Eukaryotic Expression and Purification of Recombinant Extracellular Matrix Proteins Carrying the Strep II Tag - Methods in Molecular Biology: Extracellular Matrix Protocols - page 49

Crystallization of the peptide with its ligand showed that it bound to the
same pocket as biotin, though in a more superficial position. The original
peptide WRHPQFGG only bound strongly when placed C-terminally, as the
final Gly-Gly-COO - was used to form a salt bridge with the streptavidin mol-
ecule. By mutating the penultimate glycine to a glutamic acid, which can donate a
COO - to form such a bridge, a new form of the peptide was produced that is
active both C- and N-terminally (6) . Further mutations led to the formation of
an optimized binding sequence of WSHPQFEK (Strep II) (6) . Random muta-
tion of the gene for streptavidin has since produced a form of this protein
“StrepTactin,” which binds to the Strep II tag also highly specifically but with
a greater affinity (7 ). Binding can be competed with biotin or its derivatives
leading to the elution of the Strep-tagged protein under gentle conditions giv-
ing a native protein. Whereas the biotin-StrepTactin interaction is for practical
purposes irreversible, desthiobiotin, a homolog of biotin, which can also dis-
place the Strep tag from StrepTactin, binds relatively weakly to StrepTactin. It
may then be eluted simply allowing the reuse of the StrepTactin matrix.
We have produced two vectors CMVNstrep and CMVCstrep with the same
frame at the 5' end ( Fig. 1 ) for N- or C-terminal tagging. cDNA is amplified by
PCR with primers containing modified ends. The 5' end of the sense primer
carries a Spe I, Nhe I, or Xba 1 site placed to bring the codon alignment of the
amplified fragment in frame with the signal peptide (+/- tag). On the antisense
primer a stop codon is followed by a Not 1 site in the case of CMVNstrep and a
Not 1 site, bringing the linking alanines of the tag into frame in the case of the
CMVCstrep. The choice of the primers should be checked carefully, first, to
keep frame and second, as the selection of the wrong domain borders will prob-
ably lead to misfolding of the expressed protein and possibly lead to its intrac-
ellular degradation. Where the borders are unknown, we suggest making a
series of constructs of increasing size integrating successive likely border cys-
teines and leaving a small linker region between the cysteine and the tag or
signal peptide. The PCR is carried out on preexisting cDNA using a proofread-
ing enzyme to limit polymerase-induced errors and, after cloning into the rel-
evant vector, it is sequenced in its entirety. We transfect by electroporation,
which is simple and efficient and we generally use 293-EBNA cells to obtain
episomal plasmid replication.

2. Materials
2.1. Media for 293-EBNA Cells
1.

Serum-free media; Dulbeccos MEM-nutrient mix F-12 (Gibco, Gaithersburg,
MD) with final concentrations of 2 m M glutamine (Gibco) and 100 U/mL peni-
cillin, 100 µ g/mL streptomycin (Gibco).