Research article
QM/MM–PB/SA scoring of the interaction strength between Akt kinase and apigenin analogues

https://doi.org/10.1016/j.compbiolchem.2014.07.002Get rights and content

Highlights

  • The QM/MM–PB/SA is compared with empirical scoring functions and free energy analysis.

  • The QM/MM–PB/SA is used to score the binding affinity of apigenin analogues to Akt.

  • Kinase assay is performed to solidify the results arising from theoretical analysis.

  • Four apigenin analogues are found to exhibit Akt inhibitory activity at nanomolar level.

Abstract

Identification of small-molecule compounds that can bind specifically and stably to protein targets of biological interest is a challenge task in structure-based drug design. Traditionally, several fast approaches such as empirical scoring functions and free energy analysis have been widely used to fulfill for this purpose. In the current study, we raised the rigorous quantum mechanics/molecular mechanics in combination with semi-empirical Poisson–Boltzmann/surface area (QM/MM–PB/SA) as an efficient strategy to characterize the intermolecular interaction between Akt kinase and its small-molecule ligands, although this hybrid approach is computationally expensive as compared to those empirical methods. In a round of experimental activity reproduction test based on a set of known Akt–inhibitor complexes, QM/MM–PB/SA has been shown to perform much better than two widely used scoring functions as well as the sophisticated MM-PB/SA analysis with or without improvement by molecular dynamics (MD) simulations. Next, the QM/MM–PB/SA was employed to screen for strong Akt binders from an apigenin analogue set. Consequently, four compounds, namely apigenin, quercetin, gallocatechin and myricetin, were suggested to have high binding potency to Akt active site. A further kinase assay was conducted to determine the inhibitory activity of the four promising candidates against Akt kinase, resulting in IC50 values of 38.4, 67.5, 157.1 and 25.5 nM, respectively.

Graphical abstract

The rigorous QM/MM–PB/SA is employed to screen for strong Akt binders from a set of apigenin analogues and, consequently, four compounds, namely apigenin, quercetin, gallocatechin and myricetin, are suggested to have high binding potency to Akt active site, which are further solidified by a kinase assay.

  1. Download : Download full-size image

Introduction

Flavonoids consist of a large group of polyphenolic compounds and are ubiquitously present in plants. They are synthesized by phenylpropanoid pathway and can be categorized, according to chemical structure, into flavonols, flavones, flavanones, isoflavones, catechins, anthocyanidins and chalcones. Over 4000 flavonoids have been identified, many of which occur in fruits, vegetables and beverages (Heim et al., 2002). Flavonoids have been shown to have a wide range of biological and pharmacological activities, including anti-allergic, anti-inflammatory, antioxidant, anti-microbial, anti-cancer and anti-diarrheal activities (Si et al., 2009, Chakravarthi et al., 2009, Taupin, 2009, Zhao et al., 2010, Ruela-de-Sousa et al., 2010, Ohno et al., 2013); the activities and functions depend on their structural class, degree of hydroxylation, other substitutions and conjugations, and degree of polymerization (Kumar and Pandey, 2013).

On of the most biologically important flavonoids is the apigenin, which has been shown to be a very potent anti-cancer compound. It beneficially protects against a wide variety of cancers with high selectivity for cancer cells as opposed to non-cancerous cells (Gupta et al., 2001, Patel et al., 2007, Shukla and Gupta, 2010). It also has a very high safety threshold, and active doses can be gained through consuming a vegetable and fruit rich diet. In vitro studies implicated that the apigenin inhibits the development, proliferation and invasion of tumor cells is to inactivate Akt signaling pathway by binding specifically to the Akt kinase, which triggers apoptosis in human cancer (Kaur et al., 2008). The Akt also known as protein kinase B (PKB), is a human homologue of the viral oncogene v-Akt that plays a key role in multiple cellular processes such as glucose metabolism, cell proliferation, transcription and migration (Fayard et al., 2005). However, the molecular mechanism of Akt–apigenin recognition and interaction still remain largely unexplored to date. The apigenin is a flavonoid derivative that is substituted by three hydroxyl functional groups at positions 4′, 5 and 7 of the basic flavonoid skeleton, namely, flavone. Accumulated data revealed that many flavonoids exert chemopreventive effects through acting at protein kinase signaling pathways, more than as conventional hydrogen-donating antioxidants. Recent studies show that flavonoid inhibitors can bind directly to certain protein kinases such as Fyn JAK, MEK1 and PI3K, and then alter their phosphorylation state to regulate multiple cell signaling pathways in carcinogenesis processes (Hou and Kumamoto, 2010). Calorimetric studies and theoretical investigations have revealed an essential role of substituents in flavonoid’s selectivity and affinity toward their cognate targets; minor change in substituent type, position and/or quantity can address a marked effect on the binding behavior of flavonoid ligands to protein receptors (Saija et al., 1995, Jain et al., 1999, Shi et al., 2011).

The flavonoid skeleton possesses a number of positions at which the substitutions can result in various flavonoid derivatives. Thus, empirical rule has been widely used to perform theoretical analysis of protein–flavonoid interaction and rational design of novel flavonoid ligands that can bind specifically and stably to the protein targets of biological interest. Previously, however, we demonstrated a typical non-additive contribution of hydroxyl substituents to Akt–apigenin affinity by integrating several sophisticated molecule modeling techniques, and we also found that the non-additivity can only be described accurately with rigorous QM/MM–PB/SA approach (Lu et al., 2014). Very recently, Wichapong et al. (2014) compared QM/MM–GB/SA, MM-GB/SA and MM-PB/SA in discriminating between active and inactive kinase inhibitors, and results showed that QM/MM–GB/SA performed much better than normal docking scores or MM-GB/SA and MM-PB/SA in classifying active and inactive inhibitors. In the current study, we attempted to implement QM/MM–PB/SA scoring of the relative binding strength of apigenin and its flavonoid analogues to Akt kinase. Before the analysis we have collected a number of small-molecule Akt inhibitors with known biological activities to test the QM/MM–PB/SA scoring performance, and the obtained results were compared systematically with that derived from empirical scoring functions and MM-PB/SA analysis with or without assistance of molecular dynamics (MD) simulations. The method was later used to computationally estimate the relative binding strength of apigenin and its flavonoid analogues to Akt kinase. We also performed kinase assay to test the Akt inhibitory activity of several highly promising candidates screened with the QM/MM–PB/SA scoring strategy.

Section snippets

Small-molecule Akt inhibitors as well as apigenin and its flavonoid analogues

Akt inhibitor set (Table 1). We collected eight existing small-molecule inhibitors that have been reported previously to possess both high affinity and selectivity toward Akt kinase. The biological activities, i.e., IC50 values, of these inhibitors were determined by standard kinase assay experiment; all of the values are at nanomolar level, showing a strong binding potency to Akt. These compounds are diverse in terms of various structure features they possess, including quinoline and indole

Estimation of Akt–inhibitor affinity at different theoretical levels

Four out of the eight Akt inhibitors listed in Table 1 have already been solved complex crystal structures with Akt kinase. The complex structure models of other inhibitors with Akt were manually modified from these crystal templates using a protocol described above. These crystal structures or coarse-grained structure models were relaxed via atomistic molecular dynamics (MD) simulations with AMBER03 force field (Duan et al., 2003). Because the unbound Akt structure is in the ‘DFG-in’ state

Acknowledgements

This work was supported by the National Natural Science Foundation of China (Nos. 31271272 and 31200602), the Jiangsu University for Advanced Professionals (No. 1281330021) and the China Postdoctoral Science Foundation (No. 2012M521001).

References (49)

  • M.J.S. Dewar et al.

    Development and use of quantum mechanical molecular models. 76. AM1: a new general purpose quantum mechanical molecular model

    J. Am. Chem. Soc.

    (1985)
  • Y. Duan et al.

    A point-charge force field for molecular mechanics simulations of proteins based on condensed-phase quantum mechanical calculations

    J. Comput. Chem.

    (2003)
  • E. Fayard et al.

    Protein kinase B/Akt at a glance

    J. Cell. Sci.

    (2005)
  • M.J. Frisch et al.

    Gaussian 03

    (2003)
  • N. Gulati et al.

    The antiproliferative effect of quercetin in cancer cells is mediated via inhibition of the PI3K-Akt/PKB pathway

    Anticancer Res.

    (2006)
  • D.X. Hou et al.

    Flavonoids as protein kinase inhibitors for cancer chemoprevention: direct binding and molecular modeling

    Antioxid. Redox. Signal

    (2010)
  • T. Hou et al.

    Computational analysis and prediction of the binding motif and protein interacting partners of the Abl SH3 domain

    PLoS Comput. Biol.

    (2006)
  • T. Hou et al.

    Predicting drug resistance of the HIV-1 protease using molecular interaction energy components

    Proteins

    (2009)
  • A. Jain et al.

    Reactivities of flavonoids with different hydroxyl substituents for the cleavage of DNA in the presence of Cu(II)

    Phytother. Res.

    (1999)
  • W.L. Jorgensen et al.

    Comparison of simple potential functions for simulating liquid water

    J. Chem. Phys.

    (1983)
  • P. Kaur et al.

    Plant flavonoid apigenin inactivates Akt to trigger apoptosis in human prostate cancer: an in vitro and in vivo study

    Carcinogenesis

    (2008)
  • W. Kim et al.

    Myricetin inhibits Akt survival signaling and induces Bad-mediated apoptosis in a low dose ultraviolet (UV)-B-irradiated HaCaT human immortalized keratinocytes

    J. Radiat. Res.

    (2010)
  • J. Ko et al.

    The FALC-Loop web server for protein loop modeling

    Nucleic Acids Res.

    (2011)
  • P.A. Kollman et al.

    Calculating structures and free energies of complex molecules: combining molecular mechanics and continuum models

    Acc. Chem. Res.

    (2000)
  • Cited by (11)

    • Isolation of apigenin from subcritical water extracts: Optimization of the process

      2017, Journal of Supercritical Fluids
      Citation Excerpt :

      Special focus has been put on its anticancer activities and numerous scientific investigations have focused on effects of apigenin on apoptosis, inhibition of cell growth and angiogenesis [19]. Its protective role against a wide variety of cancers with high selectivity for cancer cells has been confirmed in numerous studies [19–22]. Although apigenin is present in many herbs, fruits and vegetables, one of the most common sources of apigenin is chamomile (Chamomilla matricaria L.).

    • Structural and energetic insight into the interactions between the benzolactam inhibitors and tumor marker HSP90α

      2015, Computational Biology and Chemistry
      Citation Excerpt :

      The integration step was set to 2 fs in the MD simulations. The popular MM–GBSA method (Sun et al., 2014; Lu et al., 2011; Bai et al., 2014; Shang et al., 2014; Lu et al., 2014d) was used to assess the binding free energy (ΔGbinding) between HSP90α and each compound. The 101 snapshots were extracted from the 20–50 ns MD trajectories with an equal interval of 300 ps to the ΔGbinding∙ ΔGbinding was calculated using Eq. (1).

    • Structure-based grafting and identification of kinase-inhibitors to target mTOR signaling pathway as potential therapeutics for glioblastoma

      2015, Computational Biology and Chemistry
      Citation Excerpt :

      This nonredundant set is structurally diverse in ligand chemicals and represents a wide range of kinase–inhibitor binding affinities. A method proposed by Lu et al. (2014) was modified here to carry out computational modeling of non-cognate mTOR–inhibitor complex structures. Briefly, ligand molecule is assumed to adopt a similar binding mode to interact with different kinase receptors since the kinase family members are highly conserved.

    View all citing articles on Scopus
    View full text