Within the X CX PX RX
Within the X1CX3PX5RX7 motif of several prokaryotic FGE substrates, the residues of X3, X5, and X7 are variably found as mixtures of alanine, glycine, threonine, or serine residues , . The docked model of our identified HCTPRRP motif revealed that position X1 (H1) and position X7 (P7) were not buried deep within the FGE binding pocket (Fig. S4a), which may support existing evidence of the variability of the immediate flanking positions . The existence of the C-terminal proline at X7, found in both the 4(2) and 4(1) substrate sequence, may potentially prevent unfavorable steric interaction with the adjacent p3 coat protein used in this screening strategy; though it oxytocin receptor has been noted that aldehyde tags can be generated with equal efficacy at internal protein locations or at the C or N terminus . The existence of arginine and lysine for identified substrate 4(2) and 4(1), respectively, was initially expected to be an indication of preference for a positively charged residue at this position for electrostatic complementarity with the FGE active site; however, from our docking simulation we only observed R5 participating in hydrogen bonding with the main chain backbone of FGE. For position X3, the identified substrates 4(2) and 4(1) exhibited a threonine and serine residue respectively, with the docking simulation of 4(2) showing hydrogen bonding of T3 with the FGE binding site. Interestingly, threonine and serine also existing at substrate position X3 in native AtsG and AtsA sulfatases of M. tuberculosis, respectively . In again looking at the phage display results in retrospect to the above analysis, we can see that the preponderance of positively charged residues at position X5 of the phage screening results was likely indicative of preferential phage release. Because these particular residues could provide additional sites for cleavage by trypsin (as shown for the 5th position arginine of 4(2) in Fig. 6), these sequences are expected to have arisen as a factor of the additional selection pressure provided by the sequence dependent elution used throughout the screening process. In doing so, the trypsin cleavage unknowingly provided this additional selection pressure and should be considered another major driving force for the phage display results obtained. The use of proteases to elute the phage can be undesirable, as this provides selection pressure toward enrichment of peptide sequences that possess specific proteolytic cleavage sites. In order to avoid this, it is conceivable to use alternative elution/amplification techniques after washing away non-specific phage from the capture beads. In one incarnation it may be possible to directly add E. coli culture to the beads for transfection and amplification of the captured phage. While this could inadvertently increase the extent of transfection by non-specific phage, more stringent washing conditions or an additional bead pre-screening step could be used to mitigate the presence of non-specific phage. Alternatively, the elution/amplification of phage captured on the beads may be conceivably be carried out instead by first performing PCR amplification of the internal phagemid followed by cloning and subsequent phage growth. By using such techniques to avoid sequence dependent elution and rather focusing solely on sequence dependent capture, these alternative strategies may help provide a means for increasing the screening efficiency for identifying specific enzymatic substrates.
Conclusions In summary, a phage screening method was developed which could be used to identify a substrate capable of conversion by the FGE. The screening was facilitated by the covalent capture of viable FGE substrates displayed on a phage library. Elution by enzymatic release using trypsin is believed to provide addition selection pressure on the variable XCXPXRX library acting as an auxiliary cut site to enrich sequences recognized by trypsin. After screening, a substrate sequence, HCTPRRP, was identified by combined panning and sequence analysis. The peptide conversion by FGE was subsequently confirmed by capture assay and mass spectroscopy. It is important to note that while the HCTPRRP peptide served as an FGE substrate, our results indicate it is not the optimal sequence, which was in contrast to our expectations. Instead, we find this FGE substrate sequence was also readily cleavable by trypsin, and this is believed to be attributed to the additional selection pressure arising via the sequence dependent elution used throughout the screening. From our results, we see that this screening approach may provide a versatile strategy for screening of other potential enzymatic substrates. In looking forward, we expect this work may lead to the future generation of novel FGE variants with different sequence selectivity in order to create a range of enzymes capable of bio-orthogonal modification. Such a desirable outcome may someday allow researchers to easily incorporate multiple modifications/labels at different specific sites on a single protein.