**2. Development of bioassays for antimalarial activity**

### **2.1 Malaria parasite life cycle**

*P. falciparum* has a complex life cycle involving the *Anopheles* vector, where the sexual phase occurs, and humans where the parasite undergoes two phases of extensive asexual proliferation. After the bite by an infected *Anopheles*, the parasites at the sporozoite stage enter the blood and are carried to the liver where they traverse and invade hepatocytes initiating what is called the hepatic or exoerythrocytic phase. During this asymptomatic period, that lasts 5-7 days for *P. falciparum*, the sporozoites develop within the hepatocytes and, after several rounds of mitosis, produce several thousands of new infective forms called the merozoites that are released into the bloodstream and invade the red blood cells. During this intraerythrocytic cycle, which lasts 48 h for *P. falciparum* and causes the malaria disease, the parasite undergoes a successive development into the ring stage (0-20 h) and the trophozoite stage (20-36 h), which then undergoes several mitoses (schizont stage, 36-48 h) that lead to the differentiation of 16-24 infective merozoites. Erythrocyte lysis releases the merozoites into the bloodstream and a new intraerythrocytic cycle can be initiated. For some still not well understood reasons, some merozoites differentiate into male and female gametocytes that are taken up by the *Anopheles* during the blood meal. Gametogenesis resumes in the digestive tract of the mosquito and sexual reproduction occurs forming a zygote that produces, after a meiosis and several mitosis, thousands of haploid sporozoites under the basal lamina, which then migrate to the salivary glands and can be transmitted to humans during a bite. Different antimalarial bioassays have been developed based on the *in vitro* inhibition of parasite growth or, more recently, on the inhibition of potential parasite targets, allowing the screening of chemical compounds.

#### **2.2 Bioassays against the erythrocytic stage**

322 Malaria Parasites

Fig. 1. Global antimalarial drug development pipeline (February 2009), after (Olliaro & Wells, 2009). Artemisinin (Art) derivatives or drugs containing the trioxane ring of

targets or mechanisms of action are in brown-red.

**2.1 Malaria parasite life cycle** 

**2. Development of bioassays for antimalarial activity** 

artemisinin are illustrated in blue; aminoquinolines and structurally related compounds as well as aryl alcohols are in green; antibiotics are in orange; others drugs having different

*P. falciparum* has a complex life cycle involving the *Anopheles* vector, where the sexual phase occurs, and humans where the parasite undergoes two phases of extensive asexual proliferation. After the bite by an infected *Anopheles*, the parasites at the sporozoite stage enter the blood and are carried to the liver where they traverse and invade hepatocytes initiating what is called the hepatic or exoerythrocytic phase. During this asymptomatic period, that lasts 5-7 days for *P. falciparum*, the sporozoites develop within the hepatocytes and, after several rounds of mitosis, produce several thousands of new infective forms called the merozoites that are released into the bloodstream and invade the red blood cells. During this intraerythrocytic cycle, which lasts 48 h for *P. falciparum* and causes the malaria disease, the parasite undergoes a successive development into the ring stage (0-20 h) and the trophozoite stage (20-36 h), which then undergoes several mitoses (schizont stage, 36-48 h) that lead to the differentiation of 16-24 infective merozoites. Erythrocyte lysis releases the merozoites into the bloodstream and a new intraerythrocytic cycle can be initiated. For some still not well understood reasons, some merozoites differentiate into male and female gametocytes that are taken up by the *Anopheles* during the blood meal. Gametogenesis resumes in the digestive tract of the mosquito and sexual reproduction occurs forming a zygote that produces, after a meiosis and several mitosis, thousands of haploid sporozoites under the basal lamina, which then migrate to the salivary glands and can be transmitted to The intraerythrocytic cycle, being responsible for the symptoms of the disease, is still the main parasite stage against which drugs are tested. Initially, drug screenings were limited to the use of animal malaria models (rodent, chicken or monkey). The development of the continuous culture of *P. falciparum* on human erythrocytes (Trager & Jensen, 1976) was a critical advance, allowing drug evaluation on well established laboratory strains and on fresh isolates from patients. Typically, parasites are maintained on leucocyte-free erythrocytes at 2- 5% haematocrit, in 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES)-buffered Roswell Park Memorial Institute (RPMI) medium and supplemented with 5-10% human serum, at 37 °C, under a reduced percentage of oxygen. Hypoxanthine can be added to the culture medium to stimulate parasite growth and to sustain high parasitemia, and it is recommended for the culture of fresh isolates. Standardized protocols have been proposed e.g. (in *Methods in Malaria Research*, http://www.mr4.org/Publications/tabid/326/Default.aspx).

*Plasmodium vivax* is the most widespread species and, except in equatorial Africa, is responsible for the most prevalent malaria infection of humans, causing 70 to 80 million clinical cases per year. Although *P. vivax* infections are rarely fatal, they remain an important cause of morbidity, particularly in Asia Pacific region. Compared to *P. falciparum*, *P. vivax* can be considered as a neglected disease. Resistance of *P. vivax* to the antimalarial drug chloroquine, the reference drug for treating *P. vivax* infection, has been reported since the 90's. It is therefore of importance to develop tools for monitoring drug resistance and developing new drugs. In contrast to *P. falciparum*, establishment of continuous cultured lines of *P. vivax* has not yet been achieved limiting drug evaluation, particularly high throughput screening. *P. vivax* does not easily grow in culture, requiring removal of leucocytes and enrichment of the growth media. Parasite growth can only be performed for short periods, but maintaining cultures up to 4 weeks can be obtained by supplying reticulocytes from normal blood (Udomsangpetch et al., 2008). Drug assays developed for *P. falciparum* are transposable to fresh and cryopreserved *P. vivax* isolates (Kosaisavee et al., 2006). It is generally assumed that drugs active against *P. falciparum* blood stages will be also active against *P. vivax* blood stages; this has been shown in clinical studies for dihydroartemisinin–piperaquine and for artesunate–pyronaridine (Olliaro & Wells, 2009). Such assumption and the technical constraints to study *P. vivax* explain the limited interest for this species. For the other human malaria parasites, *P. knowlesi* has been adapted to longterm culture on monkey erythrocytes (Kocken et al., 2002) but no continuous cultivation of *P. ovale* and *P. malariae* has been set up.

Standard protocols of drug and resistance evaluation on *P. falciparum* are recommended by the World Health Organisation to facilitate comparison of data. They generally involve evaluation by using Giemsa-stained smears and counting parasitemia or parasite stage distribution in treated and non-treated cultures. These assays require minimal equipment and can be easily applied in the field. However they are time-consuming thereby preventing rapid, large-scale screening of molecules. Several methods have been developed for screening large numbers of compounds in 96-well plates, or even in 384-well plates.

Advances in Antimalarial Drug Evaluation and New Targets for Antimalarials 325

worldwide. These fluorometric assays seem sufficiently sensitive to identify complex chemical mixtures with antiplasmodial activity such as plant extracts (Abiodun et al., 2010). Although several works observed similar results between the isotopic and fluorometric assays, users must have in mind that fluorescence intensity is highly dependent on its physico-chemical environment and that false positive responses due to fluorescence

 Flow cytometry is currently used to study the biology of malaria parasites, but is still little used to evaluate parasite drug susceptibility, certainly because of the high cost of cytometers and a capacity restricted to moderate throughput assays (Grimberg, 2011). However, owing to the large range of fluorescent probes available having different biological or biochemical affinities that can be simultaneously analysed, flow cytometry allows the access to many more information than just parasite proliferation (DNA content). It opens a new way to analyse drug susceptibility by integrating the DNA content and a deeper investigation of biological and biochemical effects of the drug on the parasite. Quantum dot (QD) is a CdSe/ZnS semiconductor crystal widely used for bioimaging applications and diagnostics because of its high photostability, large stokes shift and narrow emission spectra. In a recent work, authors searched for QD chemicals that could label *P. falciparum*-infected erythrocytes and found a polyethylenglycolated cationic QD (PCQD) that specifically label erythrocytes infected by the late stage parasites and not the uninfected erythrocytes and the erythrocytes infected by the early stage parasites (Ku et al., 2011). They used PCQD to quantify the antimalarial effect of chloroquine in 96-well plates by flow cytometry and found the assays as sensitive as the pLDH assay with a comparable cost and its performance enabled highthroughput assays. Innovative advances were also recorded for the development of assays using flow cytometry for evaluating drugs active on the gametocyte stage (see section 2.3).

Discrepancies in the activity of a same compound are often observed between laboratories. They can be due to the type of antimalarial assay used or to the *P. falciparum* strains used as well as their drug resistance status. However, data may also be influenced by laboratory-

• The use of asynchronized or synchronized cultures. *P. falciparum* has an asynchronous intraerythrocytic development *in vitro*. The synchronous development that is observed *in vivo* in humans can be restored by specific treatments taking into account the biological properties of the erythrocytes infected by the different parasite stages: e.g. 5% sorbitol treatment or gelatine floatation to specifically lyse or concentrate, respectively, the erythrocytes infected by the trophozoite and schizont stages. Assays on synchronized cultures of early stage parasites aim to be generalized because they reflect the parasite synchronisation observed in humans. Synchronous cultures allow further

• The duration of the drug incubation. This can vary from the classical 48 h assays that are widely used, to assays of 24 h or 72 h or of a longer period, especially for evaluating fast or slow-acting drugs. That is particularly illustrated by drugs having a delayeddeath effect such as the antibiotics tetracyclines and clindamycin. In the presence of clinically relevant concentrations of these antibiotics, parasites grow normally, divide and invade new erythrocytes. Inhibitory effects are only observed when parasites

studies of the drug effects on specific stages of the parasite development.

specific modifications of screening conditions. One can report:

quenching by a chemical compound of the tested sample are not occasional.

**2.2.4 Interlaboratory variations** 

#### **2.2.1 Incorporation of radiolabelled precursors**

These assays are based on the incorporation of metabolic precursors by the parasite that reflects its growth. They take advantage of the fact that the red blood cell possesses a reduced basic metabolism. Uptake of [3H]-hypoxanthine, a precursor for nucleic acids, was the first described (Desjardins et al., 1979) and is still the most widely used but other precursors, such as [3H]-amino acids or [3H]-ethanolamine have also been described. Users must keep in mind that addition of hypoxanthine to the medium to sustain parasite development will interfere with the sensitivity of the isotopic assay using [3H] hypoxanthine, considered as the gold standard antimalarial drug assay. Although very sensitive and reproducible, applicable for high-throughput screening, these assays have several disadvantages. They are costly, involve radioactivity with the constraints associated to the manipulation of radioactive compounds and the treatment of radioactive wastes, and require special equipment (cell harvester, liquid-scintillation counter). They can thus be problematic to set up in locations with poor resources.

### **2.2.2 Colorimetric assays**

Non-radioactive and low-cost alternative assays have been developed but are usually less standardized and often less sensitive than the [3H]-hypoxanthine uptake method. Colorimetric assays include the detection of *Plasmodium* lactate dehydrogenase (pLDH) activity (Makler et al., 1993), by its preferential metabolism of the substrate 3-acetylpyridine adenine dinucleotide which is weakly metabolized by the human LDH, or by the tetrazolium assay (Delhaes et al., 1999). The assays require accessible equipments (spectrophotometers) and good correlations were observed with the radioactive assays. Sensitive immunocapture assays based on monoclonal antibodies directed against pLDH were developed for drug screening (Druilhe et al., 2001, Makler et al., 1998) and were miniaturized for assays in 384-well plates (Bergmann-Leitner et al., 2008). pLDH detection was also the basis of rapid immunoassays for malaria diagnostic from whole blood using strips (OptiMal®) (Palmer et al., 1998). ELISA based on the quantification of the *Plasmodium* histidine-rich protein II (HRP2) was also developed and commercialized (Noedl et al., 2002). Currently, both assays are as sensitive as the isotopic assay. However, reagents are not widely available and their stability in field conditions may be questioned; also the genetic variations encountered within these parasite proteins may lead to a decrease of the assay sensitivity.

#### **2.2.3 Fluorometric and flow cytometry-based assays**

Taking advantage that erythrocytes have no or only remnant DNA/RNA, fluorescencebased assays were developed for high-throughput drug assays using DNA-binding fluorophores. Fluorescence correlates parasitemia. The currently used fluorophores are Hoechst, DAPI, SYBRGreen I, PICO green and YOYO, the two former being less used because they exhibit excitation and emission properties not appropriate for current fluorescence plate readers and consumables, which is not the case for the latter. Protocols have been optimized and they propose one-step assays applicable to high-throughput screening and as sensitive as the isotopic and the immunocapture assays on laboratory strains or clinical isolates (Bacon et al., 2007, Baniecki et al., 2007, Bennett et al., 2004). The assays are cost-effective, requiring only a spectrofluorometer, and dyes are readily available worldwide. These fluorometric assays seem sufficiently sensitive to identify complex chemical mixtures with antiplasmodial activity such as plant extracts (Abiodun et al., 2010). Although several works observed similar results between the isotopic and fluorometric assays, users must have in mind that fluorescence intensity is highly dependent on its physico-chemical environment and that false positive responses due to fluorescence quenching by a chemical compound of the tested sample are not occasional.

 Flow cytometry is currently used to study the biology of malaria parasites, but is still little used to evaluate parasite drug susceptibility, certainly because of the high cost of cytometers and a capacity restricted to moderate throughput assays (Grimberg, 2011). However, owing to the large range of fluorescent probes available having different biological or biochemical affinities that can be simultaneously analysed, flow cytometry allows the access to many more information than just parasite proliferation (DNA content). It opens a new way to analyse drug susceptibility by integrating the DNA content and a deeper investigation of biological and biochemical effects of the drug on the parasite. Quantum dot (QD) is a CdSe/ZnS semiconductor crystal widely used for bioimaging applications and diagnostics because of its high photostability, large stokes shift and narrow emission spectra. In a recent work, authors searched for QD chemicals that could label *P. falciparum*-infected erythrocytes and found a polyethylenglycolated cationic QD (PCQD) that specifically label erythrocytes infected by the late stage parasites and not the uninfected erythrocytes and the erythrocytes infected by the early stage parasites (Ku et al., 2011). They used PCQD to quantify the antimalarial effect of chloroquine in 96-well plates by flow cytometry and found the assays as sensitive as the pLDH assay with a comparable cost and its performance enabled highthroughput assays. Innovative advances were also recorded for the development of assays using flow cytometry for evaluating drugs active on the gametocyte stage (see section 2.3).

#### **2.2.4 Interlaboratory variations**

324 Malaria Parasites

These assays are based on the incorporation of metabolic precursors by the parasite that reflects its growth. They take advantage of the fact that the red blood cell possesses a reduced basic metabolism. Uptake of [3H]-hypoxanthine, a precursor for nucleic acids, was the first described (Desjardins et al., 1979) and is still the most widely used but other precursors, such as [3H]-amino acids or [3H]-ethanolamine have also been described. Users must keep in mind that addition of hypoxanthine to the medium to sustain parasite development will interfere with the sensitivity of the isotopic assay using [3H] hypoxanthine, considered as the gold standard antimalarial drug assay. Although very sensitive and reproducible, applicable for high-throughput screening, these assays have several disadvantages. They are costly, involve radioactivity with the constraints associated to the manipulation of radioactive compounds and the treatment of radioactive wastes, and require special equipment (cell harvester, liquid-scintillation counter). They can thus be

Non-radioactive and low-cost alternative assays have been developed but are usually less standardized and often less sensitive than the [3H]-hypoxanthine uptake method. Colorimetric assays include the detection of *Plasmodium* lactate dehydrogenase (pLDH) activity (Makler et al., 1993), by its preferential metabolism of the substrate 3-acetylpyridine adenine dinucleotide which is weakly metabolized by the human LDH, or by the tetrazolium assay (Delhaes et al., 1999). The assays require accessible equipments (spectrophotometers) and good correlations were observed with the radioactive assays. Sensitive immunocapture assays based on monoclonal antibodies directed against pLDH were developed for drug screening (Druilhe et al., 2001, Makler et al., 1998) and were miniaturized for assays in 384-well plates (Bergmann-Leitner et al., 2008). pLDH detection was also the basis of rapid immunoassays for malaria diagnostic from whole blood using strips (OptiMal®) (Palmer et al., 1998). ELISA based on the quantification of the *Plasmodium* histidine-rich protein II (HRP2) was also developed and commercialized (Noedl et al., 2002). Currently, both assays are as sensitive as the isotopic assay. However, reagents are not widely available and their stability in field conditions may be questioned; also the genetic variations encountered within these parasite proteins may lead to a decrease of the assay

Taking advantage that erythrocytes have no or only remnant DNA/RNA, fluorescencebased assays were developed for high-throughput drug assays using DNA-binding fluorophores. Fluorescence correlates parasitemia. The currently used fluorophores are Hoechst, DAPI, SYBRGreen I, PICO green and YOYO, the two former being less used because they exhibit excitation and emission properties not appropriate for current fluorescence plate readers and consumables, which is not the case for the latter. Protocols have been optimized and they propose one-step assays applicable to high-throughput screening and as sensitive as the isotopic and the immunocapture assays on laboratory strains or clinical isolates (Bacon et al., 2007, Baniecki et al., 2007, Bennett et al., 2004). The assays are cost-effective, requiring only a spectrofluorometer, and dyes are readily available

**2.2.1 Incorporation of radiolabelled precursors** 

problematic to set up in locations with poor resources.

**2.2.3 Fluorometric and flow cytometry-based assays** 

**2.2.2 Colorimetric assays** 

sensitivity.

Discrepancies in the activity of a same compound are often observed between laboratories. They can be due to the type of antimalarial assay used or to the *P. falciparum* strains used as well as their drug resistance status. However, data may also be influenced by laboratoryspecific modifications of screening conditions. One can report:


Advances in Antimalarial Drug Evaluation and New Targets for Antimalarials 327

obtain large amounts of gametocytes have limited research in this area. Gametocytes have a quite long maturation period *in vitro* (more than 10 days), gametocytogenesis is stimulated by stress conditions and sexual subpopulation represents a small percentage of infected red blood cells. Furthermore, *Plasmodium* strains have a tendency to lose their capacity to produce gametocytes over time in culture and gametocyte-infected erythrocytes are difficult to separate from erythrocytes infected by asexual stages. Until recently, there had been few significant advances in techniques for producing gametocytes for drug screening. These techniques were difficult to set up and required costly equipments. New protocols have now improved our capability to produce gametocytes *in vitro* in a reproducible way, and to enrich gametocytes with a high degree of purity using MACS® magnetic affinity columns (Baker, 2007). Those improvements associated to the ability to genetically manipulate parasites have allowed the recent setting-up of protocols for evaluating drug effects on sexual conversion and gametocyte maturation with high-throughput capacities. They are based on : 1) the discrimination by flow cytometry of asexual and sexual forms using hydroethine that is taken up by the parasite and metabolized into ethidium, a nucleic acidbinding fluorochrome (Chevalley et al., 2010); 2) the use of transgenic *P. falciparum* parasites expressing a green fluorescent protein chimera of the early sexual blood stage (protein Pfs16) as a marker for commitment to gametocytogenesis; this marker associated to hydroethine allows also to measure the direct activity of drugs against the late-stages gametocytes (Peatey et al., 2009). In a same way, the stage II or later stage marker (PF10\_0164) fused to the green fluorescent protein was used associated with the nuclear dye Hoescht 33342 to quantify the drug effects on the asexual stages and on the sexual conversion and the gametocyte maturation in a same assay (Buchholz et al., 2011). In a general way, the application of transfection technology to malaria parasites paves the way to

a new generation of assays targeting specific pathways or parasite stages.

Drug development against the liver stage has two main advantages: 1) Such drugs could be used as prophylactic agents by preventing the access of the parasite to the blood and, in consequence, the development of the disease; 2) *P. vivax* and *P. ovale*, unlike *P. falciparum*, have latent forms (hypnozoites) in the liver responsible for relapsing infections, sometimes years after the primary infection. Hypnozoites are unaffected by the drugs acting on the erythrocytic stage. Thus, anti-hypnozoite drugs could be used as anti-relapse agents,

The number of available drugs that act against the hepatic phase is limited. The 8 aminoquinolines (primaquine and tafenoquine) are the main and the most efficient ones, but have serious haematological toxicity as already mentioned. Evaluation of drugs against the hepatic stage has been mainly performed using *in vivo* models of rodent parasites e.g., *P. berghei* and *P. yoelii*. However, such drug screening is time consuming and costly due to the need to reproduce the complete life-cycle of the parasite to obtain viable sporozoites. Difficulties increase greatly if screening is performed using monkey models for human malaria parasites. The ability to culture *Plasmodium* in hepatocytes has opened new perspectives in the research of drugs active against the hepatic stage (Mazier et al., 1985, Mazier et al., 2004). *In vitro* screenings are usually performed using sporozoites of rodent *Plasmodium* infecting primary rodent hepatocytes or hepatoma cell lines but assays based on

**2.4 Bioassays for the hepatic stage of** *Plasmodium*

limiting then recrudescences.

develop during the second erythrocytic cycle. This delayed-death effect has been attributed to inhibition of mitochondrion or apicoplast targets (Dahl & Rosenthal, 2008). Activity of such molecules would be underestimated using synchronized cultures and assays of less than 48 h.


#### **2.3 Bioassays against the gametocyte stages**

The emergence of resistance has highlighted the importance to develop drugs against parasite transmission (from human to *Anopheles* and from *Anopheles* to human) that has been recognized to be a priority in the efforts to eradicate malaria. Sexual stages (gametocytes) are being investigated as targets for vaccines but few drugs have been developed against these transmission stages. The only available compounds known to efficiently eliminate gametocytes are 8-aminoquinolines (primaquine and tafenoquine), but their use is restricted due to serious haematological toxicity. They cause haemolysis in patients with G6PD deficiency and, in cases of severe forms of G6PD deficiency, the risks of primaquine treatment might exceed the benefits (Beutler & Duparc, 2007). Difficulties to culture and to

• The use of human serum or serum substitutes. Variability of quality of human serum batches as well as the difficulty to have access to non-immune human sera in endemic malaria countries can be serious problems for data reproducibility and has led to the evaluation of numerous serum substitutes or sera of other species to support *P. falciparum* growth in culture (Basco, 2003). Albumax® was proposed as a substitute for human serum and is now used routinely by many laboratories. Two types exist, Albumax® I and II, both are used, depending upon the authors, generally at a concentration of 5 g/l. Their composition is not well characterized, as well as the biochemical differences between both types. They are mainly composed of chromatographically purified fatty acid-rich bovine serum albumin and have a low IgG content. In addition to their high cost, batch-to-batch variability were observed in the past and *Plasmodium* strains must be adapted to culture with Albumax® (Grellier P., personal communication). This latter point can be problematic for evaluating drug resistance of fresh isolates, which have been reported to grow poorly in the presence of Albumax® (Basco, 2004). Furthermore, contrasted results were reported when comparing antimalarial drug efficacy in assays containing Albumax® or human serum. Some authors observed differences in the IC50 of antimalarials (Ringwald et al., 1999), others observed no difference (Singh et al., 2007). In our hands and in our search of natural products with antimalarial activity, differences of IC50 may be important between both conditions for some chemicals. This can be due in part to the differences in lipid and protein content of the two culture media and to the binding properties of the antimalarial that may influence the pharmacodynamic profile of the drug, as observed for halofantrine, a highly lipophilic drug which significantly associates to

triglyceride rich plasma lipoproteins (Humberstone et al., 1998).

at 21% O2 (Briolant et al., 2007).

**2.3 Bioassays against the gametocyte stages** 

• The culture conditions. Initial parasitemia, haematocrit, as well as atmosphere under which cultures are maintained (e.g., a 5% CO2 atmosphere versus a well-defined atmosphere such as 6% O2, 3% CO2, 91% N2) are all factors that may be the cause of discrepancies. For example, it has been documented that IC50 values for chloroquine depend highly on the oxygen pressure during the assay. A *P. falciparum* strain may display chloroquine-resistance (IC50 >100nM) at 10% O2 and behave as a sensitive strain

The emergence of resistance has highlighted the importance to develop drugs against parasite transmission (from human to *Anopheles* and from *Anopheles* to human) that has been recognized to be a priority in the efforts to eradicate malaria. Sexual stages (gametocytes) are being investigated as targets for vaccines but few drugs have been developed against these transmission stages. The only available compounds known to efficiently eliminate gametocytes are 8-aminoquinolines (primaquine and tafenoquine), but their use is restricted due to serious haematological toxicity. They cause haemolysis in patients with G6PD deficiency and, in cases of severe forms of G6PD deficiency, the risks of primaquine treatment might exceed the benefits (Beutler & Duparc, 2007). Difficulties to culture and to

assays of less than 48 h.

develop during the second erythrocytic cycle. This delayed-death effect has been attributed to inhibition of mitochondrion or apicoplast targets (Dahl & Rosenthal, 2008). Activity of such molecules would be underestimated using synchronized cultures and obtain large amounts of gametocytes have limited research in this area. Gametocytes have a quite long maturation period *in vitro* (more than 10 days), gametocytogenesis is stimulated by stress conditions and sexual subpopulation represents a small percentage of infected red blood cells. Furthermore, *Plasmodium* strains have a tendency to lose their capacity to produce gametocytes over time in culture and gametocyte-infected erythrocytes are difficult to separate from erythrocytes infected by asexual stages. Until recently, there had been few significant advances in techniques for producing gametocytes for drug screening. These techniques were difficult to set up and required costly equipments. New protocols have now improved our capability to produce gametocytes *in vitro* in a reproducible way, and to enrich gametocytes with a high degree of purity using MACS® magnetic affinity columns (Baker, 2007). Those improvements associated to the ability to genetically manipulate parasites have allowed the recent setting-up of protocols for evaluating drug effects on sexual conversion and gametocyte maturation with high-throughput capacities. They are based on : 1) the discrimination by flow cytometry of asexual and sexual forms using hydroethine that is taken up by the parasite and metabolized into ethidium, a nucleic acidbinding fluorochrome (Chevalley et al., 2010); 2) the use of transgenic *P. falciparum* parasites expressing a green fluorescent protein chimera of the early sexual blood stage (protein Pfs16) as a marker for commitment to gametocytogenesis; this marker associated to hydroethine allows also to measure the direct activity of drugs against the late-stages gametocytes (Peatey et al., 2009). In a same way, the stage II or later stage marker (PF10\_0164) fused to the green fluorescent protein was used associated with the nuclear dye Hoescht 33342 to quantify the drug effects on the asexual stages and on the sexual conversion and the gametocyte maturation in a same assay (Buchholz et al., 2011). In a general way, the application of transfection technology to malaria parasites paves the way to a new generation of assays targeting specific pathways or parasite stages.

#### **2.4 Bioassays for the hepatic stage of** *Plasmodium*

Drug development against the liver stage has two main advantages: 1) Such drugs could be used as prophylactic agents by preventing the access of the parasite to the blood and, in consequence, the development of the disease; 2) *P. vivax* and *P. ovale*, unlike *P. falciparum*, have latent forms (hypnozoites) in the liver responsible for relapsing infections, sometimes years after the primary infection. Hypnozoites are unaffected by the drugs acting on the erythrocytic stage. Thus, anti-hypnozoite drugs could be used as anti-relapse agents, limiting then recrudescences.

The number of available drugs that act against the hepatic phase is limited. The 8 aminoquinolines (primaquine and tafenoquine) are the main and the most efficient ones, but have serious haematological toxicity as already mentioned. Evaluation of drugs against the hepatic stage has been mainly performed using *in vivo* models of rodent parasites e.g., *P. berghei* and *P. yoelii*. However, such drug screening is time consuming and costly due to the need to reproduce the complete life-cycle of the parasite to obtain viable sporozoites. Difficulties increase greatly if screening is performed using monkey models for human malaria parasites. The ability to culture *Plasmodium* in hepatocytes has opened new perspectives in the research of drugs active against the hepatic stage (Mazier et al., 1985, Mazier et al., 2004). *In vitro* screenings are usually performed using sporozoites of rodent *Plasmodium* infecting primary rodent hepatocytes or hepatoma cell lines but assays based on

Advances in Antimalarial Drug Evaluation and New Targets for Antimalarials 329

development of the life-cycle of *Plasmodium*. Many putative targets, initially suspected to be essential to a given parasite stage turned out finally dispensable to this stage but essential to others (see few examples such as FabI or falcipain-1 below). Furthermore, access to transcriptomic and proteomic technologies offers new opportunities to study the impact of a drug treatment on the entire parasite metabolism either at the RNA or protein levels. This leads to a better understanding of the mechanism(s) of action of drugs and the biochemical pathways involved in killing the parasites (Sims & Hyde, 2006). Recently, the achievement of random mutagenesis of *P. berghei* and *P. falciparum* genome by PiggyBac transposable element opened the route towards systematic plasmodial gene invalidation, that will obviously speed up the genetic validation process of putative targets (Crabb et al., 2011).

These recent advances have allowed development of bioassays based upon validated targets for drug screening, or targets still in the process of validation (for review see (Grellier et al., 2008, Prabhu & Patravale, 2011, Sahu et al., 2008)) : haem polymerization (O'Neill et al., 2006), pyrimidine, purine, folate (Hyde, 2007), lipid (Wengelnik et al., 2002), shikimate (McRobert et al., 2005), non-mevalonate (Wiesner & Jomaa, 2007) and other apicoplast metabolisms (Sato & Wilson, 2005), mitochondrial electron transport (Mather et al., 2007), redox homeostasis (Bauer et al., 2006), protein prenylation (Van Voorhis et al., 2007), proteases (Wegscheid-Gerlach et al., 2010), kinases (Doerig & Meijer, 2007)... Some of them

Databases such as TDR targets (http://tdrtargets.org) gather information on putative targets for several pathogens including *Plasmodium*, providing tools for their prioritization in whole genomes depending on user queries (Aguero et al., 2008). The Protein Data Bank (www.pdb.org) is also a useful resource, providing data on proteins whose 3D-structures have been solved experimentally by using either X-Rays or NMR. As of September 2011, more than 300 entries correspond to plasmodial proteins, knowing that several structures

Haeme polymerization (Figure 2) is a natural process that occurs in the acidic food vacuole of the parasite where haemoglobin, a major source of amino acids for the parasite, is degraded by specific proteases. Haeme, which is toxic for the parasite, is detoxified by polymerization into a pigment, the haemozoin. Inhibition of haemozoin formation is considered as an attractive target for antimalarial drugs and high-throughput assays are available (Biagini et al., 2003, Deharo et al., 2002, Huy et al., 2007, Ncokazi & Egan, 2005, Rush et al., 2009, Weissbuch & Leiserowitz, 2008). The 4-aminoquinolines such as chloroquine and amodiaquine act by interfering with this polymerization process. Cysteine and aspartic parasite proteases (falcipains and plasmepsins, respectively) involved in the degradation of haemoglobin are also potential targets (Wegscheid-Gerlach et al., 2010) (see

Although none of the currently marketed antimalarials is targeting plasmodial proteases, this class of enzymes, which is involved in a wide diversity of biological pathways during the parasite life cycle, has been the subject of intense investigations for the last decades

involving different ligands may have been solved for a same protein.

**2.5.1 Inhibition of haeme polymerization** 

are detailed below.

section 2.5.2).

**2.5.2 Proteases** 

primary culture of human hepatocytes infected with *P. falciparum* sporozoites were also reported. Recently, assay on liver stage of *P. vivax* was established using purified and cryopreserved sporozoites opening new perspectives for development of anti-relapse drugs (Chattopadhyay et al., 2010). The sporozoite development within the hepatocyte into a schizont stage is followed either by Giemsa staining or immunofluorescence assay using specific antibodies or a genetically-manipulated fluorescent parasite. Potential antimalarial drugs acting on the hepatic stage have emerged (Carraz et al., 2006, Mahmoudi et al., 2003, Mahmoudi et al., 2008, Parvanova et al., 2009, Singh et al., 2010, Tasdemir et al., 2010, Yu et al., 2008). A high-throughput *in vitro* screening of drug activity on *Plasmodium* liver stages was developed based on a sophisticated infrared fluorescence scanning system, which allows rapid, automatic counting of infected hepatocytes (Gego et al., 2006). The recent development of bioluminescent parasites allows now a non-invasive real time monitoring and quantitative analysis of liver stage development *in vitro* and *in vivo* in rodents that offers new tools for drug evaluation (Mwakingwe et al., 2009, Ploemen et al., 2009) (see section 4.3.2). Automated visual assay was also set up to follow the extracellular cell death of sporozoites (Hegge et al., 2010). Although promising, such assays are however greatly dependent upon the production of a large number of sporozoites in insects, a limiting step for high-throughput screening.

#### **2.5 Bioassays for parasite targets**

Our understanding of malaria parasite biochemistry has considerably increased over the past two decades and has allowed the identification of many potential targets for new drugs, even if half of the *P. falciparum* genes are still in search of a biological function (Florent et al., 2010). Progresses have been made possible thanks to the decrypted genomes of several *Plasmodium* species and other Apicomplexans (see www.eupathdb.org), that have enabled the rapid identification of putative targets that are homologous to validated targets in other organisms. Usually, target validation in *Plasmodium* requires demonstrating that a specific inhibitor, designed or identified as being active against this particular target, indeed shows antimalarial activity *in vitro* and/or *in vivo*. However, such an approach called "chemical-validation" has limits since it is often difficult to demonstrate that the inhibitor kills the parasite by indeed acting specifically on this target and not by inhibiting unrelated biochemical mechanisms. Thus, nowadays, "genetic-validation" of the selected target must also be achieved in parallel i.e., the deletion of the corresponding gene must seriously impair the parasite's growth or ideally prevent its survival.

Recent advances in genetic understanding of *Plasmodium* have greatly increased the ability to genetically validate potential drug targets (Limenitakis & Soldati-Favre, 2011). Methodologies able to transfect rodent *Plasmodium* and *P. falciparum* so as to modify or knock out genes that code for potential target proteins have been available for over a decade. They have been progressively improved and refined to now permit conditional knock out of the gene at a chosen parasite stage that may be different from the transfected stage (Lacroix et al., 2011), conditional elimination of the gene product by fusing a degradation domain and Shield ligand (Dvorin et al., 2010) or compensation of a lethal phenotype resulting from knock out of an essential gene by expression in trans of a wildtype version of this gene (Slavic et al., 2010). Phenotypic analysis of the resultant mutant parasites then allows to study the importance of a particular target in the different phases of

primary culture of human hepatocytes infected with *P. falciparum* sporozoites were also reported. Recently, assay on liver stage of *P. vivax* was established using purified and cryopreserved sporozoites opening new perspectives for development of anti-relapse drugs (Chattopadhyay et al., 2010). The sporozoite development within the hepatocyte into a schizont stage is followed either by Giemsa staining or immunofluorescence assay using specific antibodies or a genetically-manipulated fluorescent parasite. Potential antimalarial drugs acting on the hepatic stage have emerged (Carraz et al., 2006, Mahmoudi et al., 2003, Mahmoudi et al., 2008, Parvanova et al., 2009, Singh et al., 2010, Tasdemir et al., 2010, Yu et al., 2008). A high-throughput *in vitro* screening of drug activity on *Plasmodium* liver stages was developed based on a sophisticated infrared fluorescence scanning system, which allows rapid, automatic counting of infected hepatocytes (Gego et al., 2006). The recent development of bioluminescent parasites allows now a non-invasive real time monitoring and quantitative analysis of liver stage development *in vitro* and *in vivo* in rodents that offers new tools for drug evaluation (Mwakingwe et al., 2009, Ploemen et al., 2009) (see section 4.3.2). Automated visual assay was also set up to follow the extracellular cell death of sporozoites (Hegge et al., 2010). Although promising, such assays are however greatly dependent upon the production of a large number of sporozoites in insects, a limiting step

Our understanding of malaria parasite biochemistry has considerably increased over the past two decades and has allowed the identification of many potential targets for new drugs, even if half of the *P. falciparum* genes are still in search of a biological function (Florent et al., 2010). Progresses have been made possible thanks to the decrypted genomes of several *Plasmodium* species and other Apicomplexans (see www.eupathdb.org), that have enabled the rapid identification of putative targets that are homologous to validated targets in other organisms. Usually, target validation in *Plasmodium* requires demonstrating that a specific inhibitor, designed or identified as being active against this particular target, indeed shows antimalarial activity *in vitro* and/or *in vivo*. However, such an approach called "chemical-validation" has limits since it is often difficult to demonstrate that the inhibitor kills the parasite by indeed acting specifically on this target and not by inhibiting unrelated biochemical mechanisms. Thus, nowadays, "genetic-validation" of the selected target must also be achieved in parallel i.e., the deletion of the corresponding gene must seriously

Recent advances in genetic understanding of *Plasmodium* have greatly increased the ability to genetically validate potential drug targets (Limenitakis & Soldati-Favre, 2011). Methodologies able to transfect rodent *Plasmodium* and *P. falciparum* so as to modify or knock out genes that code for potential target proteins have been available for over a decade. They have been progressively improved and refined to now permit conditional knock out of the gene at a chosen parasite stage that may be different from the transfected stage (Lacroix et al., 2011), conditional elimination of the gene product by fusing a degradation domain and Shield ligand (Dvorin et al., 2010) or compensation of a lethal phenotype resulting from knock out of an essential gene by expression in trans of a wildtype version of this gene (Slavic et al., 2010). Phenotypic analysis of the resultant mutant parasites then allows to study the importance of a particular target in the different phases of

for high-throughput screening.

**2.5 Bioassays for parasite targets** 

impair the parasite's growth or ideally prevent its survival.

development of the life-cycle of *Plasmodium*. Many putative targets, initially suspected to be essential to a given parasite stage turned out finally dispensable to this stage but essential to others (see few examples such as FabI or falcipain-1 below). Furthermore, access to transcriptomic and proteomic technologies offers new opportunities to study the impact of a drug treatment on the entire parasite metabolism either at the RNA or protein levels. This leads to a better understanding of the mechanism(s) of action of drugs and the biochemical pathways involved in killing the parasites (Sims & Hyde, 2006). Recently, the achievement of random mutagenesis of *P. berghei* and *P. falciparum* genome by PiggyBac transposable element opened the route towards systematic plasmodial gene invalidation, that will obviously speed up the genetic validation process of putative targets (Crabb et al., 2011).

These recent advances have allowed development of bioassays based upon validated targets for drug screening, or targets still in the process of validation (for review see (Grellier et al., 2008, Prabhu & Patravale, 2011, Sahu et al., 2008)) : haem polymerization (O'Neill et al., 2006), pyrimidine, purine, folate (Hyde, 2007), lipid (Wengelnik et al., 2002), shikimate (McRobert et al., 2005), non-mevalonate (Wiesner & Jomaa, 2007) and other apicoplast metabolisms (Sato & Wilson, 2005), mitochondrial electron transport (Mather et al., 2007), redox homeostasis (Bauer et al., 2006), protein prenylation (Van Voorhis et al., 2007), proteases (Wegscheid-Gerlach et al., 2010), kinases (Doerig & Meijer, 2007)... Some of them are detailed below.

Databases such as TDR targets (http://tdrtargets.org) gather information on putative targets for several pathogens including *Plasmodium*, providing tools for their prioritization in whole genomes depending on user queries (Aguero et al., 2008). The Protein Data Bank (www.pdb.org) is also a useful resource, providing data on proteins whose 3D-structures have been solved experimentally by using either X-Rays or NMR. As of September 2011, more than 300 entries correspond to plasmodial proteins, knowing that several structures involving different ligands may have been solved for a same protein.

## **2.5.1 Inhibition of haeme polymerization**

Haeme polymerization (Figure 2) is a natural process that occurs in the acidic food vacuole of the parasite where haemoglobin, a major source of amino acids for the parasite, is degraded by specific proteases. Haeme, which is toxic for the parasite, is detoxified by polymerization into a pigment, the haemozoin. Inhibition of haemozoin formation is considered as an attractive target for antimalarial drugs and high-throughput assays are available (Biagini et al., 2003, Deharo et al., 2002, Huy et al., 2007, Ncokazi & Egan, 2005, Rush et al., 2009, Weissbuch & Leiserowitz, 2008). The 4-aminoquinolines such as chloroquine and amodiaquine act by interfering with this polymerization process. Cysteine and aspartic parasite proteases (falcipains and plasmepsins, respectively) involved in the degradation of haemoglobin are also potential targets (Wegscheid-Gerlach et al., 2010) (see section 2.5.2).

### **2.5.2 Proteases**

Although none of the currently marketed antimalarials is targeting plasmodial proteases, this class of enzymes, which is involved in a wide diversity of biological pathways during the parasite life cycle, has been the subject of intense investigations for the last decades

Advances in Antimalarial Drug Evaluation and New Targets for Antimalarials 331

dual activity, both able to process peptides in the food vacuole and to mature proteins imported to the apicoplast (Eggleson et al., 1999, Ponpuak et al., 2007), and the metalloaminopeptidases PfA-M1 and PfA-M17 (McGowan et al., 2009, Skinner-Adams et al., 2009, Trenholme et al., 2010). Some have emerged from studies focusing on other key biological events such as erythrocyte egress/invasion and merozoite surface antigen maturation such as the subtilisin-like proteases -1 and -2 and the cystein proteases DPAP3, SERA-5 and SERA-6 (Blackman, 2008). Other proteases for which essential roles during the parasite asexual development have been demonstrated include Plasmepsin V, involved in maturation of proteins exported to the infected-red blood cell (Boddey et al., 2010, Russo et al., 2010). During these investigations, some proteases were found dispensable for the parasite asexual development in erythrocyte but important in other stages such as gametogenesis or the development of insect stages (Sologub et al., 2011). This is the case for the cystein proteases falcipain-1 and SERA-8. Gene deletion assays were determinant to indicate that falcipain-1, initially believed to be involved in haemoglobin breakdown then in red blood cell invasion, was in fact dispensable for the parasite blood-stage development but important for oocyst production (Eksi et al., 2004). Also, the *P. berghei* ortholog of SERA-

8 appears involved in sporozoite egress from oocysts (Aly & Matuschewski, 2005).

*Plasmodium* growth in culture and *in vivo* (Seeber & Soldati-Favre, 2010).

These results stimulated efforts both to genetically validate these putative targets and to improve inhibitor discovery, by solving 3D structures and producing enzymes under recombinant active forms amenable to medium/high throughput screenings (Freundlich et al., 2007, Sato, 2011). In 2008, the genetic inactivation of FabI in *P. falciparum* and in *P. berghei* produced parasite blood stages that were growing normally and were still affected by triclosan (Yu et al., 2008). The FabI gene deletion, on the other hand, blocked the development of the parasite in the liver. Not long later, the genetic inactivation in murine

The apicoplast is a vestigial, non-phostosynthetic, plastid-like organelle inherited from the prokaryotic world by secondary endosymbiosis and found in most apicomplexan parasites including *Plasmodium* that is essential to their biology. In *P. falciparum*, it maintains a 35-kb circular genome and several particular biochemical pathways that are present in bacteria and plants but are absent in humans, thus providing many attractive targets that are extensively investigated for drug development. These pathways include the type II fatty acid biosynthesis pathway, which involves 6 distinct enzymes in *Plasmodium* while in human the type I fatty acid biosynthesis pathway involves a multifunctional enzyme, the 1 deoxy D xylulose 5 phosphate (DOXP) isoprenoid biosynthesis pathway that is mevalonateindependent in the malaria parasite contrary to humans, and apicoplast replication, transcription and translation which involve enzymes of bacterial origins (Dahl & Rosenthal, 2008, Goodman & McFadden, 2007, Grawert et al., 2011, Jayabalasingham et al., 2010). Pioneering works led to the emergence of promising antimalarials such as triclosan (believed to target the NADH-dependent enoyl ACP reductase or FabI enzyme), thiolactomycin (targeting FabH and FabB enzymes) and fosmidomycin (targeting the DOXP reductoisomerase) to name the main ones. Also, antibiotics such as tetracyclines and clindamycin, targeting prokaryotic protein synthesis, or quinolone antibiotics and rifampicin, targeting the prokaryotic DNA and RNA machinery, were shown to inhibit

**2.5.3 Apicoplast-based targets** 

(Wegscheid-Gerlach et al., 2010). The cystein and aspartyl endoproteases involved in the essential pathway of haemoglobin degradation now known as falcipain-2, 2' and -3 and plasmepsin-I, -II, -III (or HAP, for histo-aspartyl protease) and –IV have first emerged as promising protease targets. Intensive research focused on these enzymes during the last decades has yielded inhibitors active down to the nM levels on the native or recombinant enzymes, inhibiting parasite growth in culture from the nM to µM levels and capable to cure *Plasmodium*-infected mice (McKerrow et al., 2008, Rosenthal, 2010). Gene disruption studies however clearly indicated that all of these enzymes besides falcipain-3 are highly redundant. In fact, falcipain-3 is the only for which gene disruption did not yield viable blood-stage parasites (Liu et al., 2006, Sijwali et al., 2006). The development of inhibitors targeting these enzymes but displaying improved properties in term of activity, selectivity, safety to the host nonetheless pertains, facilitated by their availability under recombinant active form, allowing both high-throughput screenings and experimental determination of their 3D-structures, including for the most recently described plasmepsin I (Bhaumik et al., 2011). A strong argument for maintaining such a development against these haemoglobinases is linked to the proposal to use combined protease inhibitor therapy based on the observation that cystein and aspartyl inhibitors demonstrated synergistic antimalarial effects (Semenov et al., 1998).

Fig. 2. Mechanisms of haemoglobin degradation and haeme detoxification in *Plasmodium falciparum*-infected red blood cell, after (Mambu & Grellier, 2008)

In parallel, the last decade has seen the emergence as promising targets of several new parasite proteases, among the ~100 that are encoded in the *P. falciparum* genome (Wu et al., 2003). Some are also involved in haemoglobin breakdown such as the cystein protease DPAP1 (Klemba et al., 2004), the metallo-endopeptidase falcilysin displaying however a dual activity, both able to process peptides in the food vacuole and to mature proteins imported to the apicoplast (Eggleson et al., 1999, Ponpuak et al., 2007), and the metalloaminopeptidases PfA-M1 and PfA-M17 (McGowan et al., 2009, Skinner-Adams et al., 2009, Trenholme et al., 2010). Some have emerged from studies focusing on other key biological events such as erythrocyte egress/invasion and merozoite surface antigen maturation such as the subtilisin-like proteases -1 and -2 and the cystein proteases DPAP3, SERA-5 and SERA-6 (Blackman, 2008). Other proteases for which essential roles during the parasite asexual development have been demonstrated include Plasmepsin V, involved in maturation of proteins exported to the infected-red blood cell (Boddey et al., 2010, Russo et al., 2010). During these investigations, some proteases were found dispensable for the parasite asexual development in erythrocyte but important in other stages such as gametogenesis or the development of insect stages (Sologub et al., 2011). This is the case for the cystein proteases falcipain-1 and SERA-8. Gene deletion assays were determinant to indicate that falcipain-1, initially believed to be involved in haemoglobin breakdown then in red blood cell invasion, was in fact dispensable for the parasite blood-stage development but important for oocyst production (Eksi et al., 2004). Also, the *P. berghei* ortholog of SERA-8 appears involved in sporozoite egress from oocysts (Aly & Matuschewski, 2005).
