Large-scale in vitro cultivation of Plasmodium falciparum under atmospheric air

Background: The most important tool for the study of the malaria parasite is probably the in vitro culture that was established in 1976 by Trager and Jensen. However, the methodology is technically complex, expensive, and unsuitable for use in laboratories and regions with limited resources. Human factors such as hormones and the immune system are frequently ignored, and the similarity between the in vitro culture conditions and physiologic-like environment has been questioned. One of the most controversial factors about Plasmodium culture is the gaseous composition employed in vitro, which consists of a mixture poor in O 2 (1%–5%) and rich in CO 2 (approximately 5%). Results: The requirement of employing commercial gaseous mixtures and their suitability for use in reusable glass bottles under agitation for culturing were assessed. The methodology presented here is suitable for small-and large-scale culturing in cell culture �asks or hermetic and sterile glass bottles in agitation and provides an easier and cheaper Plasmodium culture alternative, which dispenses several sophisticated technical requirements of the classical methodology. The parasite viability, free oxygen, and drug-screening assays under these conditions were compared with those in the classic conditions. The cultivation under atmospheric air did not substantially affect the free oxygen levels and parasitic proliferation for periods longer than two years. However, it changed atovaquone, artesunate, and chloroquine e�cacy in a different ways across the different isolates. Oxygraphic experiments demonstrated that the free oxygen concentration contained in the media was slightly superior compared to both the human blood and the media employed the classical methodology. Conclusions: The results indicate that the described culture conditions are suitable for parasite maintenance and can be used to obtain high yields of parasite. Although parasites proliferate under various culture conditions, it is clear that it is necessary to evaluate the different gaseous conditions to validate an antimalarial candidate. With this, it will be possible to detect an oxygen level that encompasses the physiological oxygen levels of human blood. Since oxygen decreases through media layers, future studies should further explore the use of oxygen-rich mixtures as a strategy to expose static cultures to more physiologic-like conditions.

The cultivation under atmospheric air did not substantially affect the free oxygen levels and parasitic proliferation for periods longer than two years. However, it changed atovaquone, artesunate, and chloroquine e cacy in a different ways across the different isolates. Oxygraphic experiments demonstrated that the free oxygen concentration contained in the media was slightly superior compared to both the human blood and the media employed the classical methodology.
Conclusions: The results indicate that the described culture conditions are suitable for parasite maintenance and can be used to obtain high yields of parasite. Although parasites proliferate under various culture conditions, it is clear that it is necessary to evaluate the different gaseous conditions to validate an antimalarial candidate. With this, it will be possible to detect an oxygen level that encompasses the physiological oxygen levels of human blood. Since oxygen decreases through media layers, future studies should further explore the use of oxygen-rich mixtures as a strategy to expose static cultures to more physiologic-like conditions. Background Malaria is a prevalent parasitic disease in Africa, Southeast Asia and South America, affecting 216 million people in 2018 and has caused approximately 445 000 deaths mostly due Plasmodium falciparum infections in Africa [1]. Unfortunately, malaria is endemic in several undeveloped regions where there are few economic resources for treatment and epidemiologic control [2,3]. Moreover, malaria control presents other challenges including the major issue of drug-resistance [1,4]. That's why the discovery of new antimalarial compounds, a better understand of the parasite's metabolism, and assessment of the drug-resistance spreading is required [1,4]. In this vein, interdisciplinary efforts are need to ght against malaria at a global scale.
Central to the study of infectious diseases is the in vitro culturing of each etiologic agent. Brie y, it means to cultivate pathogens ex vivo as similarly as possible to their natural environments to increase the biological relevancy of the results. The Plasmodium culture was established in 1976 by Trager and Jensen, allowing the creating of an arti cial, continuous intraerythrocytic cycle through the infection of red blood cells by the cultured parasites [5]. This method is one of the most widely-used tools for the study of parasites [5] such as epidemiologic studies focused on isolating eld isolates [6] for in vitro drug screening the study of the parasitic metabolism and for studies requiring large amounts of parasites.
The Trager and Jensen culture methodology was based on the use of a candle jar and/or cell asks containing human red blood cells suspended in culture medium and a speci c gaseous mixture rich in CO 2 and poor in O 2 [5]. These gaseous conditions were obtained by direct gas injection or by candle combustion [5]. The culture medium used was composed of RPMI medium 1640 (Thermo Fisher Scienti c, Waltham, Massachusetts, U.S.) complemented with 0.1 g/L p, 2 g/L glucose, 0.04 g/L gentamicin, and 5 g/L of Albumax I® (Thermo Fisher Scienti c, Waltham, Massachusetts, U.S.). Some authors have attempted to optimize the primogenital culture methodology: the most important modi cations were the substitution of human plasma for commercial Lipid-Rich Bovine serum albumin [7] and the parasitic stage synchronization by the D-sorbitol techniques [8,9]. Preechapornkul et al in 2010, used bioreactors under constant agitation and controlled atmosphere to optimize the Plasmodium culture [10]. Moreover, Radfar et al combined new techniques and standardized very intricate protocols to scale-up the culturing for large scale parasite recovery [11]. Unfortunately, the last two studies describe culturing methodology that is not suitable for research groups in areas lacking technical and economic resources, a frequent situation in areas endemic for malaria.
An important requirement for in vitro culture is the ability to mimic physiologic conditions, allowing extrapolations of the biological reality. Human factors such as temperature changes and immune responses are relevant differences between Plasmodium cultures and the human infection [12]. However, the most controversial factor reported about Plasmodium cultures are the in vitro gaseous composition employed [12].
Different gaseous mixtures were employed across the studies and marginal effects on the parasitic viability [13] and drug susceptibility [14], were identi ed, however the relevance of gas mixtures remains unclear.
Here, we present a new and simple methodology, suitable for the culture of Plasmodium falciparum at a small-or large-scale. Though the techniques used we have employed have previously been established, our protocol drastically reduces the expenses, resource-allocation, and use of specialized equipment, enabling easier Plasmodium culturing. By administering atmospheric air (ATM) we show similar outcomes regarding parasite viability and some drug-screening, compared with the classic culturing methodology. Our protocol simpli es the in vitro culturing of Plasmodium and provides a new, standardized method for use by researcher in this eld. In subsequent sections, we discuss the different culturing methodologies in literature as well as researcher opinions regarding the applicability and biological relevance of the various in vitro culturing techniques with respect to the natural environment of the parasite.

Results
In uence of gas mixtures and cultivation method on parasite growth in vitro Three P. falciparum isolates were cultivated in cell asks under the classical mixture of gas described by Trager and Jensen or ATM. The parasitemia of the cultures were recorded for at least 5 days and as can be seen in Fig. 1a-d. There were no differences in P. falciparum proliferation neither the proportion on intraerythrocytic stages. This result was con rmed in the two methodologies used to measure the parasitemia. Inferior ratios of parasitic growth were only observed while employing an initial parasitemia < 0.5%. It is important to note that there were other cultures in our laboratory that have been kept for over two years in the absence of gas mixture or candle jar. In addition, other cultures were also kept in ATM but stirring in 2, 1, 0.5, 0.25 or 0.1 L glass bottles as described in materials and methods. In all cases no differences on parasitic proliferation neither the proportion on intraerythrocytic stages were observed as can be seen in Fig. 1 (a to c shows results obtained by employing 0.1 L bottles as an example) and in the representative images of Fig. 2.
Finally, oxymetric measurements of uninfected erythrocytes maintained in the same conditions of parasite culture showed that free oxygen levels in ATM (~ 180 µM) are superior in comparison to classic conditions (~ 147 µM) (Fig. 1e).
In uence of O 2 tension on antimalarial drugs activity in vitro In Fig. 3, we show that increasing O 2 levels chloroquine (CQ) become more effective in the three isolates used. Interestingly, this phenomenon was much more evident for isolates NF54 and 3D7 in which CQ IC 50 value decreases x18 and x13 fold respectively. However, under identical conditions CQ IC 50 value only decreases x3 fold in the K1 isolate. Moreover, NF54 isolate showed sensibility to CQ only when employing ATM (x10 fold more sensitive than K1, a CQ-resistant isolate) but under classical conditions it is just x1.7 fold more sensible than K1. Under classic conditions all isolates have AV IC 50 value between 0.11-0.37 nM and under ATM the IC 50 values increase but in a much different proportion: x3 fold for 3D7 isolate, x33 fold for K1 and x80 fold for NF54. Finally, under classic conditions, ART antiplasmodial effect does not show differences between the isolates but the introduction of ATM increases the effectiveness by x10 in the K1 isolate, decreases effectiveness by x10 in 3D7 and was statistically equal to classic conditions for NF54. It must be noted that under classic conditions the most ART-sensible isolate was NF54 followed by K1 and 3D7 while under ATM the most ART-sensible was 3D7 followed by NF54 and consecutively the K1 isolate.  Table 1. to the natural gaseous composition of tissues [5] and they are widely used in malaria studies. Importantly, the 5% O 2 refer to the gaseous conditions of the culture vessels and/or incubators, however, the pericellular environment of the red blood cells may not present to the same gaseous condition. Before oxygen reach the cells, it must be exchanged at the surface of the media and cross the liquid layer in a process which is dependent on several factors such as pH, cell density, pressure and medium volume and composition [15]. Notably, oxygen travels approximately 30 millimeters to reach cells in culture, whereas it travels approximately 10-30 µm in animal tissues [16]. Once the cells are consuming oxygen, its diffusion must exceed consumption to avoid hypoxia.
Branco and Francisco previously discussed [12] that is di cult to reproduce physiologic-like conditions in vitro, because of the large differences between the O 2 saturations across human tissues, where some tissues can experience 10-13% O 2 saturation. However, considering that arterial oxygen concentration is approximately 130 µM [15], our results indicate that the parasite statically cultured in vitro is exposed to slightly hyperoxia by employing both ATM and classic conditions. It should be noted that here it was assessed the free oxygen content in homogenized uninfected erythrocytes; the oxygen which reaches erythrocytes at the bottom of the ask probably is lower, quickly consumed by parasites or bounded to hemoglobin. Exemplifying, in another cells cultures in vitro, a fourfold decrease in O 2 concentration respect the gas-liquid interphase has been previously reported [19]. Similarly to Branco [14] or even microaerophilic environments [18]. Therefore, if it's unknown the parasitic natural gaseous conditions in human and if it's actually viable to maintain the parasites by employing ATM, why are we not maintaining the P. falciparum cultures in ATM? All results here presented, show that the cultivation under ATM did not affect the parasitic proliferation and parasitic stages proportion during the studied periods. Furthermore, cultures of P. falciparum 3D7 isolate have been maintained under ATM in our group for more than two years, which demonstrates the technique suitability for extended periods.
Once exposed the success by cultivating parasites in shaking and under ATM, some questions must be discussed about the cons to not follow the classic methodology. It could be said that using ATM or oxygen-rich environments occurs decreases on parasitic viability [13,14]. However, a decrease in parasitic viability under ATM has just been observed at low parasitemia levels (< 1%) [13] which could create di culties in parasitic growth tests by [ 3 H]-hypoxanthine incorporation or DNA staining [20,21].
We need to also consider the effects of antimalarial compounds on parasitic growth when modifying gaseous conditions in culture. Our ndings demonstrated that those changes in the oxygen tension strongly modulate the antimalarials effect in a very different proportion across the isolates. Considering the question of whether pericellular oxygen in culture is similar to in vivo conditions in humans, our results suggest that it is not. This supports the idea that many reported drug-effects against parasites might not be extrapolable to their behavior and/or response in their natural environments, negatively impacting numerous studies in the literature. For example, many studies which showed promissory therapeutic indices in vitro might not be accurate. Similarly, another confounding variable might be the parasitic models ( eld isolates vs. obtained by mutagenesis) used in these studies for identi cation of drug resistant/sensible. The purpose of this study was to generate a culturing methodology that had low technical di cult and improve the environmental similarity to the parasitic environment.
From a biochemical perspective, differences in the antimalarial effects under different gaseous conditions elucidate mechanisms of action. CQ induces oxidative stress to the parasites [22], and our results suggest that oxygen-rich environments potentiate reactive oxygen species, consequently damaging the cells. Alternatively, since AV acts as a mitochondrial uncoupler [23] its antiplasmodial e cacy would improve as the mitochondrial activity diminishes due to the lack of oxygen. Both hypotheses were extracted by alternating the gaseous conditions. These kinds of studies can be interesting tools to better understand the parasitic metabolism and mechanisms of action of different drugs. Our group has previously employed non-classical gaseous mixtures (0% or 20% O 2 ) for these purposes [17,18,24]. In both cases, we observed parasite viability for at least 2-3 weeks under different gaseous mixtures and the biosynthesis of some substances directly involved in the respiration or oxidative stress defense such as ubiquinone and tocopherols [17,18,25]. The changes in the biosynthesis of these metabolites were attributed to the gaseous alterations. This permitted us to better understand both the biochemical mechanism in the parasite as well as the metabolic adaptations to different environments associated with different vertebrate host localizations.
Besides the effects of gaseous composition, other aspects of the culture methodology should also be reviewed. In our opinion, one important aspect that is often neglected is the in vitro chemical environments of parasites are cultured in. In this sense, several human-blood metabolites have demonstrated the ability to interfere on the antimalarials' activity. Despite this, these metabolites are not added to media or supplemented at physiologic concentrations. For example, vitamin E is not added to RPMI but it is established that it protects the parasite from the oxidative stress produced by CQ [26]. Another example is related the folate metabolism. Studies (both in vivo and in vitro) have demonstrated how p-aminobenzoic acid (pABA) and 4-hidroxybenzoate (4-HB) can rescue the antiplasmodial effect of some antifolate antibiotics [27,28]. However, 4-HB is not added to commercial RPMI-1640 and pABA is added at approximately 7 µM, which is a much lower concentration than seen in blood (approximately 32 µM).

Conclusions
In this work, we showed a continuous culture of P. falciparum in a shaker with reusable-glass bottles, instead of using a sophisticated apparatus. With this methodology, it is possible to easily cultivate large volumes of cultures, with high parasitic yields (1 × 10 10 parasites in a few 1L bottles). Our protocol decreases the technical di culty of culturing and increases the yield of biological material for research groups lacking space and/or economic sources. Additionally, shaking cultures is another interesting option for increasing the oxygen diffusion to generate environments that more closely mimic the physiologic-like conditions. Throughout this work, we used three different P. falciparum isolates: NF54 (drug sensible clone obtained from cases of malaria imported into the Netherlands) [29], K1 (isolated in Thailand and reported to be chloroquine-resistant) [30] and 3D7 (derived from NF54 by limiting dilution) [31]. Parasites were cultured at 37 °C in RPMI-1640 medium solved in ultra-pure in Milli-Q water (Millipore, Burlington, Massachusetts, EUA) and supplemented with 25 mM HEPES, 0.5% Albumax I®, 370 µM hypoxanthine, 2 g/L glucose and 25 mg/L gentamicin sulfate [11]. The pH of the media was adjusted to 7.4 with a 10% NaHCO 3 sterile solution and then it was ltered in Millipore Express™ PLUS 0.22 µm sterile lters. The reference gaseous mixture was composed of 5% CO 2 , 5% O 2 , and 90% N 2 [11], however, ATM was also employed when indicated. To avoid culture contamination, mycoplasma testing was done regularly and optic microscopy was performed daily [32]. Leukodepleted blood was a gift from Sírio Libanês Hospital (NESTA, São Paulo, Brazil) blood bank.

Reagents
Trager & Jensen-based culture The P. falciparum intraerythrocytic stages at 2% hematocrit were cultured in vitro according to Trager & Jensen [5] with modi cations [11] in RPMI-1640 complete culture medium in 25 or 75 cm 3 sterile cell culture asks purchased from Corning® (Corning, NY, EUA). Culture changes were performed as described by Radfar et. at. [11], the reference gaseous mixture or ATM were employed when indicated. Parasitic proliferation was studied by microscopy of Giemsa-stained smears and statistically analyzed to compare the parasitic growth/stages under different culture methodologies or conditions by a one-way ANOVA/Tukey multiple comparison test.

Culture in glass bottles
Parasites were cultured in 2, 1, 0.5, 0.25, or 0.1 L Schott hermetic and sterile glass bottles (Schott Glas, Mainz, Germany) containing and 20% of its volume of 2% hematocrit culture as described. Parasites were cultivated at 37 °C and under orbital agitation (130 RPM) in a Thermo Scieni c MaxQ 6000 shaker (model 4353). In this methodology, it was employed ATM. Culture medium composition and changes were performed as described in last item.
Parasitic synchronization at ring stage Culture synchronization at ring stage (6-22 h after the invasion) was performed as described by C. Lambros & Vanderberg [8]. Brie y, cultures were centrifuged at 200 x g for 5 min and synchronization was performed by incubating the pellet in a solution D-sorbitol 5% (w/v) solution in water ( 0.5 mL red blood cells/2.5 mL D-sorbitol solution).

Drug assays
Drug experiments that required the measurement of parasite proliferation were performed in 96-well plates contained in hermetic boxes for plate incubation. The boxes were lled with the indicated gaseous mixture or maintained in ATM. All assays started at ring stage (2% parasitemia, 2% hematocrit) and parasite growth was monitored after 48 h by both Giemsa-stained smears and DNA staining [20]. The compound concentration which decreases at 50% of parasitic growth (IC 50 ) at 48 h was calculated following the Desjardins et al. methodology [33]. Brie y, 1:1 (vol:vol) serial dilution was performed to obtain a decreasing drug concentration in the 96-well plate. An initial concentration of 400 nM for AV and 1000 nM for CQ or ART was selected. Comparison with untreated controls and a solvent control was always performed. Inhibition of parasitic growth was analyzed in relation to the logarithm of the concentration of the compound using a nonlinear regression (dose-response slope/variable sigmoid equation

AVAILABILITY OF DATA AND MATERIALS
Please contact the author for data requests.

COMPETING INTERESTS
The authors declare that they have no competing interests. Parasite proliferation under distinct culture conditions. Cycle progression of parasites under different gas conditions employing cell culture asks or 0.1 L sterile glass bottles. The P. falciparum isolates (a) 3D7, (b) K1, and (c) NF54 were pre-cultivated and synchronized. 2% hematocrit were infected reaching to 1% parasitemia (day 1). Parasitemia was measured daily using optical microscopy as previously described. In (d) the culture parasitemia of 3D7 isolate was measured daily using the methodology previously described by Smilkstein et al. [20]. ATM: atmospheric air, 5% O2: low-oxygen gaseous mixture described by Trager and Jensen, R: ring stage, M: mature trophozoite/schizont forms. (d) Exposed free oxygen levels in uninfected cultures of erythrocytes maintained under different conditions.

Figure 2
Representative pictures of P. falciparum schizonts of 3D7, K1 and NF54 at day 6 post-infection using optical microscopy. ATM CF: parasites cultivated using culture asks under atmospheric air. 5% O2 CF: parasites were cultivated using culture asks under classical mixture of gas described by Trager and Jensen. ATM GB: parasites cultivated using glass bottles under atmospheric air. Cultures at 2% hematocrit were infected with 1% parasitemia (day 1). isolated 3D7, isolated K1, isolated NF54.  Sigmoidal dose-response effect of antimalarials under classic low oxygen mixture. Sigmoidal doseresponse effect of AV (Atovaquone), CQ (Chloroquine), and ART (Artesunate) in P. falciparum 3D7, K1, and NF54 isolates utilizing classic gaseous mixture established by Trager and Jensen. The parasites were cultivated for 48 h at 37ºC in 96 well plates.

Figure 5
Sigmoidal dose-response effect of antimalarial under atmospheric air. Sigmoidal dose-response effect of AV (Atovaquone), CQ (Chloroquine), and ART (Artesunate) in P. falciparum 3D7, K1, and NF54 isolates without addition of gaseous mixture. The parasites were cultivated for 48 h at 37ºC in 96 well plates.