NATURE BIOTECHNOLOGY VOLUME 30 NUMBER 5 MAY 2012 447 ARTICLES After exposure to a foreign antigen, the mammalian humoral immune response generates a diverse repertoire of antibodies through changes in the genome of B cells by V(D)J gene recombination, gene conversion (in rabbit and chicken) and somatic hypermutation 1–4 . Each B-cell clone that undergoes this process contributes a specific monoclonal antibody to the diverse polyclonal response that is critical to fending off infection. For over a hundred years, polyclonal antibodies have been used as tools for basic research and clinical diagnostics, as well as for passive immunity therapy for infectious diseases 5,6 . However, polyclonal antibodies enriched for defined properties have been his- torically difficult to produce in large scale, limiting their value for diag- nostic and therapeutic applications. The hybridoma method 7 provided for the first time a way to obtain monoclonal antibodies, and it opened the door to interrogating the complexity of the humoral immune response to an antigen. Since then, various newer technologies have been developed to obtain antigen-specific monoclonal antibodies. Some of these alternative strategies—involving B-cell immortaliza- tion, single-cell sorting and molecular cloning, or phage display—have become increasingly effective, but the antibodies they generate do not necessarily represent the actual antibody repertoire found in circula- tion and are often labor intensive and time consuming 8–15 . To directly investigate the monoclonal composition of polyclonal antibodies in circulating serum, we used a proteomics approach based on nano-flow liquid chromatography coupled to mass spectrometry (LC-MS/MS). Such an approach, however, is difficult to implement because of the high complexity of the polyclonal mixture and the lack of a reference database of the constantly evolving repertoire of antibodies generated against foreign antigens in an individual animal. To address these challenges, we used affinity purification to reduce sample complexity, and next-generation DNA sequencing to gener- ate a reference database derived specifically from the animal’s B-cell repertoire 16 (Fig. 1). To validate our approach, we generated mono- clonal antibodies with potential for diagnostic application from the serum of rabbits immunized with human progesterone receptor A/B (PR A/B) peptide antigens. We focused on PR A/B because of its clinical significance as a biomarker used in immunohistological assays for the diagnosis of breast cancer 17 . RESULTS We immunized New Zealand white rabbits with human PR A/B pep- tides conjugated to keyhole limpet hemocyanin. Next, we screened antigen-specific antibody activity in the crude serum of each animal to select the rabbit with the strongest enzyme-linked immunosorbent analysis (ELISA) and western blot analysis signals to PR A/B (data not shown). Serum from this animal was collected from 20 ml of blood, and RNA was obtained from splenic B cells. We isolated total IgG from the serum using a protein A sepharose column and purified antigen-specific polyclonal antibodies by affinity chromatography using a custom column consisting of antigen-specific peptide conju- gated to sepharose beads. Bound IgGs were washed extensively with PBS then subjected to sequential elutions with progressively acidic buffers (pH 3.5, pH 2.7 and pH 1.8) (Fig. 2a). Fractions from each elution were collected, neutralized and screened by antigen-specific ELISA (data not shown) and western blot analysis of lysate from the PR A/B-expressing cell line T47D and the PR A/B-negative cell line HT1080 (Fig. 2a). We found that specific activity to PR A/B by A proteomics approach for the identification and cloning of monoclonal antibodies from serum Wan Cheung Cheung 1,2 , Sean A Beausoleil 1,2 , Xiaowu Zhang 1 , Shuji Sato 1 , Sandra M Schieferl 1 , James S Wieler 1 , Jason G Beaudet 1 , Ravi K Ramenani 1 , Lana Popova 1 , Michael J Comb 1 , John Rush 1 & Roberto D Polakiewicz 1 We describe a proteomics approach that identifies antigen-specific antibody sequences directly from circulating polyclonal antibodies in the serum of an immunized animal. The approach involves affinity purification of antibodies with high specific activity and then analyzing digested antibody fractions by nano-flow liquid chromatography coupled to tandem mass spectrometry. High-confidence peptide spectral matches of antibody variable regions are obtained by searching a reference database created by next-generation DNA sequencing of the B-cell immunoglobulin repertoire of the immunized animal. Finally, heavy and light chain sequences are paired and expressed as recombinant monoclonal antibodies. Using this technology, we isolated monoclonal antibodies for five antigens from the sera of immunized rabbits and mice. The antigen-specific activities of the monoclonal antibodies recapitulate or surpass those of the original affinity-purified polyclonal antibodies. This technology may aid the discovery and development of vaccines and antibody therapeutics, and help us gain a deeper understanding of the humoral response. 1 Cell Signaling Technology, Inc., Danvers, Massachusetts, USA. 2 These authors contributed equally to this work. Correspondence should be addressed to R.D.P. (rpolakiewicz@cellsignal.com). Received 6 December 2011; accepted 21 February 2012; published online 25 March 2012; doi:10.1038/nbt.2167 npg © 2012 Nature America, Inc. All rights reserved.