Recombinant production of ΔC-fragment 604-isoform
Recombinant 604-isoform lacking C-terminal 32 amino acid peptide was overexpressed in E. coli strain BL21 (DE3) in a pET26b+ vector. A polyHistidine-tag was located at the C-terminal of the desired protein. PCR-sequencing of recombinant vector with T7 promoter and T7 terminator primers confirmed the accuracy of target gene. Optimum expression of both wild type and the modified recombinant proteins were reached at 1mM IPTG, and shaking for 5 hours at 20 ˚C.
In order to study the fluctuations of the 604-isoform structure, proteolytic digestion was carried out. The expressed purified proteins were digested with α-chymotrypsin for 0, 2, 4, 6, 8, 10 and 15 minutes and SDS-PAGE analyses were used to assess the resultant bands of the proteolytic digestion. As it is shown in figure 1, the protein bands lower than 35 kDa are the products of digestion, while molecular weight of WT protein is 64 kDa. As is shown on figure 1, it is well accepted that the mobility pattern of the digested two samples is different. As in the engineered protein digestion the bands occurred faster and within the first 2 minutes, much more digested fragments (MW < 35 kDa) are formed relative to that of the wild type. This could be interpreted as increased flexibility and extension of structure due to the C-terminal fragment elimination and more accessibility for α-Chymotrypsin activity.
Molecular Dynamics Simulation
Molecular dynamics simulation was utilized in order to obtain a mere understanding of the internal interactions of the C and N terminal peptides with the other domains, both in engineered and wild type 604-isoforms. Here, a dimer of proteins was extracted from a convex octamer structure obtained from a previous study in our lab . The structure of the engineered protein was constructed by eliminating 32 amino acids from the wild type 604-isoform.
In order to assess the possible interactions of the dimer structures of the proteins, the modeled structures of both 604- and the engineered isoforms were subjected to MD simulation for 20 ns. To evaluate the stability and flexibility, Root Mean Squared Deviation (RMSD) and Root Mean Squared Fluctuation (RMSF) were calculated.
The RMSD of all structures converged after 10 ns and stayed stable for the remaining time of MD simulation. This suggests that the systems, dimers of both wild type and the engineered protein, reached equilibrium state after 10 ns (Fig. 2).
Additionally, to investigate mean residual movement over the course of time, the RMSF was calculated for each residue (Fig. 3). As expected, the amino and carboxyl terminal residues of both protein dimers show higher values than the other protein sites. Although lacking the extended C-terminal peptide, the carboxy terminal (C-ter) of the engineered protein (Fig. 2, shown in red and yellow) shows higher degree of mobility compared to the carboxy terminal of the wild type protein (Fig. 2, shown in light and dark green). The higher RMSF value of the N-ter of engineered protein in comparison with the wild type protein, also suggests that in absence of C-ter peptide, the restriction of movement on N-ter is eliminated. Despite the more fluctuations observed in C- and N-ter of the engineered protein, the average mobility of the catalytic flap has declined. This observation might point to a possible internal interaction between the catalytic flaps in the absence of C-ter peptides.
Carboxyl Terminal Peptide Interactions
During the MD simulation, various interactions between terminal peptides and catalytic flaps of both 604-isoform monomers were observed. Due to the disordered nature of these peptides, the interactions between C-ter peptide of one monomer and N-ter, catalytic flap and carboxyl peptide of the other monomer was not unexpected. As shown in figure 4, the C-ter (blue) of the lower monomer, over the course of MD simulation, was extended and reached to the catalytic flap (magenta) of upper monomer, while still partly covering catalytic flap of the same monomer. This interaction was limited to the lower monomer and the extension of C-ter peptide of the upper monomer to reach the catalytic flap of adjacent monomer did not occur. Instead, this peptide covered the catalytic flap of the same monomer, limiting its mobility. This observation is consistent with the low RMSF of the catalytic flaps.
Figure 5 illustrates details of interactions between C-ter and catalytic flap of distinct 604-isoform monomers. In this interaction hydrogen bonds between Gly472, Ser473 and Asp475 of catalytic flap and Thr577, Cys576 and Ser565 of the C-ter peptide of the other monomer were formed, respectively. This interaction might highlight the possible role of C-terminal peptide in attenuation of the catalytic activity among retinal isoforms.
As mentioned in previous sections, interactions between the C- and N-ter peptides of the 604-isoform within the same monomer were observed. These terminal peptides formed two hydrogen bonds between Glu2 of the N-ter peptide and Arg592 of the C-ter peptide and a hydrogen bond between Glu2 and Gly593 (Fig. 6). This explains the lower values of RMSF in N-ter of the 604-isoform. A contact between Glu10 and Arg398 has previously been reported . Along with this finding, interactions between amino terminal peptide and the rest of the protein were not observed.
Interactions In Absence Of C-terminal Peptide
Catalytic flap, not restricted by the presence of C-ter, showed higher mobility. During the MD simulation, for short periods of time interactions between catalytic flaps of engineered monomers were observed. These interactions resulted in forming a hydrogen bond between Lys493 and Asp461 of the catalytic flaps of the upper and lower monomers (Fig. 7). This interaction lasted for the majority of simulation time, leading to lower RMSF value of the catalytic flap in comparison to the wild type isoform in which the corresponding domains have been restricted by the C-ter peptides.
Eliminating the mobility inhibition, caused by C-ter peptide on N-ter peptide, in the engineered 604-isoform, the N-ter peptide extended and formed a short helix in the hinge connecting CBS and catalytic domain. This helix which formed in the hinge of both monomers, occupied the GTP2 binding site. Even though there were no hydrogen bonds formed between the newly formed helix of N-ter and other parts of the protein, the N-ter peptide tends to occupy the binding site of GTP2 (Fig. 8). D283N mutation (equivalent to D226N in canonical numbering) which is located in this binding site, has been identified to be effective in occurrence of RP symptoms. Given the importance of this binding site, it might be suggested that in absence or engagement of C-ter peptide, the N-termainal extension could play a regulatory role in catalytic activity of the 604-isoform.