Is breaking of a hydrogen bond enough to lead to drug resistance?

Exploring the Impact of Protein Chain Selection in Binding Energy Calculations with DFT

Calculation of binding free energies between a protein and a ligand are highly desired for computer-aided drug design. Here we approximate the binding energies of ABL1, an enzyme which is the target for drugs used in the treatment of chronic myeloid leukaemia, with minimal models and density functional theory (DFT). Starting from the crystal structures of protein-drug complexes, we estimated the binding free energies having used all available individual molecules (protein chains) within each structure, not only a single one as commonly used, in order to see if the choice of the protein chain is important in such calculations. Differences were observed between chains in the same file. Energy decomposition analysis (EDA) revealed that the most important factors for binding were exchange, repulsion and electrostatics. The desolvation term varied dramatically between the inhibitors (between 4.2 and 92.3 kcal/mol). All functionals showed similar patterns in the EDA and in discriminating between the ligands. Non-covalent interactions (NCI) analysis was used to further explain the differences between protein chains and functionals. Overall, it is shown that small minimal models of a drug binding site can be useful to infer on the suitability of an initial crystal structure for further analysis such as EDA.

Trans vs. cis: a computational study of enasidenib resistance due to IDH2 mutations

Isocitrate dehydrogenase 2 (IDH2) is a homodimeric enzyme that plays an important role in energy production. A mutation R140Q in one monomer makes the enzyme tumourigenic. Enasidenib is an effective inhibitor of IDH2/R140Q. A secondary mutation Q316E leads to enasidenib resistance. This mutation was hitherto only found in trans, i.e. where one monomer has the R140Q mutation and the other carries the Q316E mutation. It is not clear if the mutation only leads to resistance when in trans or if it has been discovered in trans only by chance, since it was only reported in two patients. Using molecular dynamics (MD) simulations we show that the binding of enasidenib to IDH2 is indeed much weaker when the Q316E mutation takes place in trans not in cis, which provides a molecular explanation for the clinical finding. This is corroborated by non-covalent interaction (NCI) analysis and DFT calculations. Whereas the MD simulations show a loss of one hydrogen bond upon the resistance mutation, NCI and energy decomposition analysis (EDA) reveal that a multitude of interactions are weakened.

Structural and thermodynamic characterization of allosteric transitions in human serum albumin with metadynamics simulations

Human serum albumin (HSA) is the most prominent protein in blood plasma, responsible for the maintenance of blood viscosity and transport of endogenous and exogenous molecules. Fatty acids (FA) are the most common ligands of HSA and their binding can modify the protein's structure. The protein can assume two well-defined conformations, referred to as 'Neutral' and 'Basic'. The Neutral (N) state occurs at pH close to 7.0 and in the absence of bound FA. The Basic (B) state occurs at pH higher than 8.0 or when the protein is bound to long-chain FA. HSA's allosteric behaviour is dependent on the number on FA bound to the structure. However, the mechanism of this allosteric regulation is not clear. To understand how albumin changes its conformation, we compared a series of HSA structures deposited in the protein data bank to identify the minimum amount of FA bound to albumin, which is enough to drive the allosteric transition. Thereafter, non-biased molecular dynamics (MD) simulations were used to track protein's dynamics. Surprisingly, running an ensemble of relatively short MD simulations, we observed rapid transition from the B to the N state. These simulations revealed differences in the mobilities of the protein's subdomains, with one domain unable to fully complete its transition. To track the transition dynamics in full, we used these results to choose good geometrical collective variables for running metadynamics simulations. The metadynamics calculations showed that there was a low energy barrier for the transition from the B to the N state, while a higher energy barrier was observed for the N to the B transition. These calculations also offered valuable insights into the transition process.

Resistance to a tyrosine kinase inhibitor mediated by changes to the conformation space of the kinase

Gilteritinib is a highly selective and effective inhibitor of the FLT3/ITD mutated protein, and is used successfully in treating acute myeloid leukaemia (AML). Unfortunately, tumour cells gradually develop resistance to gilteritinib due to mutations in the molecular drug target. The atomistic details behind this observed resistance are not clear, since the protein structure of the complex is only available in the inactive state, while the drug binds better to the active state. To overcome this limitation, we used a computer-aided approach where we docked gilteritinib to the active site of FLT3/ITD and calculated the Gibbs free energy difference between the binding energies of the parental and mutant enzymes. These calculations agreed with experimental estimations for one mutation (F691L) but not the other (D698N). To further understand how these mutations operate, we used metadynamics simulations to study the conformational landscape of the activation process. Both mutants show a lower activation energy barrier which suggests that they are more likely to adopt an active state until inhibited, making the mutant enzymes more active. This suggests that a higher efficiency of tyrosine kinases contributes to resistance not only against type 2 but also against type 1 kinase inhibitors.

Therapeutic Targeting of FLT3 in Acute Myeloid Leukemia: Current Status and Novel Approaches

FMS-like tyrosine kinase 3 (FLT3) is mutated in approximately 30% of acute myeloid leukemia (AML) patients. The presence of FLT3-ITD (internal tandem duplication, 20-25%) mutation and, to a lesser extent, FLT3-TKD (tyrosine kinase domain, 5-10%) mutation is associated with poorer diagnosis and therapy response since the leukemic cells become hyperproliferative and resistant to apoptosis after continuous activation of FLT3 signaling. Targeting FLT3 has been the focus of many pre-clinical and clinical studies. Hence, many small-molecule FLT3 inhibitors (FLT3is) have been developed, some of which are approved such as midostaurin and gilteritinib to be used in different clinical settings, either in combination with chemotherapy or alone. However, many questions regarding the best treatment strategy remain to be answered. On the other hand, various FLT3-dependent and -independent resistance mechanisms could be evolved during FLT3i therapy which limit their clinical impact. Therefore, identifying molecular mechanisms of resistance and developing novel strategies to overcome this obstacle is a current interest in the field. In this review, recent studies of approved FLT3i and knowledge about major resistance mechanisms of clinically approved FLT3i's will be discussed together with novel treatment approaches such as designing novel FLT3i and dual FLT3i and combination strategies including approved FLT3i plus small-molecule agents targeting altered molecules in the resistant cells to abrogate resistance. Moreover, how to choose an appropriate FLT3i for the patients will be summarized based on what is currently known from available clinical data. In addition, strategies beyond FLT3i's including immunotherapeutics, small-molecule FLT3 degraders, and flavonoids will be summarized to highlight potential alternatives in FLT3-mutated AML therapy.

Multistep orthophosphate release tunes actomyosin energy transduction

Muscle contraction and a range of critical cellular functions rely on force-producing interactions between myosin motors and actin filaments, powered by turnover of adenosine triphosphate (ATP). The relationship between release of the ATP hydrolysis product ortophosphate (Pi) from the myosin active site and the force-generating structural change, the power-stroke, remains enigmatic despite its central role in energy transduction. Here, we present a model with multistep Pi-release that unifies current conflicting views while also revealing additional complexities of potential functional importance. The model is based on our evidence from kinetics, molecular modelling and single molecule fluorescence studies of Pi binding outside the active site. It is also consistent with high-speed atomic force microscopy movies of single myosin II molecules without Pi at the active site, showing consecutive snapshots of pre- and post-power stroke conformations. In addition to revealing critical features of energy transduction by actomyosin, the results suggest enzymatic mechanisms of potentially general relevance.

Release of the ATP hydrolysis product orthophosphate (Pi) from the myosin active site is central in force generation but is poorly understood. Here, Moretto et al. present evidence for multistep Pi-release reconciling apparently contradictory results.

Allosteric Enhancement of the BCR-Abl1 Kinase Inhibition Activity of Nilotinib by Co-Binding of Asciminib

Inhibitors that bind competitively to the ATP binding pocket in the kinase domain of the oncogenic fusion protein BCR–Abl1 are used successfully in targeted therapy of chronic myeloid leukemia (CML). Such inhibitors provided the first proof of concept that kinase inhibition can succeed in a clinical setting. However, emergence of drug resistance and dose-dependent toxicities limit the effectiveness of these drugs. Therefore, treatment with a combination of drugs without overlapping resistance mechanisms appears to be an appropriate strategy. In the present work, we explore the effectiveness of combination therapies of the recently developed allosteric inhibitor asciminib with the ATP-competitive inhibitors nilotinib and dasatinib in inhibiting the BCR–Abl1 kinase activity in CML cell lines. Through these experiments, we demonstrate that asciminib significantly enhances the inhibition activity of nilotinib, but not of dasatinib. Exploring molecular mechanisms for such allosteric enhancement via systematic computational investigation incorporating molecular dynamics, metadynamics simulations, and density functional theory calculations, we found two distinct contributions. First, binding of asciminib triggers conformational changes in the inactive state of the protein, thereby making the activation process less favorable by ∼4 kcal/mol. Second, the binding of asciminib decreases the binding free energies of nilotinib by ∼3 and ∼7 kcal/mol for the wildtype and T315I-mutated protein, respectively, suggesting the possibility of reducing nilotinib dosage and lowering risk of developing resistance in the treatment of CML.

Evaluation of residue variability in a conformation-specific context and during evolutionary sequence reconstruction narrows drug resistance selection in Abl1 tyrosine kinase

Diseases with readily available therapies may eventually prevail against the specific treatment by the acquisition of resistance. The constitutively active Abl1 tyrosine kinase known to cause chronic myeloid leukemia is an example, where patients may experience relapse after small inhibitor drug treatment. Mutations in the Abl1 tyrosine kinase domain (Abl1-KD) are a critical source of resistance and their emergence depends on the conformational states that have been observed experimentally: the inactive state, the active state, and the intermediate inactive state that resembles Src kinase. Understanding how resistant positions and amino acid identities are determined by selection pressure during drug treatment is necessary to improve future drug development or treatment decisions. We carry out in silico site-saturation mutagenesis over the Abl1-KD structure in a conformational context to evaluate the in situ and conformational stability energy upon mutation. Out of the 11 studied resistant positions, we determined that 7 of the resistant mutations favored the active conformation of Abl1-KD with respect to the inactive state. When, instead, the sequence optimization was modeled simultaneously at resistant positions, we recovered five known resistant mutations in the active conformation. These results suggested that the Abl1 resistance mechanism targeted substitutions that favored the active conformation. Further sequence variability, explored by ancestral reconstruction in Abl1-KD, showed that neutral genetic drift, with respect to amino acid variability, was specifically diminished in the resistant positions. Since resistant mutations are susceptible to chance with a certain probability of fixation, combining methodologies outlined here may narrow and limit the available sequence space for resistance to emerge, resulting in more robust therapeutic treatments over time.

Rotating between ponatinib and imatinib temporarily increases the efficacy of imatinib as shown in a chronic myeloid leukaemia model

Targeted therapies for chronic myeloid leukaemia (CML) are effective, but rarely curative. Patients typically require treatment indefinitely, which gives ample time for drug resistance to evolve. Drug resistance issues are one of the main causes of death owing to CML, thus any means of preventing resistance are of importance. Drug rotations, wherein treatment is switched periodically between different drugs are one such option, and have been theorized to delay the onset of resistance. In vitro testing of drug rotation therapy is a first step towards applying it in animal or human trials. We developed a method for testing drug rotation protocols in CML cell lines based around culturing cells with a moderate amount of inhibitors interspersed with washing procedures and drug swaps. Drug rotations of imatinib and ponatinib were evaluated in a CML specific cell line, KCL-22. The growth of KCL-22 cells was initially reduced by a drug rotation, but the cells eventually adapted to the protocol. Our results show that ponatinib in a drug rotation temporarily sensitizes the cells to imatinib, but the effect is short-lived and is eventually lost after a few treatment cycles. Possible explanations for this observation are discussed.

The molecular mechanisms behind activation of FLT3 in acute myeloid leukemia and resistance to therapy by selective inhibitors

Acute myeloid leukemia is an aggressive cancer, which, in spite of increasingly better understanding of its genetic background remains difficult to treat. Mutations in the FLT3 gene are observed in ≈30% of the patients. Most of these mutations are internal tandem duplications (ITDs) of a sequence within the protein coding region, an activation mechanism that is almost non-existent with other genes and cancers. As patients each carry their own unique set of mutations, it is challenging to understand how ITDs activate the protein, and ascertain the risk for each individual patient. Available treatment options are limited due to development of drug resistance. Here, recent studies are reviewed that help to better understand the molecular mechanism behind activation of the FLT3 protein due to mutations. It is argued that difference in mutation sequences and especially location might be coupled to prognosis. When it comes to FLT3 inhibitors, key differences between them can be attributed to the mode of inhibition (type-1 and type-2 inhibitors), effective inhibitory coefficient in the blood plasma and off-target binding. Accounting for the position and length of insertions may in the future be used to predict prognosis and rationalise treatment. Development of new inhibitors must take into account the potential for resistance mutations. Inhibitors aimed at multiple specific targets are currently being developed. These, and as well as combination therapies will hopefully lead to longer periods during which targeted FLT3 therapy will remain effective.

Preferential Binding of Lanthanides to Methanol Dehydrogenase Evaluated with Density Functional Theory

Methanol dehydrogenase (MDH) is an enzyme used by certain bacteria for the oxidation of methanol to formaldehyde, which is a necessary metabolic reaction. The discovery of a lanthanide-dependent MDH reveals that lanthanide ions (Ln³⁺) have a role in biology. Two types of MDH exist in methane-utilizing bacteria: one that is Ca²⁺-dependent (MxaF) and another that is Ln³⁺-dependent. Given that the triply charged Ln³⁺ are strongly hydrated, it is not clear how preference for Ln³⁺ is manifested and if the Ca²⁺-dependent MxaF protein can also bind Ln³⁺ ions. A computational approach was used to estimate the Gibbs energy differences between the binding of Ln³⁺ and Ca²⁺ to MDH using density functional theory. The results show that both proteins bind La³⁺ with higher affinity than Ca²⁺, albeit with a more pronounced difference in the case of Ln³⁺-dependent MDH. Interestingly, the binding of heavier lanthanides is preferred over the binding of La³⁺, with Gd³⁺ showing the highest affinity for both proteins of all Ln³⁺ ions that were tested (La³⁺, Sm³⁺, Gd³⁺, Dy³⁺, and Lu³⁺). Energy decomposition analysis reveals that the higher affinity of La³⁺ than Ca²⁺ to MDH is due to stronger contributions of electrostatics and polarization, which overcome the high cost of desolvating the ion.