Cloning antibody by Phage Display
Immunization of mice, RNA purification and cDNA synthesis Four female BALB/c mice (8 weeks old, ∼20 g body weight) were given intraperitoneal injections of 50 µg of BSA-TNP conjugated analog or BSA-PETN (2 mice received each analog) in complete Freund adjuvant (BD, NJ, USA) (Jackson and Fox, 1995). After 2 weeks, the mice were boosted with another dose of 50 µg analog in incomplete Freund adjuvant (IFA; BD). Blood samples were collected 1 week after each immunization. After a total of three boosts, two of the mice (one for each hapten-carrier) were further boosted twice at 2-week intervals with human apotransferrin-conjugated to TNP or PETN, respectively. A week after the last boost the mice were sacrificed, spleens excised, the cell solubilized and total mRNA was produced from the mice with RNeasy protect midi kit (Qiagen, Germany) according to the manufacturer's instructions. The recovered RNA was converted to cDNA using the Superscript™ III first-strand synthesis kit (Invitrogen, CA, USA). All the animal experimentation was carried out with approval of the Tel-Aviv University IRB.
Construction of the phagemid library scFv libraries were constructed essentially as described previously (Benhar and Reiter, 2002). Briefly, the library construction process consisted of three steps: (i) amplification of the VH and VL domains using framework region family-specific primers, (ii) re-amplification of each segment with primers containing the linker segment, (iii) assembly of both segments (VH and VL) by overlap extension polymerase chain reaction (PCR). All PCR reactions were carried using the Expand High Fidelity PCR System (Roche Applied Science, Germany) according to the manufacturer's instructions. The PCR primers that were used had been reported earlier (Benhar and Reiter, 2002). The PCR-assembled scFv DNA fragments from the TNP-immune repertoire were digested with NcoI and NotI restriction enzymes and were cloned in the pCC16 phagemid vector DNA (Azriel-Rosenfeld et al., 2004) that was previously digested with the same enzymes. The generated library was transformed via electroporation into freshly prepared Escherichia coli XL-1 Blue bacteria, which were spread over large 2 × YT (16 g/l Bacto-Tryptone (BD), 10 g/l yeast extract (BD), 5 g/l NaCl) agar plates supplemented with 100 μg/ml ampicillin, 25 μg/ml tetracycline and 1% (w/v) glucose. After overnight incubation, the plates were scraped into a liquid medium (2 × YT–ampicillin–tetracycline–1% glucose), and a phage rescue process was initiated as described. The electroporated bacteria were also titered by plating serial dilutions to assess the library size and the scapred bacteria were tittered similarly to assess the amplification of the library during the growth on plates.
Affinity selection (panning) of phages All the phage panning steps were carried out essentially as described previously (Benhar and Reiter, 2002). Briefly, 1 ml of 5 µg/ml of the carrier-TNP conjugate was incubated in one well of a 24-well cell culture plate (Jet Biofil, China) overnight at 4°C. The wells were washed with PBS containing 0.05% (v/v) Tween 20 (PBST, Sigma, Israel) and blocked with 3% (w/v) skimmed milk in PBS (blocking solution) for 2 h at 37°C. Meanwhile, 1012 CFU/ml of scFv-displaying M13 phages were incubated with a blocking solution for 1 h at room temperature. The coated wells were washed with PBST and incubated with the proper phages for 1 h at room temperature. The wells were washed with blocking solution diluted three times in PBST, and the remaining bound phage particles were eluted with 1 ml 0.1 M triethylamine, pH 13 (Fluka, Germany), for 30 min at room temperature on a rocking platform. The eluted phages were immediately neutralized with 200 µl of 1 M Tris-(HCl), pH 7.4. To prevent enriching phages that bind the blocking protein(s), the immune plates in each panning cycle were subjected to alternating blocking solutions consisting 3% (w/v) milk, 1% (w/v) BSA or 3% (w/v) gelatin in PBS. The phage-blocking solution was changed between 1% BSA, 1% gelatin and 1% milk in PBS. After the fourth panning cycle, individual colonies of phagemid-carrying cells were used to prepare monoclonal phages in sterile 96-well plates (Nunc, Sweden) that were screened by phage enzyme-linked immunosorbent assay (ELISA).
Phage ELISA Ninety-six-well ELISA plates (BioFil, China) were loaded with 100 µl of 5 µg/ml of the antigen in PBS and incubated overnight at 4°C. The following day the plates were washed with PBST and blocked with a 3% skimmed milk in PBS solution for 2 h at 37°C. After additional washing with PBST, 100 µl of freshly prepared phages were added per well and incubated for 1 h at room temperature. The plates were washed again with PBST, 100 µl of 1 : 5000 diluted horseradish peroxidase (HRP)-conjugated anti-M13 antibody (GE Healthcare, UK) were placed in each well for 1 h at room temperature, followed by additional washes with PBST and finally, 100 µl of 3, 3', 4, 4'– tetramethylbenzidine (TMB) (HRP substrate) solution (Dako, Denmark) were added to each well and incubated at room temperature. The reaction was stopped by the addition of 50 µl 1 M H2SO4 to each well. The plate was analyzed with an ELX-800 plate reader (Bio-Tek instruments, VT, USA) at 450 nm. Phages that yielded ELISA signals that were at least 3 SD above background were selected as potential hits and validated by repeating phage ELISAs, where phages were applied with repetitions and dilutions.
HTS and data analysis of the VH and VL domains The pool of VH and VL cDNA from splenocytes isolated from animals that had been immunized with BSA-TNP or with BSA-PETN and boosted with BSA-hapten followed by apotransferrin conjugates were recovered after the first PCR cycle of the phage library construction process. One hundred nanograms of each cDNA-derived PCR products of VH or VL domains were processed for Roche GS-FLX 454 DNA sequencing according to the manufacturer's protocol. All the 454 raw sequencing data were first processed using the sequence quality and signal filters of the 454 Roche pipeline and then subjected to bioinformatics analysis that relied on homologies to conserved framework regions and V germline gene identification using IMGT/HighV-Quest Tool (Lefranc, 2008; http://www.imgt.org/HighV-QUEST/index.action). Additional filters were applied for full repertoire database construction as follows: (i) sequences without a valid V gene and allele were removed and (ii) for the purposes of germline V(D)J, CDR3 length and amino acid composition analyses, only sequences with n > 1 reads were considered. Identical CDR3s were grouped, and the high abundance sequences were clustered according to their somatic mutations along the entire V region.
After filtering the raw data to identify full length variable domains, the CDR3 sequences in the repertoire were rank-ordered based on their relative abundance. The most abundant CDR3s were selected. The most abundant CDR3 sequences in the TNP libraries were cross-referenced for their frequency in the PETN libraries in order to exclude CDR3 sequences that are abundant in both libraries from being tagged as ‘TNP-specific’. Altogether, three VH and three Vκ sequences were chosen (Table I) and the respective genes were constructed by total gene synthesis using gene optimization for expression in E.coli (GeneArt, Germany). The most abundant VH was then fused in frame with the most abundant Vκ in the scFv format. Similarly, the second and third ranked VH and VL were paired, respectively.