In vivo Antimalarial Activity of the Hydroalcoholic Extract of Knipho a foliosa Hochst and Its Constituents


 Background: Kniphofia foliosa Hochst is endemic to Ethiopian highlands, where its rhizomes are traditionally used for the treatment of malaria, abdominal cramps and wound healing. As a continuation of our search for antimalarial compounds from Ethiopian medicinal plants, we have tested the 80% methanol extract of K. foliosa rhizomes and its constituents against Plasmodium berghei in mice. Methods: Isolation was carried out using column and preparative thin layer chromatography (PTLC). The chemical structures of the compounds were elucidated by spectroscopic methods (ESI-MS, 1D and 2D-NMR). Peters’ 4-day suppressive test against P. berghei in mice was utilized for in vivo antimalarial evaluation of the test substances. Results: Three compounds, namely knipholone, dianellin, and 12-hydroxypentadec-9-en-1-yl methyl phthalate (HPMP) were isolated and characterized from the 80% methanolic extract of K. foliosa rhizomes. The hydroalcoholic extract (400 mg/kg) and knipholone (200 mg/kg) showed the highest activity with chemosuppression values of 61.52 and 60.16%, respectively. From the dose-response plot, the median effective (ED50) doses of knipholone and dianellin were determined to be 81.25 and 92.31 mg/kg, respectively. Molecular docking study revealed that knipholone had a strong binding affinity to Plasmodium falciparum l-lactate dehydrogenase (pfLDH) target. Conclusion: Results of the current study support the traditional use of the plant for the treatment of malaria.

observed individually for any physical or behavioral changes such as loss of appetite, hair erection, lacrimation, mortality, and other signs of toxicity for 4 h. The same procedure was followed for the remaining mice for the next ve consecutive days and the results recorded. The follow-up observations was continued for all mice for 14 days.
In vivo antimalarial assay Inoculation Blood smear was prepared on microscope slides from blood lms taken from the donor (infected) mouse tail. The smear was xed with methanol and stained with Giemsa to count the parasitemia of the donor under a microscope. The mice were then inoculated on day 0 with parasitized erythrocytes obtained from the donor by cardiac puncture using a sterile syringe when the parasitaemia level was 30-40%. The blood from the donor was collected on a Petri dish containing 2% trisodium citrate and was immediately diluted with uninfected mouse blood and normal saline in such way that the nal volume contains 5 × 10 7 infected erythrocytes/ml of blood. The diluted blood (0.2 ml) was injected into all the experimental mice intraperitoneally [22,23].

4-Day suppressive test
The standard 4-day suppressive method was used for antimalarial evaluation of the test substances. The test was carried out in two phases. The extract and phenol fractions were evaluated in the rst phase followed by KFP-1 and YKFM-2 in the second phase. During the rst phase, 60 inoculated mice were randomly grouped into 12 groups each having ve mice. Groups 1 served as a negative control (distilled water,Vehicle1, 0.2 ml) for the extract and phenolic fraction 2 treated groups, while group 2 animals were used as a negative control (1% tween 80, Vehicle2, 0.2 ml) for phenolic fraction 1 treated group. The third group which served as a positive control was treated with standard pure chloroquine (25 mg/kg/day). The remaining nine groups were treatment groups and received 100, 200, and 400 mg/kg/day of the hydroalcoholic extract and the two phenol fractions. Similarly, during evaluation of KPF-1 and YKFM-2, 45 inoculated mice were randomly grouped into 9 groups, each containing ve mice. The rst two groups were negative controls (distilled water, Vehecle3, 0.2 ml) and positive controls (standard pure chloroquine, 25 mg/kg/day). The rest of the groups were treatment groups and received KFP-1 or YKFM-2 at doses of 25, 50, 100 and 200 mg/kg/day. All the test substances were administered orally using oral gavage. Treatment was started 3 h post-infection on day 0 and continued daily for the next 3 days (i.e. from day 0 to day 3). On the fth day (or day 4), two Giemsa-stained blood smears were prepared from each mice to count the number of parasites under the microscope with an oil immersion objective of 100x magni cation power [24,25,26].
Mean percent parasitaemia and percent suppression were calculated using the following formulae.
Body weight and survival time measurement Body weight of each mouse was measured on day 0 before infection and on day 4. Survival time was recorded from day 1 to day 28 post inoculation. Then, mean body weight and mean survival time were calculated for each group [19].

Molecular docking study
Docking study was carried out on two crystal structures of plasmodium enzymes plasmepsin II (Protein Data Bank; PDB: 4CKU) and l-lactate dehydrogenase (pfLDH) [PDB: 1LDG], using SeeSar10.0 software (BioSolveIT, Sankt Augustin, Germany). For plasmepsin II, the selected binding site was the binding pocket of a previously designed inhibitor P2FE-400, while for pfLDH, the cofactor nicotinamide adenine dinucleotide (NADH) binding site was selected for docking. The HYDE score was used to estimate the binding a nity of the molecules [27,28].

Statistical analysis
Data analysis was carried out using IBM SPSS (Statistical Package for Social Sciences) Statistics for Windows, Version 20.0. Armonk, NY: IBM Corp. Results were expressed as mean ± standard error of mean (M ± SEM). The statistical signi cance was determined by one-way ANOVA followed by Tukey post hoc test to compare percent suppression (activity), mean survival time and percent changes in body weight of the P. berghei infected mice among the treatment and control groups. P < 0.05 were considered signi cant.  H-5) indicated the presence of aromatic ring moiety. Moreover, three of these proton signals which are multiplets imply that they are found in close proximity (or are adjacent) and the fourth singlet aromatic proton peak at δ 7.12 (s, H-4) provides clues for the presence of a fused aromatic ring system. The presence of a disaccharide unit in compound 1 was revealed by the typical anomeric proton signals at δ 4.28 (d, J = 3.2 Hz, 1H, H-4′′) and 5.25 (d, J = 3.7 Hz, 1H, H-1′). The proton peaks from δ 5.25 to 3.06 further justify the presence of a disaccharide moiety. The 13 C spectrum region from δ 76.82 to 66.59 also con rmed that the compound contains a disaccharide moiety. In addition, the two elevated 13 C sugar signals at δ 102.84 and 100.85 indicate that the sugar units are linked through acetal bond. Furthermore, the absence of one CH signal in the sugar region (δ 76.82 -δ 66.59) suggests one of the sugar units to be rhamnose. And this was found to be in good agreement with 13 C NMR reports of similar glycosides [29,30]. Hence, the disaccharide moiety was con rmed to be rhamnose-glucose 1,6 linkage. In addition, the presence of 10 13 C signals from δ 154.71 to δ 110.5 implies that the fused aromatic ring system is naphthalene. Six of these carbon signals are absent from DEPT spectrum indicating they are quaternary aromatic carbons. Besides, two of them are elevated (δ 154.71 and δ 151.49) suggesting that they are oxygenated quaternary aromatic carbons. On the other hand, the two less elevated (δ 136.74 and δ 113.54) quaternary aromatic carbons are the bridgehead carbons of the fused aromatic system [30]. The remaining two quaternary aromatic carbon signals resonated at δ 124.73 (C-2) and δ 133.30 (C-3). Lastly, the 13 C signals at 207.7 and 41.3 are the carbonyl carbon and its acetyl methyl. Therefore, based on the above evidence and in comparison with 1 H and 13 C NMR data of the same and related compounds [30,31], the structure of compound 1 was determined to be dianellin or 1-(1-hydroxy-3-methyl-8-(((2S,3R,4S,5S,6R)3,4,5-trihydroxy-6-((((2S,3S,4S,5S,6R)3,4,5-trihydroxy-6-methyltetrahydro-2H-pyran-2-yl)oxy)methyl)tetrahydro-2H-pyran-2yl)oxy)naphthalen-2-yl) ethanone. Table 1 summarizes the NMR data of compound 1. indicate the presence of chrysophanol moiety. Besides, the singlet aromatic proton signal present at 7.28

Results And Discussion
suggests that it is found adjacent to a substituted aromatic carbon. The 13 C and DEPT spectra of compound 2 also support the presence of chrysophanol moiety [32,33]. Moreover, from the 13 [34,35,36,37]. Hence, based on the above data and in comparison with the 1 H NMR and 13 C NMR of related compounds, the structure of compound 3 was determined to be 12hydroxypentadec-9-en-1-yl methyl phthalate (HPMP). Table 2 summarizes the NMR data of compounds 2 and 3.  [21,38]. After 72 hours, the animals tolerated the administered dose although immediate mild toxicity signs such as hair erection and loss of appetite, which disappeared few hours after administration were observed. Also, there was no mortality within 14 days of observation which entails that the LD 50 s of the extract and knipholone are above 2000 mg/kg.

Antimalarial activity of the hydroalcoholic extract
The 80% methanol extract of K. folosia showed chemosuppressive effect against P. berghei in mice (Table 3). At all dose levels tested, the extract exhibited a statistically signi cant (p < 0.001) dose dependent effect. The extract showed the highest activity with 61.52 and 51.39% suppression at 200 and 400 mg/kg, respectively. Moreover, at doses of 200 and 400 mg/kg, the extract signi cantly extended the survival days of treated groups compared to the negative controls, indicating that the extract has the capacity to lower the overall pathologic effect of the parasite in mice. However, there was no signi cant difference in percent change in weight before and after treatment among groups except with the positive control group. According to Deharo et al. [39], antimalarial activity of the 80% methanol extract of K. folosia can be regarded as good since it showed greater than 50% suppression at a dose of 200 mg/kg. Previous studies demonstrated that medicinal plants rich in anthraquinones such as aloes and senna possess notable in vivo antimalarial activity [40,41].

Antimalarial activity of the phenol fractions and their constituents
The two phenolic fractions of K. folosia were also found to have activity against P. berghei in mice (Fig. 2). Compared to their respective negative controls, both factions possessed signi cant suppressive activity at all dose levels tested. They showed the highest activity at 400 mg/kg with fraction 1 and fraction 2 causing 46.32% and 47.53% suppression, respectively. Both fractions prolonged the mean survival days of the treatment groups by 2 days relative to their negative controls although it was not statistically signi cant. No signi cant difference in percent change in weight was noted in the treatment groups when compared with the positive controls. Therefore, it can be deduced that the phenolic fractions of K. folosia are moderate in their in vivo antimalarial activity, congruent with earlier reports that extracts containing phenolic compounds and their glycosides have modest levels of antiplasmodial activity [42,43,44].
Among the isolated compounds, knipholone displayed the strongest antimalarial activity against P. berghei infected mice (Table 4). Although knipholone and dianellin showed signi cant suppression at all dose levels tested, the former displayed superior activity with percent suppression values of 51.5 and 61.5% at doses of 100 and 200 mg/kg, respectively. Moreover, it signi cantly prolonged the mean survival days of the treatment groups ( Table 4). The dose-response plot (Fig. 3)   Values are presented as mean ± SEM; n = 5; a = compared to vehicle3 (distilled water), b = compared to knipholone 25 mg, c = compared to knipholone 50 mg, d = compared to knipholone 100 mg, e = compared to knipholone 200 mg, f = compared to dianellin 25 mg, g = compared to dianellin 50 mg, h = compared to dianellin 100 mg, i = compared to chloroquine; * (p < 0.001); **(p < 0.01); ***(p < 0.05); numbers refer to doses in mg/kg/day.
Perusal of literature reveals that a number of promising anthraquinones and preanthraquinones leads such as visimione, ru gallol, uveoside, aloin and phenyl anthraquinones have been isolated and/or synthesized [15,45,46,47]. These compounds are considered as oxidants like artemisinins and 4aminoquinolines. More importantly, they are catalytic oxidants that enhance the production of reactive oxygen species (ROS) inside parasitized erythrocytes or increase these cells' susceptibility to oxygen radicals. The free oxygen radicals formed interact with heme or other biomolecular targets inhibiting its tetramerization to the insoluble hemozoin (malaria pigment) [48,49]. Knipholone, being an anthraquinone derivative, is anticipated to undergo one-electron oxidation and subsequently interact with heme (or other biomolecular targets) thereby inhibiting its tetramerization (or detoxi cation of heme). Similarly, because of the structural similarity of dianellin with phlorizin, a monoglucosidechalcone, its antimalarial mechanism of action could be due to inhibition of the solute transporter of the host cell membrane induced by the parasite invasion [50,51].

Molecular docking study
To get further insight on the mechanism of action of the isolated compounds and to study their binding interaction and identify hypothetical binding motifs, a docking study of knipholone, dianellin, HPMP and the standard antimalarial drugs chloroquine and artemisin were carried out on two crystal structures of enzymes. The two Plasmodium enzymes were plasmepsin II (PDB code 4cku) involved in haemoglobin metabolism by the parasite, and P. falciparum l-lactate dehydrogenase (pfLDH) (PDB code 1ldg) involved in glycolysis (or glucose metabolism of the parasite) [52,53,54]. There is a strong suggestion that haemoglobin digesting enzymes found in the food vacuole of the plasmodium and pfLDH are potential antimalarial chemotherapeutic targets for chloroquine and related aminoquionlones, anthraquinones and other oxidative phenolic compounds [55,56,57,58,59,60]. Besides, chloroquine has been found to bind to the cofactor (NADH) binding site of pfLDH acting as a competitive inhibitor [61].
The binding modes of P2FE-400, a designed inhibitor of plasmepsin II, knipholone, HPMP and chloroquine to plasmepsin II are shown in Figure 4. P2FE-400 showed the highest and strongest a nity for the aspartic protease, plasmepsin II, with the HYDE score of -38.3 kj/mol. The aspartic protease plasmepsin II has two aspartic acid residues Asp34 and Asp214 (the catalytic dyad) that serve as proton donors and acceptors, respectively, in the amide hydrolysis of peptide bonds in proteins. As shown in the current study and also described by Jaudzems et al. [62], P2FE-400 forms four hydrogen bonds with the catalytic dyad (Asp34 and Asp214), Val78 and Ser218 amino acid residues. Chloroquine and HPMP showed a comparable binding a nity with an estimated HYDE score of -19.7 and -19.2 kj/mol, respectively. The Cl substituent of chloroquine was found to be unsuitable for binding in the hydrophobic cavity of plasmepsin II. Chloroquine forms hydrogen bonds with Gly36 and Val78 amino acid residues. Similarly, HPMP forms a single hydrogen bond with Ser118. Its methoxyl group and adjacent carbonyl oxygen to the methoxyl group are not favored in the hydrophobic region of the binding pocket. Knipholone and dianellin showed weak binding interaction with HYDE score of -6 and -4.2 kj/mol, respectively. Nonetheless, knipholone forms two hydrogen bonds with one of the catalytic dyad (Asp214) and Val78 amino acid residues.
The binding modes of knipholone, HPMP and chloroquine to pfLDH binding site are shown in Figure 5.
Its carbonyl oxygen (at C-9) and hydroxyl group in ring A (at C-1) of the anthraquinone moiety, and the carbonyl oxygen (at C-3¢) of the phloroglucinol moiety together with the meta and para hydroxyl groups (at C-1¢ and C-4¢) are not favorable for binding. For HPMP, the methoxyl group and double bond in the long aliphatic chain are not suited for binding in the hydrophobic region. It also forms four hydrogen bonds with Ile54, Gly99, Phe100 and Asn140 amino acid residues with unique thermodynamically stable conformation. Interestingly, the two hydrogen bonds that HPMP forms with Gly99 and Asn149 are similar to two of the ve hydrogen bond interactions seen in docking of NADH cofactor. From the experimental data, there were seven hydrogen bonds in pfLDH-NADH complex, of which four are observed in this study [63]. Chloroquine on its part showed two hydrogen bonds with Asp53 and Gly99 amino acid residues. One of the N-ethyl groups of chloroquine is not needed in the hydrophilic binding sites. Moreover, the actual pfLDH-chloroquine complex also showed two hydrogen bonds with Glu122 and Gly99 [61]. In contrast, dianellin did not show binding interaction with pfLDH.

Conclusion
In conclusion, K. folosia possesses in vivo antimalarial effect against P. berghei in mice. This nding in conjunction with the safety pro le obtained from the acute oral toxicity results support the traditional claim of the plant for the treatment of malaria. The current molecular docking study also identi ed the binding motifs of the isolated compounds showing that knipholone and HPMP interact with important amino acid residues in the binding site of the target enzymes.

Declarations
Ethics approval and consent to participate All the animal study procedures followed were reviewed and approved by the Institutional Review Board of the SoP, College of Health Sciences, AAU. The mice were handled were in accordance with the Guide for the Care and Use of Laboratory Animals [20].

Consent for publication
Not applicable

Competing interests
The authors declare no con ict of interest Funding Figure 2 Antimalarial activity of the 80% extract and phenol fractions of K. folosiain mice infected with Plasmodium bergheiin 4 day suppression test.Values are presented as mean ± SEM; n =5.

Figure 3
Antimalarial activity of knipholone and dianellin in mice infected with Plasmodium berghei. The ED50 was estimated from a plot of log dose against parasitaemia (expressed as a percentage of the control). Values are presented as mean ± SEM; n =5. Chloroquine is shown in ball-stick model.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. SupplementarymaterialsMJ.docx