Activity of N-acylneuraminate-9-phosphatase (NANP) is not essential for de novo sialic acid biosynthesis

BackgroundSialylation of glycoproteins and glycolipids is important for biological processes such as cellular communication, cell migration and protein function. Biosynthesis of CMP-sialic acid, the essential substrate, comprises five enzymatic steps, involving ManNAc and sialic acid and their phosphorylated forms as intermediates. Genetic diseases in this pathway result in different and tissue-restricted phenotypes, which is poorly understood.Methods and resultsWe aimed to study the mechanisms of sialic acid metabolism in knockouts (KO) of the sialic acid pathway in two independent cell lines. Sialylation of cell surface glycans was reduced by KO of GNE (UDP-N-acetylglucosamine 2-epimerase/N-acetylmannosamine kinase), NANS (sialic acid synthase) and CMAS (N-acylneuraminate cytidylyltransferase) genes, but was largely unaffected in NANP (N-acylneuraminate-9-phosphatase) KO, as studied by MAA and PNA lectin binding. NANP is the third enzyme in sialic acid biosynthesis and dephosphorylates sialic acid 9-phosphate to free sialic acid. LC-MS analysis of sialic acid metabolites showed that CMP-sialic acid was dramatically reduced in GNE and NANS KO cells and undetectable in CMAS KO. In agreement with normal cell surface sialylation, CMP-sialic acid levels in NANP KO were comparable to WT cells, even though sialic acid 9-phosphate, the substrate of NANP accumulated. Metabolic flux analysis with ¹³C₆-labelled ManNAc showed a lower, but significant conversion of ManNAc into sialic acid.ConclusionsOur data provide evidence that NANP activity is not essential for de novo sialic acid production and point towards an alternative phosphatase activity, bypassing NANP.General significanceThis report contributes to a better understanding of sialic acid biosynthesis in humans.     

α2-3 Sialic acid glycoconjugate loss and its effect on infection with Toxoplasma parasites

Recognition of sialylated glycoconjugates is important for host cell invasion by Apicomplexan parasites. Toxoplasma gondii parasites penetrate host cells via interactions between their microneme proteins and sialylated glycoconjugates on the surface of host cells. However, the role played by sialic acids during infection with T. gondii is not well understood. Here, we focused on the role of α2-3 sialic acid linkages as they appear to be widely expressed in vertebrates. Removal of α2-3 sialic acid linkages on macrophages by neuraminidase treatment did not influence the rate of infection or growth of T. gondii, nor did it affect phagocytosis in vitro. Sialyltransferase ST3Gal-I deficient mice (ST3Gal-I^−/− mice) lost α2-3 sialic acid linkages in macrophages and spleen cells. The numbers of T. gondii-infected CD11b⁺ cells in peritoneal cavities of the infected ST3Gal-I^−/− mice were relatively lower than those of the infected wild type animals. In addition, CD8⁺ T cell populations and numbers in the spleens and peritoneal cavities of the ST3Gal-I^−/− mice were significantly lower than those in the wild type animals before and after the T. gondii infection. ST3Gal-I^−/− mice had severe liver damage and reduced survival rates following peritoneal infection with T. gondii. Furthermore, adoptive transfer of immune CD8⁺ cells from wild type mice to ST3Gal-I^−/− mice increased their survival during infection with T. gondii. Our data show that parasite invasion via α2-3 sialic acid linkages might not contribute on host survival and indicate the impact that loss of α2-3 sialic acid linkages has on CD8⁺ T cell populations, which are necessary for effective immune responses against infection with T. gondii.     

Novel structural insights into rotavirus recognition of ganglioside glycan receptors

Rotaviruses ubiquitously infect children under the age of 5, being responsible for more than half a million diarrhoeal deaths each year worldwide. Host cell oligosaccharides containing sialic acid(s) are critical for attachment by rotaviruses. However, to date, no detailed three-dimensional atomic model showing the exact rotavirus interactions with these glycoconjugate receptors has been reported. Here, we present the first crystallographic structures of the rotavirus carbohydrate-recognizing protein VP8^? in complex with ganglioside G_M3 glycans. In combination with assessment of the inhibition of rotavirus infectivity by N-acetyl and N-glycolyl forms of this ganglioside, our results reveal key details of rotavirus-ganglioside G_M3 glycan recognition. In addition, they show a direct correlation between the carbohydrate specificities exhibited by VP8^? from porcine and by monkey rotaviruses and the respective infectious virus particles. These novel results also indicate the potential binding interactions of rotavirus VP8^? with other sialic acid-containing gangliosides.     

The repair of the surface structure of animal cells

Experiments were performed to determine if animal cells in culture possess specific mechanisms to repair surface molecules damaged by enzymes. The surface membranes of a primary cell culture, chick fibroblasts, a permanent hamster cell line, BHK₂₁/C₁₃, and its virally transformed counterpart, C₁₃/B₄ were damaged by exposure to trypsin or to neuraminidase. Following digestion with trypsin, the incorporation of radioactive amino acids or sugars into purified surface membrane of cells was monitored. No differences were noted in rates of incorporation when control and trypsin-damaged cells were compared. Neuraminidase damage to the surface of BHK₂₁/C₁₃ and C₁₃/B₄ cells was evidenced by altered gel filtration profiles of surface glycopeptides, i.e., delayed elution because of reduction in size. By labelling cells with ¹⁴C-L-fucose prior to neuraminidase treatment and following the incorporation of ³H-L-fucose into cell surface glycopeptides after neuraminidase digestion, we were able to monitor the synthesis and turnover of fucose-containing glycopeptides in the same cells. Gel filtration profiles indicated that little or no desialylated glycoproteins were resialylated (repaired) by specific replacement of sialic acid. Comparing neuraminidase-digested and control cells we observed no difference in rates of ³H-L-fucose incorporation or of ¹⁴C-L-fucose loss from these cells; nor did we find differences in the rate of incorporation of isotopic glucosamine into sialic acid. Neuraminidase treatment failed to alter the rate of cell growth or the pattern of isotopic incorporation into various cell surface components. These results support the suggestion that return of sialic acid (repair) was effected by turnover which serves as a non-specific repair mechanism to replace damaged cell surface molecules (Warren and Glick '68; Warren, '69).     

Stimulation of human peripheral blood mononuclear cells by the sialic acid precursor N-propanoylmannosamine

N-Propanoylmannosamine is an unnatural precursor of sialic acid, which is taken up by a variety of animal cells and metabolized to N-propanoylneuraminic acid. In several studies it has been demonstrated that application of unnatural precursors of sialic acids such as N-propanoylmannosamine (ManNProp) and homologues interfere with cell differentiation and proliferation of neuronal cells or embryonic stem cells. Since the function of the immune system is known to rely on the presence of sialic acid, we applied ManNProp to human peripheral blood mononuclear cells (PBMC). When culturing those lymphocytes with ManNProp 10 % of the natural sialic acid N-acetylneuraminic acid could be replaced by the newly formed N-propanoylneuraminic acid. This procedure resulted (a) in a marked stimulation in the rate of proliferation of PBMC, (b) a 10-fold increase of IL-2 production coupled with an up-regulation of its receptor CD25 on the cell surface and (c) a concomitant expression and regulation of the transferrin receptor with cell growth. The stimulation of PBMC by ManNProp might therefore introduce a new approach of immunomodulation.     

Cell surface changes associated with EGF and insulin-induced cell adhesion

The ability of insulin and epidermal growth factor (EGF) to restore cell surface function in cells damaged by serum deprivation has been examined. Both insulin (10^?7 M) and EGF (10^?8 M), when added for 2 h, resulted in reattachment of cells to the culture dish; with insulin, attachment was associated with increased amino acid uptake. Both hormones caused an increase in sialic acid and in free sulfhydryl groups associated with the cell membrane. The mobility of concanavalin A (Con A) on the upper cell surface was increased by EGF, and to a lesser extent by insulin. The effects of insulin and EGF on adhesion appear to be regulated by specific receptors for these hormones.     

Plasmodium berghei: modification of sialic acid on red cells from infected mouse blood

The surface membrane glycoproteins of normal mouse erythrocytes can be labeled by oxidation with either periodate or galactose oxidase in the presence of neuraminidase, followed by reduction with NaB³H₄. Without neuraminidase there is little galactose oxidase-catalyzed labeling of protein. Analysis of labeled proteins by SDS-polyacrylamide gel electrophoresis showed that both methods labeled the same set of glycoproteins. Plasmodium berghei infection dramatically reduced the sialoglycoprotein labeling of red blood cells from infected blood using the periodate/NaB³H₄ method. Provided neuraminidase was present, labeling by the galactose oxidase method gave identical results to normal erythrocytes. We conclude that the glycoprotein sialic acid of uninfected as well as infected red cells is modified during infection such that it is refractory to periodate oxidation. Acylation of the exocyclic hydroxyls of sialic acid is suggested to account for this. Lectin binding and cell agglutination experiments using Limulin, soybean and wheatgerm lectins, and concanavalin A confirmed and extended these observations. The possible implications of these results with regard to anemia induced by malaria are briefly discussed.     

Paracrine effects of a uterine agglutinin are mediated via the sialic acids present in the rat uterine endometrium

A 32 kDa estrogen-induced, sialic acid-specific agglutinin (P-SAS) was isolated from rat endometrium in its proestrus stage [1]. To investigate the functional importance of P-SAS in the uterine milieu, specific binding assays were carried out with ¹²⁵I-labeled P-SAS and different cellular components of the uterus (epithelial, stromal and myometrial cells), that were isolated from different stages of the estrus cycle. The results indicate that although the protein is secreted from the epithelial cells in the estrogenic phase, it binds specifically to the stromal cells, especially to those isolated from the diestrus stage of the estrus cycle. The specific binding, however, is seen to decrease with the progression of pregnancy. Scatchard analysis performed with varying amounts of ¹²⁵I-P-SAS in the presence of excess cold P-SAS revealed that the binding occurs with a Ka = 1.69 × 10⁸ M^-1. As P-SAS binds specifically to sialic acids on the stromal cell surface, further characterization of the sialic acid molecule to which P-SAS binds was carried out by gas liquid chromatography (GLC). The studies revealed that P-SAS preferentially binds to N-glycolylneuraminic acid, which is attached to the penultimate sugar of the stromal cell surface glycoprotein chain via 2,6 linkage. As P-SAS is further known to be mitogenic [2], the effect of P-SAS on cultured stromal cells was studied in vitro. The growth regulatory assays revealed that P-SAS induced ³H-thymidine uptake by stromal cells in culture. Thus, from the above observations, paracrine effects of P-SAS on the stromal cells and on the subsequent growth and development of the uterus can be assumed.     

The effect of valinomycin on the electrical properties of solutions of red cell lipids in n-decane

This paper reports the electrical properties of thick lipid membranes in the absence and presence of valinomycin. The thick lipid membranes were formed by placing a solution of sheep red cell lipids in decane between two cellophane partitions which formed the interfaces between the membrane and the two aqueous bathing solutions. The DC electrical resistance of these structures was found to be directly proportional to the reciprocal of the concentration of lipids in the decane (C_L). The limiting resistance, as (C_L ^-1) approached zero, was 3 x 10⁸ ohm-cm². Resistance was also found to be linearly related to membrane thickness. The limiting resistance at zero thickness was again 1–3 x 10⁸ ohm-cm². These data are interpreted to indicate that the DC resistance of thick lipid membranes comprises two surface resistances (R_S) at each interface with the aqueous bathing solutions, and a bulk resistance (R_B) of the lipid-decane solution, arranged in series. Measurements of the effect of variations of area on resistance were consistent with this interpretation. Valinomycin reduced R_S but had no effect on R_B. Under certain conditions, thick lipid membranes containing valinomycin behaved like highly selective K⁺ electrodes.     

Surface charge and hydrophobicity of wild and mutantCrithidia fasciculata

Surface charge of wild-typeCrithidia fasciculata and three drug-resistant mutants (T^R3, TFR^R1, and FU^R11) was studied by direct zeta-potential determination and ultrastructural cytochemistry. Surface tension was also investigated by measurements of the advancing contact angle formed by the protozoa monolayers with drops of liquids of different polarities. The individual zeta potential varies markedly among theC. fasciculata cells. The wild and FU^R11 mutant strains displayed lower negative surface charge (?12.5 and ?9.5 mV, respectively), as compared with the T^R3 (?14.8 mV) and TFR^R1 (?14.7 mV) mutant strains. Binding of cationized ferritin (CF) was observed at the cell surface of wild and mutant strains ofC. fasciculata. Neuraminidase treatment reduced the negative surface charge in the TFR^R1 and T^R3 mutants in about 37 and 29%, respectively, whereas no significant change was observed with the wild and FU^R11 mutant strains. These findings suggest that sialic acid residues are the major anionogenic groups on the surface ofC. fasciculata. The density of sialic acid residues per cell in wild and mutant strains ofC. fasciculata falls in a range of 1.4×10⁴ to 3.6×10⁴. Marked differences of hydrophobicity were also observed. For example, the TFR^R1, FU^R11, and T^R3 drug-resistant mutant strains showed higher contact angle values (55.4, 54.2, and 49.3, respectively) than the wild-type (35.6), as assessed by α-bromonaphtalene.     

AMPHOTERIC BEHAVIOR OF COMPLEX SYSTEMS : IV. NOTE ON THE ISOELECTRIC POINT AND IONIZATION CONSTANTS OF SULFANILIC ACID.

From the solubility minimum the value of the basic ionization constant of sulfanilic acid is shown to lie probably between the values 1.7 x 10^–15 and 3.2 x 10^–15. From solubility measurements the value of this same constant is shown to lie probably between 2.0 and 2.2 x 10^–15, and the isoelectric point of sulfanilic acid is thus at a cH of 0.056 or a pH of 1.25. From conductivity ratios the acid ionization constant of sulfanilic acid is shown to be 7.05 x 10^–4 at room temperature (21°C.). Calculations are made, from data published in preceding papers, of the ionization constants of glycine, K_a being 2.3 x 10^–10, and K_b being 2.2 x 10^–12.     

Sialic acid, electrophoretic mobility and transmembrane potentials of the Amphiuma red cell

Red cells from the giant salamander Amphiuma means are shown to contain sialic acid. The amount removed by the action of neuraminidase is equal to that released by acid hydrolysis, indicating that all of the sialic acid is present on the outer surface of the plasma membrane. These cells have a negative electrophoretic mobility and 100% enzymatic removal of sialic acid results in a 40% reduction in the mobility, suggesting that either a fraction of the sialic acid carboxyl groups are unavailable to the action of external electric fields, or other negatively charged groups contribute to the surface charge. A further reduction in mobility of normal and sialic acid-free cells is caused by an increased extracellular calcium concentration. The negative groups affected by calcium are most likely to be phosphate groups, since the isoelectric point of the cells is found to lie between the pK values for H₂PO₄⁻ groups and the carboxyl groups of sialic acid. Membrane potentials of single cells, from which 80% or more of the total sialic acid had been removed, were identical to those measured in normal cells, confirming that sialic acid plays little, if any, direct role in the maintenance of membrane potentials and ionic permeabilities.     

Accessibility of sialo components in a murine tumor cell to extracellular N-acetylneuraminate glycohydrolase (sialidase)

Lipid-bound sialic acid in the murine melanoma cell is not totally inaccessible to an exogenous macromolecular probe, as formerly believed. Roughly 30% of the sialic acid bound to lipid, and an equal proportion of the sialic acid bound to protein is cleaved by the action of Clostridium perfringens$\text{N-}\text{acetylneuraminate}$ glycohydrolase (neuraminidase, sialidase) when the purified enzyme is added to the suspension medium of intact murine melanoma cells freshly derived from the tumor. Cleavage of lipid-bound sialic acid is indifferent to the presence of Ca²⁺ in the medium. However, maximum release from protein requires a physiological concentration of this divalent cation. Variation in ionic strength has no effect on release of sialic acid. These findings show that a restricted portion of the bound sialic acid may be released from the intact murine melanama cell by the extracellularly supplied enzyme acting topographically.     

The Epstein-Barr Virus G-Protein-Coupled Receptor Contributes to Immune Evasion by Targeting MHC Class I Molecules for Degradation

Epstein-Barr virus (EBV) is a human herpesvirus that persists as a largely subclinical infection in the vast majority of adults worldwide. Recent evidence indicates that an important component of the persistence strategy involves active interference with the MHC class I antigen processing pathway during the lytic replication cycle. We have now identified a novel role for the lytic cycle gene, BILF1, which encodes a glycoprotein with the properties of a constitutive signaling G-protein-coupled receptor (GPCR). BILF1 reduced the levels of MHC class I at the cell surface and inhibited CD8⁺ T cell recognition of endogenous target antigens. The underlying mechanism involves physical association of BILF1 with MHC class I molecules, an increased turnover from the cell surface, and enhanced degradation via lysosomal proteases. The BILF1 protein of the closely related CeHV15 γ₁-herpesvirus of the Rhesus Old World primate (80% amino acid sequence identity) downregulated surface MHC class I similarly to EBV BILF1. Amongst the human herpesviruses, the GPCR encoded by the ORF74 of the KSHV γ₂-herpesvirus is most closely related to EBV BILF1 (15% amino acid sequence identity) but did not affect levels of surface MHC class I. An engineered mutant of BILF1 that was unable to activate G protein signaling pathways retained the ability to downregulate MHC class I, indicating that the immune-modulating and GPCR-signaling properties are two distinct functions of BILF1. These findings extend our understanding of the normal biology of an important human pathogen. The discovery of a third EBV lytic cycle gene that cooperates to interfere with MHC class I antigen processing underscores the importance of the need for EBV to be able to evade CD8⁺ T cell responses during the lytic replication cycle, at a time when such a large number of potential viral targets are expressed.     

Mild oxidation of starches with aqueous bromine : Part I. Kinetics and product analysis

Starch suspensions have been treated with dilute, aqueous bromine at 30° in the pH range 6–8; no adsorption of oxidant occurred. The oxidation kinetics were first-order in bromine and in accordance with the rate law d [bromine]/dt  k [starch] [bromine], except for a minor, initial rapid-phase in the oxidation of cereal starches, which is attributed to an enhanced reactivity of the granule surface. The apparent first-order rate-constants were 2.0–2.8 x 10^?3 min_?1, except for retrograded amylose oxidised at pH 8 when the value was 5.6 x 10_?3 min_?1. The i.r. spectra of the products indicated the presence of carboxylate and aldehyde groups. The functional group contents were determined quantitatively. Oxidation of the amylose at pH 6–7 introduced carbonyl groups, whereas at pH 8 carbonyl and carboxylate were found in equal amounts. For waxy-maize starch oxidised at pH 6–8, the carbonyl content was twice that of carboxylate. Acid hydrolysis of the product obtained by oxidation of amylose proceeded at pH 8 according to first-order kinetics. Chromatographic analysis of the anionic components of the hydrolysate indicated the presence of D-glucurono-6,3-lactone, D-gluconic acid, and an unidentified acidic ketose.     

THE LYSOSOME PERIPHERY: BIOCHEMICAL AND ELECTROKINETIC PROPERTIES OF THE TRITOSOME SURFACE

Normal rat liver lysosomes were isolated by the technique of loading with Triton WR-1339. Purity of the preparation was monitored with marker enzymes; a high enrichment in acid hydrolases was obtained in the tritosome fraction. In 0.0145 M NaCl, 4.5% sorbitol, 0.6 mM NaHCO₃, pH 7.2 at 25°C the tritosomes had an electrophoretic mobility of -1.77 ± 0.02 µm/s/V/cm, a zeta potential of 23.2 mV, a surface charge of 1970 esu/cm², and 33,000 electrons per particle surface assuming a tritosome diameter of 5 x 10^-7 m. Treatment of the tritosomes with 50 µg neuraminidase/mg tritosome protein lowered the electrophoretic mobility of the tritosome to -1.23 ± 0.02 µm/s/V/cm under the same conditions and caused the release of 2.01 µg sialic acid/mg tritosome protein. Treatment of the tritosomes with hyaluronidase did not affect their electrophoretic mobility, while trypsin treatment elevated the net negative electrophoretic mobility of the tritosomes. Tritosome electrophoretic mobilities indicated a homogeneous tritosome population and varied greatly with ionic strength of the suspending media. pH vs. electrophoretic mobility curves indicated the tritosome periphery to contain an acid-dissociable group which likely represents the carboxyl group of N-acetylneuraminic acid; this was not conclusively proven, however, since the tritosomes lysed below a pH of 4 in the present system. Total tritosome carbohydrate (anthrone-positive material as glucose equivalents) was 0.19 mg/mg tritosome protein while total sialic acid was 3.8 µg (11.4 nmol)/mg tritosome protein. A tritosome "membrane" fraction was prepared by osmotic shock, homogenization, and sedimentation. Approximately 25% of the total tritosome protein was present in this fraction. Analysis by gas-liquid chromatography and amino acid analyzer showed the following carbohydrate composition of the tritosome membrane fraction (in microgram per milligram tritosome membrane protein): N-acetylneuraminic acid, 14.8 ± 3; glucosamine, 24 ± 3; galactosamine, 10 ± 2; glucose, 21 ± 2; galactose, 26 ± 2; mannose, 31 ± 5; fucose, 7 ± 1; xylose, 0; and arabinose, 0. The results indicate that the tritosome periphery is characterized by external terminal sialic acid residues and an extensive complement of glycoconjugates. Essentially all the tritosome N-acetylneuraminic acid is located in the membrane and about 53% of it is neuraminidase susceptible.     

Engineering the sialic acid in organs of mice using N-propanoylmannosamine

Sialic acids play an important role during development, regeneration and pathogenesis. The precursor of most physiological sialic acids, such as N-acetylneuraminic acid is N-acetyl-d-mannosamine. Application of the novel N-propanoylmannosamine leads to the incorporation of the new sialic acid N-propanoylneuraminic acid into cell surface glycoconjugates. Here we analyzed the modified sialylation of several organs with N-propanoylneuraminic acid in mice. By using peracetylated N-propanoylmannosamine, we were able to replace in vivo between 1% (brain) and 68% (heart) of physiological sialic acids by N-propanoylneuraminic acid. The possibility to modify cell surfaces with engineered sialic acids in vivo offers the opportunity to target therapeutic agents to sites of high sialic acid concentration in a variety of tumors. Furthermore, we demonstrated that application of N-propanoylmannosamine leads to a decrease in the polysialylation of the neural cell adhesion molecule in vivo, which is a marker of poor prognosis for some tumors with high metastatic potential.     

Immobilized α2,6-linked sialic acid suppresses caspase-3 activation during anti-IgM antibody-induced apoptosis in Ramos cells

In Ramos cells, a human Burkitt's lymphoma cell line, stimulation of the B cell antigen receptor with anti-IgM antibody (Ab) induces apoptosis as indicated by a decrease in cell viability and an increase in DNA fragmentation and cell surface exposure of phosphatidylserine. Furthermore, these changes are suppressed by incubating the cells in α₁-acid glycoprotein (AGP)-coated tissue culture plates. Here, we found that, during Anti-IgM Ab-induced apoptosis in Ramos cells, caspase-3 is activated downstream of caspase-8 and the mitochondrial pathway is activated, as indicated by a loss of mitochondrial membrane potential, an increase in the release of cytochrome c to the cytoplasm, and enhanced Bax expression. Anti-IgM Ab-induced apoptosis of neuraminidase-treated Ramos cells was suppressed by incubating the cells on plates coated with AGP, which contains a high concentration of α2,6-linked sialic acid. The incubation on plates coated with AGP also suppressed anti-IgM Ab-stimulated caspase-3 activity and increased the level of X-linked inhibitor of apoptosis protein (XIAP), but it did not affect caspase-8 activity, the mitochondrial membrane potential, cytochrome c release, or Bax expression. The results indicate that the interaction of Ramos cells with immobilized α2,6-linked sialic acid enhances XIAP expression, directly or indirectly suppressing caspase-3 activity and inhibiting anti-IgM Ab-induced apoptosis.     

Tissue distribution of EDTA encapsulated within liposomes containing glycolipids or brain phospholipids

Multilamellar liposomes were prepared with various asialoglycolipids, gangliosides, sialic acid, or brain phospholipids in the liposome membrane and with ethylenediaminetetraacetic acid (EDTA) encapsulated in the aqueous compartments. The liposomes containing glycolipids or sialic acid were prepared from a mixture of phosphatidylcholine, cholesterol, and one of the following test substances: galactocerebroside, glucocerebroside, galactocerebroside sulfate, mixed gangliosides, monosialoganglioside G_M1, monosialoganglioside G_M2, monosialoganglioside G_M3, disialoganglioside G_D1a, or sialic acid. The liposomes containing brain phospholipids were mixtures of either sphingomyelin and cholesterol or a brain total phospholipid extract and cholesterol. Distribution of ¹⁴C-labeled EDTA were determined in mouse tissues from 15 min to 6 h or 12 h after a single injection of liposome prepartion. Liver uptake of encapsulated EDTA was lowest from all liposome preparations containing sialic acid or sialogangliosides regardless of the amount of sialic acid moiety present or the identity of the particular ganglioside; highest uptake of encapsulated EDTA by liver was from the liposomes containing galactocerebroside or brain phospholipids. Lungs and brain took up the largest amounts of EDTA from liposomes containing sphingomyelin and lesser amounts from liposomes containing G_D1a. Use of mouse brain phospholipid extract to prepare liposomes did not increase uptake of encapsulated EDTA by the brain. EDTA in liposomes containing monosialogangliosides, brain phospholipids, galactocerebroside, or sialic acid was taken up well by spleen and marrow. Highest thymus uptake of encapsulated EDTA was from liposomes containing G_D1a. These results demonstrate that inclusion of sialogangliosides in liposome membranes decreases uptake of liposomes by liver, thus making direction of encapsulated drugs to other organs more feasible. Liposomes containing glycolipids also have potential uses as probes of cell surface receptors.