Investigation of an Allosteric Deoxyhypusine Synthase Inhibitor in P. falciparum

Abstract

The treatment of a variety of protozoal infections, in particular those causing disabling human diseases, is still hampered by a lack of drugs or increasing resistance to registered drugs. However, in recent years, remarkable progress has been achieved to combat neglected tropical diseases by sequencing the parasites’ genomes or the validation of new targets in the parasites by novel genetic manipulation techniques, leading to loss of function. The novel amino acid hypusine is a posttranslational modification (PTM) that occurs in eukaryotic initiation factor 5A (EIF5A) at a specific lysine residue. This modification occurs by two steps catalyzed by deoxyhypusine synthase (dhs) and deoxyhypusine hydroxylase (DOHH) enzymes. dhs from Plasmodium has been validated as a druggable target by small molecules and reverse genetics. Recently, the synthesis of a series of human dhs inhibitors led to 6-bromo-N-(1H-indol-4yl)-1-benzothiophene-2-carboxamide, a potent allosteric inhibitor with an IC50 value of 0.062 µM. We investigated this allosteric dhs inhibitor in Plasmodium. In vitro P. falciparum growth assays showed weak inhibition activity, with IC50 values of 46.1 µM for the Dd2 strain and 51.5 µM for the 3D7 strain, respectively. The antimalarial activity could not be attributed to the targeting of the Pfdhs gene, as shown by chemogenomic profiling with transgenically modified P. falciparum lines. Moreover, in dose-dependent enzymatic assays with purified recombinant P. falciparum dhs protein, only 45% inhibition was observed at an inhibitor dose of 0.4 µM. These data are in agreement with a homology-modeled Pfdhs, suggesting significant structural differences in the allosteric site between the human and parasite enzymes. Virtual screening of the allosteric database identified candidate ligand binding to novel binding pockets identified in P. falciparum dhs, which might foster the development of parasite-specific inhibitors.

Keywords: allosteric inhibitor; bromobenzothiophene; chemogenomic profiling; deoxyhypusine synthase; glmS riboswitch; hypusine.

Figure 5

Amino acid alignment of P. falciparum DHS (first lane) and the human orthologue (second lane) performed with the Clustal Omega alignment tool. Gaps (-) were introduced to obtain maximum alignment. Asterisks label amino acid identities, colons (:) and dots (.) label amino acid similarities. Amino Acids marked in blue represent nucleotide binding sites in human DHS. Possible ligand binding sites are colored in green. Lys328 and Ala351 are part of the active site and shown in red color.

Figure 8

Predicted allosteric modulators and similar allosteric metabolites for binding pockets 1 and 2 in the plasmodial DHS protein obtained from the Allosteric Database and Allobase. (A) 4-[5-[3-(1-adamantyl)-4-hydroxyphenyl]-4,5-dihydro-1,2-oxazol-3-yl]benzoic acid with the top Alloscore of 8.61. (B) 1-[2-[(4benzoylphenyl)methoxy]phenyl)]ethanone is an allosteric inhibitor with an Alloscore of 8.02. (C) Xanthurenic acid (4,8-Dihydroxyquinoline-2-carboxylic acid represents a known, endogenous metabolite to the allosteric inhibitor A. (D) Data mining in the Allobase shows significant similarity of compound B to Gossypol (1,1′,6,6′,7,7′,-Hexahydroxy-3,3′-dimethyl-5,5′-(dipropan-2-yl)[2,2′-binaphtalene]-8,8′-dicarbaldehyde) as an allosteric modulator.

Figure 1

Biochemical pathway leading to the formation of hypusine in EIF5A protein: In the first step of the pathway, DHS transfers an aminobutyl moiety from the substrate spermidine to a specific lysine residue in the EIF5A precursor protein to form deoxyhypusine. In the second step, deox-yhypusine hydroxylase (DOHH) completes hypusine biosynthesis, introducing a hydroxyl group into the side chain.

Figure 2

Two-dimensional structure of compound 6-bromo-N-(1H-indol-4yl)-1-benzothiophene-2-carboxamide (PubChemCID146014943).

Figure 3

Dose–response growth inhibition curves with the allosteric inhibitor 6-bromo-N-(1H-indol-4yl)-1-benzothiophene-2-carboxamide against: (A) Transgenic Pfdhs_glmS parasites and (B) ribozyme-inactive transgenic PfDHS_M9 parasites with cotreatment of 2.5 mM glucosamine (GlcN) (blue triangles) or without co-treatment (red circles). (C) The estimated log₂ ratio of EC₅₀ between the +GlcN (blue points) and −GlcN (red points) conditions together with CI₉₅ is shown for the allosteric inhibitor and the pyrimethamine control in transgenic parasites, i.e., Pfdhs_glmS and the inactive mutant. p-Values comparing log₂ EC₅₀ (−GlcN/+GlcN) estimate: 0.58 Pfdhs_glmS (pyrimethamine): 1 × 10⁻⁴ Pfdhs_M9 (pyrimethamine); 0.81 Pfdhs_glmS (allosteric inhibitor); 4 × 10⁻⁸ Pfdhs_M9 (allosteric inhibitor).

Figure 4

Dose–response curve of the allosteric inhibitor showing the log doses (x axis) versus % inhibition (y axis) in non-linear regression.

Figure 6

(A) In silico homology model of PfDHS. Modelling was performed on the template alignment of human DHS co-crystallized with 6-bromo-N-(1H-indol-4yl)-1-benzothiophene-2-carboxamide (PDB 6pgr) and PfDHS (PDB 1rlz.1A) generated by Swiss Model. The protein consists of four identical subunits shown in rainbow, i.e., red (subunit A), blue (subunit B), green (subunit C) and brown (subunit D). (B) Graph of local quality estimate for the homology model structure of PfDHS. Q-MEAN score (y-axis) is plotted against PfDHS residue number. Subunits of a putative tetramer are plotted on the same axes (Chain A: grey; Chain B: yellow; Chain C: cyan; Chain D: green). (C) Alignment of P. falciparum DHS strain D7 (PDB:1rlz.1A) and human DHS co-crystallized with 6-bromo-N-(1H-indol-4yl)-1-benzothiophene-2-carboxamide (PDB: 6pgr). Secondary structure elements are indicated on the protein residues (oblong, alpha helix; arrow, beta-sheet).

Figure 6

(A) In silico homology model of PfDHS. Modelling was performed on the template alignment of human DHS co-crystallized with 6-bromo-N-(1H-indol-4yl)-1-benzothiophene-2-carboxamide (PDB 6pgr) and PfDHS (PDB 1rlz.1A) generated by Swiss Model. The protein consists of four identical subunits shown in rainbow, i.e., red (subunit A), blue (subunit B), green (subunit C) and brown (subunit D). (B) Graph of local quality estimate for the homology model structure of PfDHS. Q-MEAN score (y-axis) is plotted against PfDHS residue number. Subunits of a putative tetramer are plotted on the same axes (Chain A: grey; Chain B: yellow; Chain C: cyan; Chain D: green). (C) Alignment of P. falciparum DHS strain D7 (PDB:1rlz.1A) and human DHS co-crystallized with 6-bromo-N-(1H-indol-4yl)-1-benzothiophene-2-carboxamide (PDB: 6pgr). Secondary structure elements are indicated on the protein residues (oblong, alpha helix; arrow, beta-sheet).

Figure 7

Structural prediction of two allosteric binding pockets (marked by blue arrows) predicted by the PASServer employing the 3D model of human DHS co=crystallized with the allosteric inhibitor. Binding Pocket 1 (red color) has the highest probability of 59.36% and a druggability score of 0.007. Pocket 2 (brownish color) has a probability score of 58.2% and a druggability score of 0.228.

Figure 9

Dose–response curve of the allosteric metabolites xanthenuric acid and gossypol showing the log doses (x axis) versus % inhibition (y axis) in non linear regression.