Improved Plasmodium falciparum dilution cloning through efficient quantification of parasite numbers and c-SNARF detection

The P. falciparum NF54 and Dd2 laboratory lines were used in single homologous crossover recombination and CRISPR/Cas9 knock-in transfections. Blood-stage parasites were maintained in standard RPMI 1640-based media supplemented with 0.23% NaHCO₃, 0.5% AlbuminNZ Microbiological BSA (MP Biomedicals, Irvine CA) and O⁺ human erythrocytes (UVA Blood Bank) under a defined gas atmosphere (90% N₂, 5% O₂ and 5% CO₂) [2]. An identical, but HEPES-free culture medium was used with the c-SNARF pH indicator to facilitate detection of positive microplate wells. Parasite health was monitored by Giemsa-stained blood smears.

Schizont-stage NF54 cultures were used for single crossover recombination transfection using the protocol described by Hasenkamp and colleagues [12]. Beginning 2 days after transfection, 3 cycles of 2.5 nM WR99210 were applied for 15 days with intervening drug withdrawal for 15–20 days. The desired integration was then confirmed by PCR prior to initiating limiting dilution cloning. CRISPR/Cas9 knock-in transfection was performed using Dd2 parasites carrying a conditional Cas9-DDD cassette and the stabilizing agent trimethoprim (TMP). At day 2 after transfection, 2.5 µg/ml blasticidin S (Sigma) was added to select for a pL7 plasmid expressing sgRNA and a cassette for double homologous recombination knock-in. After regular media changes every 2–3 days, parasite outgrowth was detected by Giemsa smears 13–20 days after transfection. Limiting dilution cloning was then initiated.

Flow cytometry was used to more precisely estimate parasitaemia and total cell count before limiting dilution cloning. In a typical experiment, 5 µl of packed cells from culture (with mostly trophozoites) was transferred to a sterile microfuge tube with 995 ul complete medium. After mixing, 500 µl of the suspension was transferred to a second tube.

To one microcentrifuge tube, either 0.5 µg/ml of ethidium bromide or SYBR Green I nucleic acid stain (2X dilution, Invitrogen) with 1xPBS was added to stain parasite DNA. After incubation for 20 min at 37 °C, these cells were washed twice with 500 µl 1xPBS before resuspension in 500 µl 1xPBS for flow cytometry using excitation and emission wavelengths of the laser lines of 488 nm for SYBR Green and 525 nm for ethidium bromide, reading the emitted fluorescence at 521 and 600 nm, respectively, with a 100,000 of event count. A matched sample of uninfected erythrocytes was used to identify the threshold for distinguishing infected cells from healthy ones. The flow cytometry readout was then used to estimate the number of cells/µl and percentage of infected cells. These values and the second microfuge of unstained cells from the original culture were then used to set up the limiting dilution.

After determination of the number of parasites per µl by cytometry and a further dilution step in culture medium, a volume containing 20 parasites was transferred to a 15 mL Falcon tube; this volume ranged between 2 and 10 µl depending on the initial parasitaemia and haematocrit. Then, 9.5 ml complete medium and 0.5 ml fresh packed erythrocytes were added and mixed gently to establish a 5% final haematocrit. The solution was then dispensed in 100 µl aliquots in a 96-well plate using a multichannel pipet. Following this, the microplate was incubated for six days under the culture conditions described above. Medium changes occur on days 7, 11 and 14. From day 14 on, the plate was prepared for detection of positive wells by the method described by Lyko and colleagues [11]. Beginning on day 14, the medium was change to a HEPES-free medium containing 2 µM 5(and-6)- carboxy SNARF-1 (c-SNARF-1, Invitrogen, Carlsbad. CA. USA) [11].

Microplates were typically incubated in HEPES-free medium containing c-SNARF-1 for 48 h to permit accumulation of metabolic acid in wells containing viable parasites. Microplates were removed from the incubator and allowed to equilibrate at room temperature for 15 min before measuring c-SNARF-1 fluorescence (excitation 488 nm, emission 590 and 645 nm; Synergy HT plate reader, BioTek, Winooski, VT, USA).

Wells with a significant increase in the 590/645 nm emission ratio were then expanded as clonal cultures and a maximum of 30 positive wells/plate was required to ensure clonality. Depending on the parasite growth rate, positive wells were detected as early as 16 days after initiating limiting dilution; slow-growing transfectants typically required longer. Calculation of the 590/645 nm fluorescence ratio and identification of positive wells were automated using scripts written in SigmaPlot 10.0 (Systat, San Jose, CA). These SigmaPlot scripts are available upon request.

To compare flow cytometry and different DNA staining methods in limiting dilution cloning experiments after transfection, the methods outlined in Fig. 1 were applied, establishing three experimental groups. In group A, parasitaemia and cell numbers per unit volume were determined using the traditional method with Giemsa-stained smears and a haemocytometer cell counter to obtain a desired parasite load of 0.3 parasites/well. Group B employed flow cytometry and ethidium bromide staining of parasite DNA to achieve an estimated 0.3 parasites/well. Group C also used flow cytometry, but detected parasites with SYBR Green I staining and dilution to 0.2 parasites/well in limiting dilution microplates (Table 1). Assuming random sampling with Poisson distribution, dilution to 0.2 and 0.3 parasites/well will yield growth of one or more parasites in 18% and 26% of wells, respectively. These values will lead to nonclonal lines that result from two or more parasites/well in 9.7% and 14.3% of positive wells, respectively. Both of these rates are acceptable [14].

The results for Group A, which relied on manual counting before each experiment, were highly variable, with 8, 42, and 24 positive wells from three is substantial higher compared with the other groups (Table 1). Notably, PCR checks of the 8 positive wells did not yield clones with the desired integration-positive genotype in this single crossover recombination experiment (Fig. 2A). The second plate, with 42 positive wells, was considered invalid due to a high probability of seeding two or more parasites per well and, therefore, mixed, non-clonal parasites. Of the 24 positive wells recovered in plate 3, only two were integration-positive. Thus, three plates were required to obtain the desired transfectant with this approach.

Group B yielded more reproducible results, with 26, 32, and 24 positive wells from the three trials (Table 1). PCR checks revealed 5 integration-positive clones from these plates, which also used single crossover recombination (Fig. 2A).

Group C, using CRISPR/Cas9 transfection (Fig. 2B) with flow cytometry and SYBR green staining to initiate dilution cloning with 0.2 parasites/well, yielded 18, 16, and 21 positive wells from three trials for a low standard deviation. PCR revealed that 12 of 18 tested positive wells yielded the desired integrant, consistent with efficient CRISPR/Cas9-mediated gene editing [9].

Thus, when a single crossover recombination approach is preferred, careful dilution and setup of limiting dilution and facile genotype determination of positive wells is especially critical. Selection-linked integration may be considered for knock-in experiments when CRISPR/Cas9-based methods are not suitable [7].

Efficient detection of parasites with both ethidium bromide and SYBR Green I staining was obtained and other DNA-staining dyes, such as Hoechst and Syto16, may also work. Ethidium bromide is the most cost-effective, but some laboratories restrict its use on flow cytometers as it is reported to be more mutagenic than other DNA stains. SYBR Green I also permits quantification of different parasite stages (ring, trophozoite and schizonts) due to its linear correlation of fluorescent signal with DNA content [13, 14].

The shown flow cytometry method can be completed in a few minutes to improve limiting dilution experiments, which may fail due to either an excessively low number of positive wells that do not yield desired clones or an excessively high number of positives that cannot be reliably considered clonal.

Conventional detection of positive wells during limiting dilution cloning requires preparation and examination of 96 smears every 2 days to ensure that all clonal lines are identified and harvested. This process is both laborious and time-consuming. To evaluate whether positive wells can be more easily detected with the c-SNARF method [11], Group C transfections were subjected to detection with both microscopic examination and the pH-sensitive c-SNARF 1 dye. Figure 3 shows positive wells detected by both microscopic examination and an increased 590/645 nm fluorescence ratio on day 18 (red symbols). Some wells were identified as positives by microscopic examination, but did not exhibit an increased 590/645 nm fluorescence ratio on this day (blue symbols). Fluorescence measurements on these plates 2 days later revealed that these wells also became positive by the c-SNARF 1 method, consistent with a higher parasite load required for sufficient metabolic acidification to allow ratiometric detection. As these three plates yielded numbers of positive wells predicted from the initial setup at 0.2 parasites/well, leading to the conclusion that c-SNARF 1 is nontoxic to parasite cultures, as reported previously [11]. This avoids plate transfers as required for PCR and other methods, reducing cost and the risk of cross-contamination of wells. Although modestly longer cultivation of limiting dilution microplates is required for detection of positive wells, the c-SNARF method confidently identified all positive wells with dramatically reduced effort. In addition, it also permits scaling to larger numbers of microplates as positives may be identified from individual plates in ~ 2 min by most fluorescence plate readers.