A Plasmodium falciparum RING Finger E3 Ubiquitin Ligase Modifies the Roles of PfMDR1 and PfCRT in Parasite Drug Responses

ABSTRACT

Protein ubiquitination is an important posttranslational regulation mechanism that mediates Plasmodium development and modifies parasite responses to antimalarial drugs. Although mutations in several parasite ubiquitination enzymes have been linked to increased drug tolerance, the molecular mechanisms by which ubiquitination pathways mediate these parasite responses remain largely unknown. Here, we investigate the roles of a Plasmodium falciparum ring finger ubiquitin ligase (PfRFUL) in parasite development and in responses to antimalarial drugs. We engineered a transgenic parasite having the Pfrful gene tagged with an HA-2A-NeoR-glmS sequence to knockdown (KD) Pfrful expression using glucosamine (GlcN). A Western blot analysis of the proteins from GlcN-treated pSLI-HA-NeoR-glmS-tagged (PfRFULg) parasites, relative to their wild-type (Dd2) controls, showed changes in the ubiquitination of numerous proteins. PfRFUL KD rendered the parasites more sensitive to multiple antimalarial drugs, including mefloquine, piperaquine, amodiaquine, and dihydroartemisinin. PfRFUL KD also decreased the protein level of the P. falciparum multiple drug resistance 1 protein (PfMDR1) and altered the ratio of two bands of the P. falciparum chloroquine resistance transporter (PfCRT), suggesting contributions to the changed drug responses by the altered ubiquitination of these two molecules. The inhibition of proteasomal protein degradation by epoxomicin increased the PfRFUL level, suggesting the degradation of PfRFUL by the proteasome pathways, whereas the inhibition of E3 ubiquitin ligase activities by JNJ26854165 reduced the PfRFUL level. This study reveals the potential mechanisms of PfRFUL in modifying the expression of drug transporters and their roles in parasite drug responses. PfRFUL could be a potential target for antimalarial drug development.

INTRODUCTION

Malaria is one of the deadliest parasitic diseases and is responsible for millions of illnesses worldwide (1). Despite longstanding efforts to eradicate malaria, the disease remains a major threat to public health, especially in Africa (1). The emergence of multiple drug-resistant malaria parasites highlights the need for drug target discovery and translation into antimalarial therapeutics. One such target is the parasite’s ubiquitin system. Ubiquitination, like phosphorylation, is an essential posttranslational modification process that regulates a wide spectrum of cellular functions, including protein degradation through proteasomal machinery, endocytosis, cell cycle progression, DNA repair, the sorting and trafficking of transmembrane proteins, transcriptional regulation, and immune signaling (2–8). This process attaches one or more ubiquitin moieties to a target protein via an isopeptide bond between the C-terminal diglycine motif of ubiquitin and lysine (K) in the target protein. Ubiquitination is governed by an enzymatic cascade that is mediated by three different enzymes: ubiquitin (Ub) activating enzyme (E1), ubiquitin-conjugating enzyme (E2), and ubiquitin ligase (E3) (9, 10). The E1 and E2 enzymes are highly conserved within eukaryotes, whereas the E3 ligases are diverse. E3 ligases provide specificity for the substrate recognition and the recruitment of the activated ubiquitin from E2 to target proteins (4, 11).

Ubiquitination is a crucial mechanism that is necessary for the life cycle progression and survival of malaria parasites, and ubiquitin conjugates can be detected in all morphological forms of the Plasmodium falciparum erythrocytic cycle (12, 13). An E3 homologous to E6AP C-terminus (HECT) ubiquitin ligase was recently found to regulate the growth and virulence of Plasmodium yoelii parasites (14). Protein ubiquitination and proteasome-mediated protein degradation pathways have also been implicated in parasite drug responses (15). The retardation of parasite growth and the accumulation of ubiquitinated proteins occur after treatment with artemisinin (ART) and its derivatives (collectively abbreviated here as ART), suggesting a cellular stress response that engages the ubiquitin/proteasome system (16). Evidence suggests that the P. falciparum Kelch-like protein PfK13 may act to bring the E3 ligase close to its substrates (17), participate in the ubiquitination and degradation of P. falciparum proteins (18, 19), and contribute to a survival phenotype of parasite ring-stages under ART pressure (20–23). Also, a HECT ubiquitin ligase (MAL7P1.19 or PF3D7_0704600) was genetically associated with the responses of P. falciparum to quinine (QN) and quinidine (QD) (24). Altering the expression of the gene was found to modulate the length of the cell cycle (25). The mutations in a Plasmodium chabaudi deubiquitinating protein (PcUBP1) were linked to a resistance to chloroquine (CQ), mefloquine (MFQ), and ART (26, 27). Thus, proteins in the parasite ubiquitin pathways may be good targets for antimalarial drug development (3, 12, 13, 19, 28). Indeed, compounds that inhibit parasite growth via effects on the P. falciparum ubiquitin-proteasome system have been identified through high-throughput assays (29, 30). However, the exact mechanisms by which protein ubiquitination affects parasite growth and the responses to antimalarial drugs remain largely unknown.

In this study, we investigated the roles of a putative P. falciparum ring finger ubiquitin ligase (PfRFUL) (PF3D7_1004300 or PF10_0046) in parasite growth and the responses to antimalarial drugs. The Pfrful gene is located on chromosome 10 and has a 3,390 base-pair single exon encoding 1,129 amino acids with a predicted molecular weight of 124.3 kDa (PlasmoDB, release 52, 20 May 2021; https://plasmodb.org/plasmo/app). In addition to having an E3 ubiquitin ligase domain (amino acids 167 to 432), the predicted protein also has a zinc finger (RING) domain (amino acids 360 to 403). PfRFUL was previously proposed to be a homolog of the mammalian ring finger 4 (RNF4) protein that can interact with SUMOylated proteins (https://mpmp.huji.ac.il/maps/Ub_SUMO.html) and target poly-SUMO-modified proteins for degradation mediated by ubiquitin (31). However, the PfRFUL is a much larger protein than is a typical RNF4 ligase (194 amino acids for rat RFN4) (32). Peptides of PfRFUL have been identified from intraerythrocytic stages, salivary gland sporozoites, and gametocytes by proteomic and phosphor-proteomic analyses (33–37). To functionally characterize PfRFUL, we tagged the Pfrful gene in the P. falciparum Dd2 parasite line with a HA-2A-NeoR-glmS sequence (38, 39) to enable the conditional knockdown (KD) of gene expression via the addition of glucosamine (GlcN). The decreased PfRFUL protein levels reduced the ubiquitination of many parasite proteins. PfRFUL KD also rendered the parasite more sensitive to multiple drugs and altered two proteins that are known to be involved in parasite drug responses, namely, the P. falciparum chloroquine transporter (PfCRT) and the P. falciparum multiple drug resistance 1 protein (PfMDR1). Our results suggest that PfRFUL modifies many parasite proteins and can affect parasite drug responses through ubiquitination activity.

RESULTS

Generation of a parasite tagged with pSLI-HA-NeoR-glmS for the conditional knockdown.

We created a transgenic parasite from the Dd2 line, tagged with the HA-2A-NeoR-glmS sequence, using selection-linked integration (SLI) (38). To obtain parasites with the HA-2A-NeoR-glmS sequence integrated at the C terminus of the PfRFUL protein, a 495 bp 3′ end sequence of the PfRFUL coding region was amplified from the Dd2 parasite, using the forward primer CF and the reverse primer CR, and it was cloned into the pSLI-HA-NeoR-glmS plasmid (Fig. 1A; Fig. S1A). Blood-stage parasites were transfected with the plasmid and selected with WR99210. Once parasitemia reached 2 to 4%, the parasites were again selected with neomycin. The integration of the plasmid construct into the expected locus was confirmed using polymerase chain reaction (PCR) and Western blot methods. The amplifications for the 5′ and 3′ integrations, using one primer in the parasite genome (P1 for the 5′ ends and P2 for the 3′ ends) and another primer in the plasmid (P3 to pair with P1 and P4 to pair with P2), produced the expected PCR products in the PfRFUL glmS-tagged parasite (PfRFULg) but not in the wild-type (WT) Dd2 parasite (Fig. 1A and B; Fig. S1A). Primers P1 and P2 amplified a product from the Dd2 parasite, but not one from the tagged parasite, due to the insertion of the plasmid (Fig. 1B), which also indicated that there was no WT Dd2 contamination in the parasite culture with plasmid integration. The correct plasmid integration was validated after sequencing the PCR products amplified by primers P1 and P3 (Fig. S1B). The expected PCR product bands were also obtained from both the parasite DNA and mRNA, using a primer of the gene sequence (P1) and primers from the NeoR (P5) and glmS sequences (P6 and P7) (Fig. 1C; Fig. S1A). Again, the sequencing of the amplified PCR products confirmed the desired integration product having the 3′end of the PfRFUL gene tagged with the HA-2A-NeoR-glmS sequence (Fig. S2). On the Western blots, the anti-HA antibody detected a protein band with the expected protein size from the lysate of the PfRFULg parasite, but not one from the Dd2 parasite, indicating the proper expression of the tagged PfRFUL protein (Fig. 1D). Signals of immunofluorescence were also detected from the cytoplasm of the PfRFULg parasite using an anti-HA antibody (Fig. 1E). The insertion of the HA-2A-NeoR-glmS sequence did not affect the PfRFULg parasite growth in culture, showing similar parasitemia to that of Dd2 (Fig. 1F).

FIG 1.

Generation of the transgenic Plasmodium falciparum line (Dd2), having PfRFUL tagged with HA and glmS for gene expression knockdown. (A) Schematic of plasmid construct and recombination events for tagging the PfRFUL C terminus with a 3XHA-2A-NeoR-glmS ribozyme sequence. The primers CF and CR were used to amplify the 3′ end of the Pfrful gene. Primers P1 to P7 were used to confirm the integration of the intact plasmid construct. (B) PCR detection of the transgenic PfRFUL-HA-2A-NeoR-glmS by amplifying across integration sites, using primers as indicated in panel A. (C) Amplification of expected products from parasite cDNA using P1 in the parasite Pfrful gene (out of the cloned section) and primers in the NeoR gene (P5) and glmS sequence (P6 and P7). No product was amplified from the Dd2 control. (D) Western blotting showing the expression of HA-tagged endogenous PfRFUL during intraerythrocytic development (IDC), compared to the WT Dd2 parasite, anti-β-actin as a loading control, and protein molecular weights as indicated on the left. (E) Immunofluorescence assay (IFA) of parasites expressing HA-tagged PfRFUL that was detected using anti-HA antibodies. The PfRFUL protein is shown in green. Nuclei are stained with 4′,6-diamidino-2-phenylindole (DAPI, blue). (F) Dynamics of Dd2 and PfRFULg parasitemia. Parasites were synchronized with 5% d-sorbitol and placed in culture with an initial 0.1% parasitemia with a 2.5% hematocrit. The parasitemia was calculated by counting the parasites on a thin blood smear under a microscope.

PfRFUL knockdown reduces the ubiquitination levels of many proteins.

Next, we treated the PfRFULg parasite with 1, 3, 5, and 10 mM glucosamine (GlcN) for 24 h and 48 h before collecting the parasites for the protein expression and ubiquitination analyses using Western blots. The expression levels of PfRFUL were significantly reduced in a dose-dependent manner after treatment with GlcN from 3 to 10 mM for 24 h and from 1 to 10 mM for 48 h (Fig. 2A and B; Fig. S3). The efficiency of PfRFUL KD depended on the GlcN dosage and the treatment time. For example, we could achieve approximately 40 to 50% reduction in the PfRFUL level after 3 mM treatment for 48 h. The total protein ubiquitination levels were also significantly decreased with the increase of the GlcN concentration, starting from 1 mM for both the 24 h and 48 h treatments, although the total protein ubiquitination levels were more variable between experiments (Fig. 2A and C; Fig. S3), suggesting that PfRFUL ubiquitinates many parasite protein substrates. The PfRFULg and Dd2 parasites grew similarly under low dose GlcN (3 mM) for at least 4 days (Fig. S4A), and no detectable morphological change was observed between the two parasites (Fig. S4B). At higher GlcN levels (5 and 10 mM), the parasites appeared to grow slower; morphologically, no toxicity effects were visible after 48 h of GlcN treatment (Fig. S4C).

FIG 2.

The knockdown of the PfRFUL level using glucosamine (GlcN) affects protein ubiquitination. (A) PfRFUL levels at the presence of different concentrations of GlcN (1, 3, 5, and 10 mM) for two treatment periods (24 h and 48 h), as shown on a Western blot. The PfRFUL protein levels were detected using anti-HA, and the ubiquitination of the total proteins was stained using anti-ubiquitin antibodies. The Ponceau S-stained membrane immediately after protein transfer and β-actin (reprobe of the same membrane) were included as protein loading controls. (B) Plots of mean scanned PfRFUL protein band signals from repeats of Western blots, as shown in panel A. The signals from GlcN treated samples were compared to those of nontreated samples. The protein signals were normalized using the total protein signals from the Ponceau S-stained bands. (C) The same plots as in panel B, using the signals from the anti-ubiquitin antibody to measure the total ubiquitinated protein signals. (D) Western blot detecting the protein ubiquitination levels of Dd2 and the PfRFULg parasites after treatment with 5 and 10 mM GlcN for 24 h. The Ponceau S-stained membrane and β-actin were used as protein loading controls. (E) Plots of the mean scanned band signals from repeats of the Western blots from panel D. The protein signals were normalized using the total protein signals from the Ponceau S-stained bands. (F) Western blot images showing PfRFUL and the total ubiquitinated protein levels with and without 10 mM GlcN treatment for 24 h, as well as a sample taken 24 h after the removal of GlcN. (G and H) Plots of the mean scanned PfRFUL protein and the total ubiquitinated protein signals from repeats of the Western blots from panel F. For panels B, C, E, G, and H, a Mann-Whitney U test (n = 4; mean and standard error) was used. *, P < 0.05; **, P < 0.01. (I) Relative mRNA levels of the Pfrful gene with or without 3 mM GlcN treatment for 3 days. A Mann-Whitney U test (n = 6, mean and standard error) was used. ***, P < 0.001.

We also examined the effects of GlcN on the total protein ubiquitination in Dd2 WT parasites. A significant reduction in total protein ubiquitination was observed in the WT Dd2 parasite after 10 mM GlcN treatment for 24 h, but not after treatment with 5 mM GlcN (Fig. 2D and E; Fig. S5). No significant difference in ubiquitination was observed between Dd2 and PfRFULg parasites treated with 0 mM or 5 mM GlcN; however, at 10 mM GlcN, the PfRFULg parasite had significantly lower protein ubiquitination levels than did WT Dd2. Therefore, high concentrations of GlcN may impact parasite ubiquitination, even without Pfrful KD. The PfRFUL KD was reversible because the levels of PfRFUL and total protein ubiquitination increased to the original levels after the removal of GlcN (Fig. 2F–2H; Fig. S6). We also performed RT-qPCR to estimate the mRNA levels of Pfrful in parasites with and without 3 mM GlcN treatment for 3 days. A significant reduction in the Pfrful mRNA level was observed in the PfRFULg parasite after GlcN treatment, compared to nontreated PfRFULg or treated Dd2 parasites (Fig. 2I).

Effects of proteasome and E3 ubiquitin ligase inhibitors on protein expression.

The levels of PfRFUL protein and total protein ubiquitination were determined in the presence of epoxomicin (Epo), an inhibitor of proteasomal protein degradation. Using two concentrations at two treatment times, 20 nM for 18 h and 50 nM for 6 h, we observed increased levels of PfRFUL in the presence of Epo with or without GlcN (10 mM) treatment (Fig. 3A and B; Fig. S7). The addition of Epo also increased the ubiquitination levels of some other proteins (Fig. 3A and C), consistent with the inhibition of overall proteasome degradation activity by Epo.

FIG 3.

Effects of proteasome inhibitor epoxomicin (Epo) and ligase inhibitor JNJ26854165 (JNJ) on the levels of PfRFUL and other proteins. (A) Western blot showing the protein levels of PfRFUL and ubiquitinated proteins from parasites treated with glucosamine (GlcN, 10 mM) and/or Epo (20 or 50 nM). Signals were detected using anti-HA (PfRFUL) or anti-ubiquitin antibodies. The Ponceau S stain and β-actin were used as loading controls. (B and C) Plots of the scanned signals from four Western blot repeats for the PfRFUL protein (B) and the total ubiquitinated proteins (C). The signals were normalized using the total signals from the Ponceau S-stained proteins. A Mann-Whitney U test (n = 4, mean and standard error) was used. *, P < 0.05. (D) Western blot showing the protein levels of PfRFUL and the ubiquitinated proteins after treatment with glucosamine (GlcN, 10 mM) and/or E3 ligase inhibitor JNJ (1 μM). Ponceau S-stained proteins and β-actin were used as loading controls. (E and F) Plots of scanned signals from Western blot repeats for the PfRFUL protein (E) and the total ubiquitinated proteins (F). The signals were normalized using the total signals from the Ponceau S-stained total proteins. A Mann-Whitney U test (n = 4, mean and standard error) was used. *, P < 0.05.

We also evaluated the effects of an E3 ubiquitin ligase inhibitor, JNJ26854165 (JNJ), on the levels of PfRFUL and total protein ubiquitination. JNJ is known to block the development of P. falciparum at the trophozoite and schizont stages (28). Here, we treated the parasites with 1 μM JNJ for 24 h and measured the PfRFUL expression in the presence of JNJ, GlcN, or both. The results showed that JNJ decreased the PfRFUL level and that there was an additive effect on the PfRFUL protein level in the JNJ and GlcN cotreatment group (Fig. 3D and E; Fig. S8). However, the decreases in the levels of ubiquitination on total proteins were not statistically significant (Fig. 3D and F). Why JNJ can decrease the PfRFUL levels requires further investigation.

PfRFUL KD renders parasites more sensitive to antimalarial drugs.

The parasite ubiquitination system has been shown to play an important role in drug resistance (15), which raised the possibility that PfRFUL knockdown can affect parasite responses to antimalarial drugs. The in vitro drug responses of PfRFULg and Dd2 parasites with or without GlcN treatment were measured via a SYBR green 1 assay, as described previously (40). In control experiments, we compared Dd2 and PfRFULg parasites in assays using dihydroartemisinin (DHA), amodiaquine (AMQ), piperaquine (PPQ), mefloquine (MFQ), and chloroquine (CQ) without accompanying GlcN treatment; the results showed no differences between the Dd2 and PfRFULg responses to these drugs (Fig. 4A–4E). For the GlcN treatment experiments, Dd2 and PfRFULg parasites were first grown in a culture medium containing 5 mM GlcN for 48 h, and they were then switched to a medium containing 2.5 mM GlcN for the parasites before setting up the drug assays. Low levels of GlcN (2 to 3 mM) were previously shown to have minimal effects on parasite growth (39), and our results showed no significant difference in growth between Dd2 and PfRFULg at 3 mM GlcN for at least 72 h (Fig. S4A). Therefore, we used 2.5 mM GlcN in the drug assays. The results from the drug tests showed that, compared to Dd2, PfRFULg parasites were approximately twofold more sensitive to DHA, AMQ, PPQ, and MFQ, but not to CQ (Fig. 4F–J). We also tested the parasite responses to ART, artemether, artesunate, artemotil, arterolane, DHA, artemisone, and artefenomel in an automatic high-throughput assay (41). The data from this drug panel showed 1.41-fold to 2.24-fold reductions in the IC₅₀ values of PfRFULg parasites treated with 3 mM GlcN, relative to parasites that did not receive GlcN treatment (Table S1). Taken together, these results are consistent with an effect of PfRFUL on parasite responses to many antimalarial drugs, including ART.

FIG 4.

In vitro drug responses of parasites with and without glucosamine (GlcN) treatment. The antimalarial activities of different drugs were tested using a SYBR green proliferation inhibition assay, as described in Materials and Methods. (A–E) The drug dose-response curves for Dd2 and the PfRFULg parasites were tested against those of dihydroartemisinin (DHA), amodiaquine (AMQ), piperaquine (PPQ), mefloquine (MFQ), and chloroquine (CQ) without GlcN treatment. (F–J). Similar response curves as for panels A–E, but with the treatment of 5 mM GlcN for 48 h before the drug tests and of 2.5 mM GlcN during the drug assay (72 h).

Ring survival assays (RSAs) were performed to evaluate parasite recovery after a brief (6 h) exposure to 700 nM DHA. We first evaluated the effect of GlcN on parasitemia and observed lower Dd2 and PfRFULg parasitemias after 2.5 mM GlcN treatment for 3 days. No significant difference was detected between Dd2 and PfRFULg parasites (Fig. 5A). The parasites were then incubated with DHA for 6 h with or without prior GlcN treatment. The parasitemias on day 12 were similar for Dd2 and PfRFULg parasites treated with DHA, and the parasites treated with both 2.5 mM GlcN and DHA had significantly lower parasitemia than did those without GlcN treatment (Fig. 5B). Importantly, after treatment with both GlcN and DHA, PfRFULg showed markedly lower parasitemia than did Dd2, consistent with the increased DHA sensitivity from the KD of PfRFUL (Fig. 5B).

FIG 5.

Ring survival assays (RSAs) after treatments with glucosamine (GlcN) and dihydroartemisinin (DHA). The RSAs were performed as described in Materials and Methods, modified from the previously reported protocol (81). Due to the low parasitemia after the GlcN treatment, the parasitemia was monitored for a longer period than the 72 h that is often used in RSAs. (A) Ring survival after treatment with 2.5 mM GlcN for 72 h. The parasite cultures were adjusted to 1% parasitemia, and parasitemia was counted from blood smears at 72 h posttreatment. (B) Parasitemias after treatment with DHA for 6 h with or without prior GlcN treatment. Parasitemia was counted on day 12 post-DHA treatment. Mann-Whitney U tests (n = 4, mean and standard error) were used for all of the experiments. ***, P < 0.001.

Altered levels of key proteins in drug resistance and the invasion of RBCs.

To investigate the potential mechanism of PfRFUL in modifying parasite responses to antimalarial drugs, we measured the protein expression levels of PfMDR1 and PfCRT that serve as drug transporters in the membrane of the digestive vacuole (DV). Mutations and/or copy number variations of PfMDR1 and PfCRT are known to alter the responses of parasites to multiple drugs, including CQ, AMQ, MFQ, PPQ, and ART (42–51). Therefore, we performed quantitative Western blotting to investigate the effects of PfRFUL KD on the levels of PfMDR1 and PfCRT. A significantly lower PfMDR1 level was observed in the PfRFULg parasite after 3 mM GlcN treatment for 72 h, compared with those of untreated PfRFULg, untreated Dd2, or Dd2 treated with GlcN (Fig. 6A and B; Fig. S9). The PfCRT levels from the Western blots were also assessed for correlations with drug responses. Two bands of PfCRT were detected using rabbit sera against the PfCRT protein (44). PfRFUL KD significantly changed the protein ratio of the large PfCRT band (or upper band) over the smaller band (or lower band) in the PfRFULg parasite (the ratio for Dd2 was not changed significantly), and this was possibly due to the altered ubiquitination of the protein (Fig. 6C and D; Fig. S10). However, the total amount of PfCRT protein was not changed (Fig. 6E). These observed alterations of PfMDR1 and PfCRT suggest that both are targets of PfRFUL and that PfRFUL KD can consequently affect parasite responses to antimalarial drugs trafficked by these transporters. No significant change in the level of plasmepsin I, another protein associated with the DV, was detected on the Western blot after the PfRFUL KD (Fig. 6F and G; Fig. S11).

FIG 6.

PfRFUL knockdown changes the levels of the proteins associated with the digestive vacuole and parasite invasion of red blood cells (RBCs). (A) Western blot of Plasmodium falciparum multiple drug resistance 1 (PfMDR1) abundance after the knockdown of PfRFUL (detected by anti-HA antibodies). Ponceau S-stained proteins and β-actin were used as loading controls. (B) Plots of the scanned signals from four Western blot repeats of panel A and Fig. S9. The signals were normalized using the total signals from the Ponceau S-stained total proteins. (C) Western blot of the protein levels of the P. falciparum chloroquine resistance transporter (PfCRT) after the knockdown of PfRFUL. Ponceau S-stained proteins and β-actin were used as loading controls. (D) Plots of the signal ratios of larger over smaller PfCRT bands, scanned signals from four Western blot repeats of panel C and Fig. S10. The signal from each band was adjusted by the signals from the Ponceau S lane prior to the calculation of the ratio. (E) Plots of the total signals from the upper band and lower band of PfCRT. (F and G) Similar experiments as in panels A and B, but testing the P. falciparum plasmepsin I (PM-I) levels. Mann-Whitney U tests (n = 4, mean and standard error) were used for all of the experiments. *, P < 0.05.

Since the 3-day treatments with 2.5 mM GlcN decreased the Dd2 and PfRFULg parasitemia similarly in the RSAs (Fig. 5A), we investigated whether GlcN treatment could affect the efficiency of the parasite invasion of RBCs. No differences in the efficiency of the invasion of RBCs were detected between Dd2 and PfRFULg parasites after 3 mM GlcN treatment for 24 h, although continuous 3 mM GlcN treatment for another 24 h could significantly reduce the invasion efficiencies of both parasite lines (Dd2+GlcN-C and PfRFULg+GlcN-C groups) (Fig. S12A). Compared with Dd2, a significant reduction of parasitemia by PfRFULg parasites was observed after 7.5 mM GlcN treatment for 24 h that was followed by continued, low-level exposure to 2 mM GlcN. A similar reduction was observed after 10 mM GlcN treatment for 24 h with or without further exposure to 2 mM GlcN (Fig. S12B). These results show that low-dose GlcN (2 to 3 mM) affects the parasitemia of Dd2 and PfRFULg similarly, indicating that the drug sensitivity differences between Dd2 and PfRFULg in the presence of 2 to 3 mM GlcN were due to the effects of PfRFUL knockdown. GlcN was previously shown to specifically arrest the maturation of trophozoites through the inhibition of the enzyme that transfers the fatty acyl moiety to the inositol residue of glucosamine-phosphatidylinositol (GlcN-PI) during glycosylphosphatidylinositol (GPI) biosynthesis (52).

DISCUSSION

In this study, we generated a transgenic P. falciparum parasite line having an E3 ligase tagged with a HA-2A-NeoR-glmS sequence that allows for the conditional knockdown (KD) of gene expression (38, 39). The addition of GlcN induces glmS ribozyme activation, leading to the degradation of gene-specific mRNA, which can be reversed after removing the GlcN. Our results show that GlcN can significantly reduce the levels of mRNA and protein encoded by Pfrful, which in turn affect the ubiquitination and expression of multiple protein substrates. Knocking down PfRFUL renders the parasites more sensitive to many antimalarial drugs, including ART. Changes in the expression of some key drug transporters (the levels of PfMDR1 and molecular forms of PfCRT) may be involved in these altered drug sensitivities. Our observations suggest that PfRFUL plays a role in parasite responses to antimalarial drugs.

Polymorphisms and/or changes in the expression of PfCRT and PfMDR1 confer or contribute to resistance to CQ, AMQ, PPQ, ART, MFQ, and lumefantrine (LM, also known as benflumetol), as well as other drugs (42, 44, 46, 48, 49, 53–62). Further, a HECT ubiquitin ligase may act on PfCRT to modulate the QN and QD IC₅₀ responses of P falciparum by as much as twofold, depending upon the genetic background (24). Here, we found that PfRFUL KD can affect the PfMDR1 and PfCRT levels, potentially through the altered ubiquitination of the proteins, and that these changes are associated with altered responses to antimalarial drugs. Additionally, PfCRT has been shown to be a phosphorylated protein (63, 64). It is possible that PfRFUL may affect enzymes that play a role in protein phosphorylation.

PfRFULg parasites were more sensitive to various antimalarial drugs, including DHA, AMQ, PPQ, and MFQ, but not CQ, after PfRFUL KD. These results suggest potential drug resistance modulation by the activity of PfRFUL, particularly through PfMDR1 on the responses to DHA, MFQ, and LM, although we cannot rule out that other proteins involved in the parasite drug responses may also have been affected by the PfRFUL KD. Possible functional associations between PfRFUL and the HECT ubiquitin ligase have not been investigated and remain to be established.

Our observations of reduced levels of PfMDR1 after PfRFUL KD are opposite to what might have been expected from a protein substrate marked by ubiquitination for proteasome degradation. However, not all ubiquitination processes lead to protein degradation. Polyubiquitinations through the K27 and K48 residues are known to mark a protein for degradation, whereas polyubiquitination through K63 does not lead to protein degradation (65). Instead, K63-linked Ub chains act as a specific signal for sorting proteins to multivesicular bodies (66). Moreover, the addition, removal, and recognition of histone posttranslational modifications or the ubiquitination of transcription factors can affect gene expression (67, 68), which may in turn influence the expression levels of the drug transporters. More studies are required to elucidate the specific ubiquitination types on PfMDR1 and the subsequent effects on protein expression levels.

The observation that PfRFUL KD reduces the IC₅₀ values of DHA, AMQ, PPQ, and MFQ, but not CQ, is intriguing. The correlations of parasite responses to DHA, MFQ, LM, and other compounds with pfcrt and pfmdr1 polymorphisms were previously observed in 61 P. falciparum field isolates, suggesting interacting transport pathways and drug resistance mechanisms mediated by PfMDR1 and PfCRT (55). Genetic linkage studies have associated modulations of various drug phenotypes to PfMDR1 haplotypes (46, 58, 59), and work using a Xenopus oocyte expression system has suggested that WT PfMDR1 transports LM and MFQ into the digestive vacuole (DV) of the parasite, away from the cytosol, where targets of MFQ and LM may reside, thereby reducing the sensitivity of the parasite to the drugs (57). While the DV is an organelle in which many antimalarial drugs, including CQ, ART, AMQ, LM, MFQ, and QN, can promote heme toxicity by forming complexes that inhibit hemozoin formation (69, 70), the effects of PfRFUL KD and decreased PfMDR1 may not reduce the toxicity in the DV enough to offset the increased toxicity that results from the higher concentrations of these drugs in the cytoplasm (or an organelle, such as the mitochondrion), potentially explaining the greater sensitivity of the PfRFUL KD parasites to DHA, MFQ, and LM. Our findings of no detected change in the CQ IC₅₀ response after the PfRFUL KD may relate to the total intensity of the two PfCRT bands on quantitative Western blots. In this hypothesis, each PfCRT form would be active in CQ efflux, whereas the effect of PfMDR1 KD on CQ efflux would be small, with respect to that of PfCRT, in the CQ-resistant Dd2 parasites (71). One possible explanation for the change in the intensity of the PfCRT bands is that PfRFUL directly affects the ubiquitination of PfCRT. Second, PfRFUL may affect some other enzymes that catalyze protein phosphorylation, and these enzymes would in turn affect PfCRT phosphorylation. PfCRT is known to be a phosphorylated protein (72). Third, the PfCRT protein could be processed into two peptides with different lengths. Again, the influence of PfRFUL on this process would be indirect. These potential protein modifications can change the molecular weight of PfCRT, but they may not affect its CQ transport function. However, we do not have experimental data to support any of these speculations at this point.

Mutations in PfK13 have been associated with increased ring stage survival after short, high-dose exposure to ART (21). The Kelch 13 protein is predicted to be a ubiquitin E3 ligase substrate adaptor, and it has been shown to play a role in hemoglobin uptake and ring stage survival under ART pressure (17, 22, 23). ART treatment may affect the parasite protein ubiquitination and proteosome-mediated protein degradation (16, 19). DHA can kill malaria parasites by damaging proteins or by inhibiting proteasome degradation, which leads to the accumulation of damaged and polyubiquitinated proteins (19). In a recent study, it was shown that the treatment of malaria parasites with ART led to the accumulation of ubiquitinated proteins by blocking the destruction of the ubiquitinated substrates mediated by PfDdi1, a protease that hydrolyzes ubiquitinated substrates (73). Additionally, the synergistic killing of P. falciparum parasites was observed using DHA and compounds that target catalytic proteasome subunits (74). PfRFUL KD may also affect PfK13 expression; therefore, it may also influence parasite responses to ART.

Our study suggests that PfRFUL KD can significantly alter parasite responses to many antimalarial drugs through the ubiquitination of multiple proteins including PfMDR1 and PfCRT. The ubiquitination system is vital to cellular functions that support the proper growth and progression of the parasite life cycle (3, 19). In ubiquitination pathways, E3 ligases are critical for the conjugation of ubiquitin to protein substrates and thus offer potential targets for new antimalarial drug discovery (12, 13, 28). E3 ligase inhibitors, such as JNJ, HLI 373, and Nutlin-3, are active against both CQ-sensitive and CQ-resistant parasites (28), and they cause reductions in parasite numbers with little or no cytotoxic effect on mammalian cells (Vero and transformed THP1). Our findings support proposals that E3 ligase inhibitors may offer specific leads against malaria parasites and that PfRFUL may be a target for drug discovery.

MATERIALS AND METHODS

Cell cultures.

The Dd2 and PfRFULg parasites were cultivated at 5% hematocrit in RPMI 1640 medium that contained 0.5% Albumax II (Invitrogen), 0.25% sodium bicarbonate, and 0.1 mg/mL gentamicin. Human RBCs for parasite culture were purchased from the Interstate Blood Bank, Inc. (Memphis, Tennessee). The cultures were incubated at 37°C in an atmosphere of 5% oxygen, 5% carbon dioxide, and 90% nitrogen. The parasites were synchronized using 5% sorbitol, as previously described (75). Briefly, infected RBCs were treated with 5% sorbitol for 10 to 20 min at room temperature, washed twice with complete culture media, and placed back in the incubator. The level of parasitemia was calculated for each time point by counting three independent blood smears stained with Giemsa under a light microscope.

Plasmid construction, parasite transfection, and cloning.

The pSLI-HA-NeoR-glmS plasmid, designed for the C-terminal tagging of proteins and modified from the plasmid described previously (38), was obtained from Kirk Deitsch of Well Cornell Medicine. A 495 bp 3′-end sequence of the Pfrful gene was amplified using the forward primer 5′-ATGGCGGCCGCACTACATATAGTAATGCGAGT-3′ (CF) and the reverse primer 5′-ATGCCCGGGCAAATGTGATGTATGATCGTC-3′ (CR) and cloned into the plasmid at the NotI and Xmal sites. The parasites were transfected as previously described (76). Briefly, 0.2 cm electroporation cuvettes were loaded with 0.175 mL of erythrocytes and 100 μg of plasmid DNA in a cytomix solution. Electroporation was conducted at 310 V with a capacitance of 975 μF in a Bio-Rad Xcell Gene Pulser. Transfected RBCs were washed twice in culture media and cultured as described above. The parasites were initially selected with 4 nM WR99210 when the parasitemia reached 2 to 4%, and this was followed by neomycin (500 μg/mL neomycin G418) selection. After the neomycin selection, the parasites were again switched back to media containing WR99210 until parasites with correctly integrated plasmids were obtained. Parasites with plasmid integration were detected using PCR with one primer in the Pfrful gene and one primer in the plasmid. PCR was performed in a 25 μL reaction mixture that contained Q5 high fidelity DNA polymerase, dNTPs, and Q5 reaction buffer (New England Biolab inc.), along with primers and gDNA. The PCR was performed with the following conditions: 94°C for 3 min, 30 cycles of 94°C for 25 s, 58°C for 25 s, and 68°C for 60 s. The PCR products were visualized using ethidium bromide under UV light. The primer positions and sequences are shown in Fig. 1A and in Fig. S1. Selected PCR products were sequenced by a commercial company.

Immunofluorescence assay (IFA).

The IFA was performed as previously described (77). Briefly, 1 mL of parasite culture at 5% parasitemia was washed with phosphate-buffered saline (PBS) and resuspended in a fresh fixative solution containing 4% paraformaldehyde and 0.0075% glutaraldehyde in PBS for 30 min under rotating conditions at room temperature. The sample was washed with PBS and permeabilized with 0.1% Triton X-100 in PBS for 10 min at room temperature with swirling, and this was followed by blocking with 3% BSA in PBS for 1 h at room temperature. The samples were then incubated overnight with rabbit anti-HA primary antibody (1:250 dilution in 3% BSA). After 3 washes with PBS, the cells were incubated with Alexa Fluor 594 goat anti-rabbit IgG (1:500 dilution) for 1 h. All of the antibodies used in this study are listed in Table S2. Finally, thin blood smears were made on a slide, and parasite nuclei were stained with 4′, 6-diamidino-2-phenylindole (DAPI; Electron Microscopy Sciences). The cells were observed under a confocal microscope (SP8 confocal microscope, Leica).

Western blot analysis.

The Western blot was also performed as previously described (77). Briefly, parasite cultures were synchronized, and cultures were started with the same parasitemia. The iRBCs were lysed with 0.05% wt/vol saponin, and the parasite pellets were washed in cold PBS containing an EDTA-free protease inhibitor cocktail (PIC-Roche) to remove host cell soluble proteins. The parasite pellets were solubilized in LDS reducing sample buffer (Life Technologies), boiled at 95°C for 7 min, and resolved by SDS-PAGE on a 4 to 12% Bis-Tris-acrylamide gel (Invitrogen) using MES running buffer (Life Technologies). Proteins were transferred to the nitrocellulose membrane (Roche diagnostic), and the membrane was briefly stained with Ponceau S solution before blocking with 5% skim milk (wt/vol in PBS) for 1 h at room temperature. The membrane was probed with the respective primary antibodies (Table S2) overnight at 4°C, and this was followed by washing and incubation with secondary antibodies for 1 h at room temperature. Chemiluminescence was detected using an Amersham Imager 680 imaging system and SuperSignal West Dura Extended Duration Substrate (Thermo Scientific). The protein contents of the samples were first estimated using Coomassie stain (SLCM6930, Sigma-Aldrich). For the signal normalization, the total signals from the proteins on the membrane stained with Ponceau S (0.2% in 5% acetic acid) were used because many unknown parasite proteins could be potential targets of PfRFUL. Signals from anti-β-actin antibodies were also used to estimate protein loading, including both parasite and host β-actin.

Drug sensitivity assay.

Parasite responses to antimalarial drugs were measured using a SYBR green growth inhibition assay as previously described (78). Briefly, 100 μL parasites of 1% parasitemia at a 1% hematocrit were added to wells in a 96-well plate that contained 100 μL medium with twofold serially diluted drugs in triplicate and incubated at 37°C for 72 h. DNA was stained with 20× SYBR green I SYBR (Invitrogen) in lysis buffer that contained 20 mM Tris (pH 7.5), 5 mM EDTA, 0.008% (wt/vol) saponin, and 0.08% (vol/vol) Triton X-100. The plate was read in a FLUOstar Optima microplate reader (BMG Labtech) after 30 min of incubation in the dark. All of the drugs were dissolved in dimethyl sulfoxide (DMSO), except for CQ (which was dissolved in water), and stored at −80°C until use. The high-throughput SYBR assay for the responses to the ART derivatives was as previously described (41, 79).

Ring survival assays.

RSAs were performed as previously described (80), with slight modifications. Parasite cultures were synchronized twice with 5% sorbitol. After the sorbitol treatment, the parasite cultures were adjusted to 1% parasitemia with 1% hematocrit in complete culture media that contained either 0.1% DMSO or 700 nM DHA dissolved in 0.1% DMSO. After the DHA treatment, the culture was washed three times with incomplete media and further divided into two parts, with and without glucosamine (2.5 to 10 mM). Parasite growth was monitored for 72 h or longer (up to 12 days, due to the slow recovery of the parasites after treatment with both GlcN and DHA). The parasites on thin blood smears stained with Giemsa were counted every other day.

Red blood cell invasion efficiency assay.

For the erythrocyte invasion assay, cultured parasites were highly synchronized with 5% sorbitol and maintained at 1% parasitemia in complete culture media with 1% hematocrit. Ring-stage parasites were treated with 3.0, 7.5, or 10-mM GlcN for 24 h. After 24 h of GlcN treatment, the parasites were divided into two parts and cultured in media with and without 2.0 mM GlcN. The parasites were monitored for ring stage and parasitemia for 24 h.

Real-time quantitative PCR (RT-qPCR).

Total RNA was isolated from cultured parasites using the TRIzol reagent (Invitrogen) and reverse-transcribed using an iScript Advanced cDNA Synthesis Kit (Bio-Rad), according to the manufacturer’s instructions. RT-qPCR was performed using the SsoAdvanced Universal SYBR green Supermix (Bio-Rad) with a typical 3-step real-time PCR protocol in a Bio-Rad CFX qPCR instrument using the following primers: 5′-TGAAGAGGAATACAGACAAAGCA-3′ and 5′-AAGGATTGCGATTAGGAGCA-3′. The PCR cycles included initial denaturation at 95°C for 3 min, denaturing at 95°C for 15 s, annealing at 52°C for 30 s, and extension at 62°C for 30 s for 40 cycles.