Abstract. Using monoclonal antibodies we identified a group of eight polypeptides of rat liver nuclear envelopes that have common epitopes. Most or all of these proteins are structurally distinct, as shown by tryptic peptide mapping and analysis with polyclonal antibodies. While these polypeptides are relatively tightly bound to nuclear membranes, only one is an integral membrane protein. The eight antigens cofractionate with the nuclear pore complex under various conditions of ionic strength and detergent. It can be seen by immunofluorescence microscopy that the monoclonaI antibodies reacting with these antigens stain the nuclear surface of interphase cells in a finely punctate pattern. When the nuclear envelope is disassembled and subsequently reformed during mitosis, the proteins are reversibly dispersed throughout the cytoplasm in the form of minute foci. By EM immunogold localization on isolated nuclear envelopes, the monoclonal antibodies label exclusively the nuclear pore complex, at both its nucleoplasmic and cytoplasmic margins. Considered together, our biochemical and localization data indicate that the eight nuclear envelope polypeptides are pore complex components. As shown in the accompanying paper (Holt, G. D., C. M. Snow, A. Senior, R. S. Haltiwanger, L. Gerace, and G. W. Hart, J. Cell Biol., 104:1157-1164) these eight polypeptides contain a novel form of glycosylation, O-linked N-acetylglucosamine. The relative abundance and disposition of these O-linked glycoproteins in the pore complex are consistent with their having a role in nucleocytoplasmic transport.T HE nuclear envelope forms the membrane boundary of the nucleus in eukaryotic cells (for reviews see references 13 and 14). Its principal functions include compartmentalization of nuclear metabolism and control of selective macromolecular movement between nucleus and cytoplasm. The major structural components of the nuclear envelope are inner and outer nuclear membranes, nuclear lamina, and pore complexes. The nuclear lamina (reviewed in references 16 and 33) is a meshwork of intermediate filaments lining the nucleoplasmic surface of the inner nuclear membrane, which is thought to provide a framework for the regulation of nuclear envelope structure and an anchoring site at the nuclear periphery for interphase chromosomes (18). Pore complexes constitute large supramolecular structures present at regions where inner and outer nuclear membranes are joined to form pores (2, 15), and are thought to be the major or exclusive sites of molecular exchanges across the nuclear envelope (11,13,14,39).Based on cell microinjection studies, the pore complex appears to contain an aqueous channel of ,,o10 nm in diameter through which solutes and sufficiently small macromolecules (e.g., proteins <20-40 kD) can migrate by passive diffusion (35,41). However, nncleocytoplasmic transport of larger macromolecules probably involves mediated mechanisms (7, 10). Recent work has shown that certain nuclear proteins contain discrete structural regions (6) or sequences (2...

Abstract. A novel form of protein-saccharide linkage consisting of single N-acetylglucosamine (GIcNAc) residues attached in O-linkages directly to the polypeptide backbone has been described (Holt, G. D., and G. W. Hart, 1986, J. Biol. Chem., 261:8049-8057). This modification was found on proteins distributed throughout the cell, although proteins bearing O-linked GIcNAc moieties were particularly abundant in the cytosolic and nuclear envelope fractions of rat liver. In the accompanying article (Snow, C. M., A. Senior, and L. Gerace, 1987, J. Cell. Biol., 104: 1143-1156, the authors describe monoclonal antibodies directed against eight proteins localized to the nuclear pore complex. These proteins occur on the cytoplasmic and nucleoplasmic (but not lumenal) sides of nuclear membranes. In this report, we demonstrate that all members of this group of pore complex proteins bear multiple O-linked GIcNAc residues. Further, we show that the O-linked GIcNAc moieties are linked via serine (and possibly threonine) side chains to these proteins. Perturbing the O-linked GIcNAc residues either by covalently attaching galactose to them or by releasing them with 13-N-acetylglucosaminidase strongly diminishes the immunoreactivity of the proteins with all of the monoclonal antibodies. However, the O-linked GIcNAc moieties are only part of the epitopes recognized, since O-GlcNAc-containing limit pronase fragments of nuclear pore complex proteins cannot be immunoprecipitated by these antibodies. These findings, taken together with those in the accompanying article, are a direct demonstration that proteins of the cytoplasm and nucleoplasm bear O-linked GIcNAc residues. p REVAILING evidence suggests that the bulk of the wellstudied types of carbohydrate moieties on glycoconjugates are either localized to the cell surface or within lumenal compartments of intracellular organelles (e.g., lysosomes, Golgi, and endoplasmic reticulum). In contrast, relatively little evidence has been obtained suggesting the existence of glycoproteins in the cytoplasmic and nucleoplasmic compartments of cells. Early studies involving carbohydrate analyses of nuclei and nuclear matrix fractions indicated the presence of glycoconjugates in these fractions (reviewed in Furukawa and Terayanm, 1979;Stein et al., 1981;Stoddart, 1979), and the presence of cytoplasmic glycoproteins has also been suggested (Meyer and Burger, 1976). These data have received support in several more recent investigations (e.g., Kan and Pinto da Silva, 1986;Fedarko and Conrad, 1986;Nagakura et al., 1986). However, attempts to identify and characterize nucleoplasmic and cytoplasmic glycoproteins have been severely hampered by cross-contamination with other cellular compartments, and detailed biochemical analyses of most putative cytoplasmic and nucleoplasmic glycoproteins is lacking.Recently the subcellular distribution of glycoproteins bearing terminal N-acetylglucosamine (GIcNAc) ~ residues was surveyed (Holt and Hart, 1986) by employing galactosyltransferase, an enzyme that trans...

The human salivary amylase genes are associated with two inserted elements, a ~/-actin-processed pseudogene and an endogenous retroviral-like element. To test the contribution of these inserted elements to tissue specificity, 25 lines of transgenic mice carrying 10 amylase constructs were established. A 1-kb fragment of AMY1C (-1003 to +2) was found to be sufficient for parotid-specific expression of a human growth hormone reporter gene. The 1-kb fragment is entirely derived from inserted sequences. Deletion from -1003 to -826 resulted in reduced levels of transgene expression and loss of tissue specificity. The fragment -1003 to -327 was sufficient to transfer parotid specificity to the thymidine kinase promoter. The data demonstrate that the functional tissue-specific promoter of human AMY1C is derived from inserted sequences and that parotid expression can be conferred by sequences derived solely from the retrovirus. A role for retrotransposition in the evolution of gene regulation is indicated by these and other recent observations.

We have analyzed the junction regions of inserted elements within the human amylase gene complex. This complex contains five genes which are expressed at high levels either in the pancreas or in the parotid gland.The proximal 5'-flanking regions of these genes contain two inserted elements. A -y-actin pseudogene is located at a position 200 base pairs upstream of the first coding exon. All of the amylase genes contain this insert. The subsequent insertion of an endogenous retrovirus interrupted the -y-actin pseudogene within its 3'-untranslated region. Nucleotide sequence analysis of the inserted elements associated with each of the five human amylase genes has revealed a series of molecular events during the recent history of this gene family. The data indicate that the entire gene family was generated during primate evolution from one ancestral gene copy and that the retroviral insertion activated a cryptic promoter.The complete human amylase gene cluster has recently been cloned in our laboratory and by others (8,9,11,12,21). The cluster spans 230 kilobases (kb) on human chromosome 1. The most common human haplotype contains five intact genes: two pancreatic genes (AMY2A and AMY2B) and three salivary genes (AMYIA, AMYIB, and AMYIC) (8, 9). The salivary and pancreatic genes are 93% identical in the sequence of the proximal 718 base pairs (bp) of the 5'-flanking regions (12). Nonetheless, these genes are expressed with different tissue specifities. AMY2A and AMY2B transcripts are present in the human pancreas but not in the parotid salivary gland, while AMY] transcripts are present in the parotid gland but not in the pancreas (28).Two inserted elements have been identified in the 5' regions of the human amylase genes. A -y-actin pseudogene was identified 200 bp upstream of the first amylase-coding exon (4, 28). The promoter and nontranslated exon of the salivary amylase genes are derived from the y-actin pseudogene. Retroviral long terminal repeat (LTR) and gag sequences were also identified in the 5'-flanking region of an amylase gene (4). We subsequently determined the locations of a total of seven LTRs within the gene cluster (28). The mouse amylase gene cluster does not contain retroviral or -y-actin-related sequences (unpublished observations), indicating that insertion of both elements occurred subsequent to the separation of human and mouse.In the present report, we describe the junction regions of the y-actin and retroviral inserts upstream of each human amylase gene. These inserts provide unique molecular markers for the recent evolution of this gene family. The structural data suggest a specific sequence of events by which the five human amylase genes were generated from one ancestral gene within the past 40 million years. MATERIALS AND METHODSSequence analysis. Genomic DNA fragments from cosmids G6 (AMY2B), G45 Restriction fragments from cosmid G21 (9) were subcloned into pGEM 1 or 2 for preparation of riboprobe templates. The amylase probe was a 470-nucleotide HaeIII-EcoRI fragment containing exon...

We have analyzed the junction regions of inserted elements within the human amylase gene complex. This complex contains five genes which are expressed at high levels either in the pancreas or in the parotid gland. The proximal 5'-flanking regions of these genes contain two inserted elements. A gamma-actin pseudogene is located at a position 200 base pairs upstream of the first coding exon. All of the amylase genes contain this insert. The subsequent insertion of an endogenous retrovirus interrupted the gamma-actin pseudogene within its 3'-untranslated region. Nucleotide sequence analysis of the inserted elements associated with each of the five human amylase genes has revealed a series of molecular events during the recent history of this gene family. The data indicate that the entire gene family was generated during primate evolution from one ancestral gene copy and that the retroviral insertion activated a cryptic promoter.