To compare the phage of the th round results
To compare the phage of the 4th round results, we conducted isolated single rounds of FGE screening as described above for the 4 individual phage 4(1), 4(2), 4(3), and NC then determined the number of colony forming units afforded by cleavage from the support. Interestingly, we found that the 4(2) sequence provided a significantly higher amount of captured and eluted phage as compared to any of the other phage (Fig. 3a). We also see that 4(1) was significantly higher than 4(3) and the NC phage. At this point it is interesting to note that both 4(1) and 4(2) exhibit some homology in their sequences as the 2nd, 3rd, and 4th randomized position of their XCXPXRX sequence as both possess hydrophilic, basic, and proline residues, respectively. In order to reveal if the duration of FGE reaction time would have an effect on the extent of substrate sequence modification, the individual phage 4(1), 4(2), 4(3), and NC were again screened as above using different FGE reaction times of 60, 180, and 300min. From Fig. 3b, we see that indeed the longer duration of FGE reaction yielded a higher amount of captured and eluted phage, particularly for the case of 4(2) and 4(1). In contrast, the 4(3) sample did not appear to have any noteworthy improvement, and the NC phage, having no XCXPXRX motif, remained relatively constant as the level of non-specific background. In these results, we find that the 4(1) and 4(2) substrate sequences had increased conversion by FGE as a function of time. Since the 4(2) substrate showed significantly greater capture as compared to the other candidate substrates, the 4(2) sequence was examined further to ensure the enrichment did not arise due to any aldehyde-independent binding events. To do so, we carried out a control binding assay including the presence of methoxyamine prior to exposure to the amine bearing surface (Fig. 4a). If the capture of phage were truly occurring via the aldehyde formed by FGE, then oxime bond formation between aldehyde present on the ESI-09 and the methoxyamine would inhibit the covalent capture of phage onto the amine bearing beads. Indeed, our results confirm the important role of the aldehyde formed by FGE modification of 4(2) which is critical in facilitating phage capture. The observed decrease in the amount of captured phage when pre-reacted with methoxyamine indicates that the conversion of cysteine to formylglycine on 4(2) plays the key role in its covalent capture to the magnetic beads. In combination with the results obtained in the absence of FGE, which show no capture of phage onto the beads, we can say with certainty that the sequence 4(2) is responsible for covalent attachment only when modified to possess the aldehyde via FGE modification. To provide further supporting data that the phage capture mechanism was arising from the formylglycine generated at the cysteine residue of the 4(2) motif (HCTPRRP) rather than occurring at any other locations on the phage, we also conducted a set of control experiments on 4(2) and a phage mutant 4(2) C2A which has a replacement of the key cysteine residue with an inert alanine (HATPRRP). As seen in Fig. 4b, there is almost no observed phage CFUs for the case of the 4(2) C2A mutant indicating the importance of formylglycine formation from the cysteine present in the motif for effective covalent capture. Comparing 4(2) and 4(2) C2A capture for the FGE and no FGE case, we see a large reduction in the amount of 4(2) phage capture when FGE is omitted, and the amount of 4(2) C2A phage capture was almost non-existent in both the FGE and no FGE case. Interestingly, a very small amount of 4(2) phage capture in the no FGE case could still be observed which is in agreement with prior literature indicating that a promiscuous FGE-like activity exists in E. coli causing low levels of conversion in vivo. Additional confirmation of FGE conversion of the 4(2) peptide (HCTPRRP) was obtained by matrix-assisted laser desorption/ionization-time of flight (MALDI-TOF) mass spectroscopy (MS), which revealed a mass loss of ∼18.6m/z after reaction of the 4(2) peptide with FGE (Fig. S3). Such a mass loss is expected for the conversion of cysteine to formylglycine.