Clarification of the mechanism of acylation reaction and origin of substrate specificity of the serine-carboxyl peptidase sedolisin through QM/MM free energy simulations

Serine-Carboxyl Peptidases, Sedolisins: From Discovery to Evolution

Sedolisin is a proteolytic enzyme, listed in the peptidase database MEROPS as a founding member of clan SB, family S53. This enzyme, although active at low pH, was originally shown not to be inhibited by an aspartic peptidase specific inhibitor, S-PI (pepstatin Ac). In this Perspective, the S53 family is described from the moment of original identification to evolution. The representative enzymes of the family are sedolisin, kumamolisin, and TPP-1. They exhibit the following unique features. (1) The fold of the molecule is similar to that of subtilisin, but the catalytic residues consist of a triad, Ser/Glu/Asp, that is unlike the Ser/His/Asp triad of subtilisin. (2) The molecule is expressed as a pro-form composed of the amino-terminal prosegment and the active domain. Additionally, some members of this family have an additional, carboxy-terminal prosegment. (3) Their optimum pH for activity is in the acidic region, not in the neutral to alkaline region where subtilisin is active. (4) Their distribution in nature is very broad across the three kingdoms of life. (5) Some of these enzymes from fungi and bacteria are pathogens to plants. (6) Some of them have significant potential applications for industry. (7) The lack of a TPP-1 gene in human brain is the cause of incurable juvenile neuronal ceroid lipofuscinosis (Batten's disease).

QM/MM free energy Simulations of an efficient Gluten Hydrolase (Kuma030) Implicate for a Reactant-State Based Protein-Design Strategy for General Acid/Base Catalysis

It is a grand attraction for contemporary biochemists to computationally design enzymes for novel chemical transformation or improved catalytic efficiency. Rosetta by Baker et al. is no doubt the leading software in the protein design society. Generally, optimization of the transition state (TS) is part of the Rosetta’s protocol to enhance the catalytic efficiency of target enzymes, since TS stabilization is the determining factor for catalytic efficiency based on the TS theory (TST). However, it is confusing that optimization of the reactant state (RS) also results in significant improvement of catalytic efficiency in some cases, such as design of gluten hydrolase (Kuma030). Therefore, it is interesting to uncover underlying reason why a better binding in the RS leading to an increased k_cat. In this study, the combined quantum mechanical/molecular mechanical (QM/MM) molecular dynamics (MD) and free energy (PMF) simulations, pK_a calculation, and the statistical analysis such as the ANOVA test were carried out to shed light on the interesting but elusive question. By integration of our computational results and general acid/base theory, we answered the question why optimization of RS stabilization leads to a better TS stabilization in the general acid/base catalysis. In addition, a new and simplified protein-design strategy is proposed for the general acid/base catalysis. The idea, that application of traditional well-defined enzyme mechanism to protein design strategy, would be a great help for methodology development of protein design.

QM/MM Investigation of Substrate and Product Specificities of Suv4-20h2: How Does This Enzyme Generate Dimethylated H4K20 from Monomethylated Substrate?

Protein lysine methyltransferases (PKMTs) catalyze the methylation of lysine residues on histone proteins in the regulation of chromatin structure and gene expression. In contrast to many other PKMTs for which unmodified lysine is the methylation target, the enzymes in the Suv4-20 family are able to generate dimethylated product (H4K20me2) based exclusively on the monomethylated H4K20 substrate (H4K20me1). The origin of such substrate/product specificity is still not clear. Here, molecular dynamics (MD) and free energy (potential of mean force) simulations are undertaken using quantum mechanical/molecular mechanical (QM/MM) potentials to understand the substrate/product specificities of Suv4-20h2, a member of the Suv4-20 family. The free energy barriers for mono-, di-, and trimethylation in Suv4-20h2 obtained from the simulations are found to be well correlated with the specificities observed experimentally with the allowed dimethylation based on the H4K20me1 substrate and prohibited monomethylation and trimethylation based on H4K20 and H4K20me2, respectively. It is demonstrated that the reason for the relatively efficient dimethylation is an effective transition state (TS) stabilization through strengthening the CH···O interactions as well as the presence of a cation-π interaction at the transition state. The simulations also show that the failures of Suv4-20h2 to catalyze monomethylation and trimethylation are due, respectively, to a less effective TS stabilization and inability of the reactant complex containing H4K20me2 to adopt a reactive (near attack) configuration for methyl transfer. The results suggest that care must be exercised in the prediction of the substrate specificity based only on the existence of near attack configurations in substrate complexes.

Analysis of catalytic properties of tripeptidyl peptidase I (TTP-I), a serine carboxyl lysosomal protease, and its detection in tissue extracts using selective FRET peptide substrate

Tripeptidyl peptidase I (TPP-I), also named ceroid lipofuscinosis 2 protease (CLN2p), is a serine carboxyl lysosomal protease involved in neurodegenerative diseases, and has both tripeptidyl amino- and endo- peptidase activities under different pH conditions. We developed fluorescence resonance energy transfer (FRET) peptides using tryptophan (W) as the fluorophore to study TPP-I hydrolytic properties based on previous detailed substrate specificity study (Tian Y. et al., J. Biol. Chem. 2006, 281:6559-72). Tripeptidyl amino peptidase activity is enhanced by the presence of amino acids in the prime side and the peptide NH2-RWFFIQ-EDDnp is so far the best substrate described for TPP-I. The hydrolytic parameters of this peptide and its analogues indicated that the S4 subsite of TPP-I is occluded and there is an electrostatic interaction of the positively charged substrate N-terminus amino group and a negative locus in the region of the enzyme active site. KCl activated TPP-I in contrast to the inhibition by Ca(2+) and NaCl. Solvent kinetic isotope effects (SKIEs) show the importance of the free N-terminus amino group of the substrates, whose absence results in a more complex solvent-dependent enzyme: substrate interaction and catalytic process. Like pure TPP-I, rat spleen and kidney homogenates cleaved NH2-RWFFIQ-EDDnp only at F-F bond and is not inhibited by pepstatin, E-64, EDTA or PMSF. The selectivity of NH2-RWFFIQ-EDDnp to TPP-I was also demonstrated by the 400 times higher k(cat)/K(M) compared to generally used substrate, NH2-AAF-MCA and by its resistance to hydrolysis by cathepsin D that is present in high levels in kidneys.

Substrate-Assisted Catalysis in the Reaction Catalyzed by Salicylic Acid Binding Protein 2 (SABP2), a Potential Mechanism of Substrate Discrimination for Some Promiscuous Enzymes

Although one of an enzyme's hallmarks is the high specificity for their natural substrates, substrate promiscuity has been reported more frequently. It is known that promiscuous enzymes generally show different catalytic efficiencies to different substrates, but our understanding of the origin of such differences is still lacking. Here we report the results of quantum mechanical/molecular mechanical simulations and an experimental study of salicylic acid binding protein 2 (SABP2). SABP2 has promiscuous esterase activity toward a series of substrates but shows a high activity toward its natural substrate, methyl salicylate (MeSA). Our results demonstrate that this enzyme may use substrate-assisted catalysis involving the hydroxyl group from MeSA to enhance the activity and achieve substrate discrimination.

New families of carboxyl peptidases: serine-carboxyl peptidases and glutamic peptidases

Peptidases or proteinases are now classified into seven families based on the nature of the catalytic residues [MEROPS-the peptidase database (http://merops.sanger.ac.uk/)]. They are aspartic- (first described in 1993), cysteine- (1993), serine- (1993) metallo- (1993), threonine- (1997), glutamic- (2004) and asparagine-peptidase (2010). By using an S-PI (pepstatin Ac) as a probe, a new subfamily of serine peptidase, serine-carboxyl peptidase (sedolisin) was discovered in 2001. In addition, the sixth family of peptidase, glutamic peptidase (eqolisin) was also discovered in 2004. The former peptidase is widely distributed in nature from archea to mammals, including humans. One of these enzymes is related to a human fatal hereditable disease, Batten disease. In contrast, the distribution of the latter peptidases is limited, with most of them found in human or plant pathogenic fungi. One such enzyme was isolated from a fungal infection in an HIV-infected patient. In this review, the background of the findings, and crystal structures, catalytic mechanisms, substrates specificities and distribution of the new peptidase families are described.