High-throughput sequencing analysis
High-throughput sequence data was collected from the Miseq system and parsed through a Perl script to extract the coding region. Extracted sequences were further analyzed by the FastAptamer software (Alam et al., 2015) to count and rank the unique sequences based on their number of reads. We first focused on the top 50 most abundant sequences from each round of mRNA display to confirm the enrichment of FLAG epitope during the selection. Sequences from the FLAG-random library (DYKDDxxx) resulted in a fixed Asp (D) residue at the 6th position immediately after the first round of selection, followed by positively charged 7th position with Arg (R) and Lys (K) and a variable 8th position (Figure 3A). After three rounds of selection, sequences derived from the 10aa-random library converge on a conserved core FLAG motif (DYKxxD) in two different positions, in one case starting at position 2 and in the other case starting at position 5 (Figure 3B). This binding motif has been reported in the previous in vitro selection studies to have strong affinity for anti-FLAG M2 antibody (Miceli et al., 1994; Osada et al., 2009; Srila & Yamabhai, 2013; Yonezawa et al., 2003). By focusing on only those sequences with the core FLAG motif, we further found features in other variable residues such as weakly conserved Glu (E) immediately before the FLAG motif, (D/L/N) and (D/F/L/N/Y) located between the conserved Lys (K) and Asp (D) with over 80% probability followed by downstream Pro (P) residue (Figure 3C and 3D).
We next compared the enrichment rate of various FLAG motifs detected through mRNA display with that of cDNA display. The expected frequency for the full FLAG epitope sequence (DYKDDDDK) in the initial library (0th round) is 0.29 reads per million, and indeed we were not able to detect the epitope sequence from the 0th round in our sample of 1,768,526 sequences. Although as the rounds progressed, we observed continuous enrichment up to 1479 and 198 reads per million for the 3rd round mRNA and cDNA display, respectively (Figure 4A). When we focused on the consensus FLAG motif (DYKDDDxx) with fixed Asp (D) residue at the 6th position, we observed marked enrichment of 261,982 and 31,689 reads per million, suggesting that 6th position is significantly enhances binding to the anti-FLAG M2 antibody (Figure 4B). For the final 4th round, we performed competitive elution with the FLAG epitope octapeptide (DYKDDDDK) by increasing the concentration in a stepwise manner. The majority of the selected sequences were replaced by sequences derived from 10aa-random library harboring consensus FLAG motif: DYK(D/L/N)(D/F/L/N/Y)D located at two different positions (Figure 4C and 4D).
The highest enrichment of consensus FLAG motifs was achieved from sequences recovered from the eluant with high concentrations of competitor FLAG peptide (20 or 100 μg/ml) reaching total 325,917 and 135,913 RPM for mRNA and cDNA display, respectively (Table 1). Under our selection condition, mRNA display shows higher enrichment rate of the consensus FLAG motif (average 28.7 fold per round) over cDNA display (average 19.1 fold per round). Additionally, with mRNA display, the abundance of the FLAG motif among the sequences that remain bound to the beads after all of the elution was consistently greater than with cDNA display (Figure 4 and Table 1, 100 μg/ml vs beads). These results suggest that the mRNA/cDNA conjugate is more readily eluted form the beads.
Finally, the influence of the individual FLAG amino acid residues on antibody binding was evaluated through the comparison of sequences lacking one of the four residues of the core FLAG motif (DYKxxD). Enrichment of sequences lacking Tyr (Y) or Lys (K) residues showed negative enrichment in the first round indicating the importance of these residues for primary recognition of the anti-FLAG M2 antibody. The observed result is consistent with the previous selection (Miceli et al., 1994; Srila & Yamabhai, 2013; Yonezawa et al., 2003) and obtained M2 antibody structural data (Roosild et al., 2006).