Page 40 - Macromolecular Crystallography

HIGH-THROUGHPUT CLONING, EXPRESSION, AND PURIFICATION 29

of the fusion protein is enabled by adding an N-His that the cell-specific productivity remains the same
tag to the Ubl (Catanzariti et al., 2004; Malakhov under different culture conditions. Most small-scale
et al., 2004). The Ubl can be removed from the tar- expression screens are carried out in 96- or 24-well,
get protein in vitro using specific deubiquitylating deep-well plates using enriched complex media, for
enzymes now available from recombinant sources. example TB, 2YT, and GS96 (QBiogene), to ensure
The cleavage occurs after the final glycine residue maximum biomass. Typically, these media sup-
at the carboxy-terminus of the Ubl protein irrespec- port growth to optical densities (OD) of 5–10 OD 600
tive of the amino acid that immediately follows, thus units compared to 2–3 OD 600 units in standard Luria
regenerating the native N-termini on removal of the broth (LB). Two options are available for inducing
fusion partner. expression from T7 vectors either by addition of
IPTG (range 0.1–1.0 mM) or by using a formulation
2.3 Expression screening of glucose and lactose in the media which leads to
autoinduction (Studier, 2005). In the first method,
Amajor feature of HTP protein production pipelines the time of induction and therefore the growth state
is the inclusion of a screening step at a relatively of the culture (usually midlog phase) can be prede-
small scale to identify constructs suitable in terms termined, whereas in the second method induction
of soluble protein yield for scale-up and subsequent occurs usually in late log stage and cannot be con-
protein purification. Given the relatively high cost trolled. However, a major operational advantage of
of the latter steps in terms of time and resources, the autoinduction method is that once the cultures
particularly if eukaryotic expression is required, the have been set up no further manipulation is required
screening stage is seen as crucial to the overall pro- prior to harvest and expression evaluation. Further,
cess. In this section the approaches to the evaluation in our experience autoinduction can lead to over-
of expression at small scale will be reviewed. all higher levels of expression. Another important
parameter in the expression of recombinant proteins
2.3.1 E. coli in E. coli is the culture temperature; reducing the
temperature from 37 Cto20 C or even lower has
◦
◦
2.3.1.1 Choice of strains and media beenfoundtoimprovethesolubilityoftheexpressed
For the most part, B strains of E. coli are preferred protein. By carrying out small-scale expression tests
for recombinant protein expression since they are inparallel, varyingmedia, inductionregime, and/or
deficient in the ATP-dependent lon and membrane- temperature, the optimum conditions for expression
associated ompT proteases that may otherwise lead of a given target protein can be determined.
to degradation of the heterologously expressed pro-
tein. A variety of derivatives of the original BL21 2.3.1.2 Assay format
strain are available, including those carrying addi- The starting point for any expression assay is
tional plasmids, which either provide fine control lysing the cells after harvesting by centrifugation
of T7 vectors (pLysS see Section 2.2.2.1) or sup- and then separation of the soluble from insolu-
ply tRNAs for codons that are relatively rare in ble fractions. Cell lysis can be carried out using
E. coli but may be commonly used in eukaryotic standard protocols by either a freeze-thaw cycle
proteins. The presence of rare codons, particularly followed by treatment with DNAse/lysozyme or
in the first 25 positions, is often a major cause of by sonication with/without lysozyme. Alterna-
poor expression (Gia-Fen and Chen, 1990). Special- tively, commercial detergent-based lysis reagents,
ized BL21 strains include C41, which appears to for example BugBuster™ (Merck), FastBreak™
favour production of membrane proteins, and B834, (Promega), CelLytic™ (Sigma), and Poppers™
a methionine auxotroph for biosynthetic labelling (Pierce) can be used. Chemical methods lend them-
with selenomethionine (Table 2.1). selves to 96-well formats, though sonicators which
The choice of culture media determines the can accommodate a 96-well plate are available
biomass achievable in simple batch cultures and (Misonics). For HTP methods, centrifugation is
therefore the overall yield of protein. This assumes generally avoided for fractionation of the lysates,