Research Highlights

Glycomics: Drosophila's simple sweetness

Lectin staining reveals O- and N-glycans present in larval imaginal discs of Drosophila. Pictures taken from Tian and Ten Hagen (2007). Copyright © 2007 Oxford University Press.

In mice, humans and other higher animals, the activity of many glycosyltransferases is redundant: the enzymatic activities of one transferase can be replaced by others. This feature renders it difficult to attribute a specific function to a specific glycosyltransferase even when studying a single, double or even triple mouse knockout. In contrast to this, Drosophila melanogaster has much less glycosyltransferase redundancy and less complex glycan structures. Such restricted enzyme and product diversity makes the fly a highly interesting model organism for the study of glycosyltransferase knockouts and the analysis of glycan repertoires. Recently, several functional glycobiology studies have shed light on carbohydrate function and diversity in Drosophila.

In Proteomics, Schwientek et al. analyzed the mucin-type O-glycoproteome of Drosophila S2 cells, a hemocyte-like cell line. Mucin-type O-glycosylation is the most abundant type of glycan modification and occurs at protein serines or threonines. In contrast to N-glycosylation, which occurs at asparagines of N-X-S/T sequons, mucin-type O-glycosylation cannot be predicted based on peptide sequence, mainly because it occurs after protein folding. The O-glycans of Drosophila are less complex and more uniform than their mammalian counterparts and therefore permit glycan analysis based on carbohydrate recognition by only a few lectins. Schwientek et al. identified mucin-type O-glycans on 22 proteins. The disaccharide galactose-N-acetylgalactosamine and the monosaccharide N-acetylgalactosamine (T- and Tn-antigen) were the predominant structures that were identified. Among the proteins that were shown to carry mucin-type O-glycans were extracellular matrix (ECM) proteins such as peroxidasin, ECM-modifying metalloproteases, cell-surface proteins such as midline fasciclin, and serine proteases. Furthermore, an intracellular heat shock protein was shown to possess mucin-type O-glycans. The results of this study - gained by combining serial lectin affinity chromatography with mass spectrometry and 2D gel electrophoresis – are a step forward in the mapping of the entire Drosophila O-glycoproteome.

In another Drosophila glycomics study published in Glycobiology, Koles et al. studied N-glycosylated proteins from the fly brain. While the authors identified a large group of cell adhesion molecules which were N-glycosylated, the largest group of N-glycosylated proteins was composed of molecules with unknown function. In addition, the authors established that the G-protein coupled receptor protein rhodopsin is N-glycosylated at two sites. Furthermore, the authors found that Drosophila integrins and other cell adhesion molecules were glycosylated. This lends support to an earlier hypothesis that interaction between heterophilic cell adhesion molecules might involve glycan-lectin interactions. Lastly, a role for glycans in neuron function was indicated by the authors' finding that glutamate channel proteins were N-glycosylated. While the channels were shown to work without glycan modification, the N-glycans were shown to be required for their interaction with proteins that modify their function. Lastly, Koles et al. elucidated structural information about the dynamics of N-glycosylation, such as a preference of the N-X-T over the N-X-S sequence for the modification, which hints that the threonine residue may confer a conformational advantage for N-glycosyltransferase enzymatic activity. It will be interesting to integrate the data of Schwientek et al. and Koles et al. with the developmental glycomics data provided by Tian and ten Hagen and other studies.

While these studies greatly enhanced knowledge about the pool of glycoproteins in Drosophila, two recent functional studies examined the glycosyltransferase repertoire in the fly in an effort to elucidate mammalian glycosylation. Haines et al. in Molecular and Cellular Biology showed that fly embryos lacking protein O-mannosyltransferases (POMTs) exhibited undermannosylation of dystroglycan, which led to severe defects in muscle development. Thus, the Drosophila POMT knockout appears to provide a valuable model for studying human dystroglycan mannosylation deficiencies, which are known to cause muscle atrophies. Another study by Stolz et al. in the Glycoconjugate Journal focused on glycolipid synthesis in Drosophila and showed that one (4GalNAcT-B) of the three 1,4-N-acetylgalactosamine transferases (4GalNAcTs)  generated the majority of the glycolipid structures in the fly. Accordingly, a 4GalNAcT-B knock out led to a more severe phenotype than a knockout of one of the other two transferases. Thus, among the six mammalian glycolipid 1,4-galactosyltransferases that are evolutionary and genetically related to the Drosophila 4GalNAcTs might as well be one that is the chief player in glycolipid synthesis.

Taken together, the four studies presented here highlight the value of Drosophila for the study of glycosylation function and greatly enhance our knowledge about the diversity and function of glycosylated proteins in the fly.

Author: Mirko von Elstermann