(719c) An Artificial Golgi Reactor As an Alternative Method for Targeted Cell-Free Glycosylation

E Makrydaki¹, I Moya Ramirez¹, A Krueger ²,S Haslam², KM Polizzi¹ and C Kontoravdi¹.

¹ Department of chemical engineering, Imperial College London

² Faculty of Natural Sciences, Department of Life Sciences, Imperial College London

Glycosylation of therapeutically relevant proteins such as monoclonal antibodies (mAbs) is critical for high drug efficiency, efficacy and half-life. Furthermore, targeted glycosylation is of great importance in vaccine manufacturing where glycan-recognizing proteins such as lectins are used. Modern biotherapeutic production focuses on controlling the protein glycosylation profile using various methods.

Currently, the dominating method is the traditional cell-line engineering of host cells. The main goal is to produce mAbs with a desired and human-like glycosylation pattern. However, this approach often struggles due to high sensitivity to the fermentation environment making it difficult to scale up and control. The latter, can lead to structural heterogeneity amongst the products.

In addition to in vivo methods, in vitro techniques exist such as chemoenzymatic modification. However, they are often limited by the difficult and time-consuming implementation. Finally there are various in vitro enzymatic glycosylation methods but can struggle with product heterogeneity due to lack of control over the enzymatic reactions.

Glycan heterogeneity on therapeutic proteins is an important bottleneck to the control of glycosylation. One of the reasons lies with the enzymes regulating glycosylation e.g. glycosyltransferases and mannosidases, as they can compete against the same substrate. This enzyme promiscuity is responsible for a broad network of potential glycoforms in both in vivo and in vitro protein production methods. Modern cell-line engineering narrows this network however homogeneity is not ensured.

In line with the need to control glycosylation and enhance product quality, we propose an artificial Golgi reactor. The enzymes regulating a desired glycosylation pathway are spatially separated in microfluidic channels. This allows strict control over the enzymatic reactions. Furthermore, the glycan modifications can occur in a modular fashion addressing enzyme promiscuity. Such a system can be used to complement traditional cell-based production of therapeutic proteins to enhance the desired glycosylation profile.

As discussed, fabricating an artificial Golgi reactor requires expressing selected glycosyltransferases and spatially separating them. This can be achieved by immobilization on solid supports. Because enzymes are immobilized, we can achieve a one-step protein purification/immobilization. Furthermore, the modularity of our design makes the system more sustainable and easily tailored for each application.

Herein, to demonstrate that spatial separation can address enzyme promiscuity we have selected a glycosylation pathway consisting of three enzymes: Nicotiana Tabacum GnTI (NtGnTI), Drosophila Melanogaster ManII (dGMII) and human GalT (hGalT), where dGMII and hGalT compete against NtGntI’s product. We have achieved expression and in vivo biotinylation of Nicotiana Tabacum GnTI (NtGnTI) and human GalT in E. coli. dGmII was chemically biotinylated. All the biotinylated enzymes were bound to streptavidin beads in a one-step purification/immobilization. Furthermore they were successfully reacted in a sequential fashion following the enzymatic pathway of NtGnTI-dGMII-hGalT for artificial glycan synthesis. Results were confirmed with MALDI/TOF MS analysis. Currently, we are working on using the enzymes to glycosylate a therapeutically relevant protein.