Fig 1.
Construction of the scFv library with six non-combinatorially diversified CDRs.
Pools of oligonucleotides with the designed CDR sequences were array-synthesized and amplified by PCR. A single-CDR library (scFv library with only one of the six CDRs are diversified) was constructed for each CDR using the amplified CDR oligonucleotide mixtures, and panned against anti-HA antibody that bind to HA-tag at the C-terminus of scFv in order to proofread the CDR repertoires for in-frame sequences. The proofread CDRs were consolidated into the final scFv library with six diversified CDRs.
Fig 2.
Design process of the non-conbinatorial CDR diversity.
CDR sequences of thousands of natural human antibodies were compared with the human germline CDR sequences and their somatic hypermutation patterns/frequency, germline sequence usage, and length distribution were analyzed. Thousands of CDR sequences were simulated based on the analysis and sequences with undesirable post-translational modification motifs were removed from the repertoire. The resulting CDR repertoires contained the sequences that dutifully mimic the naturally produced human CDRs but without many of the deleterious sequence motifs.
Table 1.
Number of total and unique CDR sequences designed for the scFv library.
Fig 3.
Frequencies of unique CDR sequences in the designed and actual CDR repertoires of the library.
The frequencies of occurrence of each unique CDR sequence in the designed and the NGS-analyzed CDR repertoires of the actual scFv library are shown in XY-scatter plots. Each dot in the plot represents a unique CDR sequence. For the NGS-analyzed sequences, only those in the designed repertoires were included in the analysis. CDR-H3 sequences were not analyzed since most of the sequences occur only once in the designed repertoire.
Fig 4.
Dot blot analysis of random pre-selection scFv clones from the library.
Ninety-two clones were randomly chosen from the unpanned library, grown and induced with IPTG, and periplasmic extracts were obtained and blotted on a nitrocellulose membrane. The clones that express soluble, full-length scFv were detected by the presence of a HA-tag at the C-terminus. Out of 92 clones, 58 were detected by HRP-conjugated anti-HA antibody (63%).
Table 2.
Next-generation sequencing analysis of the library CDRs.
Fig 5.
Variable domain sequence redundancies of the constructed library.
The variable domain sequences of the unselected library were obtained by 300 bp paired-end sequencing on Illumina MiSeq platform, and the number of replicates (n) for the variable domain sequences was analyzed. Approximately 98% of VH and Vλ, and 88.5% of Vκ sequences were found only once (n = 1).
Fig 6.
The length distribution of the designed and the actual CDRs.
(a) CDR-H2, L1 (kappa and lambda), and L3 (kappa and lambda) contain sequences with varying lengths. The length distribution of the designed repertoire and the NGS analysis results suggest that shorter CDRs are preferentially incorporated in the constructed library. (b) The preference for the shorter CDRs was more evident in CDR-H3 which has a wider range of length variation than other CDRs. In both (a) and (b), the blue bars (“Designed”) indicate the frequency of each CDR length in the designed repertoire, and the orange bars (“Found”) indicate the frequency of each CDR length found from the NGS analysis of the constructed library. See text for details.
Fig 7.
The CDR-H3 amino acid distributions.
The amino acid distributions of the CDR-H3 of natural human antibodies (N), the designed repertoire (D), and the actual constructed library (L) are shown for each position. Each stacked bar reflects the combined frequencies of amino acids at each Kabat position of CDR-H3 of different lengths. For all CDR-H3s with different lengths, the last three residues are denoted by 100j, 101, and 102, respectively.
Table 3.
Average numbers of amino acid differences in the CDR sequences from the closest human germline CDR sequence.
Table 4.
Percentages of post-translational modification motifs in the CDRs of the library and the natural human antibodies.
Table 5.
Panning and screening of the library against antigens.
Table 6.
Binding kinetics of selected target-specific scFv clones determined by SPR.
Table 7.
Mutations in CDRs and FRs before and after the panning selection.