Protein design is usually limited to 20 amino acids. Since the chemical abilities of these amino acids are limited, a lot of interesting functions are not applicable in protein design. We aim to enable the translational incorporation of non-canonical amino acids (ncAAs) through an expanded genetic code. First, we repurposed the rarely used amber stop codon for translational incorporation. Therefore, we adjusted aminoacyl-tRNA synthetases (aaRS) which charge the ncAA to the tRNA. We designed a library of aaRS and a suitable selection system for the development of customized aaRS which specifically incorporate the ncAA of interest. As a proof of concept, we applied different ncAAs e.g. for labeling, photoswitching, and photolysis of proteins. To further improve the translational incorporation of ncAAs, we explored a new way to expand the genetic code by development of a semisynthetic organism. An additional unnatiral base pair (UBP) generates 61 additional codons. Therefore, we improved an UBP uptake system and identified and characterized a biosynthesis pathway for in vivo production of isoG in Escherichia coli. We engineered a retention system and and established detection methods for the UBP.