Phenotypic, chemical and functional characterization of cyclic nucleotide phosphodiesterase 4 (PDE4) as a potential anthelmintic drug target - PubMed (original) (raw)

. 2017 Jul 13;11(7):e0005680.

doi: 10.1371/journal.pntd.0005680. eCollection 2017 Jul.

Liliana Rojo-Arreola 1 2, Da Shi 3, Nelly El-Sakkary 3, Kurt Jarnagin 4, Fernando Rock 4, Maliwan Meewan 4, Alberto A Rascón Jr 1 2, Lin Lin 5, Katherine A Cunningham 5, George A Lemieux 5, Larissa Podust 1 2, Ruben Abagyan 3, Kaveh Ashrafi 5, James H McKerrow 1 2, Conor R Caffrey 1 2

Affiliations

Phenotypic, chemical and functional characterization of cyclic nucleotide phosphodiesterase 4 (PDE4) as a potential anthelmintic drug target

Thavy Long et al. PLoS Negl Trop Dis. 2017.

Abstract

Background: Reliance on just one drug to treat the prevalent tropical disease, schistosomiasis, spurs the search for new drugs and drug targets. Inhibitors of human cyclic nucleotide phosphodiesterases (huPDEs), including PDE4, are under development as novel drugs to treat a range of chronic indications including asthma, chronic obstructive pulmonary disease and Alzheimer's disease. One class of huPDE4 inhibitors that has yielded marketed drugs is the benzoxaboroles (Anacor Pharmaceuticals).

Methodology/principal findings: A phenotypic screen involving Schistosoma mansoni and 1,085 benzoxaboroles identified a subset of huPDE4 inhibitors that induced parasite hypermotility and degeneration. To uncover the putative schistosome PDE4 target, we characterized four PDE4 sequences (SmPDE4A-D) in the parasite's genome and transcriptome, and cloned and recombinantly expressed the catalytic domain of SmPDE4A. Among a set of benzoxaboroles and catechol inhibitors that differentially inhibit huPDE4, a relationship between the inhibition of SmPDE4A, and parasite hypermotility and degeneration, was measured. To validate SmPDE4A as the benzoxaborole molecular target, we first generated Caenorhabditis elegans lines that express a cDNA for smpde4a on a pde4(ce268) mutant (hypermotile) background: the smpde4a transgene restored mutant worm motility to that of the wild type. We then showed that benzoxaborole inhibitors of SmPDE4A that induce hypermotility in the schistosome also elicit a hypermotile response in the C. elegans lines that express the smpde4a transgene, thereby confirming SmPDE4A as the relevant target.

Conclusions/significance: The orthogonal chemical, biological and genetic strategies employed identify SmPDE4A's contribution to parasite motility and degeneration, and its potential as a drug target. Transgenic C. elegans is highlighted as a potential screening tool to optimize small molecule chemistries to flatworm molecular drug targets.

PubMed Disclaimer

Conflict of interest statement

I have read the journal's policy and the authors of this manuscript have the following competing interests: KJ, FR and MM are employees of Anacor Pharmaceuticals Inc.

Figures

Fig 1

Fig 1. Phenotypic screening of a benzoxaborole collection with S. mansoni somules suggests a PDE4 as a molecular target of interest.

The screen involving 1,085 benzoxaboroles was performed at 5 μM for 6 days with observations taken every day using a constrained nomenclature, as noted in the text. Three main phenotype response groups could be adjudicated by microscopical observation: (i) 104 compounds eliciting an early and sustained hypermotile phenotype, of which, 30% was also associated with a parasite degeneration; (ii) 94 compounds that yielded a range of phenotypic responses (e.g., rounding, darkening), including hypermotility, which were either transient (noted at 24 h only) or appeared later (on or after day 3 of the incubation), and (iii) 887 compounds that produced no phenotype. Of the 1,085 benzoxaboroles screened, 174 also had IC50 data for inhibition of huPDE4B2 that were distributed as 77, 82 and 15 compounds across the sustained hypermotile, no phenotype and transient hypermotile groups, respectively. Sixty-five of 77 compounds in the sustained hypermotile group inhibited huPDE4B2 with IC50 values of < 1 μM. In contrast, for the no phenotype group, only 16 of 82 compounds had IC50 values of < 1 μM. The association between the sustained hypermotile phenotype and sub-micromolar inhibition of huPDE4 was highly significant with a Fishers exact _p_-value of <0.0001. The 5-(3-cyanopyridyl-6-oxy) benzoxaborole scaffold, which is known to preferentially inhibit huPDE4 over other PDEs [73], was well represented in Group 1.

Fig 2

Fig 2. Protein sequence alignment of human, Caenorhabditis elegans and Schistosoma mansoni PDE4 sequences.

Shown are the sequences for huPDE4B2 (NP_001032416.1), the crystal structure of a crosslink-stabilized huPDE4B1 (4X0F) [79], C. elegans (NP_495601.1) and S. mansoni Smp_134140 (PDE4A), Smp_141980 (PDE4B), Smp_129270 (PDE4C) and Smp_044060 (PDE4D). Upstream Conserved Regions (UCR) 1 and 2, and the catalytic domain are shaded in blue, green and pink, respectively. Demarcations of these domains are according to [80]. Linker regions (LR) 1 and 2 are indicated by blue horizontal bars. If present, the PKA and ERK phosphorylation sites, in the UCR1 and the catalytic domain, respectively, are indicated by the red and blue typeface, respectively. The conserved PDE signature motif HNX2HNXNE/D/QX10HDX2HX25E is indicated with blue circles and those residues that coordinate directly with the catalytic zinc in the substrate binding pocket are also indicated by red circles [33, 93]. The residues in UCR1 and UCR2 that contribute to the dimerization interface in huPDE4 [79] are indicated by red squares and the tyrosine residue (Y471 (Y274 in [79]) in UCR2 that contributes to the high affinity binding of rolipram in the PDE4 active site is marked by the teal square. The sequence alignment is adjusted N-terminally given the very long N-termini in some cases, e.g., for SmPDE4B. The alignment was performed semi-automatically using the ICM pro (Molsoft LLC).

Fig 3

Fig 3. Purification and activity analysis of the recombinant catalytic domain of SmPDE4A.

(A) The three-step purification scheme described in the text resulted in purified His6-tagged SmPDE4A with the expected molecular mass of 44.2 kDa. Each lane (1–4) contains an increasing amount of protein (3, 6, 21 and 59 μg) demonstrating the absence of major contaminants. Molecular mass markers (in kDa) are indicated on the left. (B) Determination of Km and Vmax. Enzyme reaction rates were measured over increasing concentrations of the cAMP substrate up to 10 μM. All reactions ran for six minutes and contained 23.5 units/ml SmPDE4A. Km and Vmax values were determined by nonlinear regression analysis of the data (Prism GraphPad version. 6.03) using a Michaelis-Menten enzyme kinetics model. All data points were determined in triplicate.

Fig 4

Fig 4. Association between inhibition of PDE4 and activity against the parasite for benzoxaborole and catechol inhibitors.

Assays to determine IC50 values were performed in duplicate with the total number of assays performed indicated in parentheses: at a minimum, data from two assays are shown. Descriptions of phenotypes observed (descriptors): R = rounded; O = overactive (and perceived degrees thereof using plus and minus symbols); deg = degeneration; tegument damage = damage to the surface of the worms, dark = worms are darkened; none = no effects observed. Wormassay [102] is a digital camera based assay that detects adult worm-induced changes in the occupation and vacancy of pixels between frames (outputted as an average ± S.D.).

Fig 5

Fig 5. Inhibition of huPDE4B2 and SmPDE4A by exemplar benzoxaborole and catechol inhibitors.

Assays with each inhibitor were performed using the catalytic domains of (A) huPDE4B2 (NP_001032416.1) and (B) SmPDE4A (Smp_134140). Briefly, each reaction contained 23.5 units/ml SmPDE4A or 30 pg/ml huPDE4B2. Assays were performed in triplicate, and IC50 values were determined by non-linear regression analysis using the four-parameter logistic equation (Prism GraphPad version. 6.03). Compound structures are shown in Fig 4.

Fig 6

Fig 6. A smpde4A transgene restores wild type motility rates to _pde4_-deficient C. elegans.

Relative to wild type (WT) C. elegans, a loss of function allele of pde4, namely ce268 [103], causes hypermotility. This hypermotility is reverted back to WT rates upon transgenic expression of a full-length cDNA for smpde4a. Results for two independently generated lines, smpde4a(a) and smpde4a(b) are shown. The same transgenes do not alter the motility of otherwise WT animals. Error bars indicate the standard deviations around the mean motility in a representative experiment with at least 10 worms for each strain. The asterisks indicate significance by Student’s _t_-test (*p<0.005; **p<0.0005) relative to the hypermotility recorded for the pde4(ce268) mutant.

Fig 7

Fig 7. PDE4 inhibitors induce hypermotility in transgenic C. elegans expressing smpde4a.

C. elegans was exposed to 100 μM rolipram, roflumilast or the benzoxaborole compounds 2 and 4, and then worm motility measured. Effects on (A) WT, (B) hypermotile mutant pde4(ce268), (C) transgenic pde4(ce268);smpde4a(a) and (D) transgenic pde4(ce268);smpde4a(b) are shown. Means and standard deviations for motility are normalized over two experiments to DMSO controls; each experiment involved measuring at least 10 worms per treatment. The asterisks in each panel indicate significance by Student’s _t_-test (*p<0.005; **p<0.0005) relative to the respective DMSO controls. For Panel B, motility was normalized to that recorded for the WT control.

Fig 8

Fig 8. 2D interaction diagram of rolipram and roflumilast with huPDE4B and SmPDE4A.

Molecular models of each enzyme in complex with rolipram and roflumilast were built using ICM-pro and huPDE4B1 as a template (PDB ID: 4X0F) [79]. The amino acid residues in the huPDE4B1 and SmPDE4 binding sites that interact directly with the ligands are shown as ovals. Those that distinguish the schistosome ortholog are shown in magenta and the consequent changes in binding free energies are indicated underneath. The residue numbers are consistent with the alignment presented in Fig 2.

References

    1. Abdul-Ghani RA, Loutfy N, Hassan A. Experimentally promising antischistosomal drugs: a review of some drug candidates not reaching the clinical use. Parasitol Res. 2009;105(4):899–906. doi: 10.1007/s00436-009-1546-2 . -DOI -PubMed
    1. Dőmling A, Khoury K. Praziquantel and schistosomiasis. ChemMedChem. 2010;5(9):1420–34. Epub 2010/08/03. doi: 10.1002/cmdc.201000202 . -DOI -PubMed
    1. World Health Organization. Schistosomiasis: Fact sheet N°115: WHO; 2014. http://www.who.int/mediacentre/factsheets/fs115/en/.
    1. Caffrey CR. Schistosomiasis and its treatment. Future Med Chem. 2015;7(6):675–6. doi: 10.4155/fmc.15.27 . -DOI -PubMed
    1. Cioli D, Pica-Mattoccia L, Basso A, Guidi A. Schistosomiasis control: praziquantel forever? Mol Biochem Parasitol. 2014;195(1):23–9. Epub 2014/06/24. doi: 10.1016/j.molbiopara.2014.06.002 . -DOI -PubMed

MeSH terms

Substances

LinkOut - more resources