The stability of enzymes is critical for their application in industrial processes, which generally require different conditions from the natural enzyme environment. Both rational and random protein engineering approaches have been used to increase stability, with the latter requiring extensive experimental effort for the screening of variants. Moreover, some general rules addressing the molecular origin of protein thermostability have been established. Herein, we demonstrate the use of molecular dynamics simulations to gain molecular level understanding of protein thermostability and to engineer stabilizing mutations. Carbonic anhydrase (CA) is an enzyme with a high potential for biotechnological carbon capture applications, provided it can be engineered to withstand the high temperature process environments, inevitable in most gas treatment units. In this study, we used molecular dynamics simulations at 343, 353, and 363 K to study the relationship between structure flexibility and thermostability in bacterial α-CAs and applied this knowledge to the design of mutants with increased stability. The most thermostable α-CA known, TaCA from Thermovibrio ammonificans, had the most rigid structure during molecular dynamics simulations, but also showed regions with high flexibility. The most flexible amino acids in these regions were identified from root mean square fluctuation (RMSF) studies, and stabilizing point mutations were predicted based on their capacity to improve the calculated free energy of unfolding. Disulfide bonds were also designed at sites with suitable geometries and selected based on their location at flexible sites, assessed by B-factor calculation. Molecular dynamics simulations allowed the identification of five mutants with lower RMSF of the overall structure at 400 K, compared to wild-type TaCA. Comparison of free-energy landscapes between wild-type TaCA and the most promising mutants, Pro165Cys-Gln170Cys and Asn140Gly, showed an increased conformational stability of the mutants at 400 K.
Site-directed mutagenesis of the oxyanion-containing amino acid Q114 in the recombinant thermophilic T1 lipase previously isolated from Geobacillus zalihae was performed to elucidate its role in the enzyme's enantioselectivity and reactivity. Substitution of Q114 with a hydrophobic methionine to yield mutant Q114M increased enantioselectivity (3.2-fold) and marginally improved reactivity (1.4-fold) of the lipase in catalysing esterification of ibuprofen with oleyl alcohol. The improved catalytic efficiency of Q114L was concomitant with reduced flexibility in the active site while the decreased enantioselectivity of Q114L could be directly attributed to diminished electrostatic repulsion of the substrate carboxylate ion that rendered partial loss in steric hindrance and thus enantioselectivity. The highest E-values for both Q114L (E-value 14.6) and Q114M (E-value 48.5) mutant lipases were attained at 50°C, after 12-16h, with a molar ratio of oleyl alcohol to ibuprofen of 1.5:1 and at 2.0% (w/v) enzyme load without addition of molecular sieves. Pertinently, site-directed mutagenesis on the Q114 oxyanion of T1 resulted in improved enantioselectivity and such approach may be applicable to other lipases of the same family. We demonstrated that electrostatic repulsion phenomena could affect flexibility/rigidity of the enzyme-substrate complex, aspects vital for enzyme activity and enantioselectivity of T1.
Interests in Acinetobacter haemolyticus lipases are showing an increasing trend concomitant with growth of the enzyme industry and the widening search for novel enzymes and applications. Here, we present a structural model that reveals the key catalytic residues of lipase KV1 from A. haemolyticus. Homology modeling of the lipase structure was based on the structure of a carboxylesterase from the archaeon Archaeoglobus fulgidus as the template, which has a sequence that is 58% identical to that of lipase KV1. The lipase KV1 model is comprised of a single compact domain consisting of seven parallel and one anti-parallel β-strand surrounded by nine α-helices. Three structurally conserved active-site residues, Ser165, Asp259, and His289, and a tunnel through which substrates access the binding site were identified. Docking of the substrates tributyrin and palmitic acid into the pH 8 modeled lipase KV1 active sites revealed an aromatic platform responsible for the substrate recognition and preference toward tributyrin. The resulting binding modes from the docking simulation correlated well with the experimentally determined hydrolysis pattern, for which pH 8 and tributyrin being the optimum pH and preferred substrate. The results reported herein provide useful insights into future structure-based tailoring of lipase KV1 to modulate its catalytic activity.
Discovery of imatinib mesylate (IM) as the targeted BCR-ABL protein tyrosine kinase inhibitor (TKI) has resulted in its use as the frontline therapy for chronic myeloid leukemia (CML) across the world. Although high response rates are observed in CML patients who receive IM treatment, a significant number of patients develop resistance to IM. Resistance to IM in patients has been associated with a heterogeneous array of mechanisms of which point mutations within the ABL tyrosine kinase domain (TKD) are the frequently documented. The types and frequencies of mutations reported in different population studies have shown wide variability. We screened 125 Malaysian CML patients on IM therapy who showed either TKI refractory or resistance to IM to investigate the frequency and pattern of BCR-ABL kinase domain mutations among Malaysian CML patients undergoing IM therapy and to determine the clinical significance. Mutational screening using denaturing high performance liquid chromatography (dHPLC) followed by DNA sequencing was performed on 125 IM resistant Malaysian CML patients. Mutations were detected in 28 patients (22.4%). Fifteen different types of mutations (T315I, E255K, G250E, M351T, F359C, G251E, Y253H, V289F, E355G, N368S, L387M, H369R, A397P, E355A, D276G), including 2 novel mutations were identified, with T315I as the predominant type of mutation. The data generated from clinical and molecular parameters studied were correlated with the survival of CML patients. Patients with Y253H, M351T and E355G TKD mutations showed poorer prognosis compared to those without mutation. Interestingly, when the prognostic impact of the observed mutations was compared inter-individually, E355G and Y253H mutations were associated with more adverse prognosis and shorter survival (P=0.025 and 0.005 respectively) than T315I mutation. Results suggest that apart from those mutations occurring in the three crucial regions (catalytic domain, P-loop and activation-loop), other rare mutations also may have high impact in the development of resistance and adverse prognosis. Presence of mutations in different regions of BCR-ABL TKD leads to different levels of resistance and early detection of emerging mutant clones may help in decision making for alternative treatment. Serial monitoring of BCR-ABL1 transcripts in CML patients allows appropriate selection of CML patients for BCR-ABL1 KD mutation analysis associated with acquired TKI resistance. Identification of these KD mutations is essential in order to direct alternative treatments in such CML patients.