Identification of soluble protein fragments by gene fragmentation and genetic selection.

We describe a new method, which identifies protein fragments for soluble expression in Escherichia coli from a randomly fragmented gene library. Inhibition of E. coli dihydrofolate reductase (DHFR) by trimethoprim (TMP) prevents growth, but this can be relieved by murine DHFR (mDHFR). Bacterial strains expressing mDHFR fusions with the soluble proteins green fluroscent protein (GFP) or EphB2 (SAM domain) displayed markedly increased growth rates with TMP compared to strains expressing insoluble EphB2 (TK domain) or ketosteroid isomerase (KSI). Therefore, mDHFR is affected by the solubility of fusion partners and can act as a reporter of soluble protein expression. Random fragment libraries of the transcription factor Fli1 were generated by deoxyuridine incorporation and endonuclease V cleavage. The fragments were cloned upstream of mDHFR and TMP resistant clones expressing soluble protein were identified. These were found to cluster around the DNA binding ETS domain. A selected Fli1 fragment was expressed independently of mDHFR and was judged to be correctly folded by various biophysical methods including NMR. Soluble fragments of the cell-surface receptor Pecam1 were also identified. This genetic selection method was shown to generate expression clones useful for both structural studies and antibody generation and does not require a priori knowledge of domain architecture.

Multiplexed expression and screening for recombinant protein production in mammalian cells.

BACKGROUND: A variety of approaches to understanding protein structure and function require production of recombinant protein. Mammalian based expression systems have advantages over bacterial systems for certain classes of protein but can be slower and more laborious. Thus the availability of a simple system for production and rapid screening of constructs or conditions for mammalian expression would be of great benefit. To this end we have coupled an efficient recombinant protein production system based on transient transfection in HEK-293 EBNA1 (HEK-293E) suspension cells with a dot blot method allowing pre-screening of proteins expressed in cells in a high throughput manner. RESULTS: A nested PCR approach was used to clone 21 extracellular domains of mouse receptors as CD4 fusions within a mammalian GATEWAY expression vector system. Following transient transfection, HEK-293E cells grown in 2 ml cultures in 24-deep well blocks showed similar growth kinetics, viability and recombinant protein expression profiles, to those grown in 50 ml shake flask cultures as judged by western blotting. Following optimisation, fluorescent dot blot analysis of transfection supernatants was shown to be a rapid method for analysing protein expression yielding similar results as western blot analysis. Addition of urea enhanced the binding of glycoproteins to a nitrocellulose membrane. A good correlation was observed between the results of a plate based small scale transient transfection dot blot pre-screen and successful purification of proteins expressed at the 50 ml scale. CONCLUSION: The combination of small scale multi-well plate culture and dot blotting described here will allow the multiplex analysis of different mammalian expression experiments enabling a faster identification of high yield expression constructs or conditions prior to large scale protein production. The methods for parallel GATEWAY cloning and expression of multiple constructs in cell culture will also be useful for applications such as the generation of receptor protein microarrays.

Production of soluble mammalian proteins in Escherichia coli: identification of protein features that correlate with successful expression.

BACKGROUND: In the search for generic expression strategies for mammalian protein families several bacterial expression vectors were examined for their ability to promote high yields of soluble protein. Proteins studied included cell surface receptors (Ephrins and Eph receptors, CD44), kinases (EGFR-cytoplasmic domain, CDK2 and 4), proteases (MMP1, CASP2), signal transduction proteins (GRB2, RAF1, HRAS) and transcription factors (GATA2, Fli1, Trp53, Mdm2, JUN, FOS, MAD, MAX). Over 400 experiments were performed where expression of 30 full-length proteins and protein domains were evaluated with 6 different N-terminal and 8 C-terminal fusion partners. Expression of an additional set of 95 mammalian proteins was also performed to test the conclusions of this study. RESULTS: Several protein features correlated with soluble protein expression yield including molecular weight and the number of contiguous hydrophobic residues and low complexity regions. There was no relationship between successful expression and protein pI, grand average of hydropathicity (GRAVY), or sub-cellular location. Only small globular cytoplasmic proteins with an average molecular weight of 23 kDa did not require a solubility enhancing tag for high level soluble expression. Thioredoxin (Trx) and maltose binding protein (MBP) were the best N-terminal protein fusions to promote soluble expression, but MBP was most effective as a C-terminal fusion. 63 of 95 mammalian proteins expressed at soluble levels of greater than 1 mg/l as N-terminal H10-MBP fusions and those that failed possessed, on average, a higher molecular weight and greater number of contiguous hydrophobic amino acids and low complexity regions. CONCLUSIONS: By analysis of the protein features identified here, this study will help predict which mammalian proteins and domains can be successfully expressed in E. coli as soluble product and also which are best targeted for a eukaryotic expression system. In some cases proteins may be truncated to minimise molecular weight and the numbers of contiguous hydrophobic amino acids and low complexity regions to aid soluble expression in E. coli.

Application of phage display to high throughput antibody generation and characterization.

We have created a high quality phage display library containing over 1010 human antibodies and describe its use in the generation of antibodies on an unprecedented scale. We have selected, screened and sequenced over 38,000 recombinant antibodies to 292 antigens, yielding over 7,200 unique clones. 4,400 antibodies were characterized by specificity testing and detailed sequence analysis and the data/clones are available online. Sensitive detection was demonstrated in a bead based flow cytometry assay. Furthermore, positive staining by immunohistochemistry on tissue microarrays was found for 37% (143/381) of antibodies. Thus, we have demonstrated the potential of and illuminated the issues associated with genome-wide monoclonal antibody generation.