**4. Atovaquone resistance in malaria parasites**

#### **4.1. Predicted mode of action for atovaquone**

Atovaquone (a hydroxy-1,4-naphthoquinone derivative) is a broad-spectrum antiparasitic agent active against malaria, *Pneumocystis carinii* pneumonia, toxoplasmosis, and babesiosis [66]. The mode of action for atovaquone involves selective inhibition of parasite mtETC without affecting the host mitochondrial functions at effective doses, making it the first member of an entirely new class of antimalarial agents [52]. This drug shares structural similarity with ubiquinone, a coenzyme involved in mtETC and serves as a point of contact between energy metabolism and pyrimidine metabolism. Therefore, a potential molecular target of atovaquone can be the ubiquinol oxidation pocket, Qo site, of the cytochrome *bc*<sup>1</sup> complex [51, 67] because it may have a specific inhibitory effect on the parasite cytochrome *bc*<sup>1</sup> complex. Generally, the cytochrome *bc*1 complex is a structural and functional homodimer. The catalytic core comprises three redox active subunits, cyt *b* with two *b*-type hemes, cytochrome *c*<sup>1</sup> (cyt*c*1) with a *c*-type heme, and Rieske protein with a [2Fe-2S] cluster [68]. Cyt*b* catalyzes the transfer of electron from ubiquinol to cyt*c1*, coupled to the transmembrane proton translocation across the mitochondrial membrane [69]. There are two distinct catalytic sites on the cyt*b* protein, which are involved in the proton motive Q cycle and are proposed to account for the electron transfer and proton translocating activity through the cytochrome *bc*<sup>1</sup> complex: center o (also designated as Qo or center P) on the cytoplasmic site of the mitochon‐ drial inner membrane, where ubiquinol (QH2) oxidation occurs, and center i (Qi or center N) on the matrix site, where ubiquinone (Q) reduction occurs. Because of its structural similarity with ubiquinone, atovaquone appears to inhibit the cytochrome *bc1* complex by competitive binding with coenzyme Q for one of these sites.

#### **4.2. Emergence of atovaquone-resistant malaria parasites**

mitochondrion was to date considered as being impossible. However, recently, extramito‐ chondrial phenylalanyl-tRNA synthesis has been found in mitochondria of the erythrocytic stages, suggesting that the parasite mitochondrion can import tRNAs from the cytoplasmic tRNA pool [64]. These findings referring to the parasite rRNAs and tRNAs would make the parasite mitochondrial protein translation a desirable organelle to target as an antimalarial

**Figure 4.** Fragmentation of the mitochondrial LSU and SSU rRNA genes of *Plasmodium falciparum*. Red line indicates

rRNA regions. Purple lines indicate recently identified additional fragmented rRNA candidates[63].

In malaria parasites, mtDNA is replicated via rolling circle replication to generate the linear concatemers, similar to the replication mechanism used by some bacteriophages and plasmids [58]. This replication manner is remarkably different from that of the vertebrate mtDNA, which is replicated by a theta mechanism. Furthermore, mitochondrial DNA polymerase, which has been characterized as a γ-like DNA polymerase, is strongly resistant to 2,3-dideoxythymi‐ dine-5-triphosphate and, in this aspect, differs from its vertebrate homolog [65], suggesting structural differences between the *Plasmodium* and vertebrate DNA polymerase. Further research on translation and replication mechanisms of the parasite mtDNA may help identify

Atovaquone (a hydroxy-1,4-naphthoquinone derivative) is a broad-spectrum antiparasitic agent active against malaria, *Pneumocystis carinii* pneumonia, toxoplasmosis, and babesiosis [66]. The mode of action for atovaquone involves selective inhibition of parasite mtETC

drug.

24 An Overview of Tropical Diseases

potential targets for drug candidates.

**4. Atovaquone resistance in malaria parasites**

**4.1. Predicted mode of action for atovaquone**

Atovaquone is majorly used for treatment and chemoprophylaxis of falciparum malaria for international travelers [70], but the major problem is rapidity of emergence of drug resistance when it is used as a single agent. Thus far, proguanil, which inhibits the parasite dihydrofolate reductase, is combined with atovaquone to prevent the emergence. The combination drug, registered as Malarone® (GlaxoSmithKline group of companies), is approved for treating malaria in more than 30 countries and is used for chemoprophylaxis for international travelers. However, atovaquone-resistant parasites isolated from malaria patients have also been highly reported [71-73]. These studies demonstrate that atovaquone resistance is associated with point mutations of the amino acid residue at codon 268 of cyt*b* (*Pfcob*) constructing the cytochrome *bc*1 complex. The mutations Y268S, Y268N, and Y268C have been found in atovaquoneresistant parasites.

To mimic the situation of emergence of atovaquone-resistant parasites in a clinical setting, we chose a mouse malaria model using BALB/c mice and the *P. berghei* ANKA strain. In the first trial, we administered atovaquone intraperitoneally on seven consecutive days at doses ranging from 0.4 μg/kg/day to 4.8 mg/kg/day and obtained *P. berghei* isolates with four genetic resistance variations in cyt*b* [74] (Table 2). We did not observe the mutation of the amino acid residue at codon 268, which is observed in *P. falciparum*. The two mutations, M133I and L144S, are located in Qo1, and these code amino acids are critical for inhibitor resistance in yeast and mice [75, 76]. Moreover, in *Plasmodium*, the M133I and L144S amino acid changes appear to be structurally significant, altering the conformational structure of the ubiquinone-binding site and thus lowering the affinity of atovaquone to the Qo1 site. The mutation V284F is located in the sixth transmembrane domain adjacent to the Qo2 site, and the amino acid change by itself confers only an approximately 10-fold resistance to atovaquone. Notably, the mutation V284F has been found in all atovaquone-resistant clones [74].


**Table 2.** Mutations in the cytochrome *b* of *Plasmodium berghei* with atovaquone resistance

To obtain a better model for the biochemical and genetic studies of mutations found in *P. falciparum*, we performed further experiments to obtain *P. berghei* strains, resistant to atova‐ quone, with mutations in the Qo2 region conferring high degrees of resistance [77]. The parasite-infected mice were treated intraperitoneally for 3 consecutive days at a dose of 14.4 mg/kg/day, a higher dose than in the previous experiment. The results showed three variations of the atovaquone-resistant mutation, including mutations at codon 268 (Y268N, Y268C, and L271/K272R; Table 2). All the mutations were located in the Qo2 region, and these resistance levels were more than 500 times higher than those of the wild type, although the resistance levels of the previous isolates were more than 50 times higher. Administered doses of atova‐ quone affected the site of mutation in cyt*b* and the level of drug resistance.

As described above, our group has reported various mutations in the quinone-binding sites of the cyt*b* gene of *P. berghei*, such as M133I, L144S, L271V, K272R, Y268C, Y268S, Y268N, and V284F, using the mouse model with continuous atovaquone pressure. However, no direct evidence of a relationship between the mutations and resistance has been observed using intact mitochondria isolated from the malarial parasite, although biochemical analysis of the mutant has been reported using cell-free extract [78]. To address this point, we have further investi‐ gated the activity of dihydroorotate-cyt*c* reductase (regarding this mitochondrial pathway, see Section 2) in both atovaquone-resistant and atovaquone-sensitive *P. berghei* isolates [79]. The results showed that mutations in the quinone-binding site of the cyt*b* gene resulted in variable sensitivity to atovaquone and provided direct evidence for the atovaquone inhibitory mech‐ anism in the parasite cytochrome *bc*1 complex.

### **4.3. Cytochrome bc1 complex as an antimalarial drug target**

Recently, the X-ray crystallographic structure of the mitochondrial cytochrome *bc1* complex from *Saccharomyces cerevisiae* with atovaquone has been resolved, and it demonstrates atova‐ quone bound in the Qo site [80]. It can therefore explain the molecular basis for the broad spectrum of the antimalarial drug as well as for the species-specific differences in its effects. This would allow us to develop a drug targeting cytochrome *bc1* that would control the emergence of resistant parasites. Furthermore, the other group has reported cocrystallization of a bovine cytochrome *bc1* complex with the 4(1H)-pyridone class of inhibitors [81], which are potent antimalarial agents *in vivo* [82, 83]. The X-ray structure demonstrates that these inhibitors do not bind at the Qo site but rather at the Qi site. Differences in the inhibitor-binding site to cytochrome *bc*1 complex would aid the rational drug designing for reducing the emergence of inhibitor-resistant parasites and increasing selectivity against malaria parasites toward novel treatments. In the future, in addition to binding site analysis using modalities such as X-ray crystallography, we need to elucidate the molecular mechanisms explaining how atovaquone resistance mutation is generated in the parasite mt genome.

confers only an approximately 10-fold resistance to atovaquone. Notably, the mutation V284F

To obtain a better model for the biochemical and genetic studies of mutations found in *P. falciparum*, we performed further experiments to obtain *P. berghei* strains, resistant to atova‐ quone, with mutations in the Qo2 region conferring high degrees of resistance [77]. The parasite-infected mice were treated intraperitoneally for 3 consecutive days at a dose of 14.4 mg/kg/day, a higher dose than in the previous experiment. The results showed three variations of the atovaquone-resistant mutation, including mutations at codon 268 (Y268N, Y268C, and L271/K272R; Table 2). All the mutations were located in the Qo2 region, and these resistance levels were more than 500 times higher than those of the wild type, although the resistance levels of the previous isolates were more than 50 times higher. Administered doses of atova‐

As described above, our group has reported various mutations in the quinone-binding sites of the cyt*b* gene of *P. berghei*, such as M133I, L144S, L271V, K272R, Y268C, Y268S, Y268N, and V284F, using the mouse model with continuous atovaquone pressure. However, no direct evidence of a relationship between the mutations and resistance has been observed using intact mitochondria isolated from the malarial parasite, although biochemical analysis of the mutant has been reported using cell-free extract [78]. To address this point, we have further investi‐ gated the activity of dihydroorotate-cyt*c* reductase (regarding this mitochondrial pathway, see Section 2) in both atovaquone-resistant and atovaquone-sensitive *P. berghei* isolates [79]. The results showed that mutations in the quinone-binding site of the cyt*b* gene resulted in variable sensitivity to atovaquone and provided direct evidence for the atovaquone inhibitory mech‐

Recently, the X-ray crystallographic structure of the mitochondrial cytochrome *bc1* complex from *Saccharomyces cerevisiae* with atovaquone has been resolved, and it demonstrates atova‐ quone bound in the Qo site [80]. It can therefore explain the molecular basis for the broad spectrum of the antimalarial drug as well as for the species-specific differences in its effects. This would allow us to develop a drug targeting cytochrome *bc1* that would control the emergence of resistant parasites. Furthermore, the other group has reported cocrystallization

has been found in all atovaquone-resistant clones [74].

26 An Overview of Tropical Diseases

**Table 2.** Mutations in the cytochrome *b* of *Plasmodium berghei* with atovaquone resistance

quone affected the site of mutation in cyt*b* and the level of drug resistance.

anism in the parasite cytochrome *bc*1 complex.

**4.3. Cytochrome bc1 complex as an antimalarial drug target**
