Correlations between sequences and production (THESIS VERSION)

A set of 70 variable domains of heavy chain Camelid antibodies (VHHs) were grown under identical conditions in yeast. The production yields in Saccharomyces cerevisiae ranged between nearly 0 and 150 mg/l. To understand these large differences in production of VHHs, even those recognizing the same antigen or those with a high sequence similarity, we searched for correlations between a series of aspects of the sequences and features of 3D structures and the production yield. Several parameters on DNA sequence and protein sequence level were analysed using simple bioinformatics techniques. No trivial correlations were found. We did notice many non-optimal codons. Visual inspection of the multiple sequence alignment revealed an enrichment of certain amino acid types at certain positions in the poorly produced VHHs. These findings resulted in the studies described in more detail in chapter 5