Peripheral blood lymphocytes were isolated from 10 gastric cancer patients who initially did not receive chemotherapy. All patients provided a participant information statement and informed consent. This study was approved by the Human Ethics Committee of Nanchang University School of Medicine (Jiangxi, China). Total RNA was isolated individually from peripheral blood lymphocytes using an RNA purification kit for amplification of the scFv library. Primers were designed as previously described (11), with certain modifications for the construction of ribosome display libraries (Table I). First-strand cDNA was synthesized and V genes were amplified using suitable primers. Agarose gel electrophoresis revealed bands of the expected sizes. Splicing by splicing overlap extension polymerase chain reaction (PCR) led to a full-length human scFv repertoire linking the heavy and light chains. scFv antibody libraries in the format of V_H/κ and V_H/V_λ-Cκ were constructed for ribosome display.

The eukaryotic ribosome display was performed as described previously (12), with certain modifications. The PCR libraries were expressed in a coupled rabbit reticulocyte lysate system to generate ARM complexes. The reaction was performed in a volume of 50 μl containing 40 μl TNT T7 Quick for PCR DNA, up to 5 μg PCR DNA, 1 μl 1 mM methionine, 0.5 μl 100 mM magnesium acetate and 3.5 μl ddH₂O. The mixture was incubated in a siliconized tube at 30°C for 60 min. Forty units RNase-free DNase I were added to the mixture together with 6 μl 10X DNase I digestion buffer and ddH₂O, yielding a final volume of 60 μl. The mixture was then incubated at 30°C for an additional 15 min. Cold (4°C) phosphate-buffered saline (PBS; 60 μl) was added prior to antigen selection. Antigen-linked Dynabeads (2 μg) were then added and the mixture was incubated at 4°C for 2 h with gentle agitation to select a CAGE-specific antibody fragment. The beads were collected magnetically and then washed three times with cold washing buffer (PBS, 1% Tween-20, and 5 mM magnesium acetate), followed by washing twice with cold ddH₂O. The beads were collected by magnetic particle clutch and resuspended in 10 μl nuclease-free ddH₂O for reverse transcription (RT)-PCR to recover antigen-selected mRNA. The beads were stored at 4 or −20°C.

The selected mRNA was reverse transcribed to cDNA using a PrimeScript^® One Step RT-PCR Kit (ver. 2) for 30 min at 50°C for the reverse transcription and then for 2 min at 94°C for deactivation of the reverse transcriptase. The obtained cDNA was amplified in a 50 μl PCR mixture for 30 cycles (30 sec at 94°C, 30 sec at 55°C and 1 min at 72°C) with T7A1 (5′-GCAGCTAATACGACTCACTATAGAACA GACCACCATG-3′) and Hu1 (5′-GCTCAGCGTCAGGGT GCTGCT-3′). For nested PCR, different downstream primers for subsequent cycles were used to enrich production: Hu2 (5′-CTCTCCTGGGAGTTACC-3′) in the second cycle and Hu3 (5′-GAAGACAGATGGTGCAGC-3′) in the third cycle. Purified products were applied to the subsequent selection round.

The DNA generated using the above-described steps was digested with XhoI and XbaI and ligated into the vector pBluescript SK⁺ for periplasmic expression of the antibody fragment. Escherichia coli (E. coli) BL21 containing the expression plasmid was grown in 20 ml 2X tryptone-yeast extract (TY; 16 g tryptone; 10 g yeast extract, 5 g NaCl) medium containing 100 μg/ml ampicillin and 0.5% glucose at 37°C overnight. Following centrifugation at 300 × g for 10 min, the bacterial pellet was induced by resuspension in 50 ml fresh 2X TY containing 100 μg/ml ampicillin and 0.5 mM isopropylthio-β-galactoside (IPTG) and then grown at room temperature (20–25°C) for 5–7 h. The bacteria were then harvested by centrifugation at 300 × g for 10 min and the pellet was either stored at −20°C or used immediately. The pellet was first washed with 1 ml 20% (w/v) sucrose, 0.3 M Tris-HCl (pH 8.0) and 1 mM ethylenediaminetetraacetic acid (EDTA) to prepare a periplasmic extract. Following centrifugation, the bacteria were subjected to osmotic shock by rapidly mixing with 1 ml ice-cold 0.5 mM MgCl₂ and leaving on ice for 10 min with periodic agitation. The supernatant periplasmic fraction containing the soluble scFv was recovered by centrifugation at 13,000 × g for 10 min at room temperature. Recombinant protein was further purified by affinity chromatography using Ni-NTA spin kits. Following binding of the protein to Ni-NTA agarose, the column was washed with increasing concentrations of imidazole. The expressed protein was eluted with 200 mM imidazole and dialyzed overnight in PBS at 4°C.

The purified periplasmic extracts from selected anti-CAGE-producing clones were subjected to sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) on 12% polyacrylamide gel. Following SDS-PAGE, the gel was transferred to a nitrocellulose membrane [2% (w/v) skimmed milk in PBS]. The transblotted membrane was blocked for 1 h at room temperature with blocking solution and then incubated for 1 h at room temperature with peroxidase-conjugated mouse anti-His tag (1/1000 dilution with blocking solution). 4-Chloro-1-naphthol was used as a peroxidase substrate to visualize the immunoreactivity. Protein concentration was determined using the BCA protein assay kit and purity was assessed with SDS-PAGE analysis.

The equilibrium dissociation constant (K_D) value of scFv antibodies was determined in the solution phase by enzyme-linked immunosorbent assay (ELISA) (13,14). A 96-well plate was first coated with 5 μg/ml CAGE protein and then serially diluted scFv protein in 2% skimmed milk was added at 37°C for 1 h. The bound antibodies were detected with normal ELISA. The concentration of scFv antibodies presenting 50% of the maximum antigen-binding activity was used in competitive ELISA, which was performed on immobilized CAGE as described above, with the exception that CAGE at various concentrations was mixed with a constant amount of antibody in 100 mM PBS (pH 7.4) supplemented with 10 mg/ml bovine serum albumin (BSA). The antibody concentration was derived using ELISA as mentioned above. After overnight incubation at room temperature, 50 μl each mixture was transferred into the well, coated with CAGE and incubated for 1 h at 37°C. After washing, the bound scFv antibodies were detected as above in triplicate and their affinity was calculated using the Scatchard analysis equation.

Plasmid DNA from anti-CAGE-producing clones was isolated from E. coli BL21. The scFv DNA was sequenced using the dideoxy method with the pBluescript SK⁺ sequence primer set in an ABI Prism automated sequencing machine system.

In this study, a large non-immune ribosome display antibody library was established for routine isolation of a high-affinity human scFv antibody to the CAGE antigen. Purified lymphocytes from blood samples were obtained from 10 gastric cancer patients for amplification of the scFv library. Following total RNA isolation, first-strand cDNA synthesis and amplification of V genes with their corresponding primers were performed. Agarose gel electrophoresis revealed bands of the expected sizes (Fig. 1). Splicing by overlap extension PCR was then used to generate full-length scFv gene templates, i.e., V_H/κ and V_H/V_λ-Cκ, which were detected by gel electrophoresis (Fig. 2). Although variable regions in the light and heavy chains of the scFv genes were different, their upstream primer, downstream primer and linker were the same. Thus, these scFvs had an identical sequential pattern graph. DNA sequencing of randomly selected clones confirmed in-frame cloned V genes of human origin derived from various V-gene families (data not shown).

The scFv library was expressed in vitro in the rabbit reticulocyte coupled transcription/translation system to generate ternary RNA-ribosome-scFv complexes. The ribosome complexes were enriched by the binding of scFv antibody to recombinant CAGE protein-conjugated Dynabeads. Following RT-PCR recovery of the selected RNA in the complexes, the small DNA band was detected following three rounds of subtractive panning, as shown by the results of agarose gel electrophoresis (Fig. 3). For nested PCR, we used the downstream primer Hu2 in the second cycle and Hu3 in the third cycle in order that the recovered DNA became progressively shorter in each cycle. The shortening only affected the constant domain of the κ-light chain.

After three rounds of panning, the CAGE-specific scFv antibody genes were digested and ligated into the expression vector. The expressed scFv antibodies were collected from the periplasm of the bacteria and purified by affinity chromatography using Ni-NTA spin kits. The amount of the best-expressed scFv anti-CAGE proteins expressed in E. coli BL21 was ~30% of the total bacterial proteins. SDS-PAGE results for the full-length fusion proteins are shown in Fig. 4A. The expression of scFv was confirmed by western blot analysis using anti-His tag antibodies (Fig. 4B). The selected scFv antibody presented a single band measuring ~35 kDa in size.

According to the Scatchard analysis equation [A₀/(A₀ - A) = K_D/A_g +1, where A₀ is the absorbance when the antibody was incubated without any antigen, A is the absorbance corresponding to free antibody following incubation with antigen and A_g is the free antigen concentration that is equal to the antigen obtained for experimentation considering a pseudo-first-order reaction], the K_D of scFv for CAGE was 7.6×10⁻⁸ M. Fig. 5 shows the RT-PCR product sequence obtained following the third screening cycle and its deduced amino acid sequence. Complementarity determining regions (CDRs) are marked as per the Kabat definition.