A Miniaturized Screen of a Schistosoma mansoni Serotonergic G Protein-Coupled Receptor Identifies Novel Classes of Parasite-Selective Inhibitors

Abstract

Schistosomiasis is a tropical parasitic disease afflicting ~200 million people worldwide and current therapy depends on a single drug (praziquantel) which exhibits several non-optimal features. These shortcomings underpin the need for next generation anthelmintics, but the process of validating physiologically relevant targets ('target selection') and pharmacologically profiling them is challenging. Remarkably, even though over a quarter of current human therapeutics target rhodopsin-like G protein coupled receptors (GPCRs), no library screen of a flatworm GPCR has yet been reported. Here, we have pharmacologically profiled a schistosome serotonergic GPCR (Sm.5HTR) implicated as a downstream modulator of PZQ efficacy, in a miniaturized screening assay compatible with high content screening. This approach employs a split luciferase based biosensor sensitive to cellular cAMP levels that resolves the proximal kinetics of GPCR modulation in intact cells. Data evidence a divergent pharmacological signature between the parasitic serotonergic receptor and the closest human GPCR homolog (Hs.5HTR7), supporting the feasibility of optimizing parasitic selective pharmacophores. New ligands, and chemical series, with potency and selectivity for Sm.5HTR over Hs.5HTR7 are identified in vitro and validated for in vivo efficacy against schistosomules and adult worms. Sm.5HTR also displayed a property resembling irreversible inactivation, a phenomenon discovered at Hs.5HTR7, which enhances the appeal of this abundantly expressed parasite GPCR as a target for anthelmintic ligand design. Overall, these data underscore the feasibility of profiling flatworm GPCRs in a high throughput screening format competent to resolve different classes of GPCR modulators. Further, these data underscore the promise of Sm.5HTR as a chemotherapeutically vulnerable node for development of next generation anthelmintics.

Fig 1. Functional expression of Sm.5HTR.

(A) Western blot of myc-tagged Sm.5HTR (Genbank KF444051.1) in HEK293 cells. L, ladder. (B) Schematic of GloSensor assay depicting generation of cAMP by activation of Sm.5HTR (blue) which couples to endogenous G proteins and adenylate cyclase to increase cAMP levels. cAMP binding to the permuted luciferase construct leads to enhanced luminescence in the presence of substrate. (C) Functional expression of Sm.5HTR using this assay. In cells transfected with the biosensor (22F variant) addition of 5-HT (10μM) caused an increase in luminescence in Sm.5HTR transfected (blue symbols), but not control cells (open symbols).

Fig 2. Optimization of cAMP bioluminescent sensor for monitoring Sm.5HTR activity.

(A) Timecourse of cAMP changes following application of 5-HT (10μM) in HEK293 cells transfected with Sm.5HTR (circles) or untransfected controls (triangles). Assays are shown in the presence (open symbols) and absence (solid symbols) of a phosphodiesterase inhibitor (IBMX, 200μM). Luminescence values were recorded every 3 minutes in cells transfected with the high affinity (F20, left) or low affinity (F22, right) cAMP biosensor. Inset shows data for Sm.5HTR in the absence of IBMX on a rescaled y-axis. (B) 5-HT dose response curves from experiments such as those described in (A) for F20 (left) and F22 (right) in the presence (open) or absence (solid) of IBMX. Control data using untransfected cells are also shown (triangles). Note log scale for responses with 22F. (C) Z’ values calculated for 5-HT evoked signals over time under the conditions described previously. (D) Real time comparison of methiothepin pretreatment on 5-HT evoked changes in cAMP. One cohort of cells was preincubated with methiothepin (10μM, open circles), prior to addition of 5-HT (10μM) and IBMX (200μM, closed triangle) and then forskolin (20μM, open triangle). (E) 5-HT dose response curves for Sm.5HTR (solid) and Sm.5HTR_L (open). Inset, schematic comparison of schistosome 5HT receptor isoforms. Sm.5HTR_L is a longer isoform with additional sequence at the N-terminus and third intracellular loop. (F) Kinetics of 5-HT response (10μM) for Sm.5HTR and Sm.5HTR_L.

Fig 3. Pharmacological profiling of Sm.5HTR.

(A) Schematic of assay workflow for screening a library of known GPCR ligands against HEK293 cells expressing either Sm.5HTR or Hs.5HT7. Cells transfected with either 5HTR and F22 cAMP biosensor were plated in 96-well format and exposed to test compounds (10μM). 5-HT was added after 30mins (at an EC₈₀ concentration) after which luminescence values were recorded (time = 60min). (B) Scatter plots summarizing effects of test compounds on the Sm.5HTR response to 5-HT (dotted line highlights threshold for defining compound ‘hits’). Hits were defined at a threshold of ≥50% inhibition relative to control wells (DMSO only, open symbols). (C) Compounds were also screened against HEK293 cells expressing the F22 sensor alone (no Sm.5HTR) to screen for cAMP generation at endogenous receptors. For reference, a forskolin data point is shown in red. (D) Compounds were also screened against forskolin (20μM) evoked changes in luminescence relative to control samples (DMSO, open circles). (E) Heat map of all test compounds screened against S. mansoni Sm.5HTR (left) and human Hs.5HT7 (middle). Each colored box represents the fold change in luminescence in response to an individual test compound (253 in total) keyed by the pseudocolor scale. Compounds showing activity against endogenous receptors in cells transfected with the 22F biosensor alone (21 compounds total) were masked (black). Right, Venn diagram summarizing selectivity of antagonist ‘hits’ against either Sm.5HTR, Hs.5HT7R, or both 5-HT receptors. In total, 23 ligands were classified as potential ‘hits’ at Sm.5HTR and 31 ligands as ‘hits’ at Hs.5HTR7, with 7 in common.

Fig 4. Profiling ligand selectivity for schistosome and human 5HT7 receptors.

(A) Dose response relationships for compound antagonism of 5-HT evoked cAMP generation at human Hs.5HT7R (green) or parasite Sm.5HTR (blue) GPCRs. Illustrative data from compounds showing preferential selectivity toward Sm.5HTR (top) or Hs.5HT7R (bottom), or compounds with no selectivity (middle). Data represent mean±s.d, n = 3. (B) Schematic representation of ratio of IC₅₀s for ‘hits’ profiled against both Sm.5HTR and Hs.5HT7R expressing HEK293 cells. Compounds exhibiting poor blockade of either GPCR (*) precluded calculation of IC₅₀ values, so a minimal ratio estimate is provided. Solid circles represent compounds for which data is shown in ‘A’.

Fig 5. Effects of ergot alkaloids on Sm.5HTR.

(A) Dose response relationship for various bioaminergic ligands reveals ergotamine and dihydroergotamine act as partial agonists at Sm.5HTR. (B-D) Bromocriptine (B), metergoline (C) and LSD (D) act as competitive antagonists of Sm.5HTR. (E) PZQ lacks antagonist activity at Sm.5HTR. (F) Structure-activity-relationship for various ergoline ligands at the Sm.5HTR. Data in parentheses represent pEC₅₀ ± S.E.M or pK_B ± S.E.M.

Fig 6. Long lasting inhibition of Sm.5HT7R evoked by a subset of ligands.

Sm.5HTR displays an inactivating antagonist property reported for human 5HT7R. (A) Both the ‘inactivating antagonist’ bromocriptine (10μM, blue) and the competitive antagonist cyproheptadine (10μM, grey) acutely inhibit the effect of 5-HT (10μM) at Sm.5HTR (5-HT alone, black). (B) Sm.5HTR remains insensitive to 5-HT following washout of bromocriptine but not cyproheptadine. Cells were pre-incubated with antagonists as in (A) for 30 mins, followed by solution exchange, and the assay for 5-HT responsiveness 1hr later. (C) Inhibition of 5-HT response at Sm.5HTR by both ‘inactivating antagonists’ established at Hs.5HT7R (methiothepin, bromocriptine, lisuride, risperidone, metergoline) and competitive antagonists (clozapine, cyproheptadine). All drugs were tested at 10μM for 30mins. **, p <0.01. (D) Persistent effects of antagonists (10μM) shown in (C) after washout and subsequent assay for 5-HT response (1hr later). **, p <0.01, *, p <0.05. (E) Titration of these ‘inactivating antagonists’ revealed the dose-response relationship for Sm.5HTR inhibition after washout. Colors correspond to drug identity in C&D. Data represent mean±s.e.m., n = 3 (C-E).

Fig 7. Structure activity relationships for various drug classes against Hs.5HT7R and Sm.5HTR.

Comparison of IC₅₀s for compounds inhibiting cAMP generation via Hs.5HT7R (abscissa) or Sm.5HTR (ordinate). Compounds with no preferential selectivity for either receptor, showing similar IC₅₀s, would cluster along the solid line. Hits in the lower right quadrant (red square) show sub-μM potency at Sm.5HTR but supra-μM potency at Hs.5HT7R. Four compounds meet this criterion (bromocriptine = ‘39’, rotundine = ‘37’, tetrandrine = ‘31’ and tetrabenazine = ‘32’). Compound classes are indicated as follows: ergot alkaloids (green), isoquinolines (blue), tricyclic and tetracyclic antidepressants (magenta), sulfonyl compounds (orange), miscellaneous structures (open). Individual compounds are: 1, SB269970; 2, amisulpride; 3, SB742457; 4, olanzapine; 5, mianserin; 6, quetiapine; 7, clozapine; 8, cyproheptadine; 9, ketotifen; 10, loratadine; 11, maprotiline; 12, clomipramine; 13, desloratadine; 14, rupatadine; 15, vortioxetine; 16, amitriptyline; 17, risperidone; 18, domperidone; 19, chlorprothixene; 20, clemastine; 21, aripiprazole; 22, ketanserin; 23, ifenprodil; 24, tripelennamine; 25, fluoxetine; 26, atomoxetine; 27, orphenadrine; 28, lisuride; 29, benztropine; 30, cyclizine; 31, tetrandrine; 32, tetrabenazine; 33, berberine; 34, 6, 7-diethoxy-1, 2, 3, 4-tetrahydroisoquinoline; 35, corynoline; 36, alfuzosin; 37, rotundine; 38, fanchinoline; 39, bromocriptine; 40, metergoline; 41, LY215840; 42, nicergoline; 43, mesulergine; 44, dihydroergocristine.

Fig 8. Small molecule inhibitors of Sm.5HTR antagonize 5-HT stimulation of schistosomule contractility.

Effects of selected ligands on schistosomules. (A) 5-HT stimulates basal contractility in S. mansoni schistosomules resolved through measurements of body length over time (1 minute recording duration). A contractile cycle is defined when a deviation of ≥20% of the average body length (dashed lines) occurs. (B) Dose-response curve for 5-HT stimulation of contractility. (C) Schistosomule movement was quantified for basal mobility (i, no 5-HT addition; white bars), after addition of 5-HT ((ii, 10μM 5-HT; grey bars), and subsequent exposure to Sm.5HTR inhibitors in the presence of 5-HT (iii, indicated doses; black bars). Representative body length traces over one minute for individual schistosomules are shown for indicated conditions (right). Bar graphs represent mean±s.e.m. of independent samples, n = 3. Drugs assayed represent ligands identified as Sm.5HTR antagonists in the GPCR screen (rotundine), and follow up testing of methoxyisoquinoline compounds (tetrandrine, tetrabenazine) and the ergot alkaloid bromocriptine.

Fig 9. Sm.5HTR inhibitors antagonize basal and 5-HT stimulated movement in adult schistosomes.

(A) Movement of adult male (top) and female (bottom) schistosomes under basal conditions (no 5-HT) and in the presence of 5-HT (100μM, arrow). Traces represent one minute of recorded movement for each condition. Scale, 1cm. (B) Dose response curves showing movement of male (solid circles) and female (open circles) schistosomes exposed to increasing concentrations of 5-HT. (C) Left, effect of bromocriptine (10μM) on the basal movement of adult male and female schistosomes. Representative traces showing the movement of worms in the absence of drug (black) and the presence of bromocriptine (red). Right, quantification of basal movement for male (solid bars) and female (open bars) worms exposed to the indicated compounds. (D) Left, effect of bromocriptine (10μM) on 5-HT (100μM) stimulated movement of adult schistosomes. Representative movement of worms exposed to 5-HT alone (black) or bromocriptine and 5-HT (red). Right, quantification of 5-HT stimulated movement for male (solid bars) and female (open bars) worms exposed to the indicated compounds. n≥3 independent experiments. * p < 0.05, ** p < 0.01.

Fig 10. Long lasting inhibition of schistosome movement caused by bromocriptine.

Time course of schistosome recovery from exposure to Sm.5HTR antagonists (black = DMSO control, blue = rotundine, red = bromocriptine). Mobility of worms was recorded following exposure to antagonist (10μM, ‘drug’, 2 hour exposure) and addition of 5-HT (100μM, ‘drug + 5-HT’). Media was then exchanged, and recordings were subsequently made on the same worms stimulated with 5-HT (100μM) at the indicated timepoints (3, 6, 24hrs after drug washout). Data for males (A) and females (B) is presented normalized to basal movement of control (dmso) worms in the absence of 5-HT.