Lectin-induced deciduoma formation in the pseudopregnant rat

Decidual cell induction in the pseudopregnant rat was examined in this study using the lectin concanavalin A (ConA). The histochemical binding of the lectin to the uterine cell surface at the time of deciduomatic induction was also studied. ConA was found to induce significant deciduomata (decidual-like tissue) in the uterine horn when injected intraluminally on day 5 of pseudopregnancy (PSP). ConA-induced deciduomata appeared as a series of discrete nodules in the uterine horn, reminiscent of the anatomical appearance of normal embryo implantation sites. Deciduoma induction by ConA was greatly reduced by pre-absorption of the lectin with its competitive sugar. Lectin histochemistry revealed binding of ConA to the cell surface on day 5 of PSP. Pre-absorption of the lectin with its competitive sugar also significantly reduced surface binding of the lectin, and this finding may be correlated with the greatly reduced ability of the pre-absorbed lectin to induce deciduomata. Possible mechanisms for the induction of deciduomata by lectins are considered.     

Ultrastructural visualization of selective peanut agglutinin binding sites in rat osteoclasts

We investigated the distribution of concanavalin A (ConA)-reactive alpha-D-mannosyl and alpha-D-glucosyl groups and peanut agglutinin (PNA)-reactive beta-D-galactose-(1----3)-N-acetyl-D-galactosamine residues on the surface of osteoclasts with pre-embedment ultrastructural lectin cytochemistry after aldehyde fixation of the metaphyses of the rat tibiae. By routine morphology, the plasma membrane of the ruffled border of the osteoclast was distinguished from the rest of the cell membrane, with the exception of the membrane of coated pits, by its characteristic thick coat at its cytoplasmic surface. Cytochemistry, using ConA in combination with horseradish peroxidase (ConA-HRP) and PNA conjugated to HRP, showed that binding of ConA was distributed over the entire cell surface of osteoclasts. In contrast, intense binding of PNA was limited to the membranes of the ruffled border and coated pits, whereas the remainder of the cell membrane stained weakly or not at all. These results demonstrate that preferential PNA binding sites of the cell surface correspond to coated membranes associated with osteoclastic endocytosis.     

Lectin staining of cultured CNS microglia.

Carbohydrate binding proteins, known as lectins, bind to specific sugar groups on most membranes. We used fluorescent and light microscopy to study the interaction of various lectins with the membranes of microglia cultured from neonatal rat or fetal mouse cerebral cortices. Microglia stained intensely with GS-1, RCA, WGA, and ConA and slightly with DBA, UEA, BPA, and SBA. No staining was seen with GS-2, MPA, or PNA. Staining was specific for microglia in the mixed glial cultures and was dose dependent. In addition, microglial lectin binding could be reduced or blocked by competitive inhibition using specific sugars. Treatment of the microglia with agents such as dimethylsulfoxide (DMSO), interleukin-1 (IL-1), interferon (IFN), or lipopolysaccharide (LPS) did not eliminate lectin staining, although the degree of staining was altered. Positive staining of the microglia was also associated with a functional change for at least one lectin, i.e., ConA. Superoxide anion production by microglia was increased in the presence of ConA. Overall, binding of the lectins GS-1, RCA, WGA, and ConA can be used as an identifying tool for microglia in glial cultures, but intensity of staining varies depending on their functional state.     

Studies on fertilization in the ascidians : II. Lectin binding to the gametes of ciona intestinalis

In the present work we have compared the binding of fluorescein-conjugated lectins (concanavalin A (ConA), wheat germ agglutinin (WGA), fucose binding protein (FBP) and soybean agglutinin (SBA)) to the sperm surface and to the egg and its envelopes of Ciona intestinalis. Only WGA is bound to the follicle cells: yet this lectin has no binding sites on the sperm surface. Both ConA and FBP are bound by the chorion, the oolemma and the sperm surface. However, while ConA reacts only with the sperm head, FBP is bound both to the head and to the flagellum. Experiments on the effect of ConA and FBP on the fertilization reaction have been carried out. The role of the lectin-binding sites that are shared by the surfaces of both gametes is discussed in connection with the nature of the sperm-binding sites.     

Lectin histochemistry of epidermal glandular cells in the earthworm Lumbricus terrestris (Annelida Oligochaeta)

Carbohydrate residues were localized in the glandular cells of the epidermis of Lumbricus terrestris by lectin histochemistry. The following biotinylated lectins were used: ConA, PNA, WGA, UEA-I. Each lectin has a specific binding pattern in the epidermal glandular cells. The ConA binding is evident in the orthochromatic mucous cells; PNA in the metachromatic mucous cells; WGA in the neuroendocrine-like cells; UEA-I in the cuticle. The epidermal glandular cells possess specific sites for the different lectins in relation to their functional characteristics. Therefore, these sugar residues indicate different behaviours of the cells in epidermal functions related to ion transport, receptor-secretory processes and defence.     

Two major antigens of heart sarcolemma are Ca2(+)-binding glycoproteins that copurify with the dihydropyridine receptor.

Ca2+ binding has been studied in isolated heart sarcolemmal membranes using the 45Ca overlay technique. 45Ca bound to two sarcolemmal polypeptides of 125 kDa and 97 kDa in preparations from dog, rabbit, cow and pig. During fractionation on DEAE ion-exchange and wheat-germ lectin affinity columns, the two Ca2(+)-binding polypeptides copurified with the dihydropyridine receptor associated with the voltage gated Ca2+ channel. These polypeptides were the major proteins in the isolated fraction as judged by silver staining in SDS-PAGE. Antisera raised against purified dog heart, sarcolemma indicated that the 125 and 97 kDa polypeptides were highly antigenic components of this membrane. The antisera cross-reacted with similar polypeptides in cardiac sarcolemmal preparations from rabbit, cow and pig, but not sarcoplasmic reticulum membranes. Purified antibodies against the 125 kDa polypeptide did not cross-react with the 97 kDa polypeptide, while antibodies against the 97 kDa polypeptide did not cross-react with the 125 kDa polypeptide. Both the 125 kDa and 97 kDa polypeptides bound wheat-germ lectin, suggesting both were glycoproteins. It is unlikely that these Ca2+ binding glycoproteins represent subunits of the dihydropyridine receptor-Ca2+ channel in this membrane.     

Fertilization-induced changes in concanavalin A binding to mouse eggs

The binding of concanavalin A (ConA) to zona-free unfertilized and fertilized mouse eggs has been investigated using tritiated ConA. At low lectin concentrations (1–5 μg ml^?1) the fertilized egg shows a higher affinity for [³H]ConA than does the unfertilized egg. In saturation conditions, however, unfertilized and fertilized eggs show the same binding capacity (1.55 × 10⁸ ConA molecules/egg). The results indicate that ConA-binding sites change qualitatively following fertilization; possible connections between this change and other fertilization-induced changes in the egg surface are discussed.     

Effect of ConA—receptor interaction on the structure of cell membrane

Recently,the effect of ligand receptor interaction on the membrane structure of liposomes has been studied extensively,However,little is known about how it exists on biological membranes,In this paper,the effect of Concanavalin A(ConA) receptorinteratcion on the structure of cell membranes was studied by Circular DIchrosim(CD) and 31P Nuclear Magnetic Resonance(NMR).CD results of both the purified macrophage membranes and human erythrocyte hgosts(EG) showed that the conformation of membrane proteins changed after ConA binding.For further research,31P-NMR was used to detect the orgainzation of phosp[holipid molecules on macrophage membranes.After ConA binding,the tendercy to form non bilayer structure increased with the amount of ConA.The changes of 31P-NMR spectra of living macrophages might be partly due to the above stated reason too.In addition,ConA-receptor interaction also induced similar results of 31P-NMR spectra in EG.In contrast,wheat germ agglutinin (WGA),another kind of lectin,rarely showed the same influence.     

Cell surface interactions with concanavalin A : Location of bound radiolabeled lectin

We have developed two improved methods: (1) a procedure for coupling ¹²⁵Iodine to ConcanavalinA (ConA) that yields intensely labeled and fully active lectin; (2) a procedure that allows studies of lectin binding to be carried out with a minimum of non-specific binding to reaction vessels. We found that BALB/c 3T3 cells, SV3T3 cells, and human red blood cells have 1.3 × 10⁷, 1.5 × 10⁷, and 2.2 × 10⁶ ConA binding sites/cell. More than 99.5% of the radioactivity in the samples counted was associated with the cells; background radioactivity, in the absence of cells, was negligible. We also found that although α-methylmannopyranoside (α-MM) prevented almost all of the ConA from binding to cells, when ConA had first been allowed to bind, α-MM removed only 60 to 80% of the bound ConA. In addition, even after the removal of a portion of bound lectin by α-MM, most, if not all, of the remaining cell-associated ConA was coupled to the plasma membrane.     

Rearrangement of concanavalin A receptor sites on cells tagged with dinitrofluorobenzene: Evidence for the chemical induction of a change usually associated with malignant transformation

Normal human peripheral blood granulocytes which are tagged with 1-fluoro-2,4-dinitrobenzene (DNFB) are agglutinated by concanavalin A (ConA) in a way which resembles the pattern of reactivity displayed by leukemic cells. The present study further defines this reaction. The binding of ConA to untagged and DNP-tagged granulocytes, treated with DNFB at a ratio of 10¹¹ molecules/cell, was quantified by isotopic dilution experiments employing [³H]ConA. Similar amounts of the lectin were bound to untagged and DNP-tagged cells following incubation for 5 min at 4 °C or 30 min at 24 °C: 1.1 × 10⁵ molecules/cell, 4.6 × 10²/μ² of surface area, and 1.6 × 10³/μg of protein. The binding of [³H]ConA to both untagged and DNP-tagged cells was inhibited to the same degree by α-methylglucopyranoside (α-MG). Fixation with either glutaraldehyde or formaldehyde, which immobilizes ConA receptor sites, completely inhibited the agglutination of both untagged and DNP-tagged cells although lectin binding was unchanged. This suggests that the inhibition of agglutination was not due to the blocking of ConA-binding sites by aldehyde groups but rather to the immobilization of lectin receptors. We conclude that dinitrophenylation of normal granulocytes facilitates the rearrangement of lectin receptors in a way which resembles the ConA-induced clustering of sites which have been observed with malignant and transformed cells.     

Modulation of receptors and cytotoxic response of tumor necrosis factor-alpha by various lectins

The effects of various lectins on the interaction of the human cervical carcinoma cell line ME-180 with recombinant human tumor necrosis factor-alpha (rTNF-alpha) was investigated. rTNF-alpha is known to have cytotoxic effects on this tumor cell line and has been reported to interact with these cells through a single class of specific high affinity receptors (Kd = 0.45 nM; approximately 1790 binding sites/cell). Exposure of cells to concanavalin A (ConA) causes an approximately 2-fold increase in rTNF-alpha receptors without any significant change in their affinity constant (Kd = 0.36 nM; approximately 3662 binding sites/cell). This increase in receptor number is dependent on temperature, the time of exposure and dose of ConA, and does not require the synthesis of new proteins. In spite of an increased binding of rTNF-alpha to cells, the cell killing induced by rTNF-alpha is totally blocked by ConA. Cells are also protected by this lectin from the synergistic cytotoxic effects of rTNF-alpha and recombinant human interferon-gamma. Furthermore, it was also found that ConA decreases the rate of internalization and dramatically inhibits the release and degradation of rTNF-alpha by the cells. These results, overall, demonstrate that ConA increases total number of binding sites for rTNF-alpha but blocks the transduction of the signal for the cytotoxic response.     

Lectin binding sites in Paramecium tetraurelia cells. I. Labeling analysis predominantly of secretory components

Though all three lectins tested (ConA, RCA II, WGA) bound to the entire cell membrane, none bound selectively to the docking site of secretory organelles (trichocysts); the same results were achieved with FITC-conjugates, or, on the EM level, with peroxidase- or gold-labeling. Only WGA triggered the release of trichocysts and none of the lectins tested inhibited AED-induced synchronous exocytosis. When exocytosis was triggered synchronously in the presence of any of these three lectins (FITC-conjugates), the resulting ghosts trapped the FITC-lectins and the cell surface was immediately afterwards studded with regularly spaced dots (corresponding to the ghosts located on the regularly spaced exocytosis sites). These disappeared within about 10 min from the cell surface (thus reflecting ghost internalization with a half life of 3 min) and fluorescent label was then found in approximately 6-10 vacuoles, which are several microns in diameter, stain for acid phosphatase and, on the EM level, contain numerous membrane fragments (otherwise not found in this form in digesting vacuoles). We conclude that synchronous massive exocytosis involves lysosomal breakdown rather than reutilization of internalized trichocyst membranes and that these contain lectin binding sites (given the fact free fluorescent probes did not efficiently stain ghosts). Trichocyst contents were analyzed for their lectin binding capacity in situ and on polyacrylamide gels. RCA II yielded intense staining (particularly of "tips"), while ConA (fluorescence concentrated over "bodies") and WGA yielded less staining of trichocyst contents on the light and electron microscopic level. Only ConA- and WGA-staining was inhibitable by an excess of specific sugars, while RCA II binding was not. ConA binding was also confirmed on polyacrylamide gels which also allowed us to assess the rather low degree of glycosylation (approximately 1% by comparison with known glycoprotein standards) of the main trichocyst proteins contained in their expandable "matrix". Since RCA II binding could be due to its own glycosylation residues we looked for an endogenous lectin. The conjecture was substantiated by the binding of FITC-lactose-albumin (inhibitable by a mixture of glucose-galactose). This preliminary new finding may be important for the elucidation of trichocyst function.     

Binding of the insecticidal lectin Concanavalin A in pea aphid, Acyrthosiphon pisum (Harris) and induced effects on the structure of midgut epithelial cells

Concanavalin A (lectin from Canavalia ensiformis L., ConA) has previously been shown to act as a feeding inhibitor for Acyrthosiphon pisum, the pea aphid. In the present study a range of histochemical and biochemical techniques were used to elucidate the target tissues and binding sites of the lectin in the aphid. Diet uptake was evaluated using a radioactive tracer (¹⁴C-methylated inulin) and demonstrated that adults were capable of ingesting high quantities of the toxin (approx. 1 μg over a 48 h period). Electophoretic analysis and enzyme-linked immuno-sorbent assay of honeydew samples confirmed these results and further demonstrated that only small levels of ConA were excreted. Histofluorescence and immunolocalisation studies on nymphs revealed that the stomach was the primary target for ConA. At concentrations up to 400 μg ml⁻¹, lectin binding only occurred in the stomach region, however, at high concentrations (800 μg ml⁻¹) the whole digestive tract was stained, although there was no evidence of binding in either the oesophagus or rectum. In addition to binding, there was evidence to suggest that ConA was also causing systemic effects in that the lectin appeared to cross the intestinal epithelial barrier. Immunohistochemical and electron microscopy studies revealed that ConA induced severe cellular swelling of the epithelial cells, accompanied by hypersecretion and a progressive detachment of the apical membrane; however, the striated border itself did not appear to be directly affected. Furthermore, there was no lysis of the epithelium, nor loss of integrity of the epithelial cells themselves. Our results suggest that ConA interacts with glycosylated receptors at the surface of the stomach epithetial cells, interfering with normal metabolism and cell function, resulting in a rapid feedback response on feeding behaviour. Whilst our results provide a much greater understanding regarding the modes of action of ConA in insects, they suggest that different lectins, including other mannose binding lectins, have different modes of action at the cellular levels, and thus generalizations should be treated with caution.     

Specific modifications of hepatoma cell-surface glycoproteins with enzymes : Effects on in vitro growth as investigated by the use of lectins

The effects of enzymic treatment on the interactions between Zajdela's tumor cells and various lectins. Concanavalin A (ConA); Wheat Germ Agglutinin (WGA); Robinia lectin; have been studied. (1) The number of lectin-binding sites and the affinity constants were investigated. (2) The effects of the lectins on cell growth and [3H]thymidine incorporation were studied on untreated and enzyme-treated cells. It was observed that treatment of tumor cells with neuraminidase resulted in a change in the binding characteristics of each lectin. However, additional treatment of the cells with galactose oxidase had no further effect on lectin binding. ConA and Robinia lectin induced a decrease of the untreated tumor cell growth and a stimulation of the [3H]thymidine incorporation. This paradoxal result may be explained as a consequence of the stimulation of the [3H]thymidine uptake observed in the presence of lectins. The enzymatic treatments themselves did not change the cell growth although they did induce a change in the effect of ConA and Robinia lectin on cell growth and [3H]thymidine incorporation. As a result of neuraminidase treatment, the effects of ConA were totally suppressed but those of Robinia lectin only partially. Although WGA interacted with untreated and enzyme-treated cell surfaces, it had no effect on tumor cell growth nor [3H]thymidine incorporation. The results are discussed in terms of lectin transport.     

A comparative study of four cytochemical detection methods of concanavalin A binding sites on the cell membrane

Four different electron cytochemical methods to detect concanavalin A (ConA) binding sites on the plasma membrane of mouse fibroblasts were compared in this study. The ConA binding sites were made visible either by adding ConA, followed by horseradish peroxidase (HRP) or hemocyanin (HC), or by marking the sites with complexes of ConA with ferritin (Fer) or with micro-peroxidase (MP). HC and Fer are directly visible in the electron microscope; HRP and MP are detected by their electron-dense reaction product with diaminobenzidin and H₂O₂. Differences in sensitivity of the ConA binding sites for the different markers were found and resulted in a tentative interpretation of the labelling reactions. All experiments suggested that normal and transformed murine fibroblasts both have plasma membranes in which the binding sites can move equally well and can be induced to form clusters. These results are discussed in relation with the hypothesis that differences in clustering of ConA sites between normal and transformed cells are responsible for differences in the agglutinability by ConA of these cells.     

Lectin binding sites as markers of neonatal porcine uterine development.

To identify lectin binding sites and to determine if lectin binding patterns change with age in developing neonatal porcine uterine tissues, gilts (n = 3/day) were hysterectomized on Day 0 (birth), 7, 14, 28, 42, or 56. Lectin binding was visualized in Bouin's-fixed uterine tissues with seven biotinylated lectins (ConA, DBA, PNA, RCA-I, SBA, UEA-I, and WGA) and avidin-peroxidase staining procedures. Lectin specificities were demonstrated by pre-incubating lectins with appropriate inhibitory sugars (0.2 M). Staining intensity was evaluated visually (absent, weak, moderate, or strong) for three endometrial tissues; luminal epithelium, glandular epithelium, and stroma. Staining intensities for DBA, PNA, SBA, and WGA were not affected by neonatal age. Staining with these lectins was greater in uterine epithelium (moderate or strong) than in stroma (weak). In contrast, binding patterns for ConA, UEA-I, and RCA-I were affected by neonatal age. Strong epithelial staining associated with ConA binding was observed on all days, whereas stromal ConA staining decreased in intensity from moderate to weak after Day 14. Epithelial staining with UEA-I increased from moderate to strong after Day 28, whereas stromal UEA-I staining decreased from moderate to weak after day 28. Staining with RCA-I was homogeneous for luminal epithelium and stroma but variegated for glandular epithelium on and after Day 7. These observations indicate that a variety of lectin binding sites are present in developing neonatal porcine endometrial tissues and that developmentally related alterations in the distribution and/or orientation of glycoconjugates containing alpha-D-mannose, beta-D-galactose, beta-D-acetyl-N-galactosamine, and alpha-L-fucose residues occur between birth and Day 56 as these tissues mature.     

Thermodynamic binding parameters of individual epitopes of multivalent carbohydrates to concanavalin a as determined by "reverse" isothermal titration microcalorimetry.

The preceding paper [Dam, T. K., Roy, R., Pagé, D., and Brewer, C. F. (2002) Biochemistry 41, 1351-1358] demonstrated that Hill plots of isothermal titration microcalorimetry (ITC) data for the binding of di-, tri-, and tetravalent carbohydrate analogues possessing terminal 3,6-di-O-(alpha-D-mannopyranosyl)-alpha-D-mannopyranoside residues to the lectin concanavalin A (ConA) show increasing negative cooperativity upon binding of the analogues to the lectin. The present study demonstrates "reverse" ITC experiments in which the lectin is titrated into solutions of di- and trivalent analogues. The results provide direct determinations of the thermodynamics of binding of ConA to the individual epitopes of the two multivalent analogues. The n values (number of binding sites per carbohydrate molecule) derived from reverse ITC demonstrate two functional binding epitopes on both the di- and trivalent analogues, confirming previous "normal" ITC results with the two carbohydrates [Dam, T. K., Roy, R., Das, S. K., Oscarson, S., and Brewer, C. F. (2000) J. Biol. Chem. 275, 14223-14230]. The reverse ITC measurements show an 18-fold greater microscopic affinity constant of ConA for the first epitope of the divalent analogue versus its second epitope and a 53-fold greater microscopic affinity constant of ConA binding to the first epitope of the trivalent analogue versus its second epitope. The data also demonstrate that the microscopic enthalpies of binding of the two epitopes on the di- and trivalent analogues are essentially the same and that differences in the microscopic K(a) values of the epitopes are due to their different microscopic entropies of binding values. These findings are consistent with the increasing negative Hill coefficients of these analogues binding to ConA in the previous paper.     

Differences in lectin binding patterns of normal human endometrium between proliferative and secretory phases

Lectin binding patterns in normal human endometrium were examined by light and electron microscopy using seven different lectins (ConA, WGA, RCA, PNA, UEA-1, DBA, and SBA). For light microscopic observations, criteria based on the incidence and intensity of cells positive for the lectin staining were adopted to evaluate the different staining patterns of the proliferative and secretory endometria obtained by the avidin-biotin-peroxidase complex (ABC) technique. At the light microscopic level, ConA, WGA, and RCA stained endometrial glandular cells in both phases. The number of PNA-positive cells with the binding sites entirely limited to the apical surface tended to be reduced slightly in the secretory phase. UEA-1 weakly stained the apical surface of glandular cells in the proliferative phase but not in the secretory phase. Among the lectins used in this study, DBA and SBA displayed remarkable changes between the phases. That is, in the proliferative phase they produced only a faint or slight positive stain at the apical surface, but the incidence and intensity of DBA- and the SBA-positive glandular cells increased in the secretory phase. By electron microscopy, the reaction product of ConA was observed in the plasma membrane, endoplasmic reticulum, nuclear envelope, and the Golgi apparatus, and the binding sites of RCA and DBA were observed in the plasma and Golgi membranes. Between both phases, the reactivity of ConA and RCA showed almost no change. However, the secretory endometrial cells containing the DBA-positive Golgi apparatus were markedly increased in number compared with the proliferative ones bearing the lectin-positive organelles.(ABSTRACT TRUNCATED AT 250 WORDS)     

Concanavalin A interactions with asparagine-linked glycopeptides. Bivalency of high mannose and bisected hybrid type glycopeptides

We have previously reported that concanavalin A (ConA) is precipitated by a high mannose type glycopeptide (Brewer, C. F. (1979) Biochem. Biophys. Res. Commun. 90, 117-122; Bhattacharyya, L., and Brewer, C. F. (1986) Biochem. Biophys. Res. Commun. 137, 670-674). In the present study, we have investigated the ability of a series of high mannose and bisected hybrid type glycopeptides to bind and precipitate the lectin. The modes of binding of the glycopeptides were studied by nuclear magnetic relaxation dispersion (NMRD) techniques, and their affinities were determined by hemagglutination inhibition measurements. The stoichiometries of the precipitation reactions were investigated by quantitative precipitation analysis. The equivalence zones (regions of maximum precipitation) of the precipitin curves indicate that certain high mannose and bisected hybrid type glycopeptides are bivalent for lectin binding. From the NMRD and precipitation data, we have identified two protein binding sites on each glycopeptide: one site on the alpha(1-6) arm of the core beta-mannose residue involving a trimannosyl moiety which binds with high affinity (primary site); and the other site on the alpha(1-3) arm of the core beta-mannose residue involving an alpha-mannose residue(s), which binds with lower affinity (secondary site). These two types of sites bind to ConA by different mechanisms. Certain bisected hybrid type glycopeptides were found to possess only the primary ConA binding sites, but not the secondary sites, and hence were able to bind but not precipitate the lectin. Other related glycopeptides have only the secondary type sites and thus exhibit low affinity and are unable to precipitate the protein. The results are related to the possible structure-function properties of cell-surface glycopeptides.     

Characterization of the isolation membranes and the limiting membranes of autophagosomes in rat hepatocytes by lectin cytochemistry

The isolation membranes and the limiting membranes of autophagosomes in rat hepatocytes were characterized by lectin cytochemistry using concanavalin A (ConA), Ricinus communis agglutinin 120 (RCA-120), and wheat germ agglutinin (WGA). We found that RCA-120, ConA, and WGA bind to these membranes. The distribution of the lectins on the isolation membranes was heterogeneous, mainly found on the rims, which we referred to as the peripheral dilated portion. When the rims fused and thus formed autophagosomes the apparent sites of fusion were strongly labeled by the lectins. After autophagosomes were transformed to autolysosomes by fusion with lysosomes, the limiting membranes became more densely and homogeneously labeled with the lectins. We previously reported that cytochrome P-450 does not exist on the limiting membranes of the autophagosomes. Taken together, these results suggest that the isolation membranes may originate not from endoplasmic reticulum membranes but from some post-Golgi membranes that contain complex type N and/or O-linked oligosaccharide chains.