The light endosomes were therefore identifiable as being derived from the peripheral early endosome compartment. In contrast, the labeling of dense endosomes from the middle of the Ficoll gradient correlated with EM showing large numbers of reaction product-containing structures in the nonsinusoidal parts of the hepatocyte. In cell-free, postmitochondrial supernatants, we have previously observed that dense endosomes, but not light endosomes, interact with lysosomes. Cell-free interaction between isolated dense endosomes and lysosomes has now been reconstituted and analyzed in three ways: by transfer of radiolabeled ligand from endosomal to lysosomal densities, by a fluorescence dequenching assay which can indicate membrane fusion, and by measurement of content mixing. Maximum transfer of radiolabel to lysosomal densities required ATP and GTP plus cytosolic components, including N-ethylmaleimide-sensitive factor(s). Dense endosomes incubated in the absence of added lysosomes did not mature into vesicles of lysosomal density. Content mixing, and hence fusion, between endosomes and lysosomes was maximal in the presence of cytosol and ATP and also showed inhibition by N-ethyl-maleimide. Thus, we have demonstrated that a fusion step is involved in the transfer of radiolabeled ligand from an isolated endosome fraction derived from the nonsinusoidal regions of the hepatocyte to preexisting lysosomes in a cell-free system.

The presence and measurement of secretory component in human bile and blood.

Five monoclonal antibodies which recognized three separate epitopes on the free secretory component molecule were produced using free secretory component obtained from human colostrum. Two-site immunoradiometric assays were developed to measure free secretory component and secretory IgA. Monoclonal antibody M9 was used on coated plates as the capture antibody. Monoclonal antibody M7 was used as the labelled signal antibody for the assay of free secretory component and a commercially available monoclonal anti-IgA antibody was used as the labelled signal antibody for the assay of secretory IgA. Free secretory component was found in human serum and bile. In serum, its concentration was raised in patients with high serum alkaline phosphatase due to liver disorders but not in patients with high serum alkaline phosphatase due to non-liver disorders. In bile from bile duct drains collected during the first week after liver transplantation, free secretory component was found in concentrations of up to 33 mg/l, in vast excess of that found in bile from gallstone patients (up to 0.3 mg/l). Bile from gallstone patients but not from liver transplant patients produced proteolytic degradation of free secretory component when incubated in vitro. The finding of large amounts of free secretory component, the free cleaved fragment of the polymeric IgA receptor in human bile, further supports the existence of the blood to bile transhepatocytic pathway in humans.

The polymeric immunoglobulin A receptor is present on hepatocytes in human liver.

A monoclonal antibody raised against human colostrum secretory component produced even staining of hepatocyte plasma membranes, as well as bile duct lining cells, in all sections examined from eight normal and three abnormal human livers. Human bile samples incubated with free secretory component degraded it to varying extents, probably proteolytically; true levels of free secretory component will therefore often be higher than those reported. It seems likely that human liver resembles that of other mammals in transferring polymeric IgA through hepatocytes to the bile by means of the polymeric IgA receptor.

Reconstitution of an endosome-lysosome interaction in a cell-free system.

A cell-free model for the transfer of endocytosed material to lysosomes is described. Rat liver late endosomes, loaded in vivo with radiolabeled ligand by intravenous injection shortly before killing the animal, showed a specific interaction with lysosomes when incubated at 37 degrees C in the presence of cytosol and an ATP regenerating system. The location of the ligand, generally asialofetuin, was analyzed by isopycnic centrifugation on Nycodenz gradients. Appearance of radiolabel in the lysosomal position on such gradients was maximal after approximately 30 min at 37 degrees C and required the provision of undamaged cytosol, lysosomes, and an ATP regenerating system. It could not be accounted for by nonspecific bulk aggregation of membranes. Transfer occurred only from late endosomes; radiolabel in early endosomes was unaffected. Digestion of the asialofetuin, as shown by the appearance of TCA-soluble radioactivity, occurred on incubation at 37 degrees C and was increased by the provision of an ATP regenerating system.

Investigation of endosomal compartments involved in endocytosis and transcytosis of polymeric immunoglobulin A by subcellular fractionation of perfused isolated rat liver.

1. A gamma camera was used to monitor continuously the uptake of radiolabelled polymeric immunoglobulin A (pIgA) into the rat body after intravenous injection. Uptake into liver was fast but, since the peak of liver labelling occurred only after 9-15 min, it was not sufficiently rapid to constitute a pulse dose. A perfused, isolated rat liver system was therefore established which could be given a single pass dose of pIgA; a variety of tests showed such livers remained viable for at least 3 h and could be subsequently fractionated on Ficoll and Nycodenz gradients with normal distributions of marker enzymes. 2. Subcellular fractionation at different times after a single pass dose of pIgA showed that whilst pIgA appeared sequentially in sinusoidal plasma membrane, light endosomes, dense endosomes, very dense endosomes and lysosomes as in vivo, the predominance of pIgA in the light endosome compartment disappeared much earlier than after injection in vivo of pIgA, presumably because this compartment was not being continuously loaded over the first 10-15 min. The time course of appearance of label in bile was unchanged. A large excess of unlabelled asialofetuin did not change these patterns, indicating that the asialoglycoprotein receptor was not involved. 3. Low doses of the microtubule agent colchicine reduced the proportion of pIgA reaching the bile, but subcellular fractionation of treated liver showed that distribution of label amongst liver fractions was little changed, although the overall liver pIgA content had increased. This would suggest that pIgA did not remain in the common compartment which could have supplied bile or lysosomes but rather flowed out of it as rapidly as in untreated liver but towards those compartments supplying the lysosomes. 4. Experiments with nocodazole, which reversibly disrupts microtubules, showed that very little of the pIgA taken into an inhibited liver appeared in the bile after nocodazole was removed 30 min later, even though a second dose of pIgA, given after nocodazole removal, appeared in bile with a normal time course. The first dose of pIgA must therefore have passed beyond the compartments competent to supply the bile before nocodazole was removed. Such compartments were undamaged since the second dose of pIgA appeared in bile normally. We therefore conclude that the bulk of pIgA must be supplied to the bile from light or dense endosomes rather than from very dense endosomes and lysosomes.

The preparative isolation of endosome fractions: a review.

The endosome is an intracellular, acidic, membrane-bound, subcellular compartment to which endocytosed ligands, receptors and plasma membrane proteins are conveyed before sorting and delivery to destinations elsewhere in the cell. The preparative isolation of elements of this compartment has been achieved successfully using various appropriate combinations of density gradient ultracentrifugation, electrophoretic, gel filtration and immunoaffinity techniques. These methods for isolating endosome fractions are reviewed together with the difficulties of establishing markers for such fractions. The isolation of an endosome fraction from the pathway of polymeric IgA transcytosis in rat liver is discussed to exemplify successful isolation procedures and appropriate subcellular markers.

Rapid subcellular fractionation of the rat liver endocytic compartments involved in transcytosis of polymeric immunoglobulin A and endocytosis of asialofetuin.

The distributions of two endocytosed radiolabelled ligands (polymeric immunoglobulin A and asialofetuin) in rat liver endocytic compartments were investigated by using rapid subcellular fractionation of post-mitochondrial supernatants on vertical density gradients of Ficoll or Nycodenz. Two endocytic compartments were identified, both ligands being initially associated with a light endocytic-vesicle fraction on Ficoll gradients, asialofetuin then accumulating in denser endosomes before transfer to lysosomes for degradation.

Sources of proteins in human bile.

The proteins of 46 human bile specimens, collected by several different routes have been studied by crossed immunoelectrophoresis, by rocket immunoelectrophoresis and by radioimmunoassay. The results were analysed by plotting the variation in the bile: plasma ratio of particular proteins against molecular weight and by examination of the correlation between the concentrations of different proteins in the biles of different patients. Our results show that the majority of human bile proteins derive from plasma although bile specific proteins are always present. The majority of plasma proteins appear to enter bile by a 'sieving' mechanism which results in an inverse relationship between the bile: plasma ratio and the molecular weight. In addition there was a very high degree of correlation between the biliary concentrations of alpha 2-macroglobulin, IgG, haptoglobin, haemopexin, albumin, prealbumin, and orosomucoid. A number of other proteins namely thyroxine binding globulin, GC globulin and alpha 2HS-glycoprotein appeared in bile at concentrations greater than those expected if entry is by the sieving mechanism. These three proteins, however, are of rather low molecular weight and the reason for the lack of correlation appears to be individual variation in the 'pore size', presumably reflecting variation in the porosity of tight junction between hepatocytes. Although the majority of human bile proteins would appear to enter bile by a molecular weight-dependent pathway, four proteins, namely secretory IgA, IgM, haemoglobin and caeruloplasmin, showed significant deviation from the predicted relationship and probably enter bile at least partly by transport across cells. The concentration of beta 2-glycoprotein I was also much greater than expected from its molecular weight. The reason for this is not yet clear but may well reflect a very efficient and specific transport mechanism.