Development of a novel biological toolbox for the synthesis of glycosides, oligo- and polysaccharides from glucose and/or succrose - a focus is on the synthesis of HMOs (Glycoside production).

Saccharides and glycosides are a diverse group of naturally occurring substances with a wide range of applications and a multi-billion € market. Health-promoting functions have been identified for some specific oligosaccharides. For example, human breast milk contains oligosaccharides (human milk oligosaccharides, HMO), which support the immune system and brain development of the breast-fed child and are therefore advantageously used in infant formula. Due to their probiotic and immunostimulant effects, they are also applied as dietary supplements.

The big challenge for this product class is an economic synthesis of these compounds. Especially due to the high complexity of sugar chemistry and the necessity of protection groups to prevent the formation of side products, stereospecific, biological production systems are superior to those in the chemical industry. Two biological manufacturing processes that take advantage of the high selectivity and specificity of enzymes are possible. The oligosaccharides can be produced in vivo, i.e. fermentatively with the help of microorganisms, or in vitro using isolated enzymes.

In this project, a novel enzyme platform for glycosyltransferases (GTs) will be established with the support of innovative in silico methods, such as molecular dynamics analysis and machine learning for structure and function prediction, as well as biofoundry based automated DNA assembly and combinatorial domain assembly, together with high throughput screening. By specifically modifying natural enzymes, GTs with new properties will be created and tested for industrial use. To ensure an economic production of glycosides, suitable in vitro methods for the activation of the applied monosaccharides will be developed as another aspect of the project. Further, in vivo production strains will be metabolically engineered. Finally, the established procedures will be compared with each other.

The production of two target structures will be used in the study as a proof of concept for the different principles. The created enzyme-platform shall then be the basis for the synthesis of diverse sugar structures, which are currently not economically accessible neither from nature nor via chemical synthesis.

The Professorship of Bioinformatics of Prof. Dr. Dominik Grimm (HSWT) at the TUM Campus Straubing has been working for several years on intelligent algorithms and machine learning methods to gain a better understanding of the underlying architecture of complex processes for the bioeconomy. In this project, the research group explores novel AI methods for the analysis of complex bio-chemical data.