Glyco-engineering in plants : Fucosyltransferase characterisation and application for the production of immunomodulatory helminth glycoproteins

Summary

Over the last two decades plants have been used for the production of relative high amounts of (potential) pharmaceutical proteins. The latter includes cases where isolation from the original source was not possible, laborious and/or resulted in low yields. Currently, almost all biopharmaceutical proteins originate from humans, such as antibodies or hormones. Many of these human proteins are glycoproteins implying that they are ‘coated’ with sugars (glycans). These glycans can be important for protein folding and/or activity, and the composition of the glycans can vary between organisms. Hence, in order to produce a glycoprotein in another organism with (close to) native glycans, the glycosylation pathway has to be comparable or has to be engineered.

Various glyco-engineering studies have shown the possibilities to “humanise” the plant glycosylation pathway, enabling the production of human glycoproteins with native glycans in plants. However, other research fields could benefit from the development of this production platform as well. For example, studies on glycosylated vaccine candidates and the immunomodulatory properties of helminth secretions, show the high potential of these helminth proteins as pharmaceuticals. High amounts of native helminth protein are required for vaccine development and research on the biological and biopharmaceutical properties of helminth secretions. The helminth proteins for such research projects cannot be isolated in sufficient quantities from the helminth or its secretions. Moreover, many of these proteins are glycosylated with glycans that cannot be mimicked in current recombinant production systems. For this purpose, the focus in this thesis is on glyco-engineering of plants to establish a production system for native helminth glycoproteins.

Helminths of the genus Schistosoma infect 252 million people worldwide. Research to Schistosoma mansoni has shown that S. mansoni secretes immunomodulatory proteins during its life cycle to influence the host immune system. These secreted proteins are often glycosylated and the glycans have shown to play an important role in immune modulation. To study the immunomodulatory properties of these proteins and the effect of their glycans, high amounts of protein are required, which cannot be isolated from the worm or its secretions. Current recombinant production systems cannot mimic the highly fucosylated N-glycans of S. mansoni. To mimic helminth glycosylation in the expression host, glycosyltransferases have to be co-expressed that synthesise the required glycan motifs. Although many putative S. mansoni fucosyltransferases (SmFucTs) are known, they are poorly characterised. Therefore, we focussed on SmFucT characterisation in Nicotiana benthamiana by localisation and functional studies. We showed subtle differences in Golgi localisation between different SmFucTs and identified SmFucTs that are involved in N-glycan core α1,3- or α1,6-fucosylation or the synthesis of antennary LeX, LDN-F or F-LDN-F. These functionally characterised SmFucTs can directly be applied to synthesise complex helminth N-glycan motifs on recombinant glycoproteins in N. benthamiana, to study the immunomodulatory properties of these glycoproteins and the effect of their glycans.

Next to, N. benthamiana as enzyme characterisation platform we focussed on N. benthamiana as general protein production platform. Plants have shown to be a promising host for the production of biopharmaceuticals. However, plant N-glycans can differ from helminth or human N-glycans. Typical plant N-glycans have an α1,3-fucosylated and β1,2-xylosylated core. The enzymes that add an α1,3-fucose to the N-glycan core in N. benthamiana are core α1,3-FucTs (NbFucTs). So, in order to generate a plant without core α1,3-fucose on its N-glycans these NbFucTs need to be knocked out. N. benthamiana is an allotetraploid and therefore multiple sequence variants of one protein can be present. Studies have already focused on the production of NbFucT knockout plants by targeting all or just a few NbFucT sequence variants. However, time and effort can be saved when targeting only the NbFucT sequence variants that are active. Therefore, we looked into the sequence variation present in our N. benthamiana plants. Ten different NbFucT sequences were retrieved, of which six encode full-length proteins. Of these six proteins only three were able to introduce a core α1,3-fucose. This knowledge allows the design of specific guide RNAs, which need to target only these three sequences for the production of knockout plants with CRISPR-Cas9 or similar technologies. Thereby saving time, effort and resources.

The gathered information on enzyme characterisation was used to produce the helminth glycoproteins omega-1 and kappa-5 in N. benthamiana. High amounts of omega-1 and kappa-5 with native N-glycan motifs could be isolated and purified in/from N. benthamiana leaves. The plant glycosylation pathway was modified by co-expression of various glycosyltransferases to established production of omega-1 and kappa-5 with native N-glycan motifs. These data indicate that mimicking the complex carbohydrate structures of helminths in plants is a promising strategy to create essential tools for further development of helminth glycoproteins as (potential) therapeutic glycoproteins.

Altogether, the results presented in this thesis show the suitability of plants as both a platform for characterisation of novel glycosyltransferases and the production of helminth glycoproteins carrying native N-glycan motifs. It has been shown before that plants are a promising platform for the production of human glycoproteins, but here we show new possibilities for plant molecular farming to be applied in the field of (parasite) glycobiology, parasitology and immunology.