Drug-resistant selectable markers are convenient for transformation of expression sequences into Pichia pastoris since they do not require a specific mutation in the host strain for selection as all biosynthetic gene/auxotrophic host marker systems do. Four such drug resistant marker genes have been described: 1) the E. coli Tn903kanR kanamycin/geneticin/G418 resistance gene (KanR) (Scorer et al., 1994), 2) the Streptoalloteichus hindustanus She ble gene conferring resistance to the bleomycin-related drug Zeocin (ZeoR) (Higgins et al. 1998), 3) the Streptomyces griseochromogens gene, conferring resistance to the nucleoside antibiotic Blasticidin S (BsdR) (www.invitrogen.com), and 4) the (Tom: please add info on the NAT gene resistance gene here). Once isolated, the drug resistant strains are stable - provided one follows standard microbial storage and manipulation procedures - and do not require further drug selection, as for example, during growth of transformed strains in shake flask or fermentor cultures.
At present, vectors containing the ZeoR gene are very popular due in part to their small size, a consequence of the ZeoR gene being the selectable marker for transformations of both E. coli and P. pastoris in these vectors. In addition, ZeoR vectors are convenient for enrichment of strains containing multiple copies of the vector. With the ZeoR gene, the level of resistance to zeocin roughly correlates with vector copy number. However, it is important to note that vector copy number varies greatly, and most transformants still contain only a single vector copy, even at high levels of drug. Thus, this procedure is an “enrichment” not a “selection”. A significant number (50–100) transformants must be subjected to further analysis of copy number and/or expression level to identify the one or two with a high copy number. By this approach, strains carrying up to 30 copies of an expression cassette have been isolated (Scorer et al. 1994).