Suppression of experimental cerebral malaria by disruption of malate:quinone oxidoreductase

Female C57BL/6J mice 5- to 6-weeks old were purchased from CLEA Japan INC (Tokyo, Japan). The experiments were approved by the Experimental Animal Ethics Committee of Kyorin University School of Medicine, Tokyo, and all experimental animals were kept at the animal facility in a specific-pathogen-free unit with sterile bedding, food and water.

The SK-1 construct contained a selection cassette consisting of the green fluorescent protein gene (gfp) and a pyrimethamine resistance gene, human dihydrofolate reductase-thymidylate synthase (hdhfr) [15]. The expression of gfp and hdhfr is controlled by hsp70 (PBANKA_071190) and elongation factor-1 (PBANKA_113340) promoters, respectively. Plasmid (pLG4.10[luc2]) containing luciferase gene (luc2) was purchased from Promega (Madison, WI, USA). To generate luciferase-expressing cassette, luc2 was amplified by PCR using specific primers (see Additional files 2, 3). The PCR product of luc2 and SK-1 construct were cleaved using the NheI and BglII restriction enzymes, and the gfp of SK-1 was replaced with luc2. The gene-targeting vectors were prepared by PCR [16]. Briefly, the 5′ and 3′ flanking regions the ORF of the target genes, p230 [17], fh locus (PBANKA_082810) and mqo locus (PBANKA_111630), were amplified by PCR. The PCR products were annealed to either side of luciferase expressing cassette and amplified by PCR using gene-specific primers (see Additional file 2).

Plasmodium berghei ANKA strain is a high-virulence strain and the parasites, which had been cloned by limiting dilution, were obtained from Dr. W. P. Weidanz (University of Wisconsin–Madison, Madison, WI, USA). Erythrocytes infected with P. berghei parasites were transferred to RPMI1640 medium supplemented with 25% (v/v) FBS, 0.05 mg/mL Penicillin, 0.05 mg/mL Streptomycin and then incubated for 18 or 22 h under condition of 90% N₂, 5% CO₂ and 5% O₂. Mature schizonts and gametocytes were collected by Nycodenz density-gradient centrifugation [18]. Transformations were performed using the Amaxa Basic Parasite Nucleofector Kit (Amaxa GmbH, Cologne, Germany) according to the manufacturer’s protocol and luciferase-expressing cassette was introduced into the ORF of targeted gene by double-crossover homologous recombination [18]. Briefly, 5 × 10⁶ to 5 × 10⁷ purified P. berghei mature schizonts were mixed with 100 μL of Nucleofector solution containing 5 μg of a gene-targeting vector. Transfections were then completed using the Amaxa Nucleofector electroporation program U-33. Transfected parasites were then injected intravenously (i.v.) into naïve C57BL/6 recipient mice. At 30 h post-injection, transfected parasites were selected by addition of pyrimethamine to the drinking water of infected mice. After parasitaemia returned to detectable levels post-selection, transfected parasites were cloned by limiting dilution, after which a single parasite was injected into a mouse to ensure a clonally pure population. Cloned transfected parasites were stored as frozen stocks in liquid nitrogen. Erythrocytes infected with transfected parasites were generated in donor mice inoculated intraperitoneally with each frozen stock of parasites. The donor mice were monitored for parasitaemia daily and bled for experimental infection in ascending periods of parasitaemia. Experimental mice were infected intravenously with 1 × 10⁴ infected erythrocytes or 5 × 10⁶ to 5 × 10⁷ purified mature schizonts of a given parasite strain.

Blood was observed by microscopic examination of methanol-fixed tail blood smears stained with 3% (w/v) Giemsa diluted with Sörensen’s phosphate buffer, pH 7.2, for 45 min. The number of infected erythrocytes in 250 erythrocytes was enumerated when parasitaemia exceeded 10%, whereas 1 × 10⁴ erythrocytes were examined when mice showed lower parasitaemia. The percentage of parasitaemia was calculated as follows: [(Number of infected erythrocytes)/(Total number of erythrocytes)] × 100.

To generate luciferase-expressing cassette and confirm the introduction of Luc2-expressing cassette into target gene, genomic PCR was performed using a 25 μL PCR mixture containing 1×TaKaRa PrimeSTAR GXL buffer (TaKaRa, Shiga, Japan), 2.5 mM dNTPs, 0.5 μL of DNA, 5 U/μL TaKaRa PrimeSTAR GXL DNA polymerase (TaKaRa), and PCR primers (0.25 μM). Thirty-five cycles of PCR were performed on a C1000 thermal cycler (Bio-Rad, Hercules, CA, USA). Each cycle consisted of denaturation at 98 °C for 15 s, annealing at 55 °C for 15 s, and extension at 68 °C for 1–6 min. PCR products were then analysed on a 1% (w/v) agarose gel, and stained with ethidium bromide.

Blood was obtained from infected mice exhibiting 2–5% parasitaemia. Total RNA was isolated from blood containing 5 × 10⁶ or 1 × 10⁷ infected erythrocytes using Isogen (Nippon Gene, Tokyo, Japan) according to the manufacturer’s protocol. Total RNA was then treated with DNase and reverse-transcribed using murine leukaemia virus reverse transcriptase (Applied Biosystems, Carlsbad, CA, USA) with random hexamer primers under the following conditions: 70 °C for 10 min, 25 °C for 10 min, and 42 °C for 30 min. The reaction was terminated by heating at 99 °C for 5 min, and the resulting cDNA products were stored at −20 °C until required. All PCR reactions were run in a 25 μL volume consisting of 1×TaKaRa Ex Taq buffer, 2.5 mM dNTP, 1 μL of cDNA products, 5 U/μL TaKaRa Ex Taq DNA polymerase, and PCR primers (0.25 μM); a list of primers used for semi-quantitative RT-PCR can be found in Additional file 2. PCR reactions were performed on a C1000 thermal cycler (Bio Rad) for 30 cycles under the following conditions: denaturation at 95 °C for 30 s, annealing at 55 °C for 30 s and extension at 72 °C for 45 s. Products were resolved on a 2% (w/v) agarose gel, and stained with ethidium bromide. Samples with DNase-treated RNA template were used as the negative control.

To remove leukocytes, the blood was mixed with same volume of PBS and passed over a Plasmodipur Filter (EuroProxima, Arnhem, Netherlands). Erythrocytes were washed twice with RPMI1640 medium by centrifugation at 4 °C at 800×g, for 5 min and then transferred to RPMI1640 medium supplemented with 25% (v/v) FBS, 0.05 mg/mL Penicillin, 0.05 mg/mL Streptomycin. Then erythrocytes were incubated for 16 h under condition of 90% N₂, 5% CO₂ and 5% O₂. Stages of parasite were mainly schizonts as confirmed by Giemsa staining. Infected parasites were collected by centrifugation at 4 °C at 800×g, for 5 min. Crude mitochondria of P. berghei were prepared as described previously [19].

MQO activity assay was performed at 25 °C with V-660 spectrophotometer (JASCO) and measured in 1 mL of the reaction mixture containing 20 µg of mitochondrial fraction, 45 µM 2,6-dichlorophenolindophenol (DCIP), 100 µM ubiquinone-2 and 2 mM KCN in 50 mM potassium phosphate buffer, pH 8.0 by recording the decrease in absorbance due to DCIP reduction at 600 nm (ε600 = 21/mM/cm) after the reaction was initiated by adding 10 mM sodium malate.

FH activity was performed 37 °C with Benchmark Plus microplate spectrophotometer (Bio-Rad) and measured in 200 µL of the reaction mixture containing 20 µg of mitochondrial fraction, 0.25% (v/v) Triton X-100, 50 µg/mL 3-acetylpyridine-adenine dinucleotide, oxidized form (APAD⁺) (Oriental Yeast, Japan), 1 U/mL diaphorase, 0.2 mg/mL nitroblue tetrazolium (NBT) and 2 U/mL malate dehydrogenase from pig heart (Oriental Yeast) in 100 mM Tris–HCl buffer pH 8.0 by recording the absorbance change of NBT at 530 nm (ε530 = 3.6 × 10⁴/mM/cm) after the reaction was initiated by adding 10 mM sodium fumarate.

To monitor the viability of fh- and mqo-disrupted P. berghei by luciferase activities, luciferase-expressing cassette was introduced into fh locus (PBANKA_082810) and mqo locus (PBANKA_111630) in the genome of P. berghei (see Additional file 3). Moreover, luciferase-expressing cassette was also introduced into p230 locus (PBANKA_030600), which is not an essential gene in the complete life cycle of P. berghei [17], and the resultant was used as the control parasite in the following experiments. Blood was obtained from C57BL/6 mice exhibiting 2–5% parasitaemia, and diluted with RPMI 1640 medium to a final concentration of 1 × 10⁷ infected erythrocytes/well in a 96-well plate. Then, infected erythrocytes were cultured for 3 h in RPMI 1640 medium supplemented with 5 mM sodium fumarate (Sigma-Aldrich Co, Missouri, USA), 5 mM sodium malate (Sigma-Aldrich Co) or control PBS. After cultivation, d-luciferin (1.5 mM; Promega) was added to each well, and luminescence measured using an SH9000 luminometer (Hitachi High-Technologies Corporation, Tokyo, Japan) at 5 min after addition of d-luciferin. Results are presented as relative fluorescence unit (RFU) fold change compared with control PBS at 5 min after addition of d-luciferin. Also, blood was obtained from C57BL/6 mice on 16 h after injection of schizonts, and diluted with RPMI 1640 medium to a final concentration of 5 × 10⁶ infected erythrocytes/well in a 96-well plate. After cultivation, d-luciferin (1.5 mM; Promega) was added to each well, and luminescence measured using an SH9000 luminometer at 20 min after addition of d-luciferin. Results are presented as absolute luciferase activity in counts per second (cps).

To investigate whether mitochondrial membrane potential (MMP) in malaria parasites is compromised by deficiency of FH or MQO, MMP in trophozoites of control, Δfh and Δmqo parasites was determined using MMP-sensitive fluorochrome Mitotracker^® Red CMXRos (Invitrogen, Darmstadt, Germany). Inner membrane and nuclear DNA were then stained with the BODIPY FL C16 (Invitrogen) and Hoechst 33342 (Invitrogen), respectively. Mitotracker^® Red CMXRos was added to medium with a concentration of 100 nM and incubated for 30 min at 37 °C. Then, BODIPY FL C16 and Hoechst 33342 were added to medium with a concentration of 100 nM and of 1 µg/mL for 10 min at 37 °C. For all staining of malaria parasites, the staining medium was removed after the incubation period and the fresh medium was added. The Bright field and fluorescence microscopy images were photographed at 1000× magnification using an All-in-One Fluorescence Microscope (BZ9000; KEYENCE Japan, Osaka, Japan). Mean fluorescence intensity (MFI) of MitoTracker in the photographs was analysed using BZ-II Analyzer software (KEYENCE Japan).

It is known that breakdown of the blood–brain barrier is an indicator of ECM. When P. berghei-infected mice develop ECM, Evans blue is injected i.v. and the brain is stained as a result of extravasation of the dye [20]. Mice were injected i.v. with 0.2 mL of 1% (w/v) Evans blue (Wako, Osaka, Japan) on days 7 and 14 post-infection. Mice were euthanized and brains perfused with PBS 1 h later. After the brains were removed and photographed they were weighed and placed in 2 mL formamide (Wako, Osaka, Japan) at 37 °C for 48 h to extract the Evans blue dye. Absorbance was measured at λ = 620 nm with a Multiscan FC microplate reader (Thermo Fisher Scientific Inc., Waltham, USA). The Evans blue concentration was calculated from a standard curve and is expressed as µg of Evans blue per g of brain.

Student’s t test was performed using Statcel (OMS Ltd., Saitama, Japan). Statistically significant differences were defined as a value of p < 0.05.

To investigate the role of FH and MQO in malaria parasite growth in erythrocytes, fh- and mqo-disrupted P. berghei were generated. The luciferase mRNA level in fh-disrupted (Δfh) and mqo-disrupted (Δmqo) parasites was comparable to that in control parasites (Fig. 1a, b). Moreover, fh and mqo mRNAs were not detected in Δfh and Δmqo parasites, respectively (Fig. 1a, b). These findings demonstrate that the luciferase-expressing cassette was successfully introduced into fh and mqo locus.

In Δfh parasite-infected erythrocytes, the specific activity of FH was 0.99 ± 0.12 nmol/min/mg, which was 58.5 and 60.0% lower compared to control-infected and Δmqo parasite-infected erythrocytes (1.69 ± 0.91 and 1.65 ± 0.54 nmol/min/mg, respectively) (Fig. 1c). As FH activity was detected in uninfected erythrocytes, the FH activity in Δfh parasite-infected erythrocytes seems to be derived from host FH (Fig. 1c). The MQO activity in Δmqo parasite-infected erythrocytes was completely abrogated, although high activities in control parasite- and Δfh parasite-infected erythrocytes were detected (Fig. 1d).

The effect of deficiency of FH and MQO on parasite growth in vivo were investigated (Fig. 2a). Δfh and Δmqo parasites showed a growth pattern similar to control parasites, until 18 h post-inoculation of schizonts (Fig. 2a). However, at 24 h post-inoculation, the proportions of late trophozoites (stage 3) of Δfh and Δmqo parasites were considerably higher than in control parasites (Fig. 2a). These findings suggest that parasite growth during late trophozoites and schizont stage is delayed by deficiency of FH or MQO.

To evaluate the production of gametocytes in Δfh and Δmqo parasites, the mRNA levels of the gametocyte-specific proteins MDV1/PEG3 and g377 were examined. mdv-1/peg3 and g377 mRNAs can be detected from the early phase of gametocytogenesis [21, 22]. As results, gametocyte-specific gene expression during asexual-blood-stage Δfh and Δmqo parasites was comparable to that of control parasites (see Additional file 4A, B). Nuclear enlargement, the distribution of pigment granules throughout the cytoplasm and enlargement of cells were observed in a 22 h culture of Δfh and Δmqo parasites (see Additional file 4C, D). These three features are gametocyte-specific characteristics [23]. Taken together, these results suggest that the development of female and male gametocytes of malaria parasites was not affected by deficiency of FH and MQO.

ATP plays a central role in energy transduction in both eukaryotic and prokaryotic cells, and is produced in all metabolically active cells. The luciferase–luciferin system, in which luciferase reacts with d-luciferin in the presence of ATP, an oxygen molecule, and magnesium ions to produce luminescence, has facilitated investigation of parasite viability [24]. To examine the effect of deficiency of FH and MQO on parasite viability, schizonts of control, Δfh and Δmqo parasites were inoculated into mice. Trophozoites were obtained at 8 and 16 h post-inoculation and their luciferase activities were measured. The absolute luciferase activity values of Δfh parasites were comparable to those of control parasites (Fig. 2b). In contrast, the luciferase activities of Δmqo parasites were significantly reduced, by ~50% compared with those of control and Δfh parasites (Fig. 2b). These findings suggest that the viability of asexual-blood-stage malaria parasites is decreased by deficiency of MQO but not FH.

To investigate the effect of FH and MQO deficiency on mitochondrial function, the study compared the mitochondrial membrane potential (MMP) of control, Δfh and Δmqo parasites by monitoring MitoTracker uptake. MMP was not affected by deficiency of FH and MQO (Fig. 3), suggesting that the reduction of parasite viability caused by MQO deficiency is independent of defects in mitochondrial integrity.

Results that luciferase activities were not affected by deficiency of FH suggest cooperation of the host- and parasite-derived FH in fumarate metabolism. Therefore, the effect of addition of fumarate or malate to the culture on the viability of control, Δfh and Δmqo parasites were investigated using the luciferase–luciferin system. In the culture of control parasites, luciferase activities were increased by addition of fumarate or malate compared with the control (Fig. 4). The luciferase activities of Δfh parasites and Δmqo parasites were also increased in culture medium supplemented with fumarate or malate (Fig. 4). Therefore, enriched metabolic substitutions by host-derived enzymes, such as FH, in erythrocytes may be associated with increased viability of Δfh and Δmqo parasites in culture medium supplemented with fumarate or malate.

Plasmodium berghei strain ANKA is the causative agent of ECM, in which neurologic signs and histopathological findings, such as haemorrhages and sequestration of infected erythrocytes within cerebral microvessels, are observed [20]. However, the relationship between parasite growth and pathogenicity in ECM is unclear. Therefore, the effect of FH or MQO deficiency on ECM were examined. Mice infected with Δfh and Δmqo parasites showed lower levels of parasitaemia than mice infected with control parasites until days 5 and 7 post-infection, respectively (Fig. 5a; see Additional file 5). All mice infected with Δfh parasites showed neurologic signs and died on days 6–7 post-infection (Fig. 5b); this is similar to the effect in control mice. In contrast, the survival of mice infected with Δmqo parasites was prolonged compared to that of mice infected with control parasites or Δfh parasites (Fig. 5b). The mice infected with Δmqo parasites ultimately died via anaemia within days 25–30 post-infection.

Breakdown of the blood–brain barrier is an indicator of ECM [20]; therefore, breakdown of the blood–brain barrier in mice infected with control, Δfh or Δmqo parasites were investigated by assessing extravasation of Evans blue in the brain. The levels of extravasation of Evans blue in the brains of Δfh-infected mice on day 7 post-infection were comparable to those in mice infected with control parasites (Fig. 5c). In contrast, the levels of extravasation of Evans blue in the brains of Δmqo-infected mice on days 7 and 14 post-infection were markedly lower than those in mice infected with control or Δfh parasites, and were comparable to those of uninfected mice (Fig. 5c). These results suggest that the ECM caused by malaria parasites is suppressed by deficiency of parasite MQO, but not FH.