A previous protocol described the use of lambda exonuclease to generate overhangs similar to overlapping regions created by restriction enzyme digestion (36). KF was used to repair the ssDNA by stabilizing the junction regions without strand displacement. This method focused on limited lambda exonuclease digestion to create short ssDNA 3′-overhangs for hybridization. The protocol was not as efficient for cloning smaller fragments, however, data showed cloning of an 89 bp fragment with reduced efficiency (36). Although this method may be efficient for antibody chain shuffling of larger fragments, it is not appropriate for short complementary determining region (CDR) mutagenesis. CDR regions vary from 3 to 12 amino acids in length (37, 38), depending on their location along the antibody sequence. Binding sites for antigens are formed by six CDRs, wherein three CDRs are found in the heavy chain and light chain, respectively. These CDRs loop out from the V region backbone to form interaction regions with target antigens. Hence, mutations on CDRs can directly increase antibody affinities for antigens and vice versa. However, only CDR3 in the heavy chain has a wide range of variation in terms of length and amino acid propensity, which contributes to antigen specificity (39, 40).
In order to overcome these limitations, we developed a new method to obtain and use ssDNA for chain shuffling and mutagenesis by degenerate oligonucleotides. Although conceptually similar in terms of ssDNA generation, rather than using a limiting digestion step, our approach produces the entire ssDNA for use. The ssDNA template generated will have an overlapping region of at least 7bp in length. This is important because the extension of KF is effective if the complementary region is longer than a heptamer (41). By hybridizing complementary ssDNA templates we are then able to generate a template strand of ssDNA which can be transcribed into dsDNA using KF (Figure 1).
To demonstrate our new methodology, we use a semi-synthetic single chain Fragment variable (scFv) antibody molecule as a model. Two strategies to introduce directed sequence evolution were examined. First, we demonstrate chain shuffling between variable heavy chain (VH) against a specific variable light chain (VL) and secondly, we describe recombination of the VH framework with a synthetic CDR3 oligonucleotide. For chain shuffling, the protocol was used to introduce various light chains to a fixed heavy chain (Figure 2A) while for CDR mutagenesis, we introduced short randomized regions at the heavy chain CDR3 (Figure 2B).
Materials and methods Ethics statement
Human peripheral blood mononuclear cells (PBMCs) were collected with written informed consent according to the guidelines approved by Universiti Sains Malaysia Human Ethics Committee.PCR amplification of antibody genes
Phagemid DNA was prepared from overnight E. coli cultures of anti-eGFP scFv clones 44 and G1 using a MiniPrep Kit (Qiagen, Hilden, Germany). The final volume for all PCRs was either 20 or 50μL containing 40 or 100ng of phagemid DNA, respectively, 200 μM deoxynucleoside triphosphates (dNTPs), 0.2μM of each primer (1st Base, Kuala Lumpur, Malaysia), and 1× DreamTaq reaction buffer (Fermentas, Lithuania). The amount of polymerase used per reaction was 0.2 or 0.5 μL of 5 U/μL DreamTaq (Fermentas, Lithuania). All reactions were carried out using a MyCycler thermocycler (BioRad, Hercules, CA, USA). After heating at 95°C for 45 s, we performed 30 or 25 cycles, respectively (30 s at 95°C, 30 s at 55°C, and 30 s at 72°C) and ended after 5 min at 72°C. PCR amplification of the variable heavy and light chain regions was carried out using the primers VH-NcoI-Fw (ACATGCCATGGCCGAGGTGCAGC) and VH-phos-tom-exo-Rv (5′-phos-CAGGACGGTGACCA-GGGTTCCCTG) as well as VL-phos-tom-exo-Fw (5′-phos-CAGGG-AACCCTGGTCACCGTC) and VL-NotI-Rv (AAGGAAAAAAGCGGCCGCCCGTTTGATTTC) respectively. For CDR3 mutagenesis, primers used for VH framework were VH-NcoI-Fw (ACATGCCATGGCCGAGGTGCAGC), VH-Tom-Fr3-phos-Rv (5′-phos-ACAGTAATATACGGC) and synthetic CDR3 oligonucleotides (GACGGTGACCAGGGTTCCCTGGCCCCAGTAGTCAAAMNNMNNMNNMNNTTTCGCACAGTAATATACGGCCGT). Kappa light chain variable region gene sequence was amplified using cDNA template, synthesized from RNA that was extracted from human PBMCs according to Lim et al. (42); amplification of VK genes from family 2, 4, and 6, were conducted with the primers VK246-Sall-Fw (TGTGACAAAGTCGACGGATATTGTGMTGACBCAGWCTCC) and HscFv kappa-NotI-Rv (ATGATGATGTGCGGCCGCGAAGACAGATGGTGCAGCCACAGT). All PCR products were purified using the PCR Purification Kit (Qiagen) and eluted in 30 or 50 μL of distilled water. dsDNA concentration was determined using a NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific).