HPLC-Based Activity Profiling for Antiprotozoal Compounds in Croton gratissimus and Cuscuta hyalina

In a screening of Sudanese medicinal plants for antiprotozoal activity, the chloroform fractions obtained by liquid-liquid partitioning from ethanolic extracts of fruits of Croton gratissimus var. gratissimus and stems of Cuscuta hyalina Roth ex Schult. exhibited in vitro activity against axenically grown Leishmania donovani amastigotes. This antileishmanial activity was localized by HPLC-based activity profiling. Targeted preparative isolation afforded flavonoids 1–6, 3-methoxy-4-hydroxybenzoic acid (7), and benzyltetrahydroisoquinoline alkaloids laudanine (8) and laudanosine (9) from C. gratissimus, and pinoresinol (10), isorhamnetin (11), (-)-pseudosemiglabrin (12), and kaempferol (13) from C. hyalina. The antiprotozoal activity of 1–13 against L. donovani (axenic and intracellular amastigotes), Trypanosoma brucei rhodesiense (bloodstream forms), and Plasmodium falciparum (erythrocytic stages), and the cytotoxicity in L6 murine myoblast cells were determined in vitro. Quercetin-3,7-dimethylether (6) showed the highest activity against axenic L. donovani (IC50, 4.5 µM; selectivity index [SI], 12.3), P. falciparum (IC50, 7.3 µM; SI, 7.6), and T. b. rhodesiense (IC50, 2.4 µM; SI, 23.2). The congener ayanin (2) exhibited moderate antileishmanial (IC50, 8.2 µM; SI, 12.2), antiplasmodial (IC50, 7.8 µM; SI, 12.9), and antitrypanosomal activity (IC50, 11.2 µM; SI, 8.9). None of the compounds showed notable activity against the intramacrophage form of L. donovani.


INTRODUCTION
Parasitic protozoa are the causative agents of devastating, yet often neglected diseases. The kinetoplastids, a group of flagellated protozoa, cause neglected tropical diseases that put more than one billion people around the globe at risk (WHO|World Health Organization, n.d.;Khalid, 2012). These diseases are human African trypanosomiasis (HAT) caused by Trypanosoma brucei spp., Chagas' disease caused by Trypanosoma cruzi, and Leishmaniasis caused by Leishmania spp. (Stuart et al., 2008). The apicomplexan parasite Plasmodium falciparum is the causative agent of malaria tropica which claims more than 400,000 lives every year (Organization, W.H. 2019).
These infections are of high public health relevance and socioeconomic impact. Most of the currently available drugs have drawbacks in terms of toxicity, limited availability of oral therapeutic dosage forms, development of resistance, or non-affordability.
Natural products have in many instances provided new leads to combat neglected tropical diseases (Schmidt et al., 2012). As part of an ongoing screening project of Sudanese medicinal plants for antiprotozoal activity (Mahmoud et al., 2020a;Mahmoud et al., 2020b), the chloroform extract of Croton gratissimus var. gratissimus (Euphorbiaceae), and Cuscuta hyalina Roth ex Schult. (Convolvulaceae) showed promising activity against P. falciparum and Leishmania donovani.
The genus Croton comprises over 1300 species that are widely distributed throughout tropical and subtropical regions of the world. Croton species have been used traditionally in Africa, South Asia, and Latin America for the treatment of infections and digestive disorders (Wu and Zhao, 2004;Xu et al., 2018). In Sudan, C. gratissimus, locally known as Um-Geleigla, has been used traditionally for the treatment of hypertension and malaria (Mohamed et al., 2009). The main secondary metabolites include flavonoids, terpenoids, and essential oil (Ngadjui et al., 2002;Aderogba et al., 2011;Yagi et al., 2016). Previous studies have demonstrated that the roots of C. gratissimus possessed antiplasmodial activity in vivo (Okokon and Nwafor, 2009). Cembranolide diterpenes isolated from the leaves were found to be active when tested against P. falciparum (Langat et al., 2011).
The genus Cuscuta comprises over 200 species distributed worldwide. They are stem obligate holoparasitic plants possessing neither roots nor fully expanded leaves. The interaction between parasite and host is established through haustoria (Kaiser et al., 2015). Different Cuscuta species have been used in traditional Indian and Chinese medicine. Cytotoxic, antioxidant, and antimicrobial activities have been reported (Ahmad et al., 2017). Previous phytochemical investigations of the genus Cuscuta identified flavonoids, lignans, alkaloids, fatty acids, and essential oil (Donnapee et al., 2014;Ahmad et al., 2017). The phytochemistry and antiparasitic activity of C. hyalina has not been studied.
In an earlier screening of Sudanese medicinal plants for antiprotozoal activity, the ethanolic extracts of Croton gratissimus fruits and Cuscuta hyalina stems had been found to exhibit in vitro antiprotozoal activity against axenic L. donovani (MHOM/ET/67/L82). Subsequent liquid-liquid partitioning against petroleum ether, chloroform, and ethyl acetate located the activity in the chloroform portion (Mahmoud et al., 2020b). We here report on the targeted isolation and structure elucidation of compounds responsible for the activity, and on their in vitro activity against T. b. rhodesiense (STIB 900), axenic and intramacrophage amastigotes of L. donovani (MHOM/ET/67/L82), and P. falciparum (NF54).

General Experimental Procedures
HPLC-grade methanol and acetonitrile from Macron Fine Chemicals (Avantor Performance Materials), and water from Milli-Q water purification systems (Merck Millipore) were used for HPLC separations. For fractionation and preparative separation, technical grade solvents from Scharlau (Scharlab S. L.) were used after distillation. Silica gel 60 F 254 coated aluminum TLC plates were obtained from Merck. Silica gel (230-400 μm, Merck) and Sephadex LH-20 (25-100 μm, Sigma-Aldrich) were used for open column chromatography. Optical rotation was measured in methanol using a JASCO P-2000 digital polarimeter equipped with a sodium lamp (589 nm) and a temperaturecontrolled microcell (10 cm). UV and ECD spectra were recorded in methanol on a Chirascan CD spectrometer (Applied Photophysics) using 110 QS 1 mm path precision cells (Hellma Analytics). NMR spectra were recorded on a Bruker Avance III NMR spectrometer operating at 500.13 MHz for 1 H and 125.77 MHz for 13 C. 1 H NMR, COSY, HSQC, HMBC, and NOESY spectra were measured at 23°C in a 1 mm TXI probe with a z-gradient, using standard Bruker pulse sequences. Spectra were analyzed by Bruker TopSpin 3.5 pl 7 and ACDLabs Spectrus Processor. NMR spectra were recorded in DMSO-d 6 (99.9 atom % D; Armar Chemicals).
HPLC-PDA-ELSD-ESIMS data were recorded in positiveand negative-ion mode (scan range of m/z 200-1,500) on a Shimadzu LC-MS/MS 8030 triple quadrupole MS system, connected via a T-splitter (1:10) to a Shimadzu HPLC system consisting of degasser, binary mixing pump, autosampler, column oven, and a diode array detector and to an Alltech 3300 ELSD detector. Separation was achieved on a SunFire C 18 (3.5 mm, 150 × 3.0 mm i.d.) column equipped with a guard column (10 mm × 3.0 mm i.d.) (Waters). Data acquisition and processing were performed with LabSolution software.
Microfractionation was carried out with the same HPLC instrument connected via a T split to an FC204 fraction collector (Gilson) with only UV detection, using a SunFire C 18 (3.5 mm, 150 mm × 3.0 mm i.d.) column equipped with a guard column (10 mm × 3.0 mm i.d.) (Waters).
All handling of infectious agents (L. donovani, T. b. rhodesiense, P. falciparum) was performed under strict biosafety level 2 conditions under notification A000275 to the Swiss Federal Office of Public Health.

Plant Material
Croton gratissimus var. gratissimus fruits and Cuscuta hyalina Roth ex Schult. stems were obtained from the Herbarium of the Faculty of Pharmacy, University of Science and Technology, Omdurman, Sudan. The taxonomic identity was confirmed by the Medicinal and Aromatic Plants Research Institute, Sudan and voucher specimens (CZFCHL02 and ChSCHL 02) were deposited. Plant materials were dried at room temperature and milled before extraction.

Extraction
Powdered materials of C. gratissimus fruits and C. hyalina stems (500 g each), respectively, were extracted with 1 L of 70% ethanol and kept in a magnet rod shaker for 24 h. The extraction procedure was repeated three times for each herbal drug. Extracts were filtered and dried under reduced pressure. For each plant, the ethanolic extract was suspended in water and partitioned successively with petroleum ether, chloroform, and ethyl acetate. Three repetitive partitioning procedures, each with 500 ml of either solvent were performed. This afforded 3.5 and 1.2 g of the chloroform extracts of C. gratissimus fruits and C. hyalina stems, respectively.

Sample Preparation
Compounds were dissolved in DMSO (10 mg/ml) and warmed up to 40°C and/or sonicated if necessary. These DMSO stocks were kept at −20°C. For each assay, a fresh dilution to 100 μg/ml in medium was prepared. This was used to prepare the serial dilutions directly in the 96-well assay plates. Since DMSO is cytotoxic, the maximum DMSO concentration in the test was 1%.

Activity Against Leishmania donovani Axenic Amastigotes
Amastigotes of L. donovani strain MHOM/ET/67/L82 were grown under an atmosphere of 5% CO 2 in air in axenic culture at 37°C in SM medium (Cunningham, 1977) at pH 5.4 supplemented with 10% heat-inactivated fetal bovine serum. 50 μL of culture medium was added in the wells of a 96-well plate and serial drug dilutions of eleven three-fold dilution steps covering a final range from 100 to 0.002 mg/ml were prepared. 50 μL culture medium with 2 × 10 5 amastigotes from axenic culture were added to each well. After 70 h of incubation the plates were inspected under an inverted microscope to assure growth of the controls and sterile conditions. 10 ml of resazurin (12.5 mg resazurin dissolved in 100 ml distilled water) were added to each well and the plates incubated for another 2 h. Then the plates were read with a Spectramax Gemini XS microplate fluorometer (Molecular Devices Cooperation, Sunnyvale, CA, USA) using an excitation wavelength of 536 nm and an emission wavelength of 588 nm. Data were analyzed using the software Softmax Pro (Molecular Devices Cooperation, Sunnyvale, CA, USA). Decrease of fluorescence (= inhibition) was expressed as percentage of the fluorescence of untreated control cultures and plotted against the drug concentrations. From the sigmoidal inhibition curves the IC 50 values were calculated. Miltefosine was used as positive control drug. Assays were performed in two independent replicates at least.

Activity Against Leishmania donovani Intramacrophage Amastigotes
Macrophages were isolated from the mouse (CDI) peritoneal cavity (Zhang et al., 2008) using 2% starch solution as eliciting agent injected 2 day prior to cell harvest. Animal work was carried out according to the rules and regulations for the protection of animal rights ("Tierschutzverordnung") of the Swiss "Bundesamt für Veterinärwesen" (License number 2374). Mouse peritoneal macrophages (4 × 10 4 in 100 μL RPMI 1640 medium with 10% heat-inactivated FBS) were seeded into wells of a 96-well plate. After 24 h, 2 × 10 5 amastigote L. donovani in 100 μL were added. The amastigotes were taken from an axenic amastigote culture grown at pH 5.4. The medium containing free amastigote forms was removed after 24 h and replaced with fresh medium. The washing step was repeated and afterward the serial drug dilution was prepared with at least 6 dilution steps. Compounds were dissolved in DMSO at 10 mg/ml and further diluted in medium. After 96 h of incubation at 37°C under a 5% CO 2 atmosphere, the medium was removed, and cells were fixed by adding 50 μl 4% formaldehyde solution followed by a staining with a 5 μM DRAQ5 solution. Plates were imaged in ImageXpress XLS (MD) microscope using a 20× air objective (635 nm excitation: 690/50 emission). 9 images were collected per well. Automated image analysis was performed with a script developed on Meta Xpress Software (MD). Three outputs were provided for each well: i) number of host cell nuclei; ii) numbers of infected and non-infected host cells; iii) number of parasite nuclei per infected host cell. The IC 50 values were calculated based on the infection rate and the numbers of intracellular amastigotes. The cytotoxicity to macrophages was determined in parallel, and IC 50 values were calculated based on the numbers of surviving, uninfected macrophages. Miltefosine was used as control. Assays were performed in two independent replicates at least.

Activity Against Trypanosoma brucei rhodesiense STIB900
The stock was originally isolated from a Tanzanian patient and adapted to axenic culture conditions after several mouse passages and cloned. Minimum Essential Medium (50 μl) supplemented with 25 mM HEPES, 1 g/L additional glucose, 1% MEM nonessential amino acids (100×), 0.2 mM 2-mercaptoethanol, 1 mM Na-pyruvate (Baltz et al., 1985) and 15% heat inactivated horse serum was added to each well of a 96-well microtiter plate. Serial drug dilutions of eleven 3-fold dilution steps covering a range from 100 to 0.002 mg/ml were prepared. Then 4 × 10 3 bloodstream forms of T. b. rhodesiense STIB 900 in 50 μL were added to each well and the plate incubated for 70 h at 37°C and under a 5% CO 2 atmosphere. 10 μL resazurin solution (resazurin, 12.5 mg in 100 ml double-distilled water) was then added to each well and incubation continued for a further 2-4 h (Räz et al., 1997). Plates were read with a Spectramax Gemini XS microplate fluorometer (Molecular Devices Cooperation, Sunnyvale, CA, USA) using an excitation wavelength of 536 nm and an emission wavelength of 588 nm. Softmax Pro program (Molecular Devices Cooperation, Sunnyvale, CA, USA) was used for data analyses and IC 50 values were calculated by linear regression (Huber and Koella, 1993), and 4-parameter logistic regression from the sigmoidal dose inhibition curves. Melarsoprol (Arsobal Sanofi-Aventis, received from WHO) was used as control. Assays were performed in two independent replicates at least.

Activity Against Plasmodium falciparum
In vitro activity against the erythrocytic stages of P. falciparum was determined using a 3 H-hypoxanthine incorporation assay (Desjardins et al., 1979), using the drug sensitive NF54 strain (Ponnudurai et al., 1981). Compounds were dissolved in DMSO at 10 mg/ml and further diluted in medium before addition to parasite cultures incubated in RPMI 1640 medium without hypoxanthine, supplemented with HEPES (5.94 g/L), NaHCO 3 (2.1 g/L), neomycin (100 U/ml), Albumax R (5 g/L), and washed human red cells A + at 2.5% haematocrit (0.3% parasitemia). Serial drug dilutions of 11 three-fold dilution steps covering a range from 100 to 0.002 mg/ml were prepared. The 96-well plates were incubated in a humidified atmosphere at 37°C; 4% CO 2 , 3% O 2 , 93% N 2 . After 48 h, 50 ml of 3 H-hypoxanthine (=0.5 mCi) was added to each well of the plate. The plates were incubated for a further 24 h under the same conditions. The plates were then harvested with a Betaplate ™ cell harvester (Wallac, Zurich, Switzerland), the red blood cells transferred onto a glass fibre filter, and lysed with distilled water. The dried filters were inserted into a plastic foil with 10 ml of scintillation fluid and counted in a Betaplate ™ liquid scintillation counter (Wallac, Zurich, Switzerland). IC 50 values were calculated from sigmoidal inhibition curves by linear regression using Microsoft Excel. Chloroquine (Sigma C6628) was used as control. Assays were performed in two independent replicates at least.

In Vitro Cytotoxicity With L-6 Cells
Assays were performed in 96-well microtiter plates, each well containing 100 μl of RPMI 1640 medium supplemented with 1% L-glutamine (200 mM) and 10% fetal bovine serum, and 4000 L-6 cells (a primary cell line derived from rat skeletal myoblasts) (Ahmed et al., 1994). Serial drug dilutions of 11 three-fold dilution steps covering a range from 100 to 0.002 mg/ml were prepared 24 h post seeding L-6 cells. The plates were incubated for 70 h and inspected under an inverted microscope to assure growth of the controls and sterile conditions. 10 μl of resazurin was then added to each well and the plates incubated for another 2 h. Then the plates were read with a Spectramax Gemini XS microplate fluorometer (Molecular Devices Cooperation, Sunnyvale, CA, USA) using an excitation wavelength of 536 nm and an emission wavelength of 588 nm. The IC 50 values were calculated by linear regression and four-parameter logistic regression from the sigmoidal dose inhibition curves using SoftmaxPro software (Molecular Devices Cooperation, Sunnyvale, CA, USA). Podophyllotoxin (Sigma P4405) was used as positive control. All assays were performed in two independent replicates at least. Activities of all compounds were expressed in μM using the formula: Activity (mM) = Activity (mg=ml) Â 1000=Molecular weight:

Extraction and HPLC-Based Activity Profiling
The methanolic extracts of Croton gratissimus var. gratissimus fruits and Cuscuta hyalina Roth ex Schult. stems had been previously found to exhibit antiprotozoal activity (Mahmoud et al., 2020b). The antileishmanial activity displayed by the chloroform fractions of the two plants was tracked by HPLC-based activity profiling, a procedure combining analytical separation with on-line spectroscopy and time-based microfractionation for bioactivity testing (Potterat and Hamburger, 2013;Potterat and Hamburger, 2014). One-minute microfractions were collected and tested for L. donovani growth inhibition. The HPLC−ESIMS (positive base peak chromatograms) trace and the corresponding antileishmanial activity profiles for C. gratissimus and C. hyalina are shown in Figures 1 and 2. Major antileishmanial activity and a series of distinct peaks in the HPLC-ESIMS trace were observed in the time FIGURE 1 | HPLC-based activity profiling of the chloroform fraction of Croton gratissimus var. gratissimus against axenic amastigotes of L. donovani. The ESIMS (positive base peak chromatogram) of a separation of 300 mg of fraction on an analytical RP-HPLC column is shown. Activities of 1-min microfractions are shown with grey columns, and are expressed as percent growth inhibition compared to untreated parasites. Bold numbers in the chromatogram refer to compounds 1-9 window between 18 and 24 min for C. gratissimus, and between 13 and 17 min for C. hyalina.

Compound Isolation and Structure Elucidation
Separation of the chloroform fraction of C. gratissimus on a Sephadex LH-20 column yielded 19 subfractions (A-S). Based on the HPLC−PDA-ESIMS analysis, subfractions M, O, K, and C were found to contain peaks associated with the active time window. Further purification by semipreparative RP-HPLC afforded compounds 1-3 from subfraction M, 4-6 from subfraction O, 7 from subfraction K, and 8 and 9 from subfraction C.
Preparative chromatography on silica gel of the chloroform fraction of C. hyalina yielded 16 subfractions (A-P). Peaks associated with the active time window were detected in subfraction B. Further separation by semipreparative RP-HPLC afforded compounds 10-12 (Figure 3).
Kaempferol (13) and isorhamnetin (11) have been previously reported from different Cuscuta species (Ahmad et al., 2017), while pinoresinol has been identified in C. chinensis (Yahara et al., 1994). To the best of our knowledge, this is the first report on isolation of pseudosemiglabrin (12) from Cuscuta species.
After a first testing against axenic amastigotes, compounds were tested against L. donovani amastigotes in mouse macrophages. However, in this more elaborate and more physiological model none of the compounds showed activity ( Table 1). In general, IC 50 values for the intramacrophage form are higher than those for the axenic amastigotes (Berry et al., 2018). This loss of activity in the intracellular model could be due to poor cellular permeability of the compounds, binding to cytosolic proteins in the host cell, or metabolism in the host cell phagolysosome (Burchmore and Barrett, 2001;De Rycker et al., 2013;Berry et al., 2018).

Correlation Between Chemical Structure of Isolated Flavonoids and Antiprotozoal Activity
Of the isolated compounds, only flavones showed notable activity ( Table 1). From a comparison of flavones 1-3, 5, 6, 10, and 13 the following conclusions can be drawn: Compounds with a hydroxyl group at C-3′ (2, 5, and 6) were the most active against the three parasites, whereby a catechol moiety as in 6 further increased the activity. Free hydroxyl groups at C-3 or C-7 (as in 5, 11, and 13) had not result in significant in vitro activity. Compounds 2 and 6 exhibited the highest selectivity, while 5 showed significant cytotoxicity in L6 cells leading to a low SI. Naringenin (4) displayed the weakest antiparasitic activity among the tested flavonoids. The presence of a double bond between C-2 and C-3 has been previously found to be essential for antiparasitic activity (Tasdemir et al., 2006). Overall, our results were in agreement with previous structure-activity studies of flavonoids (Tasdemir et al., 2006).
The influence of a balance between antioxidant and prooxidant properties of flavonoids on antiparasitic activity, and a correlation with their chemical structure has been investigated with the aid of QSAR models (Baldim et al., 2017). Compounds that displayed moderate to higher antitrypanosomal activity shared structural features, such as D 2,3 unsaturation, presence of a hydroxyl group at C-3, a carbonyl group at C-4, and a catechol moiety in ring B. Our results were in line with these findings. To the best of our knowledge, the antitrypanosomal activities of quercetin-3,7dimethylether (6) and ayanin (2) are here reported for the first time.

FUNDING
This work was supported by grants to AM by the Amt für Ausbildungsbeiträge Basel (www.hochschulen.bs.ch/ueber-uns/ organisation/amtausbildungsbeitraege.html) and the Emilia Guggenheim-Schnurr Foundation (www.ngib.ch/stiftung-egs). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.