Activity and tissue distribution of splice variants of alpha6-fucosyltransferase in human embryogenesis

Fucosylation is associated with the malignant transformation of intraductal papillary mucinous neoplasms: a lectin microarray-based study

PURPOSE

Intraductal papillary mucinous neoplasm (IPMN) is an intraductal mucin-producing pancreatic neoplasm with the potential for malignant transformation. Changes in glycans expressed on the cell surface and glycotransferases play important roles in malignant transformation. We conducted this study to analyze glycan alterations in IPMNs by using a lectin microarray and to identify the factors associated with altered glycans and their relationships with malignant transformation.

METHODS

Using a lectin microarray, we evaluated glycan expression in 22 samples of IPMN with carcinoma, obtained from curative resections performed in our department. We also used immunohistochemistry to investigate fucosyltransferase 8 (Fut 8) protein expression, which is associated with glycan alterations in IPMNs.

RESULTS

The lectin microarray demonstrated that only two lectins, Aleuria aurantia lectin (AAL) and Aspergillus oryzae L-fucose-specific lectin (AOL), which bind to fucose, exhibited significant sequential increases from normal pancreatic duct to adenoma and carcinoma. Similarly, Fut 8 protein expression, which is associated with AAL and AOL, sequentially and significantly increased from the normal pancreatic duct to adenoma and carcinoma.

CONCLUSIONS

Lectin microarray analysis suggested that fucosylation is associated with the malignant transformation of IPMNs.

Molecular cloning, characterization, genomic organization and promoter analysis of the α1,6-fucosyltransferase gene (fut8) expressed in the rat hybridoma cell line YB2/0

Background

The rat hybridoma cell line YB2/0 appears a good candidate for the large-scale production of low fucose recombinant mAbs due to its lower expression of fut8 gene than other commonly used rodent cell lines. However, important variations of the fucose content of recombinant mAbs are observed in production culture conditions. To improve our knowledge on the YB2/0 fucosylation capacity, we have cloned and characterized the rat fut8 gene.

Results

The cDNAs encoding the rat α1,6-fucosyltransferase (FucT VIII) were cloned from YB2/0 cells by polymerase chain reaction-based and 5' RNA-Ligase-Mediated RACE methods. The cDNAs contain an open reading frame of 1728 bp encoding a 575 amino acid sequence showing 94% and 88% identity to human and pig orthologs, respectively. The recombinant protein expressed in COS-7 cells exhibits a α1,6-fucosyltransferase activity toward human asialo-agalacto-apotransferrin. The rat fut8 gene is located on chromosome 6 q and spans over 140 kbp. It contains 9 coding exons and four 5'-untranslated exons. FISH analysis shows a heterogeneous copy number of fut8 in YB2/0 nuclei with 2.8 ± 1.4 mean copy number. The YB2/0 fut8 gene is expressed as two main transcripts that differ in the first untranslated exon by the usage of distinct promoters and alternative splicing. Luciferase assays allow defining the minimal promoting regions governing the initiation of the two transcripts, which are differentially expressed in YB2/0 as shown by duplex Taqman QPCR analysis. Bioinformatics analysis of the minimal promoter regions upstream exons E-2 and E-3, governing the transcription of T1 and T2 transcripts, respectively, evidenced several consensus sequences for potential transcriptional repressors. Transient transfections of Rat2 cells with transcription factor expression vectors allowed identifying KLF15 as a putative repressor of T1 transcript in Rat2 cells.

Conclusion

Altogether, these data contribute to a better knowledge of fut8 expression in YB2/0 that will be useful to better control the fucosylation of recombinant mAbs produced in these cells.

Reaction mechanism and substrate specificity for nucleotide sugar of mammalian alpha1,6-fucosyltransferase--a large-scale preparation and characterization of recombinant human FUT8

FUT8, mammalian alpha1,6-fucosyltransferase, catalyzes the transfer of a fucose residue from the donor substrate, guanosine 5'-diphosphate (GDP)-beta-L-fucose, to the reducing terminal GlcNAc of the core structure of asparagine-linked oligosaccharide via an alpha1,6-linkage. FUT8 is a typical type II membrane protein, which is localized in the Golgi apparatus. We have previously shown that two neighboring arginine residues that are conserved among alpha1,2-, alpha1,6-, and protein O-fucosyltransferases play an important role in donor substrate binding. However, details of the catalytic and reaction mechanisms and the ternary structure of FUT8 are not understood except for the substrate specificity of the acceptor. To develop a better understanding of FUT8, we established a large-scale production system for recombinant human FUT8, in which the enzyme is produced in soluble form by baculovirus-infected insect cells. Kinetic analyses and inhibition studies using derivatives of GDP-beta-L-fucose revealed that FUT8 catalyzes the reaction which depends on a rapid equilibrium random mechanism and strongly recognizes the base portion and diphosphoryl group of GDP-beta-L-fucose. These results may also be applicable to other fucosyltransferases and glycosyltransferases.

Characterization of the LARGE family of putative glycosyltransferases associated with dystroglycanopathies

The Large(myd) mouse has a loss-of-function mutation in the putative glycosyltransferase gene Large. Mutations in the human homolog (LARGE) have been described in a form of congenital muscular dystrophy (MDC1D). Other genes (POMT1, POMGnT1, fukutin, and FKRP) that encode known or putative glycosylation enzymes are also causally associated with human congenital muscular dystrophies. All these diseases are associated with hypoglycosylation of the membrane protein alpha-dystroglycan (alpha-DG) and consequent loss of extracellular ligand binding. Hence, they are termed dystroglycanopathies. A paralogous gene for LARGE (LARGE2 or GYLTL1B) may also have a role in DG glycosylation. Using database interrogation and reverse-transcriptase polymerase chain reaction (RT-PCR), we identified vertebrate orthologs of each of these LARGE genes in many vertebrates, including human, mouse, dog, chicken, zebrafish, and pufferfish. However, within invertebrate genomes, we were able to identify only single homologs. We suggest that vertebrate LARGE orthologs be referred to as LARGE1. RT-PCR, dot-blot, and northern analysis indicated that LARGE2 has a more restricted tissue-expression profile than LARGE1. Using epitope-tagged proteins, we show that both LARGE1 and LARGE2 localize to the Golgi apparatus. The high similarity between the LARGE paralogs suggests that LARGE2 may also act on DG. Overexpression of LARGE2 in mouse C2C12 myoblasts results in increased glycosylation of alpha-DG accompanied by an increase in laminin binding. Thus, there may be functional redundancy between LARGE1 and LARGE2. Consistent with this idea, we show that alpha-DG is still fully glycosylated in kidney (a tissue that expresses a high level of LARGE2 mRNA) of Large(myd) mutant mice.

Fucosyltransferase substrate specificity and the order of fucosylation in invertebrates

Core alpha1,6-fucosylation is a conserved feature of animal N-linked oligosaccharides being present in both invertebrates and vertebrates. To prove that the enzymatic basis for this modification is also evolutionarily conserved, cDNAs encoding the catalytic regions of the predicted Caenorhabditis elegans and Drosophila melanogaster homologs of vertebrate alpha1,6-fucosyltransferases (E.C. 2.4.1.68) were engineered for expression in the yeast Pichia pastoris. Recombinant forms of both enzymes were found to display core fucosyltransferase activity as shown by a variety of methods. Unsubstituted nonreducing terminal GlcNAc residues appeared to be an obligatory feature of the substrate for the recombinant Caenorhabditis and Drosophila alpha1,6-fucosyltransferases, as well as for native Caenorhabditis and Schistosoma mansoni core alpha1,6-fucosyltransferases. On the other hand, these alpha1,6-fucosyltransferases could not act on N-glycopeptides already carrying core alpha1,3-fucose residues, whereas recombinant Drosophila and native Schistosoma core alpha1,3-fucosyltransferases were able to use core alpha1,6-fucosylated glycans as substrates. Lewis-type fucosylation was observed with native Schistosoma extracts and could take place after core alpha1,3-fucosylation, whereas prior Lewis-type fucosylation precluded the action of the Schistosoma core alpha1,3-fucosyltransferase. Overall, we conclude that the strict order of fucosylation events, previously determined for fucosyltransferases in crude extracts from insect cell lines (core alpha1,6 before core alpha1,3), also applies for recombinant Drosophila core alpha1,3- and alpha1,6-fucosyltransferases as well as for core fucosyltransferases in schistosomal egg extracts.