Membrane transport in the endocytic pathway.

Despite controversies and debates, some fundamental properties of endosomes become apparent when comparing results from in vivo and in vitro strategies used to study endosomal membrane traffic. In addition, recent studies are starting to unravel the complex organization of early endosomes, in particular along the route followed by recycling receptors.

Rme-1 regulates the distribution and function of the endocytic recycling compartment in mammalian cells.

RME-1 is an Eps15-homology (EH)-domain protein that was identified in a genetic screen for endocytosis genes in Caenorhabditis elegans. When expressed in a CHO cell line, the worm RME-1 protein and a mouse homologue are both associated with the endocytic recycling compartment. Here we show that expression of a dominant-negative construct with a point mutation near the EH domain results in redistribution of the endocytic recycling compartment and slowing down of transferrin receptor recycling. The delivery of a TGN38 chimaeric protein to the trans-Golgi network is also slowed down. The function of Rme-1 in endocytic recycling is evolutionarily conserved in metazoans as shown by the protein's properties in C. elegans.

Fibrin regulates neutrophil migration in response to interleukin 8, leukotriene B4, tumor necrosis factor, and formyl-methionyl-leucyl-phenylalanine.

We have examined the capacity of four different chemoattractants/cytokines to promote directed migration of polymorphonuclear leukocytes (PMN) through three-dimensional gels composed of extracellular matrix proteins. About 20% of PMN migrated through fibrin gels and plasma clots in response to a gradient of interleukin 8 (IL-8) or leukotriene B4 (LTB4). In contrast, < 0.3% of PMN migrated through fibrin gels in response to a gradient of tumor necrosis factor alpha (TNF) or formyl-methionyl-leucyl-phenylalanine (FMLP). All four chemoattractants stimulated PMN to migrate through gels composed of collagen IV or of basement membrane proteins (Matrigel), or through filters to which fibronectin or fibrinogen had been adsorbed. PMN stimulated with TNF or FMLP adhered and formed zones of close apposition to fibrin, as measured by the exclusion of a 10-kD rhodamine-polyethylene glycol probe from the contact zones between PMN and the underlying fibrin gel. By this measure, IL-8- or LTB4-treated PMN adhered loosely to fibrin, since 10 kD rhodamine-polyethylene glycol permeated into the contact zones between these cells and the underlying fibrin gel. PMN stimulated with FMLP and IL-8, or FMLP and LTB4, exhibited very little migration through fibrin gels, and three times as many of these cells excluded 10 kD rhodamine-polyethylene glycol from their zones of contact with fibrin as PMN stimulated with IL-8 or LTB4 alone. These results show that PMN chemotaxis is regulated by both the nature of the chemoattractant and the composition of the extracellular matrix; they suggest that certain combinations of chemoattractants and matrix proteins may limit leukocyte movements and promote their localization in specific tissues in vivo.

Regulation of leukocyte locomotion by Ca2+.

Neutrophils migrate towards sites of inflammation and infection by chemotaxis. Their motility is dependent on the actin cytoskeleton and on adhesion to extracellular substrates, but how these are regulated in response to stimuli is not clear. This review focuses on the potential role of Ca(2+) as a second messenger in neutrophil motility. Several effects of Ca(2+) and Ca(2+)-binding proteins on the stability and crosslinking of actin polymers have been demonstrated in vitro. Nevertheless, the complex mechanism by which Ca(2+) regulates actin in neutrophils is not fully understood. In addition, intracellular Ca(2+) regulates the intergin-mediated adhesion of neutrophils to extracellular matrix.

Weak bases and ionophores rapidly and reversibly raise the pH of endocytic vesicles in cultured mouse fibroblasts.

It has been shown that endocytic vesicles in BALB/c 3T3 cells have a pH of 5.0 (Tycko and Maxfield, Cell, 28:643-651). In this paper, a method for measuring the effect of various agents, including weak bases and ionophores, on the pH of endocytic vesicles is presented. The method is based on the increase in fluorescein fluorescence with 490-nm excitation as the pH is raised above 5.0. Intensities of cells were measured using a microscope spectrofluorometer after internalization of fluorescein-labeled alpha 2-macroglobulin by receptor-mediated endocytosis. The increase in endocytic vesicle pH was determined from the increase in fluorescence after addition of various concentrations of the test agents. The following agents increased endocytic vesicle pH above 6.0 at the indicated concentrations: monensin (6 microM), FCCP (10 microM), chloroquine (140 microM), ammonia (5 mM), methylamine (10 mM). The ability of many of these agents to raise endocytic vesicle pH may account for many of their effects on receptor-mediated endocytosis. Dansylcadaverine caused no effect on vesicle pH at 1 mM. The observed increases in vesicle pH were rapid (1-2 min) and could be reversed by removal of the perturbant. This reversibility indicates that the vesicles themselves contain a mechanism for acidification. The increase in vesicle pH due to these treatments can be observed visually using an SIT video camera. Using this method, it is shown that endocytic vesicles become acidic at very early times (i.e., within 5-7 min of continuous uptake at 37 degrees C).

Yolk platelets in Xenopus oocytes maintain an acidic internal pH which may be essential for sodium accumulation.

Yolk platelets constitute an embryonic endocytic compartment that stores maternally synthesized nutrients. The pH of Xenopus yolk platelets, measured by photometry on whole oocytes which had endocytosed FITC-vitellogenin, was found to be acidic (around pH 5.6). Experiments on digitonin-permeabilized oocytes showed that acidification was due to the activity of an NEM- and bafilomycin A1-sensitive vacuolar proton-ATPase. Proton pumping required chloride, but was not influenced by potassium or sodium. Passive proton leakage was slow, probably due to the buffer capacity of the yolk, and was dependent on the presence of cytoplasmic monovalent cations. In particular, sodium could drive proton efflux through an amiloride-sensitive Na+/H+ exchanger. 8-Bromo-cyclic-AMP was found to increase acidification, suggesting that pH can be regulated by intracellular second messengers. The moderately acidic pH does not promote degradation of the yolk platelets, which in oocytes are stable for weeks, but it is likely to be required to maintain the integrity of these organelles. Furthermore, the pH gradient created by the proton pump, when coupled with the Na+/H+ exchanger, is probably responsible for the accumulation and storage of sodium into the yolk platelets during oogenesis.

Attachment to fibronectin or vitronectin makes human neutrophil migration sensitive to alterations in cytosolic free calcium concentration.

Transient increases in cytosolic free calcium concentration, [Ca2+]i, appear to be required for the migration of human neutrophils on poly-D-lysine-coated glass in the presence of dilute serum (Marks, P. W., and F. R. Maxfield. 1990. J. Cell Biol. 110:43-52). In contrast, no requirement for [Ca2+]i transients exists when neutrophils migrate on albumin-coated glass in the absence of serum. To determine the mechanism that necessitates [Ca2+]i transients on poly-D-lysine in the presence of serum, migration was examined on substrates consisting of purified adhesive glycoproteins. In the absence of external Ca2+, a treatment which causes the cessation of [Ca2+]i transients, migration on fibronectin (fn) and vitronectin (vn) was significantly inhibited. Migration was also inhibited in Ca2(+)-buffered cells on these substrates, indicating that this effect was the result of an alteration of [Ca2+]i. In the absence of external Ca2+, the inhibition of migration on fn or vn was more pronounced when soluble fn or vn was added to cells migrating on these substrates. This effect of soluble adhesive glycoprotein was specific: in the absence of external Ca2+, soluble fn did not affect the migration of cells on vn, and soluble vn did not affect the migration on fn. No additional inhibition of migration was observed in Ca2(+)-buffered cells with the addition of soluble adhesive glycoprotein. These data indicate that [Ca2+]i transients are involved in continued migration of human neutrophils on fn or vn, proteins which are part of the extracellular matrix that neutrophils encounter in vivo.

Fc and C3bi receptors and the differentiation antigen BH2-Ag are randomly distributed in the plasma membrane of locomoting neutrophils.

Reports from several laboratories suggest that neutrophils arrested during locomotion preferentially bind immune complexes at the front of the cell. Such asymmetry of binding has been interpreted as indicating an active modulation of phagocytic receptors to the anterior of the cell. To investigate this further, we have used digital analysis of fluorescence images to determine the binding patterns of mAbs directed against the Fc receptors, the receptors for the C3bi fragment of C3, and a neutrophil-specific antigen. We found that all three proteins are distributed nearly identically along the length of migrating neutrophils, and their distribution very closely parallels the anterior to posterior distribution of the plasma membrane. The use of mAbs offered an important advantage in that the binding of antireceptor antibodies, unlike the binding of ligands, should be independent of potential changes in the affinity of the receptors. We conclude that the anterior distribution of the phagocytic receptors in the plasma membrane of locomoting neutrophils parallels the overall increase in membrane area at the front of a migrating cell and that specific translocation of phagocytic receptors does not occur.

Cytosolic free calcium increases before and oscillates during frustrated phagocytosis in macrophages.

When macrophages and neutrophils are allowed to settle onto an appropriate surface, they attach and spread in a frustrated attempt to phagocytose the substrate. Spreading is associated with extensive rearrangements of the actin cytoskeleton which resemble those occurring during phagocytosis. We have previously shown that spreading in human neutrophils is preceded by an increase in cytosolic-free calcium concentration [( Ca2+]i) (Kruskal, B. A., S. Shak, and F. R. Maxfield. 1986. Proc. Natl. Acad. Sci. USA. 83:2919-2923). To assess the generality of this signal, we measured [Ca2+]i in single thioglycollate-elicited mouse peritoneal macrophages as they spread on an immune complex-coated surface, using fura-2 microspectrofluorometry. A [Ca2+]i increase always precedes spreading. This increase can involve several (up to 8) [Ca2+]i spikes, with an average peak value of 387 +/- 227 nM (mean +/- SD, n = 92 peaks in 24 cells), before spreading is detected. Neither spreading nor the magnitude of these spikes is significantly altered by removal of extracellular calcium. Many of the spreading macrophages exhibit periodic [Ca2+]i increases before and during spreading. The proportion which does so varies among experiments from 0 to 90%, but it is frequently greater than 40%. The largest number of cells (approximately 25%) exhibited only a single peak. In 13 cells that showed more than 10 peaks, the median period was 29 s (range 19-69 s). The average peak [Ca2+]i was 385 +/- 266 nM (mean +/- SD, n = 208 peaks in 14 cells). The calcium producing these increases is derived from intracellular pools. The oscillations occur with spreading on either opsonized or nonopsonized surfaces. The function of these oscillations is not clear, but the large number of cells which exhibit them suggest that they may be important to macrophage function.

Intracellular fusion of sequentially formed endocytic compartments.

A polyclonal anti-fluorescein antibody (AFA) which quenches fluorescein fluorescence has been used to distinguish between two models of intracellular vesicle traffic. These models address the question of whether sequentially endocytosed probes will mix intracellularly or whether they are carried through the cell in a sequential, isolated manner. Using transferrin (Tf) as a recycling receptor marker, we incubated Chinese hamster ovary (CHO) cells with fluorescein-Tf (F-Tf) which is rapidly endocytosed. After the F-Tf was completely cleared from the surface, AFA was added to the incubation medium and entered endocytic compartments by fluid phase endocytosis. Fusion of a vesicle containing AFA with the compartment containing F-Tf results in binding of AFA to fluorescein and the quenching of fluorescein fluorescence. When AFA was added to the culture medium 2 min after clearance of F-Tf from the surface, time dependent fluorescence quenching occurred. After 20 min, 67% saturation of F-Tf with AFA was observed. When the interval between F-Tf clearance and AFA addition was increased to 5 min only 41% saturation of F-Tf was found. These data indicate that there are some compartments which are accessible for mixing with subsequently endocytosed molecules, but the efficiency of mixing falls off rapidly as the interval between pulses is increased. In CHO cells Tf swiftly segregates to a collection of vesicles or tubules in the para-Golgi region, and at steady state most of the F-Tf is in this compartment. Using digital image analysis to quantify quenching in this region, we have found that F-Tf/AFA mixing is occurring either within this compartment or before transferrin enters it.

Transient increases in cytosolic free calcium appear to be required for the migration of adherent human neutrophils.

Human neutrophils exhibit multiple increases in cytosolic free calcium concentration [( Ca2+]i) spontaneously and in response to the chemoattractant N-formyl-L-methionyl-L-leucyl-L-phenylalanine (Jaconi, M. E. E., R. W. Rivest, W. Schlegel, C. B. Wollheim, D. Pittet, and P. D. Lew. 1988. J. Biol. Chem. 263:10557-10560). The function of these repetitive increases in [Ca2+]i, as well as the role of Ca2+ in human neutrophil migration, remain unresolved. We have used microspectrofluorometry to measure [Ca2+]i in single fura-2-loaded human neutrophils as they moved on poly-D-lysine-coated glass in the presence of serum. To investigate the role of Ca2+ in human neutrophil migration, we examined cells in the presence and absence of extracellular Ca2+, as well as intracellular Ca2(+)-buffered and Ca2(+)-depleted cells. In the presence of extracellular Ca2+, multiple increases and decreases in [Ca2+]i were frequently observed, and at least one such transient increase in [Ca2+]i occurred in every moving cell during chemokinesis, chemotaxis, and phagocytosis. In addition, neutrophils that extended pseudopodia and assumed a polarized morphology after plating onto a surface were always observed to exhibit [Ca2+]i transients even in the absence of chemoattractant. In contrast, a [Ca2+]i transient was observed in only one of the nonpolarized stationary cells that were examined (n = 15). Although some cells exhibited relatively periodic increases and decreases in [Ca2+]i, resembling the regular oscillations that have been observed in some cell types, many others exhibited increases and decreases in [Ca2+]i that varied in their timing, magnitude, and duration. Buffering of [Ca2+]i or removal of extracellular Ca2+ damped out or blocked transient increases in [Ca2+]i and reduced or inhibited the migration of neutrophils. Under these conditions, polarized cells were often observed to make repeated attempts at migration, but they remained anchored at their rear. These data suggest that transient increases in [Ca2+]i may be required for the migration of human neutrophils on poly-D-lysine-coated glass in the presence of serum by allowing them to release from previous sites of attachment.

Fusion accessibility of endocytic compartments along the recycling and lysosomal endocytic pathways in intact cells.

A fluorescence assay developed for the quantitation of intracellular fusion of sequentially formed endocytic compartments (Salzman, N. H., and F. R. Maxfield. 1988 J. Cell Biol. 106:1083-1091) has been used to measure the time course of endosome fusion accessibility along the recycling and degradative endocytic pathways. Transferrin (Tf) was used to label the recycling pathway, and alpha2-macroglobulin (alpha 2 M) was used to label the lysosomal degradative pathway. Along the degradative pathway, accessibility of vesicles containing alpha 2M to fusion with subsequently formed endocytic vesicles decreased with apparent first order kinetics. The t12 for the loss of fusion accessibility was approximately 8 min. The behavior of Tf is more complex. Initially the fusion accessibility of Tf decayed rapidly (t1/2 less than 3 min), but a constant level of fusion accessibility was then observed for 10 min. This suggests that Tf moves through one fusion accessible endosome rapidly and then enters a second fusion accessible compartment on the recycling pathway. At 18 degrees C, fusion of antifluorescein antibodies (AFA) containing vesicles with F-alpha 2M was observed when the interval between additions was 10 min. However, if the interval was increased to 1 h, no fusion with incoming vesicles was observed. These results identify the site of F-alpha 2M accumulation at 18 degrees C as a prelysosomal late endosome that no longer fuses with newly formed endosomes since no delivery to lysosomes is observed at this temperature.

Sorting of membrane components from endosomes and subsequent recycling to the cell surface occurs by a bulk flow process.

A central question in the endocytic process concerns the mechanism for sorting of recycling components (such as transferrin or low density lipoprotein receptors) from lysosomally directed components; membrane-associated molecules including receptors are generally directed towards the recycling pathway while the luminal content of sorting endosomes, consisting of the acid-released ligands, are lysosomally targeted. However, it is not known whether recycling membrane receptors follow bulk membrane flow or if these proteins are actively sorted from lysosomally directed material because of specific protein sequences and/or structural features. Using quantitative fluorescence microscopy we have determined the endocytic route and kinetics of traffic of the bulk carrier, membrane lipids, to address this issue directly. We show that N-[N-(7-nitro-2,1,3-benzoxadiazol-4-yl)-epsilon-aminohexanoyl]- sphingosylphosphorylcholine (C6-NBD-SM) in endocytosed as bulk membrane, and it transits the endocytic system kinetically and morphologically identically to fluorescently labeled transferrin in a CHO cell line. With indistinguishable kinetics, the two labeled markers sort from lysosomally destined molecules in peripherally located sorting endosomes, accumulate in a peri-centriolar recycling compartment, and finally exit the cell. Other fluorescently labeled lipids, C6-NBD-phosphatidylcholine and galactosylceramide also traverse the same pathway. The constitutive nature of sorting of bulk membrane towards the recycling pathway and the lysosomal direction of fluid phase implies a geometric basis of sorting.

Functional expression of the human transferrin receptor cDNA in Chinese hamster ovary cells deficient in endogenous transferrin receptor.

Transferrin (Tf) receptor-variant Chinese hamster ovary cells have been isolated by selection for resistance to two Tf-toxin conjugates. The hybrid toxins contain Tf covalently linked to ricin A chain or a genetically engineered diphtheria toxin fragment. The Tf-receptor-variant (TRV) cells do not have detectable cell-surface Tf receptor; they do not bind fluorescein-Tf or 125I-Tf. TRV cells are at least 100-fold more resistant to the Tf-diphtheria toxin conjugate than are the parent cells. The TRV cells have retained sensitivity to native diphtheria toxin, indicating that the increased resistance to the conjugate is correlated with the loss of Tf binding. The endocytosis of fluorescein-labeled alpha 2-macroglobulin is normal in TRV cells, demonstrating that the defect does not pleiotropically affect endocytosis. Since these cells lack endogenous Tf receptor activity, they are ideally suited for studies of the functional expression of normal or altered Tf receptors introduced into the cells by cDNA transfection. One advantage of this system is that Tf binding and uptake can be used to monitor the behavior of the transfected receptor. A cDNA clone of the human Tf receptor has been transfected into TRV cells. In the stably expressing transfectants, the behavior of the human receptor is very similar to that of the endogenous Chinese hamster ovary cell Tf receptor. Tf binds to cell surface receptors, and is internalized into the para-Golgi region of the cell. Iron is released from Tf, and the apo-Tf and its receptor are recycled back to the cell surface. Thus, the TRV cells can be used to study the behavior of genetically altered Tf receptors in the absence of interfering effects from endogenous receptors.

Delivery of ligands from sorting endosomes to late endosomes occurs by maturation of sorting endosomes.

After endocytosis, lysosomally targeted ligands pass through a series of endosomal compartments. The endocytic apparatus that accomplishes this passage may be considered to take one of two forms: (a) a system in which lysosomally targeted ligands pass through preexisting, long-lived early sorting endosomes and are then selectively transported to long-lived late endosomes in carrier vesicles, or (b) a system in which lysosomally targeted ligands are delivered to early sorting endosomes which themselves mature into late endosomes. We have previously shown that sorting endosomes in CHO cells fuse with newly formed endocytic vesicles (Dunn, K. W., T. E. McGraw, and F. R. Maxfield. 1989. J. Cell Biol. 109:3303-3314) and that previously endocytosed ligands lose their accessibility to fusion with a half-time of approximately 8 min (Salzman, N. H., and F. R. Maxfield. 1989. J. Cell Biol. 109:2097-2104). Here we have studied the properties of individual endosomes by digital image analysis to distinguish between the two mechanisms for entry of ligands into late endosomes. We incubated TRVb-1 cells (derived from CHO cells) with diO-LDL followed, after a variable chase, by diI-LDL, and measured the diO content of diI-containing endosomes. As the chase period was lengthened, an increasing percentage of the endosomes containing diO-LDL from the initial incubation had no detectable diI-LDL from the second incubation, but those endosomes that contained both probes showed no decrease in the amount of diO-LDL per endosomes. These results indicate that (a) a pulse of fluorescent LDL is retained by individual sorting endosomes, and (b) with time sorting endosomes lose the ability to fuse with primary endocytic vesicles. These data are inconsistent with a preexisting compartment model which predicts that the concentration of ligand in sorting endosomes will decline during a chase interval, but that the ability of the stable sorting endosome to receive newly endocytosed ligands will remain high. These data are consistent with a maturation mechanism in which the sorting endosome retains and accumulates lysosomally directed ligands until it loses its ability to fuse with newly formed endocytic vesicles and matures into a late endosome. We also find that, as expected according to the maturation model, new sorting endosomes are increasingly labeled during the chase period indicating that new sorting endosomes are continuously formed to replace those that have matured into late endosomes.(ABSTRACT TRUNCATED AT 400 WORDS)

Kinetics of endosome acidification in mutant and wild-type Chinese hamster ovary cells.

Acidification of endocytic compartments is necessary for the proper sorting and processing of many ligands and their receptors. Robbins and co-workers have obtained Chinese hamster ovary (CHO) cell mutants that are pleiotropically defective in endocytosis and deficient in ATP-dependent acidification of endosomes isolated by density centrifugation (Robbins, A. R., S. S. Peng, and J. L. Marshall. 1983. J. Cell Biol. 96:1064-1071; Robbins, A. R., C. Oliver, J. L. Bateman, S. S. Krag, C. J. Galloway, and I. Mellman. 1984. J. Cell Biol. 99:1296-1308). In this and the following paper (Yamashiro, D. J., and F. R. Maxfield. 1987. J. Cell Biol. 105:2723-2733) we describe detailed studies of endosome acidification in the mutant and wild-type CHO cells. Here we describe a new microspectrofluorometry method based on changes in fluorescein fluorescence when all cellular compartments are equilibrated to the same pH value. Using this method we measured the pH of endocytic compartments during the first minutes of endocytosis. We found in wild-type CHO cells that after 3 min, fluorescein-labeled dextran (F-Dex) was in endosomes having an average pH of 6.3. By 10 min, both F-Dex and fluorescein-labeled alpha 2-macroglobulin (F-alpha 2M) had reached acidic endosomes having an average pH of 6.0 or below. In contrast, endosome acidification in the CHO mutants DTG 1-5-4 and DTF 1-5-1 was markedly slowed. The average endosomal pH after 5 min was 6.7 in both mutant cell lines. At least 15 min was required for F-Dex and F-alpha 2M to reach an average pH of 6.0 in DTG 1-5-4. Acidification of early endocytic compartments is defective in the CHO mutants DTG 1-5-4 and DTF 1-5-1, but pH regulation of later compartments on both the recycling pathway and lysosomal pathway is nearly normal. The properties of the mutant cells suggest that proper functioning of pH regulatory mechanisms in early endocytic compartments is critical for many pH-mediated processes of endocytosis.

Acidification of morphologically distinct endosomes in mutant and wild-type Chinese hamster ovary cells.

In the preceding paper (Yamashiro, D. J., and F. R. Maxfield. 1987. J. Cell Biol. 105:2713-2721), we have shown that there is rapid acidification of endosomal compartments to pH 6.3 by 3 min in wild-type Chinese hamster ovary (CHO) cells. In contrast, early acidification of endosomes is markedly reduced in the CHO mutants, DTF 1-5-4 and DTF 1-5-1. Since these CHO mutants are pleiotropically defective in endocytosis (Robbins, A. R., S. S. Peng, and J. L. Marshall. 1983. J. Cell Biol. 96:1064-1071; Robbins, A. R., C. Oliver, J. L. Bateman, S. S. Krag, C. J. Galloway, and I. Mellman. 1984. J. Cell Biol. 99:1296-1308), our results are consistent with a requirement for proper acidification of early endocytic compartments in many pH-regulated endocytic processes. In this paper, by measuring the pH of morphologically distinct endosomes using fluorescence microscopy and digital image analysis, we have determined in which of the endocytic compartments the defective acidification occurs. We found that the acidification of both the para-Golgi recycling endosomes and lysosomes was normal in the CHO mutants DTG 1-5-4 and DTF 1-5-1. The mean pH of large endosomes containing either fluorescein-labeled alpha 2-macroglobulin or fluorescein-isothiocyanate dextran was only slightly less acidic in the mutant cells than in wild-type cells. However, when we examined the pH of individual large (150-250 nm) endosomes, we found that there was an increased number of endosomes with a pH greater than 6.5 in the CHO mutants when compared with wild-type cells. Heterogeneity in the acidification of large endosomes was also seen in DTF 1-5-1 by a combined null point pH method and digital image analysis technique. In addition, both CHO mutants showed a marked decrease in the acidification of the earliest endosomal compartment, a diffusely fluorescent compartment comprised of small vesicles and tubules. We suggest that the defect in endosome acidification is most pronounced in the early, small vesicular, and tubular endosomes and that this defect partially carries over to the large endosomes that are involved in the sorting and processing of ligands. The proper step-wise acidification of the different endosomes along the endocytic pathway may have an important role in the regulation of endocytic processes.

Evidence for nonvectorial, retrograde transferrin trafficking in the early endosomes of HEp2 cells.

We have previously characterized the trafficking of transferrin (Tf) through HEp2 human carcinoma cells (Ghosh, R. N., D. L. Gelman, and F. R. Maxfield, 1994. J. Cell Sci. 107:2177-2189). Early endosomes in these cells are comprised of both sorting endosomes and recycling compartments, which are distinct separate compartments. Endocytosed Tf initially appears in punctate sorting endosomes that also contain recently endocytosed LDL. After short loading pulses, Tf rapidly sorts from LDL with first-order kinetics (t1/2 approximately 2.5 min), and it enters the recycling compartment before leaving the cell (t1/2 approximately 7 min). Here, we report a second, slower rate for Tf to leave sorting endosomes after HEp2 cells were labeled to steady state with fluorescein Tf instead of the brief pulse used previously. We determined this rate using digital image analysis to measure the Tf content of sorting endosomes that also contained LDL. With an 11-min chase, the Tf in sorting endosomes was 24% of steady-state value. This was in excess of the amount expected (5% of steady state) from the rate of Tf exit after short filling pulses. The excess could not be accounted for by reinternalization of recycled cell surface Tf, implying that either some Tf was retained in sorting endosomes, or that Tf was delivered back to the sorting endosomes from the recycling compartment. The former is unlikely since nearly all sorting endosomes contain detectable Tf after an 11-min chase, even though more than one third of the sorting endosomes were formed during the chase time. Furthermore, while observing living cells by confocal microscopy, we saw vesicle movements that appeared to be fluorescent Tf returning from recycling compartments to sorting endosomes. The slow rate of exit after steady-state labeling was similar to the Tf exit rate from the cell, suggesting an equilibration of Tf throughout the early endosomal system by this retrograde pathway. This retrograde traffic may be important for delivering molecules from the recycling compartment, which is a long-lived organelle, to sorting endosomes, which are transient.

Endocytic sorting of lipid analogues differing solely in the chemistry of their hydrophobic tails.

To understand the mechanisms for endocytic sorting of lipids, we investigated the trafficking of three lipid-mimetic dialkylindocarbocyanine (DiI) derivatives, DiIC16(3) (1,1'-dihexadecyl-3,3,3',3'-tetramethylindocarbocyanine perchlorate), DiIC12(3) (1,1'- didodecyl-3,3,3',3'-tetramethylindocarbocyanine perchlorate), and FAST DiI (1,1'-dilinoleyl-3,3,3', 3'-tetramethylindocarbocyanine perchlorate), in CHO cells by quantitative fluorescence microscopy. All three DiIs have the same head group, but differ in their alkyl tail length or unsaturation; these differences are expected to affect their distribution in membrane domains of varying fluidity or curvature. All three DiIs initially enter sorting endosomes containing endocytosed transferrin. DiIC16(3), with two long 16-carbon saturated tails is then delivered to late endosomes, whereas FAST DiI, with two cis double bonds in each tail, and DiIC12(3), with saturated but shorter (12-carbon) tails, are mainly found in the endocytic recycling compartment. We also find that DiOC16(3) (3,3'- dihexadecyloxacarbocyanine perchlorate) and FAST DiO (3, 3'-dilinoleyloxacarbocyanine perchlorate) behave similarly to their DiI counterparts. Furthermore, whereas a phosphatidylcholine analogue with a BODIPY (4,4-difluoro-4-bora-3a,4a-diaza-s-indacene) fluorophore attached at the end of a 5-carbon acyl chain is delivered efficiently to the endocytic recycling compartment, a significant fraction of another derivative with BODIPY attached to a 12-carbon acyl chain entered late endosomes. Our results thus suggest that endocytic organelles can sort membrane components efficiently based on their preference for association with domains of varying characteristics.

Rapid acidification of endocytic vesicles containing asialoglycoprotein in cells of a human hepatoma line.

Acidification of endocytic vesicles has been implicated as a necessary step in various processes including receptor recycling, virus penetration, and the entry of diphtheria toxin into cells. However, there have been few accurate pH measurements in morphologically and biochemically defined endocytic compartments. In this paper, we show that prelysosomal endocytic vesicles in HepG2 human hepatoma cells have an internal pH of approximately 5.4. (We previously reported that similar vesicles in mouse fibroblasts have a pH of 5.0.) The pH values were obtained from the fluorescence excitation profile after internalization of fluorescein labeled asialo-orosomucoid (ASOR). To make fluorescence measurements against the high autofluorescence background, we developed digital image analysis methods for estimating the pH within individual endocytic vesicles or lysosomes. Ultrastructural localization with colloidal gold ASOR demonstrated that the pH measurements were made when ligand was in tubulovesicular structures lacking acid phosphatase activity. Biochemical studies with 125I-ASOR demonstrated that acidification precedes degradation by more than 30 min at 37 degrees C. At 23 degrees C ligand degradation ceases almost entirely, but endocytic vesicle acidification and receptor recycling continue. These results demonstrate that acidification of endocytic vesicles, which causes ligand dissociation, occurs without fusion of endocytic vesicles with lysosomes. Methylamine and monensin raise the pH of endocytic vesicles and cause a ligand-independent loss of receptors. The effects on endocytic vesicle pH are rapidly reversible upon removal of the perturbant, but the effects on cell surface receptors are slowly reversible with methylamine and essentially irreversible with monensin. This suggests that monensin can block receptor recycling at a highly sensitive step beyond the acidification of endocytic vesicles. Taken together with other direct and indirect estimates of endocytic vesicle pH, these studies indicate that endocytic vesicles in many cell types rapidly acidify below pH 5.5, a pH sufficiently acidic to allow receptor-ligand dissociation and the penetration of some toxin chains and enveloped virus nucleocapsids into the cytoplasm.