Protoberberine alkaloids berberine and palmatine employ Pf Pgh-1 transporter to enter the digestive vacuole and kill the malaria parasite by inhibiting heme biocrystallization

With Plasmodium falciparum devising means of evading antimalarial agents and in the absence of a vaccine that can be used across all age groups, malaria continues to ravage large parts of the tropical and sub-tropical regions of the world. Hence improved drugs with novel scaffolds are constantly needed to replenish the antimalarial chemotherapeutic arsenal. Medicinal plants have afforded some of the best antimalarials of which members of the alkaloid class are the ones that have made the most significant contributions. Protoberberine alkaloids are berberine-like alkaloids present in a number of medicinal plants traditionally used for the management of malaria. These alkaloids have been confirmed to have in vitro antiplasmodial activity. However, their development as antimalarials has been marred by their poor bioavailability. Among other features expected of drugs, understanding their mechanism of antiplasmodial action can spur research into designing of orally active and pharmacokinetically robust analogues. To investigate the anti-plasmodial activity of protoberberine alkaloids, Berberine and Palmatine were chosen for the present study. The mechanistic study of both alkaloids including intra-parasite localization was achieved by taking advantage of their inherent fluorescence property. Herein, we show that the protoberberine alkaloids Berberine and Palmatine, interact with vacuolar P-glycoprotein homologue-1 (Pf Pgh-1) transporter, enter the digestive vacuole and inhibit heme to hemozoin detoxification leading to death of the parasite.

Malaria, a hemoprotozoan parasite infection claimed over 610,000 lives with a recorded morbidity of 282 million in 2024 suggesting an increase in both mortality and morbidity than the 597, 000 deaths and 263 million cases recorded in 2023 (World Malaria Report, 2024 and World Malaria Report, 2025). In the 1950s–60s, the worldwide launching of the Global Malaria Eradication Program by the WHO saw the record of massive success in controlling malaria leading to reduction in the global malaria incidence map. These gains have been reversed by the evolution of Plasmodium falciparum parasites that are resistant to all categories of drugs used to battle the disease. These include aminoquinolines (Holmgren et al., 2006; White et al., 2004; Plowe, 2003; Sidhu et al., 2005), aryl amino alcohols (Cowman et al., 1994; Duraisingh and Cowman, 2005; Price et al., 2004; Reed et al., 2000), anti folates (Plowe, 2003; Wang et al., 1997), the naphthoquinone atovaquone (Korsinczky et al., 2000), and, most recently, members of the endoperoxide class (artemisinin and its derivatives), which form the backbone of the current frontline antimalarials viz. The Artemisinin-based combination therapies (ACTs) (Dondorp et al., 2009; Noedl et al., 2008; White, 2015). Indeed, the resistance to ACTs which was restricted to South Asia has now emerged in Africa as well (Balikagala et al., 2021; Uwimana et al., 2020). As such, continuous efforts are required for the discovery and development of novel antimalarial drugs.

Medicinal herbs have provided the most effective antimalarial medicines over time. Thus, quinine from Cinchona spp. provided the quinoline backbone which became the inspiration for the design of chloroquine. Likewise artemisinin from Artemisia annua, and atovaquone, a synthetic naphthoquinone based on lapachol, a prenylnaphtoquinone isolated from Tabebuia impetiginosa (Achan et al., 2011; De Paula et al., 2014) illustrate the great potential of medicinal herbs as providers of new antimalarial drugs.

Protoberberine alkaloids are alkaloids related to berberine. The tetracyclic ring structure (7,8,13,13a-tetradehydro-9,10-dimethoxy-2,3, methylene dioxy berberinium) in members of this sub-group of alkaloids is based on the 5,6-dihydrodibenzo [a, g] quinolizilium skeleton (Leitao da-Cunha et al., 2005). Berberine (BBR) and palmatine (PAL) are the most important members of the protoberberine alkaloids (Figure 1), which are also the most abundant subgroup of the isoquinoline class of alkaloids. BBR is found in many plant families and genera, including Annonaceae (Annickia, Coelocline, Rollinia, and Xylopia), Berberidaceae (Berberis, Caulophyllum, Jeffersonia, Mahonia, Nandina, and Sinopodophyllum), Menispermaceae (Tinospora), Papaveraceae (Argemone, Bocconia, Chelidonium, Corydalis, Eschscholzia, Glaucium, Hunnemannia, Macleaya, Papaver, and Sanguinaria), and Ranunculaceae (Coptis, Hydrastis, and Xanthorhiza), and Rutaceae (Evodia, Phellodendron, and Zanthoxyllum) (Neag et al., 2018). BBR is known to show several pharmacological traits including antidiabetic (Hardie, 2011), antimicrobial (Tillhon et al., 2012), anti-inflammatory (Tillhon et al., 2012; Zou et al., 2017), antiplasmodial (Wright et al., 2000) as well as anticancer (Maiti and Kumar, 2010).

Both BBR and PAL have been in traditional therapeutic use for millennia. For instance, PAL found in the roots, rhizomes, and stem barks of Berberidaceae, Papaveraceae, Fumaraceae, Menispermaceae, Rutaceae and Ranunculaceae (Maiti and Kumar, 2010) has been found useful in the treatment of jaundice, diarrhoea, hypertension, liver disorders, and inflammation, to name a few. BBR’s specific cellular localization pattern and cell cycle effects in mammalian cells are well documented (Pinto-Garcia et al., 2010; Serafim et al., 2008). Easy to see via alkaloidal auto fluorescence, BBR was found to concentrate in the mitochondria of mouse melanoma cells at low concentrations (20–50 µM), but at high concentrations (100 µM), the entire cytoplasm, including the nucleus, became fluorescent green (Serafim et al., 2008). Using human pancreatic cancer cells, it was found that this specific cellular localization pattern is not just concentration but also time dependent (Pinto-Garcia et al., 2010). The fundamental mechanism that allows BBR to move from one organelle to another is unknown. However, Serafim et al. (2008) suggested that at high concentrations and upon saturation of mitochondrial BBR binding sites, it accumulates in the cytoplasm from where it migrates to the nucleus. More recently, Li et al. (2019), reported that the mitochondrial membrane potential (MMP) is responsible for BBR’s mitochondrial localization and that the alkaloid first localizes in the mitochondria. They then went on to suggest that BBR functions as a respiratory inhibitor at high doses (Serafim et al., 2008), jeopardizing the maintenance of MMP’s energy-dependent functions. Inability to maintain MMP causes BBR to be re-distributed from mitochondria to the cytoplasm and thence to the nucleus (Li et al., 2019) where it is known to bind DNA via base pair intercalations and electrostatic interactions (Tian et al., 2008; Kundu et al., 2016). Although BBR’s antiplasmodial effect has been recognized for decades (Aldous et al., 1998; Vennerstrom and Klayman, 1988), the specific mechanism of action against the malaria parasite has remained unknown to date. Some mechanistic studies on the antiplasmodial activity of BBR have implicated its ability to bind DNA in addition to a telomerase inhibiting action found by Sriwilaijareon et al. (2002). However, while BBR’s antiplasmodial IC 50 was 2.4 µM, its telomerase inhibitory action was observed at 75 μM, with total inhibition only at 370 µM (Sriwilaijareon et al., 2002). This shows that BBR primarily acts against Plasmodium in a way that isn't related to telomerase inhibition. Understanding the mechanism of antiplasmodial action of candidate drug-like molecules not only aids in the development of better analogues (Burgess et al., 2010), but also aids in the selection of appropriate partners for combination therapy during clinical development (Fidock et al., 2004; Erhunse and Sahal, 2020), the elucidation of resistance mechanisms, and may reveal novel druggable targets in the parasite (Baragaña et al., 2015; Kato et al., 2016; Ng et al., 2016).

To explore the antiplasmodial activity of protoberberine alkaloids therefore, two members; BBR and PAL were chosen for the present study. The mechanistic study of both alkaloids including intra-parasite localization was achieved by taking advantage of their inherent fluorescence property.

Chemicals

Berberine chloride (Cat# B3251), Palmatine chloride hydrate (Cat# 361615), Verapamil hydrochloride (Cat# V4629), Tariquidar (Cat# SML1790), Hemin (Cat# 51280), human Haemoglobin (Cat# H7379), and trans -Epoxysuccinyl-L-leucylamido (4-guanidino) butane (E64) (Cat# E3132) were obtained from Sigma Aldrich, India. MitoTracker TM Red CMXRos (Cat# M7152), Hoechst 33342 (Cat# 62249) and JC-1 dye (Cat# T3168) were obtained from ThermoFisher scientific. All other chemicals used were of analytical grade.

Parasite strains

The chloroquine sensitive (Pf 3D7), chloroquine resistant (Pf INDO) and the multi-drug resistant (Pf Dd2) strains were obtained from Malaria Research and Reference Reagent Resource (MR4) Centre.

Parasite growth inhibition study

Parasites were cultivated in vitro as described by Trager and Jensen (Trager and Jensen, 1976) except that Albumax was used in place of human serum. Cultures were maintained in fresh O + human erythrocytes at 4% hematocrit in complete medium (RPMI 1640 with 0.2% sodium bicarbonate, 0.5% Albumax II, 45 mg/L hypoxanthine, and 50 mg/L gentamicin) at 37 °C under reduced O 2 (gas mixture 5% O 2, 5% CO 2, and 90% N 2). Stock solutions of chloroquine (CQ) were prepared in Milli-Q water, while test compounds were dissolved in DMSO. All stocks were then diluted with complete culture medium to achieve the required concentrations ensuring that in all cases the final concentration of DMSO in the parasite culture was the non-toxic 0.4% (v/v). CQ and test compounds were then placed in triplicate wells of 96-well flat-bottom tissue culture plates with drug concentrations ranging from 0 to 17 μM in a final volume of 100 μL. Chloroquine (1.6 µM) was used as a zero-growth positive control in all experiments. Parasite culture was synchronized at ring stage with 5% sorbitol (Lambros and Vanderberg, 1979). Synchronized culture was aliquoted to drug containing 96-well plate at 2% hematocrit (H) and 1% parasitemia (P). After 48 h of incubation under standard culture conditions, plates were treated with SYBR Green I lysis-causing buffer and read (0.1 s, bottom reading) for fluorescence (Smilkstein et al., 2004) using a 96-well fluorescence plate reader (Victor, Perkin-Elmer), with excitation and emission wavelengths of 485 and 535 nm respectively. The fluorescence readings were plotted against drug concentration, and the 50% inhibitory concentration (IC 50) was determined by analysis of dose–response curves via non-linear regression using the IC Estimator-version 1.2 software (http://www.antimalarial-icestimator.net/MethodIntro.htm), Free Software Foundation, Boston, MA, United States. This was thereafter validated by microscopic evaluation of at least 3000 cells for all concentrations tested. Parasitized and non-parasitized cells were counted by using Plasmodium autocount 0.1 software developed by Ma et al. (2010).

Intra-parasite alkaloid Localization study

To determine the intracellular site of accumulation of these alkaloids in the malaria parasite, we chose to investigate co-localization of each of the two alkaloids with mitochondrial and genomic DNA markers. For the co-localization investigations, we employed Hoechst 33342, which emits blue fluorescence (Enomoto et al., 2012) upon binding to DNA, as well as mitotracker red, which specifically stains the mitochondrion red (Purfield et al., 2009). After treating parasite’s different cell cycle stages (ring, trophozoite and schizont) with 25 µM of either BBR or PAL for 9 h at 37 °C, cells were washed twice with 1X PBS, incubated with a mixture of Hoechst 33342 + mitotracker for 30 min at 37 °C, rinsed thrice, and confocal microscopic analysis was performed. Digestive vacuolar localization was studied by using Tariquidar which is an inhibitor of the vacuolar transport protein, Pgp-1.

Time points fluorescence study to investigate time dependent loss of alkaloid fluorescence upon incubation with Plasmodium falciparum

To evaluate intracellular fluorescence loss of BBR and PAL, trophozoite-stage parasites (4% parasitemia, 2% hematocrit) were treated with IC 90 and 25 μM of BBR and PAL for 3 h. Post-treatment, cultures were washed thoroughly with complete RPMI media to remove excess drugs and incubated further under standard conditions. Wet mounts were prepared at 3, 6, 9, and 21 h, and fluorescence was tracked via microscopy using appropriate filter settings. Images were captured under identical exposure conditions and analysed using ImageJ software.

Determination of Plasmodium falciparum mitochondrial membrane potential (ΔΨm)

Plasmodium falciparum mitochondrial membrane potential (ΔΨm) was determined using JC-1 dye. Cell-permeable cationic carbocyanine dye JC-1 (Invitrogen), emits green fluorescence (525 nm) in its monomeric form. However, upon transfer to the membrane environment of a functionally active mitochondrion, it exhibits a membrane potential induced aggregation dependent orange-red fluorescence (emission at 590 nm). IC 10, IC 50, and IC 90 of the test alkaloids (0.23, 1.24, and 2.77 µM of BBR) and (0.21, 1.24, and 2.73 µM of PAL) were added to P. falciparum cultures for 24, 48, and 72 h (ring stage culture), 18 and 24 h (trophozoite stage culture), and 8 h (schizont stage culture). The alkaloid treated parasites were incubated (20 min, 37 °C) with JC-1 dye and the wet mounts of stained cultures were observed using a fluorescence microscope (Nikon 50i).

Determination of Plasmodium falciparum DNA fragmentation by hoechst 33342 staining

DNA fragmentation and condensation were detected by Hoechst 33342 (20-[4-ethoxyphenyl]-5-[4-methyl-1-piperazinyl]-2,50-bisbenzimidazole trihydrochloride trihydrate, Molecular Probes, USA). Hoechst 33342 gives a much brighter fluorescence with the condensed chromatin of apoptotic cells than it does with the native chromatin of healthy cells. Hoechst 33342 was used as directed by the manufacturer. In a 96 well microtiter plate, synchronized ring, trophozoite, and schizont stage cultures (1% P, 2% H) were incubated with IC 90 of BBR (2.77 µM)/PAL (2.73 µM) for 12, 24, 48 h (rings), 6, 12, 24 h (trophozoites), and 8 h (schizonts). The cultures were then stained with 0.2 µL Hoechst 33342 dye (max emission 460 nm) for 20 min. Cells were transferred to microfuge tubes and centrifugally washed with 1X PBS (200 μL, twice). Thereafter, wet mount slides were made and a fluorescence microscope (Nikon 50i) was used to examine the slides.

Effect of inhibiting Plasmodium falciparum P-glycoprotein homologue-1 (Pgh-1) on intracellular localization of protoberberine alkaloids

Tariquidar and Verapamil were used to inhibit Pgh-1; the P. falciparum multi-drug resistance gene-1 (pf mdr1) gene product (Rohrbach et al., 2006; Reiling, and Rohrbach, 2015). Trophozoite stage parasites (4% P, 2% H) were incubated with a either tariquidar (3.5 μM) or verapamil (30 μM) and BBR or PAL (25 µM) for 9 and 24 h respectively. The concentration and time durations chosen were based on the parameters when these alkaloids lost their fluorescence following incubation with the parasite. Control cultures were treated with BBR/PAL as above and grown without inhibitors to compare them with inhibitor-treated cultures. Cultures were washed twice with 1X PBS and re-suspended in 500 µL of PBS. Wet mounts were made and images acquired using a high-resolution fluorescence microscope (Nikon Eclipse, Japan).

Spectrofluorimetric study of interaction of berberine and palmatine with heme and hemoglobin

Hemin or hemoglobin (0, 6.25, 12.5 and 25 µM) (50 µL) were dissolved in a 0.1 M HEPES buffer (pH 7.2) and mixed with 50 µL of BBR or PAL (25 µM in methanol). This was followed by incubation (2 h, 37° C) of the parasite + alkaloid mixtures (now in 50% methanol). Thereafter, excitation was done at 465 nm and emission was recorded over 400–600 nm using a spectrofluorometer (Spectra Max M3, USA).

Heme-mediated quenching Assay within malaria parasites

To validate whether heme released from hemoglobin quenched alkaloid fluorescence, trophozoite-stage parasites (4% parasitemia, 2% hematocrit) were treated with 25 μM BBR and PAL for 9 h in the presence or absence of E64 at varying concentrations (0, 1, 2, 10, 20, and 40 μM). After 9 h, the cultures were washed with complete RPMI, centrifuged (1800 rpm, 5 min), and re-suspended in 500 μL 1X PBS. Wet mounts were prepared, and images captured using a fluorescence microscope. Fluorescence intensity was quantified from the microscopy images using ImageJ software.

Homology modelling, validation and docking of berberine and palmatine to Pgh-1

To get an insight into the interaction of protoberberine alkaloids with the P. falciparum P-glycoprotein homologue-1 (Pgh-1) transporter, the structural model of Pgh-1was made using the tools available in the Schrödinger suite. This homology model was based on mouse Pgp (PDB: 3G5U) as previously described by Patel et al. (2013). The clinical candidate ACT-451840 and the lead molecule ACT-213615; which have been reported to interact with Pgh-1 (Brunner et al., 2013; Brunner et al., 2012; Ng et al., 2016) were docked to the transporter to validate the model. Thereafter, dockings of BBR, PAL and coralyne {a synthetic protoberberine alkaloid with antiplasmodial activity (Erhunse and Sahal, 2024)} were studied for in silico binding to the modelled Pgh-1.

Bloated food vacuole phenotype assay for evaluation of hemoglobin uptake

The bloated food vacuole phenotype assay as described by Jonscher et al. (2019) was adopted with minor modifications. In this assay, parasites at 24 h post invasion (p.i.) (an age corresponding to voracious consumption of host hemoglobin) are treated with E64 to inhibit the proteolytic degradation of hemoglobin. This results in the increased concentration of hemoglobin in the parasite’s DV giving it a bloated or swollen appearance. In contrast, in the presence of inhibitors of hemoglobin uptake the largely “empty” parasite DV appears shrunken. Briefly, early ring stage parasites (0–3 h p.i.) were sorbitol synchronized. At 24 h p.i., the synchronized Trophozoite culture (4% P, 2% H) was treated with E64 (33 μM) to inhibit hemoglobin degradation. Cultures were thereafter either untreated (control) or exposed to 1, 2.5 & 5 x IC 90 of test alkaloids for 15 h at the end of which the parasites reached an age of 39–42 h p.i. At the end of the incubation period, DIC and Giemsa-stained images were captured to score % of cells with bloated and non-bloated DVs.

Study of alkaloid mediated inhibition of Hemozoin formation

Synchronized ring-stage cultures of P. falciparum 3D7 (6% parasitaemia) were treated for 30 h with BBR and PAL at concentrations in the range IC 50 to 20×IC 90. Three controls were used: untreated infected cultures (Control 1), cultures treated with chloroquine at varying concentrations (Control 2) and uninfected, untreated red blood cells (RBCs) (Control 3). Following treatment, cultures were lysed and centrifugally washed twice with 2.5% sodium dodecyl sulfate (SDS) in 0.1 M sodium bicarbonate (pH 8.8) and incubated at room temperature for 30 min. The lysates were centrifuged (1300 rpm, 10 min), and the supernatant carefully discarded. The resulting pellet was re-suspended in 0.2 mL of 50 mM NaOH containing 5% SDS and incubated (30 min, room temperature) to solubilize hemozoin into free heme. After incubation, samples were transferred to a 96-well microplate, and the amount of monomeric heme quantitated at 405/750 nm (i.e. A 405 nm minus A 750 nm) using a microplate reader (Men et al., 2012).

Effect of E64 on IC 50 of alkaloids

To validate inhibition of hemozoin formation as cause of parasite killing by BBR and PAL, we probed the potentially antagonistic relationship between protease inhibitors (E64, to block Hb catabolism) and the antiplasmodial action of protoberberine alkaloids (BBR and PAL). Thus, IC 50 of the alkaloids were assessed in untreated control parasite vs. in E64 treated parasite culture. Briefly, parasite culture was synchronized at ring stage with 5% sorbitol. The synchronized culture was aliquoted as test samples in 96-well plate at 2% Hematocrit (H) and 1% Parasitemia (P). The alkaloids (0–65 µM) were tested in triplicates and their IC 50 values determined in the presence versus absence of E64. The inhibition percentages were determined using the SYBR Green lysis method (Smilkstein et al., 2004). The lysis buffer [ Tris (20 mM; pH 7.5), EDTA (5 mM), saponin (0.008%; w/v), and Triton X-100 (0.08%; v/v)], was freshly prepared by dissolving 121 mg of Tris, 73 mg of EDTA disodium salt, and 4 mg of saponin in 400 μL of Triton 10X solution, which was itself prepared by diluting a Triton 100 X stock. The final volume was adjusted to 50 mL with autoclaved Milli-Q water, and the solution was mixed thoroughly. Just before use, 3 μL of SYBR-Green I dye (light sensitive) was added to 15 mL of this buffer. After incubation, 100 μL of the dye-containing buffer was added to each well, and the plates were incubated (in dark, room temperature, 1.5 h). Fluorescence was then measured using a microplate reader at an excitation wavelength of 485 nm and an emission wavelength of 520 nm. Chloroquine (CQ) (40 μM) was used as zero growth positive control whereas untreated cells served as 100% growth control.

Statistical analysis

Assays were performed in triplicates. Concentrations producing 50% inhibition (IC 50) were obtained by non-linear regression using the IC Estimator-version 1.2 software (http://www.antimalarial-icestimator.net/MethodIntro.htm), free software foundation, Boston, MA, United States. Data has been expressed as mean ± standard error of mean (SEM).

In spite of structural similarity, BBR and PAL exhibit strikingly different resistance indices against the parasite

Using the SYBR Green I-based fluorescence assay, the in vitro antiplasmodial activity of the alkaloids was assessed against drug sensitive (Pf 3D7) and drug resistant (Pf INDO and Pf Dd2) parasite strains in order to assess their activity against different parasite strains. Microscopy of Giemsa-stained smears was used to confirm the SYBR green results (>3000 cells were evaluated for each alkaloid concentration tested). Although both alkaloids demonstrated similar in vitro antiplasmodial activities against Pf 3D7 {IC 50 (µM) BBR 1.24; PAL 1.02}, their IC 50 values against the CQ resistant INDO and the multi-drug resistance Dd2 strains were markedly different indicative of the peculiar structure function relations of these two alkaloids. Thus, while BBR {IC 50 (µM) Pf INDO 0.16; Pf Dd2 0.54} was found to be more potent against the resistant strains than the sensitive strain, PAL {IC 50 (µM) Pf INDO 3.09; Pf Dd2 2.58} turned out to be more potent against the sensitive strain (Figures 2A,B). As a result, while BBR’s resistance indices for Pf INDO and Pf Dd2 are <1, PAL’s resistance indices are >2 (Figures 2C,D). The conflicting microscopy vs. SYBR green data for both alkaloids (Figures 2C,D) may be appreciated by our observation of some extruded parasites as seen under the microscope. Such extracellular parasites will certainly be picked up by the SYBR green stain resulting in higher % parasitemia by SYBR green lysis assay as against Giemsa based microscopy where % Parasitemia is estimated based on intracellular parasite alone.

BBR and PAL lose their fluorescence upon incubation with Plasmodium falciparum

To assess the parasite compartment which the alkaloids localize to, P. falciparum 3D7 (2% H, 4% P) was treated with different concentrations (IC 90 and 25 µM) of either BBR or PAL for 1, 3, 6, 9, 12, 24, 48 and 72 h. Using the low concentration of 25 µM of each alkaloids, at the earliest time point of 1 h, while staining due to PAL was quite weak, BBR stained the parasite more strongly (Supplementary Figure S1). However, at the subsequent time point of 3 h, while the already weak BBR staining faded, staining due to PAL became distinctly stronger (Figure 3A and Supplementary Figure S1). In the time window of 3–12 h, while PAL staining remained strong, the staining due to BBR faded. Interestingly, at 24–72 h time points (Supplementary Figure S1), staining due to both the alkaloids was found to vanish completely. As compared to the study with IC 90 values of the alkaloids, the loss of fluorescence took longer for the two alkaloids taken at 25 µM (Figure 3B). This loss of fluorescence with time upon incubation with parasite suggested that the inherently fluorescent alkaloids were localizing in a compartment within the parasite and possibly interacting with an entity that was quenching their fluorescence.

Confocal microscopic co-localization of PAL with mitotracker

The rather low quantum yields of these alkaloids necessitated very high concentrations (25 µM for PAL and 135 µM for BBR) for the subcellular co-localization analysis. Due to the significant fluorescence observed for both alkaloids when treated with these concentrations at 3–9 h time points (Figure 3A; Supplementary Figure S1), mixed stage Pf 3D7 parasites (4% P, 2% H) were treated for 9 h with either PAL (25 µM) or BBR (135 µM). It is worth recalling that BBR has been shown to localize into the mitochondria and nucleus of mammalian cells (Li et al., 2019; Pinto-Garcia et al., 2010; Serafim et al., 2008), so the nucleus and mitochondria of malaria parasites in alkaloid-treated red blood cells were stained for 30 min with a mixture of nucleus staining Hoechst 33342 (1 μg/mL) and mitochondrion staining mitotracker (25 nM). Samples were centrifugally washed thrice with 1XPBS in microfuge tubes and cells re-suspended in 300 µL of 1XPBS. Parasite wet mounts were made, and images acquired using a confocal microscope (Nikon confocal A1r). A tricolour experiment was designed where DNA was dyed blue with Hoechst 33342, mitochondria stained red using mitotracker, and alkaloids were expected to generate green fluorescence. The degree of dye co-localization was determined using Pearson’s correlation coefficient (which ranges from −1 to 1). In summary, (+1) indicates perfect correlation while (−1) denotes perfect exclusion. Despite the use of an antifade, we could not collect BBR’s fluorescence owing to photo-bleaching induced by the microscope’s high laser intensity. With the exception of the ring stage parasites which did not pick up PAL’s stain, PAL (25 µM) was found to strongly co-localize with mitochondria of schizont (Pearson’s correlation for PAL and mitotracker: 0.502, vs. that of PAL and Hoechst: 0.273) and trophozoite (Pearson’s correlation for PAL and mitotracker: 0.917, vs. that of PAL and Hoechst: 0) (Figure 4). This staining of various organelles could possibly be a result of the very high concentration of the alkaloid used (Pinto-Garcia et al., 2010; Serafim et al., 2008).

BBR and PAL localize to Plasmodium falciparum’s digestive vacuole (DV)

In light of previous studies in human cells finding stably enhanced fluorescence for both alkaloids when bound to mitochondria, nucleosomes, and cytoplasm (Li et al., 2009; Li and Lu, 1998), it was unclear why BBR and PAL showed a time dependent loss of fluorescence (Figure 3; Supplementary Figure S1) after incubation with the malaria parasite. It occurred to us that the malaria parasite is special in that it has a DV in which hemoglobin is digested and the redox active toxic heme released is detoxified by conversion of heme to hemozoin (malaria pigment). Given that amphipathic cations such as BBR and PAL are substrates of human Permeability glycoprotein (Pgp-1) (Lewis, 2001; Tarabasz and Kukula-Koch, 2020), we wondered if the P. falciparum Multi Drug Resistance (MDR) transporter P-gp homologue 1 (Pgh-1) (Lewis, 2001; Shitan et al., 2007; Stermitz et al., 2000), which is primarily expressed on the membrane of the DV and sequesters drugs inside the DV (Rohrbach et al., 2006) might also bind to these alkaloids. Notwithstanding the findings of Li et al. (2019) that BBR did not localize to Lysosomes (an acidic compartment like the parasite’s DV) in HepG2 cells, the possibility of these alkaloids localizing to the malaria parasite’s DV was investigated. The drug binding pocket in the transporter Pgh-1 is formed by 12 transmembrane (TM) helices (Ferreira et al., 2011). Polymorphisms in the human Pgp-1 transmembrane Helix 11 (TM11) have been shown to inhibit drug transport (Kajiji et al., 1993). In fact, mutations of residues 1034 and 1042 on TM11 of Pf Pgh-1 have been found to cause reduced binding for the aminoquinoline antimalarials chloroquine, quinine, and mefloquine (Patel et al., 2013), leading to the hypothesis that these two residues are part of the Pgh-1 drug binding pocket (Ferreira et al., 2011; Lekostaj et al., 2008; Patel et al., 2013; Sidhu et al., 2005).

To see if the transport of these alkaloids to the DV is facilitated by Pgh-1, we looked at the role of Pgh-1 in the antiplasmodial activity of these compounds. Aside from Pf Pgh-1, the parasite has another transporter on the surface of the DV, the P. falciparum chloroquine resistance transporter (Pf CRT). However, in contrast to Pgh-1which facilitates entry of diverse substrates into the DV, Pf CRT carries substrates out of the DV. In order to have unambiguous answers, we decided to use Verapamil (1 µM) which has been shown to block Pf CRT and reverse CQ resistance (Hayward et al., 2005; Martin et al., 1987; Martiney et al., 1995). Further to find the reason for differing potency of the two protoberberine alkaloids under study against chloroquine resistant strains than against a chloroquine sensitive strain (Figure 2) and to determine the role of Pf MDR1 in the antiplasmodial activity of these alkaloids, the resistance indices of the alkaloids were determined in the presence vs. absence of verapamil (1 µM) (Figure 5) (Hayward et al., 2005; Martin et al., 1987; Martiney et al., 1995). Indeed verapamil (1 µM) was found to reverse chloroquine resistance as evidenced by lower resistance indices for verapamil treated parasites compared to untreated parasites (Figure 5). It is worth noting that verapamil’s intrinsic anti-Plasmodial action, with IC 50 values ranging from 6.5 to 28.5 µM against several parasite strains (Wirjanata et al., 2017), is observed at concentrations several times higher than the 1 µM used here to inhibit Pf CRT.

With resistance indices <<1 (Figure 2), BBR is more effective against chloroquine resistant parasites than against the chloroquine sensitive strains used in this investigation. As shown in Figure 2, the already low resistance indices of BBR, were not significantly influenced by Verapamil (Figure 5) indicating that as a favourable substrate of Pf Pgh-1, the parasite’s increased copy number of Pgh-1 (4 in Pf Dd2 versus 1 in Pf 3D7 (Lim et al., 2009) may likely result in more of BBR localizing to the digestive vacuole (DV) of the resistant strains. The possibility of BBR’s localization to the DV in parasite strains together with greater potency against the resistant strains (Figure 2) motivated us to find intravacuolar targets for these alkaloids. Furthermore, because mutations in Pf CRT did not result in BBR becoming less potent against the CQ resistant Pf INDO (Figures 2, 5), in all probability BBR has no interactions with CRT.

In contrast to BBR (RI < 0.5), PAL (RI > 2) was found to be less effective against the chloroquine resistant strains than against the drug sensitive strain. Further, PAL’s RI against Pf INDO decreased much more (from 1.99 to 1.14) after treatment with verapamil than was the case with Pf Dd2 (from 2 to 1.87) (Figure 5). The partial reversal of resistance by PAL-treated Pf INDO parasites with verapamil suggests that unlike BBR, there may be a possibility of PAL’s exit through the CRT. Further confirmation of role of Pf Pgh-1 mediated transport of these alkaloids into the parasite’s DV by us was obtained by blocking Pf Pgh-1. This was accomplished by treating trophozoite stage parasites (Pf 3D7 and Pf Dd2 at 2% H and 4% P) with BBR/PAL (25 µM) for 9 h or for 24 h. The times chosen correspond to the times when BBR and PAL show loss of their fluorescence respectively upon incubation with the parasite (Figure 3). Tariquidar (3.5 µM) (which inhibits the ATPase activity of Pgh-1) and verapamil (30 µM), a competitive inhibitor of Pgh-1, were employed to block this transporter in the study. Wet mounts were made and images captured using a high-resolution fluorescence microscope (Nikon Eclipse, Japan); (Figures 6, 7).

Verapamil (30 µM) on incubation with the parasite was unable to prevent the loss of BBR fluorescence (Figure 8). Rohrbach and co-workers (Rohrbach et al., 2006) made a similar observation while exploring the orientation of Pgh-1 on the parasite’s DV using the Fluo-4 dye, which is a known substrate of a Pgp-1. Verapamil was unable to prevent Fluo-4 from entering the parasite’s DV in their testing, implying that it is either unable to inhibit Pgh-1 or that it is a weaker substrate of the transporter than Fluo-4. Although the primary target of Verapamil is CRT, at the high concentration of 30 µM it has been found to inhibit Pgp-1 (Martiney et al., 1995). This has led to Verapamil’s designation as a first-generation inhibitor of Pgp-1. Incubation of parasite with Tariquidar (3.5 µM), on the other hand, prevented BBR from losing its fluorescence. In the absence of Tariquidar, BBR (25 µM) totally loses its fluorescence after 9 h of incubation (Figure 6). However, in presence of Tariquidar, we observed (Figure 6) that BBR stained the Pf Dd2 cytosol much more strongly than was the case with Pf 3D7. This indicates that in the absence of Tariquidar, BBR localizes into the parasite’s DV where it loses its fluorescence. On the contrary, in presence of Tariquidar which inhibits the DV-resident Pgh-1 transporter, the alkaloid remains restricted to the cytosol retaining its full fluorescence. However, it remained unclear how the alkaloid’s intrinsic fluorescence was quenched in the DV.

Similar to BBR, only Tariquidar (3.5 µM) but not verapamil (30 µM) prevented PAL’s fluorescence loss in Pf 3D7 parasites (Figure 9). However, in Pf Dd2 parasites treated with verapamil, a significant cytosolic fluorescence was observed. The reason for this is not quite clear to us. However, a recent study (Si et al., 2023) reports that when Asparagine 86 on the transporter is replaced by Tyrosine as seen in Pf Dd2, the ATPase activity is lowered (Si et al., 2023). Tariquidar acts by inhibiting ATPase activity of the transporter. Hence, our observation of stronger fluorescence in Pf Dd2 may be due to Tariquidar inhibiting the ATPase activity of the transporter which already has a lowered ATPase activity in Pf Dd2 as compared to Pf 3D7.

Parasites (2% H and 4% P) were incubated with biologically relevant IC 90 of BBR (2.77 µM)/IC 90 of PAL (2.73 µM) for 3 h with or without Pgp-1 inhibitor Tariquidar (3.5 µM) to examine if these alkaloids failed to localize to the DV selectively in the presence of inhibitor. Following that, wet mounts were made and images taken using a high-resolution fluorescence microscope (Nikon Eclipse, Japan). Although in absence of Tariquidar both alkaloids showed no fluorescence after 3 h of incubation with Pf 3D7, addition of Tariquidar caused a distinct fluorescence from the parasite’s cytoplasm (Figure 8). This demonstrates that both alkaloids localize to the parasite’s DV which alone has the quencher. BBR-treated Pf Dd2 were similarly fluorescence positive only in presence of Tariquidar (Figure 8). Interestingly as previously observed with high concentrations of the alkaloids, unlike BBR, while PAL treated Pf 3D7 was fully bereft of fluorescence, the PAL treated Dd2 did not completely lose its fluorescence after 3 h of incubation.

Loss of mitochondrial membrane potential

With the time and dose-dependent localization pattern of BBR in mammalian cells (Li et al., 2019; Pinto-Garcia et al., 2010; Serafim et al., 2008) in mind, we looked into the possible loss of parasite MMP caused by BBR or PAL. The cationic JC-1 dye that stains metabolically active mitochondria and produces a red colour, was used to accomplish this. The red colour changes to green when MMP is lost, resulting in the release of JC-1 monomers into the cytosol leading to a drop in the red-to-green fluorescence ratio. Our results (Figure 9; Supplementary Figure S2, S3) indicated that IC 10, IC 50 and IC 90 of BBR (corresponding to 0.23, 1.24, and 2.77 µM of BBR) did not result in loss of MMP at any of the cell cycle stages or time durations. On the other hand, only schizont stage parasites treated with IC 50 (1.02 µM), and IC 90 (2.73 µM) of PAL for 8 h showed loss of MMP (Figure 10). Although we did not observe any MMP loss, it is possible that other mitochondrial pathways may be targeted by the alkaloids.

Fluorescence quenching due to complexation with heme

Spectroscopic study

Both BBR and PAL have been reported to be weakly fluorescent (Maiti and Kumar, 2010), with BBR’s fluorescence increasing dramatically (by 8-fold) when bound to DNA (Li and Lu, 1998) and upon binding to the hydrophobic regions of the polarized mitochondrial membranes (Mikes and Dadak, 1983). However, when the parasite was incubated with the alkaloids in this investigation, at all concentrations tested, there was a time-dependent complete loss of fluorescence (Figure 3). In malaria parasite, heme can be found in three different forms (hemoglobin, hemin and hemozoin) in the DV of P. falciparum. The possibility of losing fluorescence due to contact with one or more of the parasite’s heme forms was examined because both alkaloids were found to lose their fluorescence upon entering the parasite’s DV. The likelihood of interaction with hemoglobin or hemin was examined by incubating the alkaloids (25 µM) at 37 °C for 2 h with increasing quantities of hemoglobin (0–25 µM) or hemin (0–25 µM). The fluorescence spectra of the alkaloids were obtained at the end of the incubation period. The samples were excited at 465 nm and emission recorded at 400–600 nm. The fluorescence intensity of both alkaloids increased when they interacted with hemoglobin but decreased upon interaction with hemin (Figure 11).

Heme mediated fluorescence quenching within malaria parasites in culture

Within cultured parasites, the concentration of free heme was reduced by using E64 that was used to inhibit hemoglobin degradation. This was done in the presence of the alkaloids following which alkaloidal fluorescence was acquired. Treatment with increasing doses of E64 (0–10 µM) led to a dose-dependent increase in fluorescence intensity of both alkaloids (Figure 12). However, at higher doses (>10 µM), fluorescence intensity of both alkaloids decreased. It is not clear why this was the case as while other protease inhibitors at high concentrations resulted in the parasite becoming pyknotic, use of E64 doses as high as 40 µM did not result in crisis forms of the parasite (Prasad et al., 2013). However, given that E64’s IC 50 for parasite killing is <1 µM (Wijayanti et al., 2010), it is possible that at high concentrations, E64 can quench the fluorescence of the alkaloids.

Homology modelling assessment and docking

The 1419 amino acid residues long sequence of Pf Pgh-1 (accession number XP_001351787) was retrieved from NCBI protein database. A blastp homology search over PDB database identified several structural templates of which Mus musculus Pgp-1 (PDB ID: 3g5u chain A) as previously described by Patel et al. (2013) was used as the structural model for homology modelling of Pf Pgh-1. Sequence alignment revealed that both proteins had 27.9% identical residues at identical positions and 35.8% similar residues at identical positions between them (Supplementary Figure S4, S5). Because the template protein was a low homology protein (Supplementary Figure S6), Prime STA was used to build the model. This was thereafter assessed by the test of Ramachandran plot (Supplementary Figure S7) The selected model had a structure-based sequence alignment score of 0.004 and an RMSD of 0.301 Å when compared to the template protein used. It was thereafter subjected to loop refinement before preparing it for docking studies using the protein preparation wizard.

In a Ramachandran plot, the distribution of phi (ɸ) and psi (ψ) torsion angles of the protein back bone conformation are used for the assessment of the quality of the modelled protein. These plots were assessed on the VADAR (Volume Area Dihedral Angle Reporter) web server. Statistical reports showed that while a significant number of residues were in the allowed phi psi core (71% of the 1419 amino acid residues long model vs. 68% of 1276 amino acid residues long template), the model also had slightly more residues outside the allowed phi psi core than the template protein (0% and 1% for template and model respectively). Residues 1034 and 1042 on TM 11 of the Pgh-1 transporter are thought to be part of its drug binding pocket. This is so because mutations of these residues have been reported to alter the anti-plasmodial activity of drugs that are Pgh-1 substrates (Lekostaj et al., 2008; Sidhu et al., 2005). Therefore, the centroids of the functional residues Ser1034 and Asn1042 were used for grid generation with the following dimensions {inner box (20 × 20 × 20 Å), outer box (40 × 40 × 40 Å)}.

To validate the homology model, ACT-213615 (the lead molecule upon which the clinical candidate ACT-451840 was designed) reported to directly interact with partially purified Pgh-1 (Brunner et al., 2013) was docked to the modelled protein. This compound bound to the drug binding site of the transporter with a docking score of −9.747 kcal/mole (Figures 13, 14). Docking experiments of the two alkaloids with Pgh-1 revealed that BBR had a more promising binding energy (−7.32 kcal/mole) than PAL (−5.931 kcal/mole), indicating that the former is a better substrate of the transporter. Furthermore, docking studies revealed that the alkaloids interacted with the transporter in a manner that was comparable to that of the therapeutic antimalarial ACT451840, which is thought to inhibit transporter function and cause parasite death by binding to Pf Pgh-1 (Vanaerschot et al., 2017).

Protoberberine alkaloids kill the malaria parasite by inhibiting hemozoin formation

Pf Pgh-1 has been implicated to play a key role in the antimalarial of action of ACT-451840. In a similar manner, Pf Pgh-1 also plays a crucial role also in the action of Hexa-hydroquinolines (HHQs) (Vanaerschot et al., 2017). This is evident from the fact that a Pf Pgh-1 mutant (Y290F) resistant to HHQs was also resistant to ACT- 451840, which made the researchers suggest a similar mechanism of action for ACT-451840 and HHQs (Vanaerschot et al., 2017). Although ACT-451840 is thought to kill the parasite by targeting Pf Pgh-1 (Diagana and Jones, 2017; Ng et al., 2016; Vanaerschot et al., 2017), its mechanism of parasite killing is unclear since the natural substrates of the Pf Pgh-1 transporter are unknown. On the other hand, HHQs were observed to decrease hemozoin levels by interfering with hemoglobin uptake and as such acting upstream of hemoglobin catabolism. This made the researchers suggest that Pf Pgh-1 may play a role also in the uptake of hemoglobin (Vanaerschot et al., 2017). In view of this and with the observation of comparable interactions between the protoberberine alkaloids (BBR and PAL) and ACT-451840 with Pf Pgh-1 in our in silico study, we studied the hemoglobin endocytosed by the control vs. test molecule treated parasite by setting up a bloated food vacuole phenotype assay (Figure 15A,B) that enabled us to find if these alkaloids, could impede the parasite’s hemoglobin transport en route the DV. To achieve this, ring stage Pf 3D7 parasites (0–3 h p.i.) were cultured at 37° C for 24 h at 2% Parasitemia and 2% Hematocrit. They were then exposed to 1, 2.5 & 5 X IC 90 of the alkaloids. E64 (33 µM) was used to prevent hemoglobin breakdown. Cells were washed free of drugs with 1X PBS after 15 h of drug treatment (corresponding to the time when the parasite consumes host hemoglobin). Following that, wet mounts were made and Differential Interference Contrast (DIC) images were acquired using a Confocal microscope (Nikon confocal A1r). Images stained with Giemsa were also captured (Figure 15A). Intravacuolar hemoglobin uptake was inhibited only when the parasite was exposed to very high concentrations of the alkaloids (5 x IC 90) (Figure 15A) while no inhibition was seen at 1 x IC 90 & 2.5 x IC 90 (Supplementary Figure S8). Multiple images corroborating Figure 15A are provided in Supplementary Figure S9.

Since the alkaloids were homing to the DV and interacting with heme (Figures 8–12), we explored the possible inhibition of heme complexation by these alkaloids in cultured malaria parasites. As shown (Figure 16), while chloroquine, a confirmed inhibitor of hemozoin formation showed only a two-fold increment of IC 50 to 78 nM against heme bio crystallization as compared to its IC 50 for parasite killing (40 nM) the IC 50 (µg/mL) for heme bio crystallization by BBR (0.50) and PAL (0.80) was not different from those for parasite killing {IC 50 (µg/mL) BBR = 0.40; PAL = 0.80}.

The inhibition of parasite growth by the alkaloids was monitored in the presence versus in-absence of E64. In the presence of E64 (Figure 17), an increased concentration of both alkaloids (IC 50 ∼ 55 vs. 2 μM; i.e. a 28-fold increase) is required for parasite killing. This indicates that the antiplasmodial activity of the alkaloids is dependent on presence of free heme and thus suggests that the alkaloids kill the malaria parasite primarily by localizing into the parasite’s DV via DV resident Pf Pgh-1 transporter where they interact with heme and inhibit its bio crystallization. A schematic representation of the transport of BBR and PAL into the DV of malaria parasites harbouring wildtype and mutant Pgh-1 is depicted in Figure 18. In the absence of Pgh-1 inhibitors, both alkaloids lose their fluorescence upon incubation with the parasite due to heme binding. Whereas, fluorescence loss is prevented for both alkaloids upon inhibition of the transporter by Tariquidar.

Berberine and Palmatine are structural analogues with similar physicochemical and biological properties (Bhadra and Kumar, 2011). Presence of a 1,3 dioxalane cyclopentane ring in BBR and the absence of the same in PAL is the only structural difference between these two alkaloids. Their almost identical IC 50 values against the parasite’s chloroquine-sensitive strain Pf 3D7 {IC 50 (µM) BBR 1.24; PAL 1.03} could be a consequence of this structural closeness. However, when tested against drug-resistant malaria parasite strains, these two alkaloids exhibited significant differences in potency. Thus BBR {IC 50 (µM) Pf Dd2 0.54, Pf INDO 0.16; } was 2–7 times more potent against resistant strains than against the drug sensitive Pf 3D7 (IC 50 (µM) 1.24). On the other hand, PAL, showed a reversal of resistance index with lower IC 50 against the sensitive strain (IC 50 (µM) 1.03) than against the resistant strains {IC 50 (µM) Pf Dd2- 2.58, Pf IND O - 3.09}. It is thus clear that while the sensitive strain overlooks the subtle structural difference between BBR and PAL, the resistant strains are quite particular about the same. The time dependent quenching of the inherent fluorescence associated with these two alkaloids by the malaria parasite made us surmise that (a) the alkaloids home into the DV where heme and hemozoin may quench their fluorescence, (b) there is likely a transporter on the surface of DV which can ferry the alkaloids into the DV. In more concrete terms, we hypothesized that at low lethal doses (e.g. 1 μM), these alkaloids bind Pgh-1 and are transported into the parasite’s DV where their low-intensity fluorescence is muted by free heme or hemozoin present in the DV. However, at higher concentrations (e.g. 25 μM), the substrate binding site in Pf Pgh-1 transporters may become saturated, resulting in the apparent relocation of the alkaloids to both the parasite’s DV and cytosol, where they are subsequently found to stain also cytoplasmic organelles such as mitochondria. Pf Pgh-1 continues to bind extra vacuolar cytoplasmic alkaloid molecules and transport them into the DV, eventually taking all alkaloid molecules into the DV, where they may interact with heme/hemozoin and have their fluorescence quenched. Previously, BBR has been seen to first stain the mitochondria in mammalian cells before gradually colouring the nucleus at either a longer time with low concentration or a shorter time with greater concentration (Pinto-Garcia et al., 2010; Serafim et al., 2008). This time and dosage dependent labelling of the mitochondria followed by staining of the nucleus was thought to be due to a saturation of binding sites in the mitochondria (Li et al., 2019; Pinto-Garcia et al., 2010; Serafim et al., 2008). However, in our investigation with malaria parasites utilizing a high concentration (25 µM) of PAL, a strong colocalization of PAL fluorescence (green) with the mitochondrial mitotracker fluorescence (red) (Pearson’s correlation value 0.917) was seen while a similar colocalization with the nuclear DNA was poor (Pearson’s correlation coefficient 0.256). Furthermore, the alkaloids showed neither signs of loss of Mitochondrial Membrane Potential (MMP) nor nuclear DNA binding at the relevant concentrations (IC 10, 1C 50, and IC 90) for parasite suppression. Only schizonts treated with PAL (at IC 50 and IC 90) for a minimum of 8 h demonstrated MMP loss as seen by green fluorescence with the JC-1 stain. In contrast, the schizont stage parasites are killed as early as 30 min of treatment with IC 90 of these alkaloids (Erhunse et al., 2024).This sixteen fold difference in time kinetics for parasite killing and loss of MMP implies that while the alkaloids (12.5–150 µM) cause the death of mammalian cells by decreasing mitochondrial potential and inhibiting DNA synthesis (Pinto-Garcia et al., 2010; Serafim et al., 2008), parasite killing must be via another route. Two factors prompted us to investigate whether the alkaloids are transported into the DV by a transporter: (a) BBR and PAL are substrates of the multi-drug resistance transporter P-gp1 (Shitan et al., 2007), and (b) Pf Pgh-1, a homologue of P-gp1, is expressed on the parasite’s DV and is known to transport xenobiotics into the secluded ambience of DV away from cytoplasmic content (Rohrbach et al., 2006). Indeed, when the alkaloids were blocked from entering the DV using bonafide inhibitors of Pgh-1, their fluorescence was seen to originate from their cytosolic position. We however suggest that a biochemical assay for binding of the alkaloids to the Pf Pgh-1 receptor and a genetic assay such as the overexpression of Pf Pgh-1 in cells via an expression system could allow us assess the stimulation of Pf Pgh-1 ATPase activity by the alkaloids and the phosphate released upon ATP hydrolysis could have been monitored to further strengthen our finding.

Drug resistance associated with mutations in Pf MDR1 has previously been linked to mutations at N86, Y184, S1034, N1042, and D1246. Of these, Pf MDR1 computational models have implicated S1034 and N1042 in the substrate-binding pocket (Ferreira et al., 2011; Patel et al., 2013). Apart from the fact that Pf MDR1 has four copies in Pf Dd2 against only a single copy in Pf 3D7 (Lim et al., 2009), there is also a polymorphism at the 86 th residue. While residues 184, 1034, 1042, and 1246 are conserved between Pf 3D7 and Pf Dd2. Asparagine in Pf 3D7 is replaced by phenylalanine/tyrosine at the 86 th amino acid position in Pf Dd2. This N86F/Y mutation in TM1 has previously been linked to a considerable shift in drug affinity (Sanchez et al., 2008) and recently shown to reduce the ATPase activity of the transporter (Si et al., 2023). While this may explain PAL’s higher IC 50 against Pf Dd2 than against Pf 3D7 if the target of these alkaloids is found in the parasite’s DV, it does not explain BBR’s better activity against Pf Dd2 and thus suggests that both parasite strains may differ in other aspects. These limitations can be surmounted by artificially introducing point mutations and/or increasing copy numbers to the MDR1 gene of parasite lines.

Docking studies suggested that the alkaloids interact with the transporter in a manner that is comparable to the binding of the therapeutic antimalarial ACT451840, which is thought to inhibit transporter function and cause parasite death by binding to Pf Pgh-1 (Vanaerschot et al., 2017). While the natural substrates of this transporter remain unknown, Pf Pgh-1 has been implicated in drug resistance to mefloquine (Price et al., 2004), halofantrine (Wilson et al., 1993), chloroquine and quinine (Duraisingh and Cowman, 2005), ACT451840 (Ng et al., 2016) and HHQs (Vanaerschot et al., 2017). Further, members of Pgp-1-like ATP binding superfamily of proteins are involved in transport of molecules such as amino acids and proteins (Cowman et al., 1991). Thus recently, Pf Pgh-1 has been shown to play a role in the translocation of Pf Atg18; an autophagy-related protein to the DV (Sudhakar et al., 2021). Pf Pgh-1 has also been touted to play a role in vesicle mediated endocytosis of hemoglobin into the developing DV (Vanaerschot et al., 2017). With the suggested involvement of Pf Pgh-1 in vesicle mediated endocytosis of hemoglobin (Vanaerschot et al., 2017), and the observation of similar binding interaction of ACT451840 and the alkaloids to the Pf Pgh-1 transporter, we investigated the possibility of hemoglobin uptake inhibition by the alkaloids. However, shrunken DV upon treatment of E64 treated P. falciparum parasites with the tested alkaloids was seen only when parasites were treated with concentrations as high as 5 x IC 90 of the protoberberine alkaloids. As this observation was made only at high concentration of the alkaloid, we speculate that at high concentrations, the binding site of Pf Pgh-1 must have been saturated making the alkaloids localize in the cytosol from where they inhibited a player in the parasite’s hemoglobin uptake process. During the malaria parasite erythrocytic growth stages, hemoglobin is imported via cytostomes in what is referred to as host cell cytosol uptake (HCCU) (Aikawa et al., 1966) after which it is packaged in hemoglobin containing vesicles (HCVs) and transported into the DV (Milani et al., 2015). Thus, the transport of HCVs into the parasite’s DV may be compromised only at very high concentrations (5 x IC 90) (Figure 15A; Supplementary Figure S9) while low concentrations (1 x IC 90 and 2.5 x IC 90) of both alkaloids showed no effect on hemoglobin uptake (Supplementary Figure S8). In view of the fact that, actin and dynamin-mediated processes are key players in malaria parasite HCV transport (Yi et al., 2017), the demonstration that BBR inhibits actin polymerization in Jurkat cells (Yi et al., 2017) assumes great significance.

Heme is known to quench fluorescence of closely bound fluorescent compounds (Takeda et al., 2003) and our fluorescence and spectroscopic studies revealed that the fluorescence of the alkaloids is quenched following their interaction with heme. Thus, we turned our attention to the possibility of alkaloid mediated interference with conversion of heme to hemozoin. Our result suggests that indeed both alkaloids inhibit the process of heme detoxification (Figure 16). Based on our findings, BBR and PAL are substrates of Pf Pgh-1 and they localize to the parasite’s DV via the Pgh-1 transporter where they kill the parasite by inhibiting heme bio crystallization.

Our study has shown that BBR and PAL localize into the parasite’s DV and kill the malaria parasite by inhibiting hemozoin formation. Interestingly BBR does not share cross resistance with chloroquine; a classical antimalarial agent against which the parasite has generated resistance. Our observation of the localization of the alkaloids into the parasite’s DV via the Pf Pgh-1 seems to be crucial as it may guide the design of clinically-relevant antiplasmodial alkaloids of the protoberberine class. That is the optimized versions must continue to be good substrates of Pf Pgh-1 but be poor substrates of human P-gp1. Further, should parasite’s cytosolic targets of BBR and PAL be identified, then our observation of the inhibition of intravacuolar hemoglobin uptake has significant implications for the design of analogues of these alkaloids that are immune to resistance generation by the parasite. For example, the inhibition of intravacuolar uptake of hemoglobin by the alkaloids only at very high concentrations suggest that Pgh-1 binding sites were likely saturated and as such, some of the alkaloid molecules remained bound to the parasite’s cytosolic target(s) that resulted in inhibition of hemoglobin uptake. As such, should the cytosolic target(s) be unearthed, then, robust analogues of the alkaloids with potent inhibitory ability against the postulated cytosolic target(s) as well as enhanced ability for heme detoxification can be synthesized. By so doing, the parasite may have a tough time generating resistance mutations against such analogues since such compounds will have the potential of forcing the parasite into an evolutionary trap. In essence, in a bid to concentrate such drugs in its DV compartment, the parasite will have to either risk getting killed by heme toxicity or risk dying from possible hemoglobin starvation should it generate Pf Pgh-1 mutations that can interfere with the localization of such drugs into its DV. Our future directions are in line with these thoughts.