A new C–C linked benzophenathridine–2-quinoline dimer, and the antiplasmodial activity of alkaloids from Zanthoxylum holstzianum

Abstract The CH2Cl2/MeOH (1:1) extract of Zanthoxylum holstzianum stem bark showed good antiplasmodial activity (IC50 2.5 ± 0.3 and 2.6 ± 0.3 µg/mL against the W2 and D6 strains of Plasmodium falciparum, respectively). From the extract five benzophenanthridine alkaloids [8-acetonyldihydrochelerythrine (1), nitidine (2), dihydrochelerythine (3), norchelerythrine (5), arnottianamide (8)]; a 2-quinolone alkaloid [N-methylflindersine (4)]; a lignan [4,4′-dihydroxy-3,3′-dimethoxylignan-9,9′-diyl diacetate (7)] and a dimer of a benzophenanthridine and 2-quinoline [holstzianoquinoline (6)] were isolated. The CH2Cl2/MeOH (1:1) extract of the root bark afforded 1, 3–6, 8, chelerythridimerine (9) and 9-demethyloxychelerythrine (10). Holstzianoquinoline (6) is new, and is the second dimer linked by a C-C bond of a benzophenanthridine and a 2-quinoline reported thus far. The compounds were identified based on spectroscopic evidence. Amongst five compounds (1-5) tested against two strains of P. falciparum, nitidine (IC50 0.11 ± 0.01 µg/mL against W2 and D6 strains) and norchelerythrine (IC50 value of 0.15 ± 0.01 µg/mL against D6 strain) were the most active. Graphical Abstract


Introduction
The continuous emergence of resistance against antimalarial drugs means that there will always be need for new drugs to heal from malaria (Timothy and Wells 2011; Tajuddeen and Van Heerden 2019). In fact, if the recently reported resistance to artemisinin derivatives (Mbengue et al. 2015) escalates and spreads globally, the population at malaria risk area will be defenseless against this deadly parasite (Grimberg and Mehlotra 2011;Tse et al. 2019). In this regard, plants, especially those used in traditional medicinal practice, remain a potential source of new lead compounds. Amongst the different natural products known for their antiplasmodial activity over the years, the Cinchona alkaloids, particularly quinine from C. officinalis L. (as well as its synthetic derivatives), have demonstrated to be amongst the most important antimalarial products until apparition of parasitic resistance against them (Achan et al. 2011).
Some plants of the genus Zanthoxylum from Kenya and Uganda, for instance Z. chalybeum, are used to treat malaria (Bbosa et al. 2014;Waiganjo et al. 2020). These plants show antipyretic activity in either crude form or pure compounds (Jullian et al. 2006;Were et al. 2010). Besides the antiplasmodial activity attributed to the alkaloids of Zanthoxylum species, these plants show a wide range of biological activities including, antinociceptive, anticancer, anti-inflammatory, anti-oxidant and antimicrobial activities (Patino et al. 2012;Nooreen et al. 2019). Seven Zanthoxylum species are known to occur in Kenya (Beentje 1994). Investigation of some of these species have resulted in the isolation of phenanthridine alkaloids, alkamide derivatives, lignans, b-carboline alkaloids, triterpenes and sterols (Kato et al. 1996;Nyakobe et al. 2018;Kaigongi et al. 2020;Omosa et al. 2021a;Omosa et al. 2021b). Malaria being endemic to Africa, the importance of identifying a bioactive plant in the region cannot be over emphasized, since most people in malaria-endemic areas are likely to take herbal medicine before seeking treatment in the formal health facilities (Ocloo et al. 2014). On the basis of the African traditional medicine precedingly exposed, we have performed the quest of antiplasmodial activity of the stem bark and the root bark of Zanthoxylum holstzianum by means of isolation, structural characterization and antimalarial activity evaluation of the alkaloid content of the species.
The new compound (6 Figure 1) was isolated as colourless crystals with a green fluorescence on silica gel TLC plates under UV (366 nm) light. The compound gave a positive Dragendorff's reagent test. The HRMS showed a [M] þ peak at m/z 588.2244 suggesting a molecular formula of C 36 H 32 O 6 N 2 , and indicating a dimeric alkaloid. In the EI-MS, the fragment ions at m/z, 348 ([C 21 H 18 NO 4 ] þ ) and 226 ([C 14 H 12 NO 2 ] þ ), suggest that compound 6 could be a dimer of dihydrochelerythrine (3) and N-methylflindersine (4), co-metabolites. The 1 H NMR (600 MHz, Figure S3) and 13 C NMR (150 MHz, Figure S4) in DMSO-d 6 showed duplication of signals, typical of diastereomeric mixtures. The 1 H NMR spectrum shows that one half of compound 6 is indeed dihydrochelerythrine ( 3 Hz) suggest that it is coupled to a methylene protons of a substituent at C-8 (Section 3.5, Table S1) as in 8-acetonyldihydrochelerythrine (1) (Supporting Materials).
The C 8 -C 1 0 linkage between the two moieties was established from the HMBC spectrum ( Figure S6), where the methyl protons at d H 1.407/1.583 (Me-2 0 ) showed 3 J HMBC correlation with C-1 0 (methylene carbon at d C 43.19/44.82), C-3 0 (d C 126.49/127.28), and 2 J with C-2 0 (d C 80.39/80.42). The connectivity between the two moieties was further supported by long-range HMBC correlation between H-8 and C-2 0 , and between the methylene protons CH 2 -1 0 and C-8 (Table S1). The H,H-COSY spectrum ( Figure S5) showed a cross peak between the benzylic proton (H-8) and the two methylene protons (CH 2 -1 0 ), supporting the C 8 -C 1 ' linkage. In agreement with this, the NOESY spectrum ( Figure S8) showed correlation of H-8 (d H 4.764/4.863) with CH 2 -1 0 and CH 3 -2 0 . The N-methyl group (d H 2.459/2.584) also showed HMBC correlation with C-8 and C-6 as in compound 1. Complete assignment was done using HMBC ( Figure S6) and HSQC ( Figure S7) spectra, as well as comparing with the spectra data with those for compounds 1 and 4 (Supporting Materials). The data of the new compound (6), trivial name holstzianoquinoline, was further compared with that of simulanoquinoline (Wu and Chen 1993), the only related dimeric alkaloid isolated from Zanthoxylum simulans, and was in good agreement.
According to Batista et al. (2009), a compound with IC 50 >200 mg/mL is considered to be inactive; IC 50 100-200 mg/mL, low activity; IC 50 20-100 mg/mL, moderate activity; IC 50 1-20 mg/mL, good activity; and IC 50 <1mg/mL, excellent or potent antiplasmodial activity. Among the isolated compounds, nitidine (against both W2 and D6 strains) and norchelerythrine (5) (against D6 strain) showed potent activity. 8-Acetonyldihydrochelerythrine (1), dihydrochelerythine (3), norchelerythrine (5) and Nmethylflindersine (4) showed good activity (Table S2). The activity of nitidine against W2 and D6 and the activity of norchelerythrine (5) against D6 strain was higher than that of the crude extract, and similar to the activity of chloroquine, an antimalarial drug (Table S2). Gakunju et al. (1995), reported the activity of nitidine ranging from 0.009-0.11 mg/mL, which is in agreement with results obtained in this research. The relatively low activity of the 2-quinoline N-methylflindersine (4) compared to the benzophenanthridine alkaloids suggests that, the benzophenathridine nucleus appears to be important for the antiplasmodial activity of the extract.

General
The NMR spectra were recorded using Varian-Mercury 200 MHz and/or Bruker Avance 500 and 600 MHz spectrometers. COSY, HSQC, NOESY and HMBC spectra were obtained using standard Bruker software. Chemical shifts were measured in ppm in d values relative to the residue solvent signals. EI-MS were recorded at 70 eV on GC-TOF micromass spectrometer (micromass Wythenshawe, waters Inc. UK). ESIMS at 70 eV on a Micromass GC-TOF micromass spectrometer (Micromass, Wythenshawe, Waters, Inc. UK). Melting points were obtained on SMP 10 apparatus. Optical rotations were measured on a PerkinElmer 341-LC Polarimeter. The solvents used for chromatography were glass distilled.

Plant material
Zanthoxylum holstzianum (stem bark and root bark) was collected from Diani Veminant Forest, Coastal Province of Kenya in June 2012, and identified by Mr. Patrick Mutiso, a botanist from the Herbarium, Biology Department, University of Nairobi, where a specimen (voucher number, AD-001-2012) was deposited. Each of the plant material was air-dried and ground using a Willy mill.

Extraction and isolation of compounds from stem bark of Zanthoxylum holstzianum
The air dried and ground stem bark of Zanthoxylum holstzianum (3.4 kg) was defatted (4 L Â 4) using n-hexane for 24 hours, in each case by cold percolation. This yielded 170 g of oily extract. The marc was then dried and further extracted with CH 2 Cl 2 / MeOH, 1:1 (4 L Â 6) at room temperature, yielding 324 g (9.5%) of gummy extract. This extract was then acidified with 2 M HCl solution and extracted with dichloromethane followed by EtOAc, yielding 138 g of combined extract after concentration. The remaining aqueous layer was basified with 37% aqueous ammonia and extracted with CH 2 Cl 2 followed by EtOAc two times, yielding 4 g of combined extract after concentration.

Reference Plasmodium falciparum strains and reference antimalarial drugs
The following reagent was obtained through BEI Resources, NIAID, NIH: Plasmodium falciparum, Strain D6, MRA-285, and Strain W2, MRA-157 contributed by Dennis E. Kyle.
The reference antimalarials drugs chloroquine diphosphate (CQ) and Quinine (QN) were donated by the Worldwide Antimalarial Resistance Network External Quality Assurance Programme, Bangkok (Lourens et al. 2010).

Antiplasmodial assay
The crude stem bark extract and the pure compounds were assayed using a nonradioactive assay technique (Smilkstein et al. 2004), with modifications according to Johnson et al. (2007); Cheruiyot et al., 2016) to determine 50% growth inhibition of the cultured parasites. Two Plasmodium falciparum parasite strains, chloroquine-sensitive Sierra Leone I (D6) and chloroquine-resistant Indochina I (W2) were maintained in continuous culture as described by Johnson et al. (2007), to attain a parasitemia of 3-8% based on counts infected versus uninfected red blood cells, confirming successful in vitro culture-adaptation. The crude extract, pure compounds and the reference drug were dissolved in 99.5% DMSO and diluted in complete Roswell Park Memorial Institute 1640 series of cell culture medium (RPMI 160) prepared as described by Akala et al. (2011). The basic culture medium was prepared from RPMI 1640 powder and the complete RPMI 1640 media was stored at 4 C and used within 2 weeks.
Two-fold serial dilutions of chloroquine, the reference drug and test samples were prepared on a 96-well plate, making sure that the amount of DMSO is equal or less than 0.0875%. The culture-adapted P. falciparum at 3-8% parasitemia were first adjusted to 2% hematocrit and 1% parasitemia, then added onto the assay plate containing a range of drug doses and incubated in gas mixture (5% CO 2 , 5% O 2 and 90% N 2 ) at 37 C. The termination of the assay was done 72 hours later by addition of lysis buffer and incubating at room temperature in dark for 24 hours as described by Cheruiyot and coworkers (Cheruiyot et al., 2016). This duration allowed SYBR green I, a cyanide dye with reactive groups on nitrogen ends to get chemically linked to the P. falciparum nucleic acids resulting in a DNA-dye-complex that absorbs 485 nanometer blue light (k max ¼ 497 nm) and emits 535 nanometer green light (k max ¼ 520 nm). More interactions occurred in low drug-dose wells of the assay plate where more parasites replication was maximal during the 72 hours incubation giving more DNA than in higher dose wells. Per drug activity was quantified by analyses of dose response curves generated from the drug-dose dependent relative fluorescence units generated by the Genios tecan fluorescence reader from Baldwin Park, California, United States of America. The parasite growth quantified as mean ± standard deviation (Mean IC 50 ± SD) as described by Johnson et al. (2007).