Hit identification of novel heparanase inhibitors by structure- and ligand-based approaches

https://doi.org/10.1016/j.bmc.2013.01.033Get rights and content

Abstract

Heparanase is a key enzyme involved in the dissemination of metastatic cancer cells. In this study a combination of in silico techniques and experimental methods was used to identify new potential inhibitors against this target. A 3D model of heparanase was built from sequence homology and applied to the virtual screening of a library composed of 27 known heparanase inhibitors and a commercial collection of drugs and drug-like compounds. The docking results from this campaign were combined with those obtained from a pharmacophore model recently published based in the same set of chemicals. Compounds were then ranked according to their theoretical binding affinity, and the top-rated commercial drugs were selected for further experimental evaluation.

Biophysical methods (NMR and SPR) were applied to assess experimentally the interaction of the selected compounds with heparanase. The binding site was evaluated via competition experiments, using a known inhibitor of heparanase. Three of the selected drugs were found to bind to the active site of the protein and their KD values were determined. Among them, the antimalarial drug amodiaquine presented affinity towards the protein in the low-micromolar range, and was singled out for a SAR study based on its chemical scaffold. A subset of fourteen 4-arylaminoquinolines from a global set of 249 analogues of amodiaquine was selected based on the application of in silico models, a QSAR solubility prediction model and a chemical diversity analysis. Some of these compounds displayed binding affinities in the micromolar range.

Graphical abstract

Novel inhibitors of heparanase have been discovered by the combined use of pharmacophore and docking strategies. The most relevant one was amodiaquine, a known antimalarial compound. The application of different biophysical techniques such as NMR and SPR confirmed the interaction of these compounds with the catalytic site of heparanase.

  1. Download : Download full-size image

Introduction

Heparanase is an endoglycosidase involved in several physiological activities such as embryo development, hair growth and wound healing.1 Heparanase is also implicated in tumour angiogenesis and metastasis, and a direct correlation between the overexpression of heparanase and the invasiveness of tumour cells has been shown.2, 3, 4 The implication of heparanase in cancer progression makes it a very attractive target for anti-angiogenic, anti-metastatic and/or anti-inflammatory therapies. Nevertheless, although a number of compounds with some inhibitory effect have been described, only inhibitor PI-88 has so far reached clinical trials.5, 6

The putative active form of heparanase is a heterodimer constituted by a 45 kDa glycosylated subunit non-covalently bound to a smaller 8 kDa polypeptide.2 It is widely accepted that there are two essential acidic residues (Glu225 and Glu343) involved in the catalytic mechanism, acting as a proton donor and a nucleophile, respectively.7 Furthermore, two potential heparan sulfate binding domains have been identified around the regions comprising residues Lys158-Asp171 and Gln270-Lys280, both located in the 45 kDa subunit, and it has been shown that heparanase constructs lacking these two regions are devoid of enzymatic activity.8 In a recent paper we described the cloning, expression and purification of smaller constructs of recombinant heparanase in Escherichia coli.9 A construct of 29 kDa containing the two catalytically active glutamic acids and the two binding sites for heparan sulfate was validated via NMR and SPR techniques for its use as surrogate of full length heparanase in drug discovery projects.

The three-dimensional structure of heparanase is unknown to this date. The absence of structural information has led several groups to propose structural models of the enzyme based on sequence homology.7, 10, 11 Recently, we developed a structural model on full-length heparanase,9 which was in good agreement with the information concerning the nature and position of the residues allegedly located in the active site.

In the work presented here, this model has been applied to the virtual screening of a database of structures we recently used to develop a heparanase pharmacophore model,12 comprising compounds with reported heparanase inhibitory activity6 and a commercial collection of drugs and drug-like compounds. The docking strategy was evaluated by its accuracy in ranking known inhibitors of heparanase in the context of the total database. The scores resulting from the docking approach were combined with those from the ligand-based approach with the purpose of ranking the compounds according to their putative ability to bind heparanase and selecting the best candidates for further experimental evaluation.

An initial experimental evaluation of the interaction of the selected drugs with heparanase was carried out using NMR methods. WaterLOGSY and Saturation Transfer Difference (STD) measurements were performed with those compounds previously selected in silico. Positive binders in the NMR assays were subjected to competition experiments in the presence of suramine, a known ligand of heparanase (KD = 0.5 μM), in order to evaluate the interaction of the positive compound at the binding site. In a second step, KD values were determined for those compounds showing binding to the heparanase active site using Surface Plasmon Resonance (SPR).

Using this strategy we identified the known antimalarial amodiaquine as a good candidate to inhibit heparanase, displaying affinity values for the protein on the low-micromolar range, which is on the order of other inhibitors described for this enzyme. A SAR study by NMR and SPR was performed with a subset of fourteen 4-arylaminoquinolines structurally related to amodiaquine to explore the influence of significant chemical variations on the binding affinity of this family of compounds for heparanase.

Section snippets

Docking and scoring

Docking of the compounds was performed using a 3D model of full-length heparanase previously developed in our group.9 This structure was modeled using as template a catalytically active construct of α-l-arabinofuranosidase from Geobacillus stearothermophilus T-6.13, 14 Two sets of compounds were used for the docking: (i) an ensemble of 27 known inhibitors of heparanase6 and (ii) the commercially available ‘Prestwick Chemical Library’® (http://www.prestwickchemical.com/), a collection of 1120

Reliability of the docking approach combined with the pharmacophore model

Recently, a pharmacophore model able to search for heparanase inhibitors has been developed based on a set of 1147 molecules (27 known inhibitors and the 1120 structures from the Prestwick Chemical Library).12 In the present study, the same set of chemicals was used to evaluate the ability of a docking approach to retrieve the known inhibitors of heparanase and to rank them accordingly to their reported affinities. As shown in Table 1, a high correlation was found between the binding affinities

Conclusions

This paper describes a combination of structure- and ligand-based in silico strategies to selectively narrow down the number of chemicals to be tested in a screening campaign. In this particular example, novel potential inhibitors of heparanase, a very relevant target involved in cell invasion and metastasis, have been discovered by the combined use of pharmacophore and docking strategies. Furthermore, the application of different biophysical techniques such as NMR and SPR confirmed the

Acknowledgement

The authors wish to thank the Spanish Ministerio de Ciencia e Innovación (SAF2011-28350) for its economic support.

References and notes (27)

  • N. Ilan et al.

    Int. J. Biochem. Cell Biol.

    (2006)
  • F. Levy-Adam et al.

    J. Biol. Chem.

    (2005)
  • G. Abboud-Jarrous et al.

    J. Biol. Chem.

    (2005)
  • D.D. Spellmeyer et al.

    J. Mol. Graphics Modell.

    (1997)
  • T.-L. Hwang et al.

    J. Magn. Reson.

    (1995)
  • S. Delarue-Cochin et al.

    Eur. J. Med. Chem.

    (2008)
  • S. Delarue-Cochin et al.

    Eur. J. Med. Chem.

    (2008)
  • R. Gozalbes et al.

    Bioorg. Med. Chem.

    (2010)
  • N.J. Nasser

    Cell. Mol. Life Sci.

    (2008)
  • I. Vlodavsky et al.

    Clin. Invest.

    (2001)
  • I. Vlodavsky et al.

    Curr. Pharm. Des.

    (2007)
  • L.M. Khachigian et al.

    Cardiovasc. Drug Rev.

    (2004)
  • E. Hammond et al.
  • Cited by (21)

    • Antitumor activity and structure-activity relationship of heparanase inhibitors: Recent advances

      2020, European Journal of Medicinal Chemistry
      Citation Excerpt :

      Similar to the symmetrical ureas, methyl substitution favored HPSE inhibition while replacement of CH3O- group abolished the anti-HPSE activity [88]. Fifteen 4-arylamodiaquine derivatives were selected to explore the SAR of the amodiaquine scaffold and anti-HPSE activity due to the highest affinity of amodiaquine with HPSE (Compound 42, Fig. 8) (IC50 = 0.20–0.29 μM) using in silico models, a QSAR solubility prediction model and a chemical diversity analysis [89]. The selected analogues 43–56 bearing a 4-aminophenyl-chloroquinoline scaffold displayed the anti-HPSE activities (IC50 = 0.23–25 μM), presenting the chemical diversities in the aminophenyl moiety.

    • Targeting Heparanase in Cancer: Inhibition by Synthetic, Chemically Modified, and Natural Compounds

      2019, iScience
      Citation Excerpt :

      This study suggested that targeted inhibition of heparanase may prove to be an effective therapy against cervical cancer (Song et al., 2016). Before the resolution of the heparanase crystal structure (Wu et al., 2015), a structural model using sequence homology was developed and applied for virtual screening of known heparanase inhibitors, commercially available drugs, and drug-like compounds (Gozalbes et al., 2009, 2013; Mosulen et al., 2011). Besides, biophysical methods (nuclear magnetic resonance [NMR] and surface plasmon resonance [SPR]) revealed higher affinity (KD = 52.6 μM) of amodiaquine (56) (Figure 9) toward the target.

    • Recent advances in the discovery of heparanase inhibitors as anti-cancer agents

      2016, European Journal of Medicinal Chemistry
      Citation Excerpt :

      A structural model on full-length heparanase built from sequence homology was applied to the virtual screening of a library made up of heparanase inhibitors and a commercial collection of drugs and drugs-like compounds [67,68]. Based on the application of in silico models, a QSAR solubility prediction model and a chemical diversity analysis, Rafael et al. [69] selected a subset of fourteen 4-arylaminoquinolines from a global set of 249 analogues of amodiaquine (compound 30, Fig. 7). The amodiaquine was chosen as the most promising candidate to search for analogs with similar or increased affinity for heparanase according to the NMR and SPR experiments.

    • Mannich bases in medicinal chemistry and drug design

      2015, European Journal of Medicinal Chemistry
      Citation Excerpt :

      Candidates 280 and 330 are essentially devoid of inhibitory activity of cyclin-dependent kinases; on the other hand, baicalein Mannich bases 329 were 6–20 times more potent than the parent flavone, suggesting that the presence of a nitrogen atom at C-8 is crucial for the inhibition of cyclin-dependent kinases, while the presence of a second heteroatom in the amine moiety (e.g., oxygen in morpholine, sulfur in thiomorpholine, nitrogen in N-methylpiperazine) further increases the potency of these candidates. Virtual screening of a commercial library containing drugs and drug-like molecules against a 3D model of heparanase led to the identification of several candidates with high scores for binding affinity; they were subsequently tested in NMR competitive inhibition experiments using suramine as a known heparanase inhibitor [486]. One of these candidates is amodiaquine 201 (Fig. 38), and a subset of fourteen representative compounds that retained the characteristic 4-phenylamino-chloroquinoline scaffold of amodiaquine, but presented different functional groups in the phenylamino moiety, was further selected for NMR competitive inhibition experiments.

    View all citing articles on Scopus

    Current address: ProtoQSAR SL, Vivero de Empresas Creix, 46008 Valencia, Spain.

    These authors contributed equally to this work.

    View full text