The structural, biosynthetic and genetic basis of anti-horseradish peroxidase carbohydrate epitopes in Drosophila melanogaster and Caenorhabditis elegans

Research project P 13810 Neutral carbohydrate epitopes in invertebrates Iain Benedict HOWE WILSON 28.06.1999 For many years, there have been reports that antibodies raised against plant glycoproteins either IgG or IgE cross- react with some insect glycoproteins. In particular, the anti-horseradish peroxidase has become an accepted means to allow the staining of neurons in Drosophila and grasshopper. More recently, it has become apparent that a proportion of neurons in Caenorhabditis elegans can also be stained by the same method. However, in neither case has the basis of this cross-reaction been explained, even though a very high proportion of anti-horseradish peroxidase antibodies are anti-carbohydrate. To determine the basis of this specific neuronal staining, we propose to study the structure and biosynthesis of N-linked oligosaccharides in Drosophila and Caenorhabditis, with a special emphasis on seeking the glycans, and the relevant enzymes, that could account for this cross-reaction, the primary candidate being core alpha1,3-linked fucose, a modification reported in plants, insects and helminths. To this end, using immunoaffmity chromatography, we will select for the glycoproteins expressing the anti-horseradish peroxidase epitope and analyse their N-linked oligosaccharides and peptide maps by HPLC and MALDI mass spectrometry. In addition, we will seek Drosophila DNAs encoding fucosyltransferases and clone, sequence and express them, as well as assay extracts of Drosophila and Caenorhabditis for fucosyltransferase activities. We will also examine the cross-reaction between anti-horseradish peroxidase and Drosophila and Caenorhabditis proteins by inhibition with specifically-synthesised alpha1,3-fucosylated neoglycoconjugates.

Fruitflies (Drosophila melanogaster) are considered as an important model for the study of animal development. We performed some experiments with them with a view to seeing whether a protein-linked carbohydrate expressed in their neural tissue has the same or similar structure to that found in plants - the antibody (anti-horseradish peroxidase) which recognises this structure also binds some allergenic proteins of plant and insect origin. We found the structure we expected to find using analytical techniques and found an enzyme (a fucosyltransferase) encoded by one of the fly`s genes which has the capability of making the structure. This structure is recognised by another fly protein that was identified by collaborators of ours and whose properties were examined using carbohydrates that we made in the lab. We looked at a Drosophila neuronal cell line and found we could abolish the binding to anti-horseradish peroxidase by adding double-stranded RNA (encoding a bit of the fucosyltransferase) to the cells. We also looked at the protein-linked carbohydrates in the nematode Caenorhabditis elegans and found that these are rather complicated (some of the nematode`s proteins are also recognised by anti-horseradish peroxidase), a finding which was unexpected, since the worm`s glycans seem more complex than those of the fly, which is a bigger organism. In addition other Drosophila glycosyltransferases were examined, for instance a xylosyltransferase which the fly uses to start making some developmentally-important carbohydrates known as proteoglycans.