Mechanism and Significance of Post-Translational Modifications in the Large (LS) and Small (SS) Subunits of Ribulose-1,5 Bisphosphate Carboxylase/Oxygenase

Abstract

This project focused on a molecular and biochemical characterization of the protein methyltransferases responsible for methylation of the LS and SS in Rubisco, and the associated functional consequences accompanying these modifications. Our results provided some of the most informative structural and mechanistic understandings of SET domain protein methyltransferases. These results also positioned us to provide the first unambiguous assignment of the kinetic reaction mechanism for SET-domain protein methyltransferases, and to design and engineer an alternative substrate for Rubisco LSMT, enabling substrate specificity and functional significance studies. We demonstrated that the minimal substrate recognized by Rubisco LSMT is free lysine as well as monomethyllysine, an observation corroborated both by structural analyses as well as enzymatic activity and subsequent product distribution analyses. Ternary complexes between Rubisco LSMT and free lysine compared to complexes with monomethyllysine demonstrated that the structural basis for multiple methyl group additions is a consequence of hydrogen-bond driven spatial shifts in the amino group of Lys-14, which maintains the direct in-line geometry necessary for SN2 nucleophilic attack. The structural observations are also consistent with the previous proposal that the multiplicity of methyl group additions takes place through a processive mechanism, with successive methyl group additions to an enzyme protein complex which does not disassociate prior to the formation of trimethyllysine. This mechanism has important implications, since the regulation of gene expression by SET domain histone methyltransferases is not only dependent on site-specific lysine methylation, but also the degree of methylation. We examined the kinetic reaction mechanism for three different types of SET domain protein methyltransferases, each under conditions supporting mono-, di-, or trimethyllysine formation corroborated by product analyses. Additionally, the tight initial binding of Rubisco LSMT to Rubisco also allowed us to design a novel immobilized complex between Rubisco and Rubisco LSMT, which allowed for an unambiguous demonstration of the requirement for trimethyllysine formation prior to disassociation of the Rubisco LSMT:Rubisco complex, and therefore proof of the processive mechanism for methyl group transfer. These kinetic studies also demonstrated that an important factor has been overlooked in all kinetic analyses of SET domain protein methyltransferases reported to date. This factor is the influence of the low turnover number for SET domain protein methyltransferases and how, relative to the time-frame of kinetic enzyme assays, this can generate changes in kinetic profiles shifting reciprocal plot patterns from random/ordered bi-bi to the real kinetic reaction mechanism plots of ping-pong. Although the ternary complexes of Rubisco LSMT with S-Adenosylhomocysteine and lysine and monomethyllysine were informative in regard to reaction mechanism, they were not helpful in identifying the mechanism used by Rubisco LSMT for determining substrate specificity. We were unsuccessful at obtaining ternary complexes of Rubisco LSMT with bound synthetic polypeptide substrates, as has been reported for several histone methyltransferases. However, we were able to model a polypeptide sequence corresponding to the N-terminal region of the LS of Rubisco into the apparent substrate binding cleft in Rubisco LSMT. Knowledge of the determinants of polypeptide substrate specificity are important for identifying possible alternate substrates, as well as the possibility of generating more desirable substrates amenable to site-directed mutagenesis experiments unlike Rubisco. We determined that Rubisco LSMT is capable of methylating synthetic polypeptide mimics of the N-terminal region of the LS, both free as well as conjugated to keyhole limpet hemacyanin, but with considerable less efficiency than intact holoenzyme.