Analysis of the Physicochemical Properties of Anti-Schistosomal Compounds to Identify Next-Generation Leads

Abstract

To investigate the physicochemical properties of anti-schistosomal compounds reported between 2008 and 2023, a simple but extensive literature scrutiny was conducted. Keywords were searched in Chemical Abstracts Service (CAS) SciFinder and primary medicinal chemistry and pharmacology literature to locate publications with compounds displaying ex vivo and/or in vivo anti-schistosomal activity. A total of 57 repurposed U.S. Food and Drug Administration (FDA)-approved drugs, hits and their derivatives were manually extracted, curated and compared to known anti-schistosomal oral drugs in view of establishing trends of calculated critical molecular properties. From this analysis, it was determined that more than 65% of the compounds display cLogD_7.4 > 3 values, whereas oxamniquine, metrifonate and praziquantel (PZQ), previous and currently used oral anti-schistosomal drugs, possess lower cLogD_7.4 values (≤2.5). Furthermore, the lipophilicity associated with PZQ corresponds to a highly permeable and sparingly soluble compound, characteristics that favor drug absorption and compound penetration in the parasite. These physicochemical properties together with PZQ’s anti-schistosomal activity make PZQ an essential medicine for the treatment of schistosomiasis and demonstrate the importance of finding the right balance among potency (e.g., EC₅₀ < 5 and 0.5 μM), cell permeability (e.g., P_app > 2 × 10⁶ cm/s) and kinetic aqueous solubility (e.g., >10 μM) to provide high-quality hits and/or leads for the discovery of new oral anti-schistosomal therapeutics.

Schistosomiasis, also known as bilharzia, is a parasitic disease caused by blood flukes (trematode worms) of the genus Schistosoma.¹ The most clinically relevant species are Schistosoma mansoni, Schistosoma haematobium and Schistosoma japonicum which are prevalent throughout Africa (S. mansoni, S. haematobium), South America (S. mansoni) and South-East Asia (S. japonicum).² Today, treatment and control of human schistosomiasis rely solely and extensively on praziquantel (PZQ), to such an extent that it is estimated that almost 252 million people were in need of treatment in the framework of preventive chemotherapy programs in 2021.¹ Nevertheless, PZQ is only efficacious against the adult form of the parasite, it requires a relatively high single oral dose of 40 mg/kg and emergence of drug-resistant parasites in the field may soon be exacerbated by years of mass drug administration campaigns.³ These shortcomings highlight the need to develop new therapeutic interventions that are more effective at treating and preventing human schistosomiasis.^4,5

In contrast to other neglected tropical diseases such as malaria and tuberculosis, support for the discovery of new oral anti-schistosomal therapeutics has, until now, been limited and during the past four decades, no new chemical entity has progressed into clinical development.⁶ Testing compounds against molecular targets such as histone deacetylases (HDACs)^7,8 and histone acetyltransferases (HATs),⁹ venus kinase receptors (VKRs),¹⁰ lysine-specific demethylases (LSDs)¹¹ and transient receptor potential (TRP) channels¹² among others has shown some success in identifying potential leads with ex vivo and/or in vivo activity. In light of limited validated drug anti-schistosomal targets, phenotypic screening of FDA-approved drugs, natural products and structurally diverse chemical libraries from pharmaceutical companies, charitable organizations and commercial sources has been the preferred approach to generate new chemical starting points for developing alternative therapies.¹³ To generate hits, the scientific community largely relies on ex vivo (newly transformed schistosomula (NTS) and juvenile and adult worms) assays; however, obtaining worms in large numbers imposes a technical limitation that reduces the number of compounds that can be screened at a given time.^14,15 In addition, translation of ex vivo activity to in vivo efficacy via the validated S. mansoni mouse model (Supporting Information) is typically undertaken with limited successes, generally because of compounds’ suboptimal absorption, distribution, metabolism, excretion and toxicity (ADMET) properties.¹⁶ In light of these considerations, chemical entities should be carefully selected to give them the best chance of progressing effectively along the drug discovery and development process, with the ultimate goal of producing new oral anti-schistosomal drugs.

Compounds from the literature search were selected based on the following criteria: NTS half-maximal effective response (EC₅₀) of ≤5 μM with ex vivo activity against juvenile and/or adult worms (EC₅₀ ≤ 10 μM; incubation time up to 5 days) and/or efficacy in the murine challenge model (infection with 40–120 cercaria, ≤400 mg/kg single or multiple doses using oral, po, or intraperitoneal, ip, routes of administration). Molecules displaying moderate NTS potency, EC₅₀ > 5–10 μM, with no juvenile and adult worms killing activity and/or in vivo efficacy were excluded from this analysis.

How Drug-like Are Anti-Schistosomal Compounds?

In total, 57 repurposed FDA-approved drugs, hits and their derivatives stemming from corresponding structure–activity relationship (SAR) studies met the above-mentioned criteria. In addition, two azepanylcarbazole amino alcohol compounds (58 and 59) originally developed as anti-plasmodials were included as promising anti-schistosomal leads (Table S1).¹⁷

The compound collection encompasses a range of molecular property values, i.e., hydrogen bond donors (HBDs) = 0–3, hydrogen bond acceptors (HBAs) = 1–10, molecular weight (MW) = 187–645 Da, topological polar surface area (TPSA) = 7–159 Ų and a calculated distribution coefficient cLogD_7.4 = −1.6–6.4, and, although small, provides a reasonable representation of the chemical matter under investigation for the development of new oral anti-schistosomal drugs.

A total of 46 out of 59 (>80%, Table S1) compounds, and hence not surprisingly most of the selected FDA-approved therapeutics, satisfy the most commonly used molecular property values, i.e., HBA ≤ 10, HBD ≤ 5, TPSA < 140 Ų, MW < 500 Da and LogD < 5, that define the minimum requirement for a compound to have high likelihood of oral absorption.^18,19 Molecules that reside outside this property space (<20% of the listed compounds, Table S1) have frequently shown poor ADMET properties.²⁰

To investigate possible correlations between TPSA and MW and lipophilicity, the values of the subset of 46 anti-schistosomal compounds were plotted and oxamniquine, metrifonate and PZQ, the three most well-known oral anti-schistosomal drugs, were included for comparison (Figure 1). It is noteworthy that none of the 59 compounds exceeded the HBD and HBA cutoff; thus, these molecular descriptors were not included in the investigation.

Figure 1.

cLogD_7.4 vs TPSA and MW of compounds within the property space with the highest probability of achieving appropriate oral absorption (HBA ≤ 10, HBD ≤ 5, TPSA < 140 Ų, MW < 500 Da and LogD < 5). Oxamniquine, metrifonate and PZQ in green (cLogD_7.4 ≤ 2); azepanylcarbazole, 58 and 59 in red (cLogD_7.4 > 4) (Table S1).

Figure 1a indicates that there is no direct correlation between TPSA and cLogD_7.4. TPSA exhibits a broad spectrum of values, ranging from 7 to 125 Ų. This variability is expected because TPSA accounts for the intricate arrangement and distribution of functional groups within a molecule and the selected compounds represent a diverse array of chemical structures.

The range of MW values is narrower (200–500 Da) and because MW only considers the total mass of the atoms in a molecule, lipophilicity is expected to increase as compounds grow (Figure 1b). Furthermore, it can be observed that the majority of the selected anti-schistosomal compounds sit within the higher end of the cLogD_7.4 range (greater than 3). Lipophilicity is a crucial physicochemical property that plays a significant role in individual ADMET parameters.^20,21 For example, to achieve adequate oral absorption, a compound must strike a balance between aqueous solubility and cell permeability, with LogD being a crucial metric and a determinant of cell passive diffusion for small molecules. Both low (less than 1) and high LogD (greater than 4) values have detrimental effects on permeability.²² The suggested lipophilicity for achieving optimal solubility and permeability typically falls between 1 and 3 LogD values. Therefore, to assess the frequency of occurrence of this range, the cLogD_7.4 values of the subset of 46 compounds were plotted (Figure 2).

Figure 2.

Number of compounds within the suggested optimal cLogD_7.4 range. cLogD_7.4 ranges of values are grouped into three bins, with the (1, 3] bin considered optimum lipophilicity for appropriate oral bioavailability.

Figure 2 illustrates that more than 65% of the compounds (31 out of 46) exhibit cLogD_7.4 values that fall outside the suggested range for optimal solubility and permeability, which could have a detrimental effect on various factors related to ADMET properties and the developability of the molecule. Oxamniquine, metrifonate and PZQ⁶ display cLogD_7.4 values within this range (1.1, 1 and 2.4, respectively). Moreover, the lipophilicity associated with PZQ corresponds to a highly permeable compound (Caco-2 permeability, P_app(A→B) = 39 × 10^–6 cm/s; efflux ratio = 1) with low aqueous solubility, parameters that categorize PZQ as a class II in the biopharmaceutical classification system (BCS).²⁴ Nevertheless, these properties promote drug absorption and facilitate the penetration of the compound in the parasite. This is evident from ¹⁴C experiments, which demonstrate that PZQ is rapidly absorbed and evenly distributed throughout the parasite tissues with minimal biotransformation.²⁵

Analysis of the propensity of compounds to achieve a desirable balance between aqueous solubility and cell permeability within the suggested optimal cLogD_7.4 range is challenging because of the limited availability of published ADMET data. However, the efficacy (400 mg/kg single dose) shown in the murine model by 5 compounds that fall within this LogD range (Table S1) highlights the importance of achieving an equilibrium among potency (e.g., EC₅₀ < 5 and 0.5 μM), cell permeability (e.g., P_app >2 × 10⁶ cm/s) and aqueous solubility (e.g., >10 μM) when developing novel anti-schistosomal hits and/or leads.

Why Do Most Anti-Schistosomal Leads Display High Lipophilicity?

In a typical drug discovery and development screening cascade, bioactivity is employed as the main filter for compound exclusion.²⁰ Hits with the highest in vitro or ex vivo activity tend to show higher lipophilicity values and, therefore, less desirable ADMET properties.²⁰ The challenge then becomes identifying or synthesizing molecules within the optimal lipophilicity range that are sufficiently potent and display the appropriate ADMET attributes for designing libraries or optimizing leads during the drug discovery process. A clear example of such a compound is PZQ, which displays high ex vivo and in vivo activity and high absorption and compound penetration in the parasite.^25,26

Representative Series Displaying High Lipophilicity

To further illustrate the underlying issues of highly lipophilic compounds (cLogD_7.4 > 4), we disclose here for the first time an azepanylcarbazole amino alcohol series (58 and 59, Table S1), originally developed as anti-plasmodials.¹⁷58 and 59 were first screened against ex vivo adult S. mansoni worms, where they demonstrated complete killing at 11 μM following 72h incubation. Despite their moderate ex vivo activity, both compounds display high efficacy against juvenile and adult worms in the murine model. Single 100 mg/kg oral doses administered to S. mansoni-infected mice reduced juvenile and adult worm burden by 100 and 91.5%, and 95.3 and 78.7%, respectively (p < 0.05). The observed efficacy was probably due to the high permeability (Caco-2 permeability of 58, P_app(A→B) = 120 × 10^–6 cm/s; efflux ratio = 1) and moderate anti-schistosomal activity that these two compounds display. Nonetheless, this promising series was discontinued due to its low solubility (FaSSIF pH 6.5 = 10 μM and PBS pH 7.4 = 1.5 μM for compound 58) and cardiovascular toxicity.

At the beginning of this endeavor, there were low expectations that any clear trends could be observed from such a small but diverse collection of active compounds. However, the data suggests that the most successful anti-schistosomal drugs fall within the optimal lipophilicity (LogD_7.4 between 1 and 3) for suitable aqueous solubility and cell permeability.^22,23 In addition, PZQ serves as a prime example of the importance of striking the right balance between potency and these two key physicochemical properties for the development of high-quality leads that could be more effective in treating and preventing schistosomiasis.

Therefore, what are the key recommendations for the discovery of the next generation of anti-schistosomal leads?

ADMET

• ADMET experiments should be performed early in the drug discovery and development process. The cost of these experiments is justified because the number of compounds screened in the schistosomiasis field is smaller than in other neglected tropical diseases.

• It is important to select promising leads with the lowest lipophilicity possible that show the right balance among aqueous solubility, cell permeability and potency and establish trends between the former physicochemical properties and chemical descriptors to reduce the number of unnecessary ADMET experiments.

Parasitology

• NTS activity does not always correlate with killing of juvenile and adult worms but can be used as a high-throughput tool for phenotypic screening of large compound libraries, prior to the application of medicinal chemistry optimization.

• Different incubation times (from 24h to 120h), Schistosoma species in ex vivo assays and dose regiments and routes of administration in the in vivo murine experiments are being used by the scientific community. Therefore, standardization of assays in the anti-schistosomal arena to allow comparison of data between research groups will aid and enhance the discovery and development of new therapeutics. Shorter incubation times (e.g., 12h, 24h, 48h) will be preferred to increase the likelihood of translation from ex vivo to in vivo activity. Fast-acting compounds, detected by low incubation times, show stronger translation between ex vivo and in vivo if they exhibit efficient penetration into target and acceptable activity and rely less on pharmacokinetics parameters such as long exposure times.

• Leads should show multistage activity (S. mansoni, NTS - if correlation between schistosomula and worm activity is observed; juvenile worms, ∼3 weeks, and adult worms, ∼6–10 weeks) and display activity against other clinically relevant species such as S. hematobium.⁵

Glossary

Abbreviations

CAS

Chemical Abstracts Service

FDA

U.S. Food and Drug Administration

PZQ

praziquantel

EC₅₀

half-maximal effective response

P_app

apparent permeability

S. mansoni

Schistosoma mansoni

S. hematobium

Schistosoma hematobium

S. japonicum

Schistosoma japonicum

HDAC

histone deacetylase

HAT

histone acetyltransferase

VKR

venus kinase receptor

LSD

lysine-specific demethylase

TRP

transient receptor potential

NTS

newly transformed schistosomula

ADMET

absorption, distribution, metabolism, excretion and toxicity

SAR

structure–activity relationship

HBD

hydrogen bond donor

HBA

hydrogen bond acceptor

MW

molecular weight

Da

dalton (unit)

TPSA

topological polar surface area

BCS

biopharmaceutical classification system

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsmedchemlett.4c00026.

• Table S1, showing data of 57 anti-schistosomal compounds supporting the findings of this study including structures, names or codes, sources, molecular descriptors and ex vivo and in vivo activities (XLSX)

• Material and methods for in silico, ex vivo and in vivo experiments (PDF)

The authors declare the following competing financial interest(s): T.S. is an employee of Ares Trading SA, Eysins, Switzerland, an affiliate of Merck KGaA, Darmstadt, Germany. This work was funded by Merck Healthcare KGaA, Darmstadt, Germany (CrossRef Funder ID: 10.13039/100009945).