The Recombinant Antibody Network (RAN) is a consortium of three expert centers at the University of Chicago, University of California San Francisco (UCSF), and the University of Toronto that are unified under a common set of modern technologies to generate highly validated recombinant antibodies (rABs) to the human proteome (http://recombinant-antibodies.org). The RAN pipeline starts with antigen production followed by phage display to generate high quality affinity reagents to folded protein epitopes. The RAN reagents are well characterized both in vitro and in a cellular-like environment before they are distributed and released to the scientific community.
A significant number of the TF domains used in the phage display selection pipeline are sourced from Stephen Anderson's laboratory at Rutgers University. We also use alternative antigen production schemes for a broader range of targets. RAN is currently undertaking a three-pronged approach to generate high quality and natively folded antigens using (1) an E.coli expression system, (2) an in vitro transcription and translation (IVTT) protein expression system, and (3) a Yeast Surface Display (YSD) pipeline for rapid expression of target antigens on the surface of yeast. The IVTT approach is easily automated and has a number of advantages over standard E. coli expression especially for antigens that are difficult to express. Preliminary data suggest that IVTT can salvage targets that are problematic in bacterial expression. The YSD pipeline has the attribute that the biotinylated antigen is displayed on the yeast surface where it can be used directly in the phage display selection pipeline, eliminating the antigen expression and purification steps that normally precede the selection step. Our expectation is that most antigens will be accessible for phage display selections by at least one of the antigen expression platforms that are currently used in the RAN pipeline.
An overview of the phage display cycle utilized in our pipeline is shown in the top right Figure. There are several key elements that distinguish our approach. First we use a reduced genetic code library that has been shown to perform extremely well in generating rABs to many different types of antigens including both soluble and membrane-associated proteins. Second, our approach provides for exquisite control of the phage display selection conditions that enable generation of rABs that have a high level of specificity and tunable affinities. The system is currently fine-tuned to target folded protein epitopes. The antigens used in the phage display pipeline are fused to an AVI/biotin tag that allows them to be immobilized to streptavidin coated magnetic beads, which is the first step in the phage selection process. Next, the antigen-loaded magnetic beads are mixed with a synthetic and highly diverse rAB phage display library, containing about 3x1010 members. The rAB-phage pool is usually optimally enriched for antigen binding clones after 3 to 4 rounds of selection. Finally, the rAB-phages are collected and reformatted for sequencing and expression. The rAB scaffold used here consists of one heavy chain and one light chain that are co-expressed by the phagemid system. The two chains assemble in the periplasm of E. coli to form a stable IgG1 molecule.
The primary validation aims to identify rABs that are sequence unique and bind with high affinity to their target. The antigen-binding activity is measured by high-throughput robotic ELISA assays, in a rAB-phage and/or in a protein format. The pass criteria is <50 nM. Each rAB DNA clone (<50 nM) is sequenced to determine its identity and also to assure that only the unique rAB clones are moved forward in the validation pipeline.
The secondary validation aims to identify high quality rABs that bind both tightly and specifically to their target in a cellular-like environment. We have developed a high throughput assay termed Spiked IP (Spiked Immunoprecipitation) that tests the ability of the rABs to bind and pull-down exogenously added antigen in the presence of complex human cell lysates compared to buffer alone. Biotinylated rABs are immobilized to streptavidin magnetic beads and allowed to bind the antigen in the presence or absence of a human HEK cell lysate as monitored by flow cytometry. To be considered a "passing reagent" for the Protein Capture Reagents Program, a rAB must pass both the ELISA (<50nM) and the Spiked IP test.
A subset of the rABs is also tested for its ability to recognize their targets in vivo in a complex cellular milieu by preforming IP-MS analysis of human HEK cells. Over 300 RAN reagents have passed the IP-MS test. Moreover, we determine the in vivo antigen-binding cellular compartment by high-throughput immuno-fluorescence (IF) analysis of six different human cell lines. Analyses of over 1000 RAN rABs by IF have revealed the cellular localization of more than 300 DNA regulatory proteins.
The rABs that passed both primary and secondary validation are tested for their ability to ChIP-seq by outsourcing them to the Myers ENCODE laboratory at HudonAlpha. A sub-set of the rABs has also successfully been validated in-house by ChIP-WB and ChIP-qPCR.
The RAN rABs are highly valuable tools for research that aims to capture the natively folded protein, but are less likely to work in assays like WB that denature the target protein. RAN reagents that pass both the affinity (<50 nM) and the specificity test (Spiked IP and/or IPMS) are available at low cost from either DNASU (rAB DNA) or DSHB (rAB protein). The rAB plasmids can also be requested directly from the RAN web site (http://recombinant-antibodies.org/).