Modeling

A modular dockerin–cohesin-based surface-display system was constructed to co-localize lytic polysaccharide monooxygenase (LPMO), laccase (Lac), and versatile peroxidase (VP) on the cell surface of Saccharomyces cerevisiae. Each enzyme was fused to a dockerin module originating from a different organism; the Enzyme-Dockerin structure was then assembled onto a single protein scaffold through highly specific cohesin–dockerin interactions, thereby enabling the presentation of the three enzymes.

To obtain three cohesin–dockerin pairs, the cellulosome scaffoldins CipA, ScaB, and CipC were noticed, which are derived from Acetivibrio thermocellus, Ruminococcus flavefaciens, and Ruminiclostridium cellulolyticum, respectively. Each of these scaffoldins contains multiple cohesin domains that recognize their corresponding dockerins in a species-specific manner.

Prior to any wet-lab work, AlphaFold3 was employed to perform a comprehensive structural assessment of the intended fusions and their pairwise interactions, thereby evaluating the molecular feasibility of cohesin domains within CipA, ScaB, and CipC:

1. Evaluation of cohesin–dockerin interactions

AlphaFold3 models of individual cohesin and dockerin domains, as well as their heterodimeric complexes, were generated and analysed. The typical interaction interface—comprising α-helices and β-sheet core—was inspected for preservation of key structures and polar interactions. Retention of these structures indicated that the proposed scaffold can probably support stable and specific modular assembly.

2. Assessment of the structure of the enzyme–dockerin

Fusion proteins linking each enzyme to its respective dockerin were modelled and compared with the corresponding native enzymes. The analysis of probable substantial changes of the fold or active-site architecture was conducted in order to ensure that the fusion of eventually chosen cohesins with the enzymes should not interfere with the functions of the enzymes.

On the basis of these analyses, the second cohesin domain of CipA (CipAcoh2), the fourth cohesin domain of ScaB (ScaBcoh4), and the second cohesin domain of CipC (CipCcoh2) were selected as the cohesins used in the project.

The resulting triplet of cohesin domains (CipAcoh2, ScaBcoh4, CipCcoh2) is expected to direct the self-assembly of the three enzyme–dockerin conjugates on the yeast surface without influences on their structure or activity.

This modeling exercise provided a pre-experimental, structure-based validation of the design, minimizing developmental risk and reducing the number of iterative trials. Beyond the immediate project, the established workflow offers a general blueprint for the rational construction of future multi-enzyme surface-display platforms.