Anti-malarial activity of geldanamycin derivatives in mice infected with Plasmodium yoelii

Chloroquine phosphate was a kind gift from BDH Industries LTD., Mumbai, India. Protease inhibitor cocktail (cat no. P2714) was obtained from Sigma-Aldrich. Swiss mice (4-6 weeks old) were provided by the animal house facility at Tata Institute of Fundamental Research, Mumbai, India. All chemicals used were of Analar grade. HRP conjugated anti-mouse IgG was from Sigma-Aldrich.

All reagents and solvents were purchased from commercial sources and used without further purification. ¹H NMR (300 and 500 MHz) and ¹³ C NMR (75 MHz) spectra were recorded in CDCl₃ solution on a 300 MHz spectrometer. Chemical shifts were referenced to δ 7.26 and 77.0 ppm for ¹H and ¹³ C spectra, respectively. High-resolution mass spectra were generated at the Purdue Mass Spectrometry Facility. Thin-layer chromatography (TLC) was performed on 250 μ M and 1000 μ M silica gel plates. Flash chromatography was run using RediSep normal-phase flash columns (230-400 mesh). Geldanamycin was isolated from fermentation of Streptomyces hygroscopicus var. geldanus that was provided by Dr. David Newman, NCI-Frederick. The production of geldanamycin was modified from a previously established method [14]. Briefly, 100 mL of production medium in 500 mL tribaffled flasks with silicon closures were inoculated with confluent oatmeal slants and agitated at 150 rpm in the dark at 28°C. Initial metabolite productions were monitored from days 4.0 to 7.0 for harvest using HPLC [24]. Harvest on day 5.5 generally achieved productions of 1.01 +/- 0.29 nmoles per 5 mL. A sample of 17-N-allyl-17-demethoxygeldanamycin (17-AAG) was prepared after the established procedure from geldanamycin [25].

To a solution containing 1 g (10.2 mmol) of 4-pentynoic acid in 20 mL anhydrous CH₂Cl₂ was added 2.1 g (10.2 mmol) of N,N'-Dicyclohexylcarbodiimide (DCC) and 3.1 g (30.5 mmol) of triethylamine at 25°C under N₂. The reaction mixture was stirred at 25°C for 10 min, and then 6.7 g (30.5 mmol) 3,3'-(2,2'-oxybis(ethane-2,1-diyl)bis(oxy))dipropan-1-amine in 20 mL anhydrous CH₂Cl₂ was added. The reaction mixture was stirred at 25°C for another 3 h. The solution was filtered and concentrated under reduced pressure. The residue was loaded into 80 g flash silica gel column eluting with two volumes of 95:5 CH₂Cl₂-methanol. The purified product was eluted using step gradients of 10:1:1 followed by 10:1:2 methanol:NH₄OH:10% NH₄OAC in H₂O to yield alkyn-PEG-amine 2.6 g (85%) as a colorless sticky liquid. TLC (80:10:10 methanol: NH₄OH:10%NH₄OAc in H₂O) followed by Ninhydrin staining showed R_f = 0.37. ¹H NMR (CDCl3): δ 1.32 (dt, 4H), 1.74 (m, 2H), 1.88 (m, 2H), 1.95 (m, 2H), 3.0 (t, 1H), 3.35 (t, 2H), 3.57 (s, 12H), 7.33 (s, 2H), 7.55, (s, 1H) Mass spectrum (300.2), m/z 301.19 (M + H)⁺.

To a solution containing 390 mg (1.3 mmol) of Alkyn-PEG-amine in 8 mL anhydrous CH₂Cl₂ was added 81 mg (0.14 mmol) of GA at 25°C under N₂. The reaction mixture was stirred at 25°C for 4 h and to the resulting solution 30 mL CH₂Cl₂ was added, and washed with three 10-mL portions of H₂O, three 10-mL portion of saturated brine. The organic layer was dried (Na₂SO₄) and concentrated under diminished pressure. The residue was purified by chromatography on a flash column, Eluted with 98:2 methylene dichloride-methanol gave 17-PEG-Alkyn-GA as purple solid: yield 42 mg (36%); Mass spectrum, m/z 851.03 (M + Na)⁺(C₃₄H₄₈O₂ requires 828.45). A complete set of ¹H and ¹³ C NMR data are provided as Additional file 1.

A lethal mouse malarial parasite, P. yoelli 17XL was cultured in six-week old Swiss mice and parasite infected red blood cells (PRBCs) were used for infecting fresh mice by intra-peritoneal injection (~10⁶ PRBC). Parasitaemia was scored everyday by tail bleeding and preparing thin blood smears from infected mice. The infected blood smears were stained with Giemsa; about 300-400 RBCs were examined by microscopy and the infected erythrocytes were reported as the percent of the total. The pharmacological agents were dissolved in 10% DMSO or water and injected intra-peritoneal. The fractional distribution of various intra-erythrocytic asexual stages of parasites were determined by counting rings, trophozoites and schizonts and expressed in terms of percentage of total infected or parasitized RBCs (PRBC).

The infected mice that survived the malaria after drug treatment (17-AAG, 17-Alkyn-PEG-GA and chloroquine) were allowed to recuperate for one month after parasite clearance. Each surviving mouse was re-challenged by injecting with ~10⁶P. yoelii 17XL PRBCs and the parasites were allowed to grow. Thin blood smears were made every day to estimate percentage parasitaemia. Some of these mice did not show disease symptoms and cleared parasitaemia completely after 21 days of parasite infection. Approximately 0.1 to 1.0 ml of blood samples were collected using capillaries and allowed to clot for 30 min at room temperature and then subjected to centrifugation for 10 min at 3000 × g. The supernatant (serum) was collected and stored at -80°C until further analysis.

To obtain parasite sensitive serum, mice were injected with lower doses of parasite (~10⁴) to sustain the viability of mice. After 21 days of post-parasite injection, serum samples were prepared as mentioned above. Naïve serum was collected from fresh mice.

Plasmodium yoelii cells were prepared as described earlier [26], with slight modification. Briefly, the mice were infected with P. yoelii 17XL (lethal strain) and the parasites were allowed to grow until the infected red blood cells reached ≥ 30%. At this stage, 1-2 mL of blood was collected in equal volume of anti-coagulant solution (136 mM glucose, 42 mM citric acid and 75 mM Sodium citrate). Red blood cells (RBCs) were collected by centrifugation (1500 × g for 10 minutes) and washed three times with phosphate buffer saline (PBS) (137 mM NaCl, 2.7 mM KCl, 10 mM Na₂HPO₄, 1.8 mM KH₂PO₄, pH 7.4). The RBCs were re-suspended in PBS containing 1 mM PMSF and appropriate amounts of protease inhibitor cocktail as recommended by the supplier (Sigma-Aldrich). To this suspension of infected RBCs, 0.05% saponin was added and allowed to incubate for 1 min at 37°C. The solutions were then kept at room temperature (~20°C) for 30 minutes to release the parasite from the infected RBCs. Parasite cells were collected by centrifugation at 18000 × g for 10 min and the pellets washed with PBS to remove all the hemoglobin (as judged by red color). The cell pellet was stored at -80°C until further analysis.

Parasite cell pellets (~200 μg) were suspended in 200 μL of PBS containing 5 mM EDTA, 1 mM PMSF and protease inhibitor cocktail (Sigma-Aldrich). After incubation on ice for 10 minutes, the cells were subjected to freeze-thaw (six cycles) by freezing in liquid nitrogen (2 minutes) and thawing at room temperature (2 minutes). These cells were then subjected to ultrasonification for 10 seconds at constant duty cycle by using Branson Sonifier 450 and then the sample was incubated on ice for 1 minute. This process was repeated six times. The cell extract was centrifuged at 100,000 × g for 30 minutes and the supernatant collected was the cytosolic fraction. Protein concentrations of the samples were estimated by measuring OD_{280 nm}. To prepare RBC extract, ~1 mL blood was collected from mice with 0% (uninfected), ~3% or ~30% parasitaemia in equal volume of anticoagulant. RBCs were collected by centrifugation and were lysed by using 0.05% saponin as mentioned above. The supernatant obtained by centrifuging of the lysed RBCs at 14000 × g was collected as RBC extract. Protein concentration of sample extracts was measured at OD_{280 nm}.

SDS-PAGE and Western blotting was performed as described earlier [27]. Typically 20 μg of cellular (parasite or RBC) protein extracts were analyzed using a 12% SDS-gel and visualized with silver stain or transferred to a PVDF membrane for Western blotting using Bio-Rad Trans-Blot Semi-Dry Transfer Cell. The blots were probed by using various anti-sera (1:1,000 dilution in 1× PBS), followed by secondary HRP conjugated anti-mouse IgG (Sigma-Aldrich) used at 1:1,000 dilutions.

Two different derivatives of geldanamycin, namely 17-allylamino-17-demethoxygeldanamycin (17 AAG) and a highly water soluble pegylated derivative of GA, 17-N-(3-(2-(-2(3-aminopropoxy)ethoxy)propyl)pent-4-ynamide-17-demethoxygeldanamycin (17-PEG-Alkyn-GA) (Figure 1) were tested on P. yoelii infected mouse malaria model system. Four groups (each group having four animals) of mice were infected with the parasite (intra-peritoneal injection of ~10⁶ PRBCs). After 6 days when the average parasitaemia reached ~8-12% and all the animals displayed characteristic symptoms of malaria, the control group was injected with vehicle control (10% DMSO). For drug administration, 300 nmoles of each agent constituted a single dose. The second group of mice was injected with 300 nmoles of 17-AAG (MW 585.31; 0.18 mg/mouse/dose; 7.2 mg/Kg body weight) and the third group was injected with 300 nmoles of 17-PEG-Alkyn-GA (MW 828.45; 0.25 mg/mouse/dose; 10.2 mg/Kg body weight) (dissolved in 10% DMSO). The fourth group of mice was injected with 300 nmoles chloroquine phosphate dissolved in water. Parasitaemia was monitored every day until either the mice died or could clear the parasites. Figure 2(A-D) shows the parasitaemia profiles for control (untreated), 17-AAG, 17-PEG-Alkyn-GA and chloroquine treated mice. The arrowheads mark the 6^th and the 12^fth day when the drugs were injected. Figure 2E presents the average parasitaemia for different groups of mice. In control groups, the parasitaemia reached almost ~60% and all the animals died by day 14 post-infection (Figure 2F). Single dose treatment with chloroquine on 6^th day post infection was adequate to clear the parasites and cure the mice of the disease. Injection of GA derivatives on day 6 post-infection did result in control of parasitaemia. However, a second dose was needed for complete clearance of the parasites. Out of four animals in two groups treated with GA derivatives, one animal succumbed to infection. These data support the conclusion that geldanamycin derivatives can cure malaria. However, the current data indicates that the GA analogs are not very effective in a single dose (non-optimized) as compared to chloroquine.

The blood smears that were prepared for each mouse on daily basis were examined for the evaluation of invasion of normocytes and reticulocytes. Before the injection of drugs (i.e. until the day 6 post-infection), parasitaemia was exclusively restricted to normocytes. Mice that were subjected to treatment with 17-AAG and 17-PEG-Alkyn-GA, preferential invasion of reticulocytes was observed. However, blood smears from control and chloroquine treated mice continued to show normocyte invasion until either the death of the animal (control group) or clearance of the parasites (chloroquine group) (Figure 3). Depending on the strain, P. yoelii is known to invade both, normocytes, as well as reticulocytes. Normocytic invasion is lethal while reticulocyte invasion is rather benign [28, 29]. The P. yoelii 17XL strain is known to invade normocytes [28] and the same is observed here. However treatment with GA-derivatives resulted in alteration of this specificity from normocytes to reticulocytes. It is likely that this change in invasion specificity renders the parasite benign and may contribute, in part to why the host system eventually succeeds in clearance of the pathogen.

Daily blood smears prepared for each animal were also scored for fractional distribution of various intra-erythrocytic asexual stages (rings, trophozoites and schizonts) of the parasite. In each smear, rings, trophozoites and schizonts were counted and plotted as the percent of total infected cells. Data are presented in Figure 4. In the control group of mice, initially (1-3 days post infection), most parasites are in the ring stage. With time, there is a gradual decrease in the ring population with concomitant increase of trophozoites (Figure 4A), suggesting constant transitions of rings to trophozoites. However, on treatment with 17-AAG or 17-PEG-Alkyn-GA on day 6 post infection, population of rings increased and stabilized at much greater fraction (Figure 4B, day 7-10 post infection) as compared to the control group (Figure 4A). Such a distribution can arise if there is a blockade or reduction in transition from ring to trophozoite stage. The profile of stage specific distribution of the parasites was rather similar for the both the geldanamycin derivatives used (17-AAG or 17-PEG-Alkyn-GA). These observations are in agreement with the conclusions arrived at by Bhanumathy et al. [19] where it was demonstrated that GA blocks the progression of rings into trophozoites in in vitro cultures of P. falciparum. In the blood smears examined here, pycnotic bodies were never observed. Thus, it is unlikely that there is any large scale death and disintegration of the parasite in response to GA treatment as reported earlier [20].

In GA-derivatives treated mice, parasite persisted for a prolonged period (Figure 2E) in the host and eventual clearance was through reticulocyte invasion. It is likely that such prolonged exposure to parasite may result in development of immunity to subsequent challenges of P. yoelii. To test this hypothesis, the mice that were cured by GA-derivative drug treatment were allowed to recover and live a healthy life for 30 days and then challenged with a fresh dose of P. yoelii 17XL. For comparison, second group of mice that had been cured from malaria symptoms by treatment with chloroquine and allowed to recover for 30 days, were also challenged for the second time. To ensure that the parasite is lethal, a control set consisting of four fresh naïve mice was also included. Parasitaemia profiles in these three groups of mice were monitored daily and are shown in Figure 5. As expected control mice had high parasitaemia (40-60%) that resulted in their death between days 5 to 8 (Figure 5A). The group treated with geldanamycin derivatives showed very low parasitaemia (< 4%) that peaked on day 2 and got cleared by day 9 (Figure 5B). Chloroquine-treated mice had intermediate profile with parasitaemia reaching around 8-13% that did clear by day 16 (Figure 5C). Average parasitaemia profiles of these three groups are shown in Figure 5D. These results suggest that mice treated with HSP90 antagonists developed a robust immunity against subsequent challenge with the parasite. In order to examine the profile of antibodies generated in different groups of mice, serum samples were collected from each of these mice and pooled together for each group.

To examine the antibody profiles of serum samples collected from above mentioned three groups of mice, proteins from the whole cell parasite extracts were separated using a 12% SDS-PAGE and transferred to a PVDF membrane. These blots were subjected to western analysis using different serum samples. Results of such western analysis are shown in Figure 6. Lane 1 is a silver stained protein profile of the whole cell P. yoelii extract. Western blots made using the sera collected from naïve (lane 2) and parasite sensitive mice sera (lane 3) did not show any reactivity towards the parasite proteins. In contrast to these, serum samples collected from 17-AAG or 17-PEG-Alkyn-GA treated mice had antibodies against vast majority of parasite proteins (lane 4). Sera collected from the chloroquine treated group showed antibodies against a subset of the parasite proteins (lane 5). These data are consistent with a hypothesis that drug-induced antibody response mounted by the host against the drug-attenuated parasite leads to protection against a subsequent parasite challenge.