Immunocytochemical localization of mutant low density lipoprotein receptors that fail to reach the Golgi complex.

In the low density lipoprotein (LDL) receptor system, blocks in intracellular movement of a cell surface receptor result from naturally occurring mutations. These mutations occur in patients with familial hypercholesterolemia. One class of mutant LDL receptor genes (class 2 mutations) produces a receptor that is synthesized and glycosylated in the endoplasmic reticulum (ER) but does not reach the cell surface. These receptors contain serine/threonine-linked (O-linked) carbohydrate chains with core N-acetylgalactosamine residues and asparagine-linked (N-linked) carbohydrate chains of the high mannose type that are only partially trimmed. To determine the site of blockage in transport, we used electron microscope immunohistochemistry to compare the intracellular location of LDL receptors in normal human fibroblasts with their location in class 2 mutant fibroblasts. In normal cells, LDL receptors were located in coated pits, coated vesicles, endosomes, multivesicular bodies, and portions of the Golgi complex. In contrast, the mutant receptors in class 2 cells were almost entirely confined to rough ER and irregular extensions of the rough ER. Metabolic labeling studies with [3H]glucosamine confirmed that these mutant receptors contain core O-linked sugars, suggesting that the enzymes that attach these residues are located in the rough ER or the transitional zone of the ER. These studies establish that naturally occurring mutations in cell surface receptors can cause the receptors to remain trapped in the ER, thereby preventing their normal function and producing a genetic disease.

Cloning, expression, purification, and characterization of the murine lysosomal acid alpha-mannosidase.

Catabolism of alpha-linked mannose residues on eukaryotic glycoproteins is accomplished by a broad specificity lysosomal alpha-mannosidase (EC 3.2.1.24). Based on regions of protein sequence conservation between the lysosomal alpha-mannosidase from Dictyostelium discoideum and the murine Golgi glycoprotein processing alpha 1,3/1,6-mannosidase, alpha-mannosidase II, we have cloned a cDNA encoding the murine lysosomal alpha-mannosidase. The longest of the clones was 3.1 kb in length and encoded a polypeptide of 992 amino acids containing a putative NH2-terminal signal sequence and 11 potential N-glycosylation sites. The deduced amino acid sequence was 76.5% identical to the human lysosomal alpha-mannosidase and 38.1% identical to the lysosomal alpha-mannosidase from D. discoideum. Expression of the cDNA in Pichia pastoris resulted in the secretion of an alpha-mannosidase activity into the culture medium. This recombinant expression product was purified and was shown to have enzymatic characteristics highly similar to the enzyme purified from mammalian sources and to the human lysosomal alpha-mannosidase cDNA expressed in Pichia. These characteristics include a similar pH optimum, Km, Vmax, inhibition by swainsonine, and activity toward natural substrates. Northern blots identified a major 3.5 kb RNA transcript in all murine tissues tested. A minor transcript of 5.4 kb was also detected in some murine tissues similar to the alternatively spliced transcripts that have been previously identified in human tissues.

Asparagine-linked oligosaccharides containing poly-N-acetyllactosamine chains are preferentially bound by immobilized calf heart agglutinin.

We have investigated the carbohydrate-binding specificity of a mammalian lectin, calf heart agglutinin, by determining the interaction of the immobilized lectin with a variety of complex-type Asn-linked oligosaccharides. Our results demonstrate that calf-heart agglutinin binds with high affinity to oligosaccharides containing the repeating disaccharide (3Gal beta 1-4GlcNAc beta 1)n or poly-N-acetyllactosamine sequence and that the presence of terminal beta-linked galactosyl residues is neither sufficient nor necessary for high affinity interactions.

Relationship of the terminal sequences to the length of poly-N-acetyllactosamine chains in asparagine-linked oligosaccharides from the mouse lymphoma cell line BW5147. Immobilized tomato lectin interacts with high affinity with glycopeptides containing long poly-N-acetyllactosamine chains.

To investigate the factors regulating the biosynthesis of poly-N-acetyllactosamine chains containing the repeating disaccharide [3Gal beta 1,4GlcNAc beta 1] in animal cell glycoproteins, we have examined the structures and terminal sequences of these chains in the complex-type asparagine-linked oligosaccharides from the mouse lymphoma cell line BW5147. Cells were grown in medium containing [6-3H]galactose, and radiolabeled glycopeptides were prepared and fractionated by serial lectin affinity chromatography. The glycopeptides containing the poly-N-acetyllactosamine chains in these cells were complex-type tri- and tetraantennary asparagine-linked oligosaccharides. The poly-N-acetyllactosamine chains in these glycopeptides had four different terminal sequences with the structures: I, Gal beta 1,4GlcNAc beta 1,3Gal-R; II, Gal alpha 1,3Gal beta 1,4GlcNac beta 1,3Gal-R; III, Sia alpha 2,3Gal beta 1,4GlcNAc beta 1,3Gal-R; and IV, Sia alpha 2,6Gal beta 1,4GlcNAc beta 1,3Gal-R. We have found that immobilized tomato lectin interacts with high affinity with glycopeptides containing three or more linear units of the repeating disaccharide [3Gal beta 1,4GlcNAc beta 1] and thereby allows for a separation of glycopeptides on the basis of the length of the chain. A high percentage of the long poly-N-acetyllactosamine chains bound by immobilized tomato lectin were not sialylated and contained the simple terminal sequence of Structure I. In addition, a high percentage of the sialic acid residues that were present in the long chains were linked alpha 2,3 to penultimate galactose residues (Structure III). In contrast, a high percentage of the shorter poly-N-acetyllactosamine chains not bound by the immobilized lectin were sialylated, and most of the sialic acid residues in these chains were linked alpha 2,6 to galactose (Structure IV). These results indicate that there is a relationship in these cells between poly-N-acetyllactosamine chain length and the degree and type of sialylation of these chains.

Localization of adenylate cyclase during development of Dictyostelium discoideum.

Cyclic AMP is known to function as the chemotactic signal during aggregation of single-celled amoebae of the cellular slime mold Dictyostelium discoideum. Evidence from several laboratories has accumulated suggesting that cAMP also acts as a regulatory molecule during Dictyostelium multicellular differentiation. We have used ultramicrotechniques and a sensitive radioimmunoassay in the localization of adenylate cyclase, the cAMP synthetic enzyme, during the development of Dictyostelium. We demonstrate that adenylate cyclase activity is localized in the prespore cells of the culminating individual with no activity detectable in the prestalk region. We show that this lack of activity in the stalk may be due to a masking by an endogenous inhibitor of the enzyme. Within the spore mass we found an increasing gradient of enzyme activity toward the base. These data, along with that from the localization of cyclic nucleotide phosphodiesterase, indicate that an enzymatic potential exists for the creation of cAMP gradients during development in the organism. Such a gradient may provide positional information necessary to direct the terminal differentiation of spore and stalk cells.

Enzymatic sulfation of N-glycosidically linked oligosaccharides by endothelial cell membranes.

Homogenates of human vascular endothelial cells catalyze the transfer of 35SO4 from 3'-phosphoadenosine 5'-phospho[35S]sulfate into macromolecular endogenous acceptors with properties of both glycoproteins and glycosaminoglycans. Analysis of the 35S-glycoprotein products by both DEAE-cellulose chromatography and sodium dodecyl sulfate-polyacrylamide gel electrophoresis indicates that these 35S-glycoproteins correspond to the major sulfated glycoproteins observed when endothelial cell cultures were labeled with 35SO4 [A. Heifetz, C. Watson, A. R. Johnson, and M. K. Roberts (1982) J. Biol. Chem. 257, 13581-13586]. The pH optimum of the glycoprotein sulfotransferase is 7.0-7.5, which is distinctly different than that of endothelial glycosaminoglycan sulfotransferase(s), whose pH optimum is 6.0. The 35S-glycoprotein products were characterized as sialated, triantenary-branched, N-linked 35SO4-oligosaccharides containing an endo-beta-N-acetylglucosaminidase D-sensitive site. The site of sulfation was characterized as the GlcNAc residue on the reducing end of the N-linked oligosaccharide. Thus, these sulfotransferases in a cell-free homogenate appear to preferentially add sulfate moieties to a specific class of glycoprotein acceptors at a specific site on sialated oligosaccharides.

The effect of swainsonine and castanospermine on the sulfation of the oligosaccharide chains of N-linked glycoproteins.

MDCK (Madin-Darby canine kidney) cells infected with the NWS strain of influenza virus incorporate 35SO4 into complex types of oligosaccharides of the N-linked glycoproteins. On the other hand, when these virus-infected MDCK cells are incubated in the presence of swainsonine, an inhibitor of the processing mannosidase II, approximately 40-80% of the total [35S]glycopeptides were of the hybrid types of structures. Thus, these sulfated, hybrid types of glycopeptides were completely susceptible to digestion by endoglucosaminidase H, whereas the sulfated glycopeptides from infected cells incubated without swainsonine were completely resistant to endo-beta-N-acetylglucosaminidase H. When virus-infected MDCK cells were incubated in the presence of castanospermine, an inhibitor of the processing glucosidase I, the N-linked glycopeptides contained mostly oligosaccharide chains of the Glc3Man7-9GlcNAc2 types of structures, and these oligosaccharides were devoid of sulfate. Structural analysis of these abnormally processed oligosaccharides produced in the presence of swainsonine or castanospermine indicated that they differed principally in the processing of one oligosaccharide branch as indicated by the structures shown below. They also differed in that only the swainsonine-induced structures were sulfated. These data indicate that removal of glucose units and perhaps other processing steps are necessary before sulfate residues can be added. (Formula: see text).

Localization and levels of cyclic AMP during development of Dictyostelium discoideum.

Cyclic AMP functions as the chemotactic signal during aggregation of amoebae of the cellular slime mold Dictyostelium discoideum. Evidence suggests that cAMP also acts as a regulatory molecule during Dictyostelium multicellular differentiation. We have used ultramicrotechniques and a sensitive radioimmunoassay to measure the levels of cAMP within the culmination stage individual. We show that there is a peak of cAMP at the culmination stage of development and that in the individual at this stage the molecule is localized in a gradient within the spore mass.

The dysfunctional LDL receptor in a monensin-resistant mutant of Chinese hamster ovary cells lacks selected O-linked oligosaccharides.

The Chinese hamster ovary (CHO) cell line Monr31, which is resistant to the cytotoxic ionophore monensin, produces a receptor for the low density lipoprotein (LDL) that has a lowered binding affinity for LDL and is approximately 5 kDa smaller in size than the receptor from parental CHO cells. It has been proposed that the reduced size and affinity for LDL are associated with a reduced level of O-glycosylation of Ser/Thr residues in the receptor. To examine this possibility in more detail, both parental CHO and Monr31 cells were metabolically radiolabeled with [3H]glucosamine, and the labeled LDL receptors were purified by immunoprecipitation and identified by SDS-PAGE-fluorography. The Ser/Thr-linked oligosaccharides in the receptors from both parental CHO and Monr31 cells are mono- and desialylated species having the common core structure Gal beta 1-3GalNAc. The receptor from Monr31 cells, however, contains about one-third fewer Ser/Thr-linked oligosaccharides than the receptor from parental CHO cells. Analysis of the glycopeptides derived from the Monr31 cell LDL receptors indicates that they contain Ser/Thr-linked oligosaccharides only in the clustered domain and are missing Ser/Thr-linked oligosaccharides in the unclustered regions of the protein. Additionally, analysis of a human LDL receptor lacking the domain for attachment of the clustered Ser/Thr-linked oligosaccharides and expressed in both parental CHO and Monr31 cells indicated that the truncated human receptor from Monr31 cells is devoid of Ser/Thr-linked oligosaccharides. In contrast, the truncated human receptor produced by parental CHO cells contains Ser/Thr-linked oligosaccharides contributing approximately 5 kDa to its apparent size. Collectively, these results demonstrate that the LDL receptor produced by the Monr31 cells contains Ser/Thr-linked oligosaccharides in the clustered domain but is missing Ser/Thr-linked oligosaccharides in the unclustered, NH2-terminal domains of the receptor.

Characterization of an S-type lectin purified from porcine heart.

We have purified a carbohydrate-binding protein from porcine heart by affinity chromatography on asialofetuin-Sepharose and have characterized this protein with respect to its size, amino acid composition, partial amino acid sequence, and carbohydrate-binding specificity. Porcine heart lectin (PHL) has a subunit molecular mass of 14,700 and is immunologically cross-reactive with a polyclonal antibody raised against a lectin isolated from calf heart. The amino acid composition of PHL is similar to that of lectins that have been isolated from calf heart, bovine brain, and rat lung. Moreover, the primary sequences of four tryptic fragments (52 amino acids total) derived from PHL are closely related to sequences previously determined for 10 other vertebrate-derived lectins. The ability of PHL to agglutinate rabbit erythrocytes was inhibited only by oligosaccharides containing terminal beta-galactosyl residues. These data indicate that PHL is a vertebrate "S-type" lectin and provide further evidence that the structures and carbohydrate-binding specificities of these lectins are highly conserved across diverse vertebrate genera.

gp160 of HIV-I synthesized by persistently infected Molt-3 cells is terminally glycosylated: evidence that cleavage of gp160 occurs subsequent to oligosaccharide processing.

The envelope glycoprotein of HIV-I in infected, cultured human T cells is synthesized as a precursor of apparent Mr 160 kDa (gp160) and is cleaved to two glycoproteins, gp120 and gp41, which are the mature envelope glycoproteins in the virus. Neither the temporal and spatial features of glycosylation nor the oligosaccharide processing and proteolytic cleavage of the envelope glycoprotein are well understood. To understand more about these events, we investigated the glycosylation and cleavage of the envelope glycoproteins in the CD4+ human cell line, Molt-3, persistently infected with HIV-I (HTLV IIIB). The carbohydrate analysis of gp160 and gp120 and the behavior of the glycoproteins and glycopeptides derived from them on immobilized lectins demonstrate that both of these glycoproteins contain complex- and high-mannose-type Asn-linked oligosaccharides. In addition, the N-glycanase-resistant oligosaccharides of gp120 were found to contain N-acetyl-galactosamine, a common constituent of Ser/Thr-linked oligosaccharides. Pulse-chase analysis of the conversion of [35S]cysteine-labeled gp160 showed that in Molt-3 cells it takes about 2 h for gp120 to arise with a half-time of conversion of about 5 h. At its earliest detectable occurrence, gp120 was found to contain complex-type Asn-linked oligosaccharides. Taken together, these results indicate that proteolytic cleavage of gp160 to gp120 and gp41 occurs either within the trans-Golgi or in a distal compartment.

Cloning, expression, purification, and characterization of the acid alpha-mannosidase from Trypanosoma cruzi.

The acid alpha-mannosidase of Trypanosoma cruzi is a broad-specificity hydrolase involved in the catabolism of glycoconjugates, presumably in the digestive vacuole. We have cloned the alpha-mannosidase gene from a T.cruzi epimastigote genomic library. The alpha-mannosidase gene was determined to be single copy by Southern analysis, and similar sequences were not detected in genomic digests of either Trypanosoma brucei or Leishmania donovani. The coding region was subcloned into the Pichia pastoris expression vector pPICZ, and alpha-mannosidase activity was detected in the medium of induced cultures. The recombinant alpha-mannosidase demonstrated a pH optimum, inhibition by swainsonine, Km, and substrate specificity consistent with the characteristics of the alpha-mannosidase previously purified from T.cruzi epimastigotes. The recombinant enzyme was purified 103-fold from the culture medium of Pichia pastoris and had a native molecular mass of 359 kDa by gel filtration. A combination of SDS-PAGE, deglycosylation with endo H, and NH2-terminal sequencing indicates that the enzyme is originally synthesized as a homodimeric polypeptide that is subsequently cleaved to form a heterotetramer composed of 57 and 46 kDa subunits. A polyclonal antibody raised to the recombinant enzyme was shown to immunoprecipitate the alpha-mannosidase from T.cruzi cell extracts and will be used in future immunolocalization studies.

Complex-type N-linked oligosaccharides of gp120 from human immunodeficiency virus type 1 contain sulfated N-acetylglucosamine.

The major envelope glycoproteins gp120 and gp41 of human immunodeficiency virus type 1, the causative agent for human AIDS, contain numerous N-linked oligosaccharides. We report here our discovery that N-acetylglucosamine residues within the complex-type N-linked oligosaccharides of both gp120 and its precursor, gp160, are sulfated. When human Molt-3 cells persistently infected with human T-cell leukemia virus IIIB were metabolically radiolabeled with 35SO4, gp160, gp120, and to some extent gp41 were radiolabeled. The 35SO4-labeled oligosaccharides were quantitatively released by N-glycanase treatment and were bound by immobilized Ricinus communis agglutinin I, a lectin that binds to terminal beta-galactosyl residues. The kinetics of release of sulfate upon acid hydrolysis from 35SO4-labeled gp120 indicate that sulfation occurs in a primary sulfate ester linkage. Methylation analysis of total glycopeptides from Molt-3 cells metabolically radiolabeled with [3H]glucosamine demonstrates that sulfation occurs at the C-6 position of N-acetylglucosamine. Fragmentation of the gp120-derived 35SO4-labeled glycopeptides by treatment with hydrazine and nitrous acid and subsequent reduction generated galactosyl-anhydromannitol-6-35SO4, which is the expected reaction product from GlcNAc-6-sulfate within a sulfated lactosamine moiety. Charge analysis of the [3H]galactose- and [3H]glucosamine-labeled glycopeptides from gp120 and gp160 indicates that approximately 14% of the complex-type N-linked oligosaccharides are sulfated.

Isolation and characterization of a class II alpha-mannosidase cDNA from lepidopteran insect cells.

Lepidopteran insect cells are used routinely as hosts for foreign glycoprotein expression by recombinant baculoviruses, but the precise nature of their N-glycosylation pathway remains poorly defined. These cells clearly have processing glucosidases and mannosidases that can convert precursors to Man3GlcNAc2 structures and fucosyltransferases that can add fucose to the oligosaccharide core. However, their ability to extend these structures to produce complex side chains like those found in mammalian cells remains to be determined. To begin to examine this pathway at the molecular genetic level, we isolated and characterized a class II alpha-mannosidase (alpha-mannosidase II) cDNA from Sf9, a lepidopteran insect cell line. In mammalian cells, this enzyme catalyzes the committed step in the pathway converting N-linked carbohydrates to complex forms. Degenerate primers against conserved regions in known class II alpha-mannosidase protein sequences were used to generate an alpha-mannosidase II-specific PCR product from Sf9 cell DNA. Sequence information from this product was used to isolate a partial cDNA clone, the 5' end was isolated by ligation-anchored PCR, and the full length alpha-mannosidase II cDNA was assembled. This cDNA contained a long open reading frame predicted to encode an 1130 amino acid protein with 37% identity to human Golgi alpha-mannosidase II and with a type II membrane topology, a feature of all known Golgi processing enzymes. Southern blotting indicated that alpha-mannosidase II is a single copy gene in Sf9 cells. Other Lepidoptera had related alpha-mannosidase II genes, but there was variation among different genera, and the Sf9 alpha-mannosidase II cDNA did not cross-hybridize with DNA from animals outside Lepidoptera. Steady-state levels of alpha-mannosidase II RNA were low in uninfected Sf9 cells and even lower after baculovirus infection. The in vitro-translated Sf9 alpha-mannosidase II protein had the expected size and was translocated and N-glycosylated by microsomal membranes. Expression of the Sf9 alpha-mannosidase II cDNA in the baculovirus system produced large amounts of a protein with the expected size and swainsonine-sensitive alpha-mannosidase II activity towards an aryl-alpha-mannoside substrate. These results demonstrate that Sf9 cells encode and express an alpha-mannosidase II with properties similar to those of the mammalian enzyme.