P. falciparum genomic DNA (gDNA) was extracted from a Dd2 parasite culture, while P. chabaudi gDNA was obtained from a blood sample of infected mice from the collaborator Dr. Pedro Cravo from Global Health and Tropical Medicine (GHTM), Instituto de Higiene e Medicina Tropical (IHMT), Universidade Nova de Lisboa. Both samples were extracted using the NZY Blood gDNA isolation (NZYTech). Concentration and purity were measured using a NanoDrop™1000 spectrophotometer (Thermo Scientific), and the gDNA was stored at -20°C.

To construct a Plasmodium spp. interspecies chimeric gene, Selection-Linked Integration (SLI) technique was used (Birnbaum et al., 2017). To mitigate potential structural constraints, genetic regions of interest were chosen according to the protein’s 3D model. The Swiss Model automated protein structure homology-modelling server was employed to generate this 3D model, using the structure of the 26S proteasome regulatory subunit rpn2 from Plasmodium falciparum (isolate 3D7) with the repository database code Q8IKH3 (Q8IKH3_PLAF7) (https://swissmodel.expasy.org/repository)”.

Then, the regions of interest of each species were amplified through a Polymerase Chain Reaction (PCR). Primers were designed considering the gene sequence, being the reverse primer from P. falciparum (p2 – Table 1 ) and the forward primer from P. chabaudi (p3 – Table 1 ) complementary allowing subsequent fusion. Samples were analyzed by electrophoresis in an agarose gel (1%), using a molecular weight marker (GeneRuler 1kb DNA Ladder, Thermo Fisher Scientific) to confirm that the obtained fragments had the expected size.

A second PCR was performed to fuse both fragments using the forward primer of the P. falciparum part (p1 – Table 1 ) and the reverse primer of the P. chabaudi part (p4 – Table 1 ) with iProof HF DNA Polymerase (Bio-Rad Laboratories). Expected PCR band size was extracted using GRS PCR & Gel Band Purification kit (GRiSP, Lda) and sequenced (STAB Vida).

The previous product was posteriorly inserted in the JET plasmid (CloneJET PCR Cloning kit, Thermo Fisher Scientific) applying the manufacturer´s protocol. Reaction products were then transformed into chemically competent Escherichia coli cells (XL-10) by heat shock. Posteriorly, cells were transfered to lysogeny broth (LB) medium [10 g/L bacto-tryptone, 5 g/L yeast extract, and 10 g/L NaCl] plates supplemented with ampicillin antibiotic (100 µg/mL) and incubated overnight at 37°C. The next day, two of the emerged colonies were grown and plasmid DNA was extracted using the Zyppy™ Plasmid Miniprep Kit (Zymo Research). Concentration and purity were measured using a NanoDrop™1000 spectrophotometer (Thermo Scientific). Finally, to confirm the DNA sequence of the construct, sequencing was carried out by STAB Vida.

The plasmid with the 738K variant was constructed using the Selection-linked Integration (SLI) technique (Birnbaum et al., 2017). This method makes it possible to obtain parasite lines with genomic integrations in P. falciparum with a greater speed and success rate. Therefore, the previous fragment was inserted in the Selection-Linked Integration plasmid (pSLI), being first needed to digest both the pSLI ((Plasmid #85791) - Addgene) and the fragment with NotI and SalI (New England Biolabs). Electrophoresis was used to verify the digestion, and the desired bands were extracted with the GRS PCR & Gel Band Purification kit (GRiSP, Lda). Then, ligation of the fragment to the plasmid was carried out using T4 DNA Ligase (Thermo Fisher Scientific). The ligation product was transformed into competent E. coli cells and, posteriorly, the plasmid was extracted using the NZYMiniprep kit (NZYTech, Lda.).

A descendent plasmid was then constructed with the E738 variant using the site-directed mutagenesis technique to replace a lysine for a glutamic acid on the 738 position (Liu and Naismith, 2008). Using primers p5 and p6 ( Table 1 ), a PCR was performed. The amplified product was then digested with DpnI endonuclease (New England Biolabs) at 37°C for 3 hours to select the mutation-containing synthesized DNA, digesting the parental DNA template. The mutated plasmid was then transformed into chemically competent E. coli. Posteriorly, expanded selected colonies were extracted using the Zyppy™ Plasmid Miniprep Kit (Zymo Research). After plasmid DNA extraction, concentration and purity were measured using a NanoDrop™ 1000 spectrophotometer (Thermo Scientific) and sent for sequencing at STAB Vida.

Both plasmids were then transfected into Dd2 P. falciparum parasites cell lines, a multi-drug resistant strain derived from an Indochina isolate (Reilly et al., 2007). Transfection was performed by electroporation in uninfected erythrocytes (Hasenkamp et al., 2012). First, erythrocytes are washed with cytomix [10 mM/L K²HPO⁴/KH²PO⁴, 120 mM/L KCl, 0.15 mM/L CaCl₂, 5 mM/L MgCl₂, 25 mM/L HEPES, 2 mM/L EGTA, adjusted with 10 M/L KOH to pH 7.6]. Afterward, 300 µL of erythrocytes were added to the plasmid solution, which contains 50 µg of plasmid and up to 200 µL of cytomix. This was transferred to a Gene Pulser/MicroPulser™ Electroporation Cuvettes, 0.2 cm gap (Bio-Rad Laboratories, Inc) and eletric shocked on the Gene Pulser Xcell™ (Bio-Rad Laboratories) electroporator. The mixture was washed twice in 5 mL of malaria culture medium (MCM) [RPMI, 1640 (Gibco) with 2 mM L-glutamine, 200 µM hypoxanthine, 0.25 µg/mL gentamycin, 25 mM HEPES, 0.2% NaHCO3, and 0.25% Albumax II (Life Technologies)] to remove lysed erythrocytes. The transfected erythrocytes were then inoculated with P. falciparum Dd2 strain parasite erythrocytes at the trophozoite stage and 5 mL of MCM.

Parasite cultures were maintained under a controlled atmosphere of 5% O₂/5% CO₂/90% N2 at 37°C in a CO₂ incubator (Thermo Fischer Scientific). Cultures on a t25 flask were maintained at 4% hematocrit in 5 mL of MCM. Medium was changed every other day, and cultures’ healthiness and parasitemia were regularly monitored by microscopy. This microscopic analysis was made using a blood smear, fixed with methanol (100%), and colored with 10% Giemsa’s eosin methylene blue solution (Merck) for 20 min. Parasitemia was calculated by dividing the number of parasitized erythrocytes per total counted erythrocytes.

To select the genetically modified parasites, first, the parasites with episomal pSLI were selected with WR99210, followed by a second selection with neomycin (G418), which allows the selection of parasites with the plasmid genome integrated, since a promoterless resistance cassette is present in it.

After the P. falciparum drug selection and culture growth, parasites were genotyped and evaluated for successful genetic manipulation. For that, gDNA was extracted from parasite cultures using the NZY Blood gDNA Isolation kit (NZYTech). To perform the genotyping, a PCR was performed using the forward primer (p7 - Table 1 ) located in the P. falciparum region and the reverse primer (p8 - Table 1 ) in the P. chabaudi region, forming a band of, 2370 bp, being posteriorly digested with the restriction enzyme NheI (New England Biolabs). Then, DNA sequencing was carried out.

P. falciparum growth was evaluated using a fluorometric assay. For that, a parasite culture with ring-stage parasites starting at 0.5% parasitemia and 4% hematocrit was prepared. Throughout 14 days, 100 µL of culture was retrieved every day to a 96-well plate and stored at -20°C. Culture’s health was daily evaluated by blood smear.

To carry out the fluorimeter readings, plates were stained with SYBR Green. For this, 4 µM of SYBR green were added to the lysis buffer (15.76 g Tris-HCl, adjust pH to 7.5, 2% w/v EDTA, 0.016% w/v saponin, 1.6% v/v Triton X-100). In a new 96-well plate, 25 µL of the lysis buffer + SYBR Green mixture were added to each well. Afterward, 50 µL of the homogeneized cultures from the assay plates were added. Then, after an incubation of 3 hours at RT on a dark chamber, plates were analyzed in the fluorimeter (Thermo Scientific Varioskan Flash) with excitation and emission wavelength bands at 485 and 530 nm, respectively.

The number of merozoites per schizont analysis was performed using a thin film Giemsa stain as described. Firstly, parasites were tightly synchronized using sorbitol (5%) into ring-stage. Then, after approximately 36 hours, the parasite stage was evaluated by a blood smear. Mature schizonts were observed, and merozoites were counted and normalized per number of schizonts.

Tightly synchronized ring-stage parasites at 0.5% parasitemia and 2% hematocrit were added to the plate previously prepared with several dilutions of DHA, 100 µL of the mixture (iRBCs, RBCs, and MCM) to each well. Plates were then incubated for 72 hours in the CO₂ incubator. After the incubation, plates were sealed and frozen at -20°C. Subsequently, fluorimeter readings were executed as previously mentioned. IC₅₀ values were calculated by nonlinear regression analysis using Graphpad^®.

To perform a ring survival assay (RSA) and a trophozoite-stage survival assay, tightly synchronized early ring-stage and trophozoite-stage parasites, respectively, at 0.5% parasitemia and 2% hematocrit, were exposed to 2.5, 5, 10, 20, 50, 250, and 700 nm of DHA. After 6 hours of incubation in a CO₂ incubator, parasites were centrifuged for 5 min at, 1500 xg, the supernatant was removed and washed three times with phosphate-buffered saline 1x (PBS). Finally, 1 mL of MCM was added to each pit and incubated for 66 hours. Plates were then stored at -20°C. Fluorimeter readings were carried out as previously mentioned. Survival (%) represents the parasitemia normalized to untreated (100%) survival and kill-treated (0% survival) controls, where kill-treated refers to samples treated with 700 nm DHA.

To analyze the proteasome activity, trophozoite-stage parasite cultures at 3% hematocrit and 8% parasitemia were first washed once with PBS (1x) and, consequently, incubated with 0.05% (w/v) saponin for 10 min on ice. Then, parasite pellets were washed with ice-cold PBS (1x) three times and lysed at -20°C overnight. It was then cleared by centrifugation at, 1000 xg for 10 min.

Cell lysates were quantified using the Bradford method. To perform the quantification, a 96-well plate was used, and 10 µL of the sample was mixed with 200 µL of 1x Bradford’s reagent (Bio-Rad Laboratories, Inc). After a 5 min incubation, the absorbances were read at 590 nm. Then, the concentration was determined using a calibration line for bovine serum albumin (BSA) concentration absorbances, which started with a, 2000 µg/mL concentration and continued with dilutions of 1:2.

The assay was then performed using the Proteasome Activity Fluorometric Assay Kit II (Ubiquitin-Proteasome Biotechnologies) and following its recommendations. For that, in a 96-well plate, 10 µL of cell lysate was mixed with 20x proteasome assay buffer and 0, 3.125, 6.25, 12.5, and 25 µM of DHA, for a total volume of 50 µL and incubated for 30 min at 37°C. Meanwhile, the proteasome substrate was prepared, mixing 50 µL of 20x proteasome assay buffer with 950 µL H₂O, incubating for 10 min at 37°C and, posteriorly, adding 2 µL of, 1000x Suc-LLVY-AMC substrate (50 mM in DMSO (dimethyl sulfoxide)) to measure the chymotrypsin-like activity, and stored at 37°C after vortexing for 10 seconds. After the 30 min incubation, 50 µL of proteasome substrate was added to each pit. Fluorescence was measured at 360/40 nm and 460/30 nm for excitation and emission, respectively, on the spectrophotometer (Thermo Scientific Varioskan Flash), previously heated at 37°C, with readings at every 5 min for 3 hours. For proteasome activity analysis, the slope value of each curve was calculated between 20 and 90 min by using the formula (Y90-Y20)/(X90-X20), in which Y90 and Y20 were AMC fluorescence readings from the Y-axis at 90 min and 20 min, respectively. Values are normalized to the untreated sample.

A Western Blot was used to evaluate the differences in the levels of ubiquitinated proteins. For that, trophozoite-stage parasites were first incubated with 5 mL of MCM treated with 0, 0.01, 0.1, 1, 10 µM of DHA for 90 min at 37°C with a controlled atmosphere. Then, trophozoites were isolated using the AutoMacs^® Pro Separator (Miltenyi Biotec), separating different stage-parasites using magnetic means. After separation, the eluate was centrifuged at, 1500 xg for 5 min and washed 3 times with PBS 1x, forming a dark pellet with the parasites. The supernatant was discarded, and 50 µL of H₂O was added, storing at -20°C to lyse the parasites. Later on, lysates were cleared by centrifugation for 15 min at, 5000 xg at 4°C and supernatant was collected. Protein quantification was assessed using the Bradford method previously described.

Electrophoresis was performed using a running gel of 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and a stacking gel of 5% SDS-PAGE. Protein samples were prepared by adding 20 µg of total protein to 10 µL of loading buffer, previously prepared by adding 50 µL of 2-mercaptoethanol to 950 µL of 2x Laemmli Sample Buffer (Bio-Rad Laboratories). Next, proteins were denaturated for 5 min at 98°C. The samples were loaded onto the gel and separated by electrophoresis (100V). Gels were then transferred to a nitrocellulose membrane (Bio-Rad Laboratories) in a Trans-Blot Turbo Transfer Buffer (Bio-Rad Laboratories) using the Trans-Blot Turbo Transfer System (Bio-Rad Laboratories). To prevent unspecific connections, the membranes were blocked with 5% of BSA in Tris-Buffered Saline (TBS)/0,1% Tween 20 for 1h at RT and then incubated overnight, at 4°C, with the primary antibodies: ubiquitin (1:1000, source: rabbit, Invitrogen) and actin (1:1000, source: mouse, Invitrogen). Actin was used as the loading control. On the next day, the membranes were washed in TBS/0,1% tween three times, each of 5 min, and then incubated for 1 hour with fluorescent secondary antibodies, respectively: anti-rabbit (1:1000, source: goat, conjugate: DyLight 488, Invitrogen) and anti-mouse (1:1000, source: goat, conjugate: DyLight 680, Invitrogen). Sapphire Biomolecular Imager (Azure Biosystems) was used for protein detection. ImageJ software was used for quantitative analyses.

Statistical analyzes were performed using the GraphPad Prism 8 software. Results were presented as means ± standard error of the mean (SEM). Comparisons between different conditions were made using t-tests. Differences were considered significant in the statistical analysis when p<0.05. At least three biological replicates were performed for each assay.

To assess the impact of the E738K variant on the response to DHA, in vitro drug susceptibility assays were performed. The number of live parasites previously exposed to serial dilutions of DHA was quantified through fluorimeter readings, and IC₅₀s were calculated. The 738K variant led to a near four-fold significant increase in the IC₅₀ compared to the E738 variant ( Figures 2C, D ), with values of 2.25 nM and 0.8 nM, respectively.

Typically, the IC₅₀ values for DHA vary between 0 and 8 nM in P. falciparum (Fairhurst and Dondorp, 2016). Both parasite lines are within the strain IC₅₀ interval usually observed. However, there is a significant difference regarding artemisinin response. Nevertheless, in vivo parasite clearance half-lives correlate poorly with DHA 50% inhibitory concentrations in vitro, therefore parasite survival assays was used to evaluate the difference in the artemisinin response (Fairhurst and Dondorp, 2016).

Artemisinin resistance was originally difficult to investigate due to poor correlation with the existing standard in vitro measures of susceptibility, which complicated the understanding of the biology behind the slow-clearing P. falciparum in Cambodia (Noedl et al., 2008; Fairhurst and Dondorp, 2016). It was later discovered that parasite resistance to ARTs did not emerge across the whole parasite life cycle, and it was only identifiable in the ring stage of development (Khoury et al., 2020). Given the need for a more reliable and standardized in vitro measure of delayed clearance that correlates with the in vivo resistance phenotype, the RSA was developed (Witkowski et al., 2013). This method consists of tightly synchronizing cultured parasites at the early ring stage and subjecting them to a 6-hour pulse of DHA. After 72 hours, the parasite survival rate is evaluated (Zhang et al., 2017; Sutherland et al., 2021).

RSA was performed to confirm the impact on the response to artemisinin. It was possible to observe a significant difference at 2.5, 5, and 10 nM of DHA between both lines, with the 738K variant demonstrating increased parasite survival ( Figure 2E ). The assay was also executed in trophozoites. In spite of the fact that rpn2^738K line present higher parasite survival, there was no significant difference between both lines at trophozoite stages ( Figure 2F ). This difference between the assay in rings and trophozoites corroborates that later parasite stages are highly susceptible to ART, unlike early ring-stage parasites that are able to survive pulses of DHA (Davis et al., 2020). This profile coincides with the peak period of hemoglobin uptake and degradation, during which Fe2+-heme is liberated and activated ART (Stokes et al., 2019).

DHA reacts with susceptible groups in biomolecules in its activated state, causing cellular damage. These damaged proteins are subsequently marked with ubiquitin, rapidly increasing the accumulation of ubiquitinated proteins in DHA-treated parasites. This rapid accumulation may clog the proteasome and lead to the inhibition of its activity (Bridgford et al., 2018).

Taking this into account, the chymotrypsin-like activity of the proteasome was analyzed with different DHA concentrations. In this essay, the proteasome cleaves the Suc-LLVY-AMC substrate, a fluorogenic substrate, while the released AMC fluorescence is monitored. In the baseline (without DHA), it was possible to observe that the rpn2^E738 and rpn2^738K proteasome have similar activity ( Figure 3A ). When DHA was added, proteasome activity was inhibited in both lines, however the P. falciparum cell line with the 738K variant presented a significantly lower proteasome activity inhibition, with the difference increasing at the highest concentrations ( Figure 3B ).

In agreement with proteasome inhibition by DHA in P. falciparum (Bridgford et al., 2018), the accumulation of polyubiquitinated proteins increased, causing oxidative stress and inducing parasite death. The rpn2^738K parasite line presented a stabilization in the proteasome’s activity, which should have an impact at the level of polyubiquitinated protein accumulation. Therefore, a western blot experiment was performed to assess the level of polyubiquitinated proteins in DHA-treated parasites ( Figure 3C ). We observed an accumulation of polyubiquitinated proteins in the rpn2^E738 parasites line compared with the baseline (without DHA). In the rpn2^738K parasite line, the relative ubiquitin presented a slight decrease at 0.1 µM concentration and was significantly smaller compared to the rpn2^E738 parasites line at 1µM and 10µM concentrations ( Figure 3D ).

To further examine the proteasome activity in parasites treated with DHA, we tried to use transfectants expressing GFP coupled to a destabilization domain (DD), which is targeted for degradation by the proteasome in the absence of the protective ligand, Shield-1 (Bridgford et al., 2018). We were able to generate a parasite line carrying the E738 variant. However, after several attempts, we were unable to generate the parasite line bearing the 738K variant. We believe that the reason was possibly due to its slower growth and parasitemia characteristics, compared to the parasite line with the E738 variant. This difference in transfection capability was tested by attempting to transfect a plasmid with a GFP-only plasmid and similar result was observed.

DHA also disrupts proteasome-dependent degradation, preventing the removal of damaged proteins, which was observed with the accumulation of polyubiquitinated proteins. This is a double threat to parasites and further induces oxidative stress, leading to parasite death. Nevertheless, it was also shown that parasites with the 738K variant have less inhibition of the proteasome activity, as well as a consequent decrease in the accumulation of polyubiquitinated proteins. This suggests that this rpn2 variant antagonizes the physiological impact of DHA controlled by the 26S proteasome.

This mutation in the rpn2 subunit is a scaffold to ubiquitin-processing factors and is involved in stabilizing the 19S RP, responsible for the recognition of ubiquitylated substrates (Kors et al., 2019; Mao, 2021). A study in the human proteasome demonstrated that kinase p38 MAPK (Mitogen-Activated Protein Kinase) phosphorylated the rpn2 subunit, being Thr-273 of rpn2 the major phosphorylation site affected, stabilizing the poly- and non-ubiquitinated substrates and reducing all three proteolytic activities of the proteasome, caspase-, trypsin- and chymotrypsin-like activities. Additionally, this study also demonstrated that a T273A mutation in the rpn2 blocked the p38 MAPK-mediated proteasome inhibition. Other studies demonstrated that rpn2 interacts with rpt3, rpt4, and rpt6 of the proteasome ATPase complex, which are involved in regulating the 20S CP gate opening (Lee et al., 2010). Therefore, our results suggest that alterations in rpn2 protein have a major impact downstream of the 26S proteasome-mediated protein degradation process in P. falciparum.