Research Article

Validation of the protein kinase PfCLK3 as a multistage cross-species malarial drug target

See allHide authors and affiliations

Science  30 Aug 2019:
Vol. 365, Issue 6456, eaau1682
DOI: 10.1126/science.aau1682

Targeting parasite's protein kinase

Malaria elimination goals are constantly eroded by the challenge of emerging drug and insecticide resistance. Alam et al. have taken established drug targets—CLK protein kinases involved in regulation of RNA splicing—and investigated how inhibition of the parasite's enzymes blocks completion of its complex life cycle. They identified an inhibitor of the parasite's CLK protein kinase that was 100-fold less active against the most closely related human protein kinase and effective at clearing rodent malaria parasites. Not only does this compound halt the development of sexual stages but it also limits transmission to the mosquito vector of the parasite, a key requirement for malaria drugs.

Science, this issue p. eaau1682

Structured Abstract


Despite the positive effects of intervention strategies that include insecticide-impregnated bed nets and artemisinin-based drug therapies, malaria still kills nearly 500,000 people per year and infects more than 200 million individuals globally. This, together with the emerging resistance of the parasite to frontline antimalarials, means that there is an urgent need for novel treatments that not only offer a cure for malaria but also prevent transmission. We show that by inhibiting an essential protein kinase that is a key regulator of RNA processing, we are able to kill the parasite in the blood and liver stages as well as prevent the development of the sexual-stage gametocytes, thereby blocking transmission to the mosquito.


Our group has previously published a list of 36 protein kinases that are essential for blood-stage survival of the most virulent form of the human malaria parasite, Plasmodium falciparum. Here, we focused on one of these protein kinases from the P. falciparum CLK (cyclin-dependent–like kinase) family, PfCLK3, and reasoned that inhibition of this protein kinase by a small drug-like molecule would be effective at killing blood-stage parasites. We further hypothesized that because PfCLK3 plays a key role in RNA splicing, inhibition of this kinase would be effective at killing the parasite at all stages of the life cycle where RNA splicing is required. This would include blood, liver, and sexual stages.


By screening a focused library of nearly 30,000 compounds, we identified a probe molecule that selectively inhibited PfCLK3 and killed blood-stage P. falciparum. Using a combination of evolved resistance and chemogenetics, we established that our probe molecule had parasiticidal activity by inhibition of PfCLK3. We further showed that inhibition of PfCLK3 in parasites resulted in a reduction in more than 400 gene transcripts known to be essential for parasite survival. The finding that the vast majority of the genes down-regulated by PfCLK3 inhibition contained introns supported the notion that inhibition of PfCLK3 killed the malaria parasite by preventing the splicing of essential parasite genes. Because there is a high degree of homology between orthologs of CLK3 in other Plasmodium species, it might be expected that our probe molecule would both inhibit CLK3 contained in other malaria parasite species and have effective parasiticidal activity in these parasites. This was indeed found to be the case, with our molecule showing potent inhibition of CLK3 from P. vivax and P. berghei as well as killing the blood stages of P. berghei and P. knowlesi. Furthermore, we demonstrated that CLK3 inhibition also kills liver-stage P. berghei parasites and prevents P. berghei infection in mice. Finally, we showed that inhibition of PfCLK3 prevents the development of P. falciparum gametocytes, thereby blocking the infection of mosquitoes.


We found that inhibition of the essential malaria protein kinase CLK3 can kill multiple species of malaria parasites at the blood stage as well as killing liver-stage parasites and blocking transmission of the parasite to mosquitoes by preventing gametocyte development. In this way, we validate Plasmodium spp. CLK3 as a target that can offer prophylactic, curative, and transmission-blocking potential.

PfCLK3: A new drug target for malaria.

Inhibition of the malaria parasite protein kinase CLK3 with our probe molecule TCMDC-135051 inhibits the development of liver-stage parasites, kills asexual blood-stage parasites at the trophozoite and schizont stages of the erythrocytic cycle, blocks the development of sexual gametocytes that infect mosquitoes, and blocks exflagellation that results in male gametes.


The requirement for next-generation antimalarials to be both curative and transmission-blocking necessitates the identification of previously undiscovered druggable molecular pathways. We identified a selective inhibitor of the Plasmodium falciparum protein kinase PfCLK3, which we used in combination with chemogenetics to validate PfCLK3 as a drug target acting at multiple parasite life stages. Consistent with a role for PfCLK3 in RNA splicing, inhibition resulted in the down-regulation of more than 400 essential parasite genes. Inhibition of PfCLK3 mediated rapid killing of asexual liver- and blood-stage P. falciparum and blockade of gametocyte development, thereby preventing transmission, and also showed parasiticidal activity against P. berghei and P. knowlesi. Hence, our data establish PfCLK3 as a target for drugs, with the potential to offer a cure—to be prophylactic and transmission blocking in malaria.

Despite artemisinin-based combination therapies offering effective frontline treatment for malaria, there are still more than 200 million cases of malaria worldwide each year, resulting in an estimated 500,000 deaths. This, combined with the fact that there is now clear evidence for the emergence of resistance not only to artemisinin (1, 2) but also to partner drugs including piperaquine and mefloquine (3, 4), means that there is an urgent need for novel therapeutic strategies to cure malaria while also preventing transmission. Global phospho-proteomic studies on the most virulent species of human malaria, Plasmodium falciparum, have established protein phosphorylation as a key regulator of a wide range of essential parasite processes (58). Furthermore, of the 65 eukaryotic protein kinases in the parasite kinome (9), more than half have been reported to be essential for blood-stage survival (812). These studies, together with the generally accepted potential of targeting protein kinases in the treatment of numerous human diseases (13, 14), suggest that inhibition of parasite protein kinases might offer a viable strategy for the treatment of malaria (6, 15)

To directly test this hypothesis, we focused on one of the four members of the P. falciparum cyclin-dependent–like (CLK) protein kinase family, PfCLK3 (PF3D7_1114700), a protein kinase essential for maintaining the asexual blood stage of both P. falciparum (8, 12) and P. berghei (10, 11). In mammalian cells, the CLK protein kinase family and the closely related SRPK family are crucial mediators of multiple phosphorylation events on splicing factors, including serine-arginine–rich (SR) proteins, which are necessary for the correct assembly and catalytic activity of spliceosomes [reviewed in (16)]. A key member of the human CLK family is the splicing factor kinase PRP4 kinase (PRPF4B), which homology-based studies have identified as the closest related human kinase to PfCLK3 (17, 18). PRPF4B plays an essential role in the regulation of splicing by phosphorylation of accessory proteins associated with the spliceosome complex (19). The finding that PfCLK3 can phosphorylate SR proteins in vitro (20) supports the notion that PfCLK3, like the other members of the PfCLK family (17), plays an essential role in parasite pre-mRNA processing (18).

High-throughput screen identifies selective PfCLK3 inhibitors

We established high-throughput inhibition assays for two essential members of the PfCLK family, PfCLK1 and PfCLK3 (fig. S1). Both of these protein kinases were purified as active recombinant proteins (fig. S2A) and were used in a high-throughput time-resolved florescence resonance energy transfer (TR-FRET) assay, showing robust reproducibility in 1536-well assay format (Z′ > 0.7) (fig. S2, B to H). This assay was used to screen 24,619 compounds, comprising 13,533 compounds in the Tres Cantos Anti-Malarial Set (TCAMS) (21), 1115 in the Protein Kinase Inhibitor Set (PKIS) (22), and 9970 in the MRCT index library (23), at a single dose (10 μM). Hits were defined as those compounds that were positioned >3 standard deviations from the mean of the percent inhibition distribution curve (Fig. 1, A and B) and that also showed >40% inhibition. This identified 2579 compounds (consisting of MRCT = 250, PKIS = 4, TCAMS = 2325), which, together with the 259 compounds identified as “the kinase inhibitor set” from within the Medicines for Malaria Venture (MMV) box, a collection of 400 antimalarial compounds (24), were used to generate concentration inhibition curves (Fig. 1C and table S1). On the basis of the selectivity criterion of a difference of more than 1.5 log units in the negative logarithm of the half-maximal inhibition (pIC50), 28% of the hits showed specific inhibition of PfCLK1 and 13% specifically inhibited PfCLK3, whereas 23% of the compounds inhibited both PfCLK3 and PfCLK1; the remainder (36%) were inactive (Fig. 1, C and D, and table S1). Exemplar molecules from each of these three classes are shown in fig. S3.

Fig. 1 High-throughput screen identifies inhibitors of PfCLK1 and PfCLK3.

(A) Percent inhibition distribution pattern of compounds screened against PfCLK3, binned in 5% intervals. Active “hit” compounds were defined as those that were positioned >3 SDs from the mean. (B) Pie chart summary of the primary single-dose screen. (C) Hit compounds were used in concentration response curves. Shown is a comparison of pIC50 values for inhibition of PfCLK3 versus PfCLK1. TCMDC-135051 (structure shown) is highlighted as the most potent and selective PfCLK3 hit. (D) The same data as shown in (C) but in pie chart format of compounds designated as inactive, pan-active (active against both PfCLK1 and PfCLK3), or selective for either PfCLK1 or PfCLK3.

Highlighted in Fig. 1C is TCMDC-135051, which showed the highest selectivity and potency for inhibition of PfCLK3. TCMDC-135051 also showed lower activity against the closely related human kinase CLK2 (29% sequence identity with PfCLK3) by a factor of ~100 (table S1). Similarly, TCMDC-135051 showed no evidence of interacting with the human ortholog of PfCLK3, PRPF4B. This was seen in thermostability assays, using differential scanning fluorimetry, where staurosporine, acting as a positive control, increased the melting temperature of PRPF4B by 2.40° ± 0.14°C. In contrast, TCMDC-135051 showed no change in PRPF4B thermostability (fig. S4A). Furthermore, in a mass spectrometry–based PRPF4B activity assay, the published inhibitor Compound A (25) showed inhibition of PRPF4B, whereas TCMDC-135051 at concentrations up to 50 μM showed no inhibitory activity (fig. S4B). In further counterscreens, TCMDC-135051 showed no activity against the P. falciparum protein kinases PfPKG and PfCDPK1 (fig. S5, A to C). Thus, TCMDC-135051 showed selective inhibition of PfCLK3 when compared against the closely related human kinases PRPF4B and CLK2, as well as the closest parasite kinase, PfCLK1, and other parasite kinases (PfPKG, PfCDPK1).

TCMDC-135051 is a member of a series of molecules that were contained in the high-throughput screen with the same chemical scaffold. This series showed similar inhibitory activity against PfCLK3 (fig. S6). Note that the TCMDC-135051 is part of the TCAMS and has previously been shown to have antiparasiticidal activity (half-maximal response EC50 = 320 nM); the structure has been published (21). However, resynthesis of TCMDC-135051, together with nuclear magnetic resonance (NMR) analysis, has determined the correct structure for TCMDC-135051 to be the one shown in Fig. 1C and fig. S3.

Parasite strains resistant to TCMDC-135051 show mutations in PfCLK3

We next sought to confirm that PfCLK3 was the target of TCMDC-135051 parasiticidal activity. Exposing P. falciparum Dd2 parasites to increasing concentrations of TCMDC-135051 resulted in the emergence of three independent lines that showed decreased sensitivity to TCMDC-135051 but no change in sensitivity to chloroquine or artemisinin (Fig. 2A and Table 1). Whole-genome sequencing of the three resistant lines revealed mutations in PfCLK3 (lines TM051A and TM051C) and a mutation in the putative RNA processing protein PfUSP39 (PF3D7_1317000) (line TM051B; Fig. 2B and Table 1). The resistant clone TM051A, which contained the mutation Pro196 → Arg (P196R) in the N-terminal region outside the PfCLK3 kinase domain (Fig. 2B), showed the smallest change in sensitivity to TCMDC-135051 (factor of 4.2 shift in EC50 relative to parental Dd2 parasites). Examination of the in vitro enzymatic properties of the P196R mutant found in TM051A did not detect any changes in enzyme kinetics or sensitivity to inhibition by TCMDC-135051 relative to the wild-type kinase; this finding suggests that this mutation could potentially stabilize the protein or be otherwise involved in the interaction between PfCLK3 and its substrates or regulatory proteins.

Fig. 2 Parasites adapted to become less sensitive to TCMDC-135051 harbored mutations in the pfclk3 gene.

(A) To generate TCMDC-135051–resistant parasite lines, we cultured Dd2 parasites with increasing concentrations of TCMDC-135051 over a 2-month period. This protocol resulted in three lines that were less sensitive to TCMDC-135051 but displayed unchanged sensitivity to artemisinin (ART) and chloroquine (CQ). (B) Illustration of the position of the mutations in the pfclk3 gene and pfusp39 gene in the drug-resistant mutant lines. (C) The line showing the greatest change in sensitivity to TCMDC-135051, TM051C, expressed a mutant form of PfCLK3 (H259P) (illustrated). Shown are death curves for parental and TM051C lines. MFI, mean fluorescence intensity. (D) Enzyme activity of recombinant PfCLK3 and the H259P mutant determined at varying ATP concentrations to derive a Km for ATP. (E) TCMDC-135051 kinase inhibition curves for PfCLK3 and the H259P mutant at Km ATP concentrations for each enzyme. Data are means ± SEM of at least three independent experiments.

Table 1 Adaptive resistance to TCMDC-135051.

Dd2 parasites were exposed to subthreshold concentrations of TMDC-135051, and three lines were isolated that were less sensitive to TCMDC-135051 but had unchanged sensitivity to artemisinin and chloroquine. Shown are nucleotide changes and associated amino acid changes in genes from the resistance lines as well as the identity and annotated function of the mutant genes. EC50 values associated with each line for artemisinin, chloroquine, and TCMDC-135051 are shown, as well as the relative (fold) change in IC50 for TCMDC-135051 compared to Dd2 parent parasites. Data are means ± SEM of three experiments.

View this table:

The line TM051C, containing a His259 → Pro (H259P) mutation in PfCLK3, showed the largest degree of resistance to TCMDC-135051, with a shift in EC50 by a factor of >11 in the death curve (Fig. 2C and Table 1). Evaluation of the enzymatic properties of the H259P mutant revealed that the mutant kinase possessed ~3 times the activity of the wild-type kinase, whereas the Michaelis constant Km for adenosine triphosphate (ATP) was similar between mutant and wild-type kinases (Fig. 2, D and E). The fact that His259 resides outside of the kinase domain suggests that this amino acid is within a regulatory region that controls enzymatic activity.

In contrast to the other two resistant lines, TM051B did not contain a mutation in PfCLK3 but rather contained a mutation, Phe103 → Ile (F103I), within the putative zinc-finger ubiquitin-binding domain of ubiquitin-specific peptidase–39 (PfUSP39). The human and yeast orthologs of PfUSP39 [small nuclear ribonucleoprotein (snRNP) assembly-defective protein–1 (Sad1)] are members of the deubiquitinase family that are essential components of the U4/U6-U5 tri-snRNP complex necessary for spliceosome activity (2628). The position of the F103I mutation within the zinc-finger ubiquitin-binding domain of PfUSP39 may be of importance because this domain has been implicated in the interaction of USP39/Sad1 with the spliceosome (27). Hence, the involvement of PfUSP39 in the same pathway/function as PfCLK3, together with the mutations found in PfCLK3 itself in the other two resistant lines, supports the notion that the parasiticidal activity of TCMDC-135051 is via inhibition of PfCLK3.

Genetic target validation of PfCLK3

To further confirm PfCLK3 as the target of TCMDC-135051 parasiticidal activity, we designed a variant of PfCLK3 that showed reduced sensitivity to TCMDC-135051. To generate this variant, we took advantage of the highly selective inhibition of PfCLK3 over PfCLK1 shown by TCMDC-135051. By exchanging amino acids within subdomain V of the PfCLK3 kinase domain with equivalent residues in the kinase domain of PfCLK1 (fig. S1), a variant of PfCLK3 was generated where Gly449 in PfCLK3 was substituted by a proline residue (G449P) (Fig. 3A). This recombinant variant protein showed a factor of ~3 log shift in sensitivity for inhibition by TCMDC-135051 (Fig. 3, B and C) [PfCLK3 pIC50 = 7.35 ± 0.12 (IC50 = 0.04 μM), G449P pIC50 = 4.66 ± 0.16 (IC50 = 21.87 μM)]. The G449P variant also showed a slightly lower enzymatic activity (Fig. 3D) [PfCLK3 maximal rate of reaction (Vmax) = 1.24, G449P Vmax = 0.88] but higher Km for ATP (fig. S7A) (PfCLK3 Km = 6.29, G449P Km = 81.3) relative to wild-type PfCLK3. Single-crossover homologous recombination targeting the PfCLK3 locus with a construct designed to insert the coding sequence for the G449P mutant (Fig. 3E) generated two independent clones (A3 and A8) that expressed the G449P mutant in place of the wild-type PfCLK3 (Fig. 3, F and G). Integration of the plasmid at the target locus was verified by polymerase chain reaction (PCR) of genomic DNA (Fig. 3F), and Western blotting confirmed the expression of the G449P mutant, which was epitope-tagged with a hemagglutinin (HA) tag at the C terminus (Fig. 3G). The growth rate of the G449P-expressing mutant parasites was not different from that of control 3D7 parasites (fig. S7B). The activity of TCMDC-135051 in parasite viability assays was significantly reduced by ~1.5 log units in both clones of G449P (Fig. 3H) [negative logarithm of the half-maximal effect (pEC50) of TCMDC-135051 in the Dd2 wild type, 6.35 ± 0.038 (EC50 = 0.45 μM); in the A3 strain, 4.86 ± 0.13 (EC50 = 13.80 μM); in the A8 strain, 4.94 ± 0.051 (EC50 = 11.48 μM)], providing further evidence that TCMDC-135051 kills parasites via inhibition of PfCLK3.

Fig. 3 Chemogenetic validation of PfCLK3 as a target for the parasiticidal activity of TCMDC-135051.

(A) Schematic of the primary amino acid sequence of PfCLK3 showing the 11 kinase subdomains and the sequence of subdomain V of PfCLK1 and PfCLK3. A, Ala; C, Cys; E, Glu; F, Phe; G, Gly; H, His; I, Ile; K, Lys; L, Leu; M, Met; N, Asn; P, Pro; R, Arg; S, Ser; T, Thr; V, Val; W, Trp; Y, Tyr. (B) Gel-based assay of the phosphorylation of myelin basic protein (MBP) by PfCLK3 and a Gly449 → Pro variant (G449P). The top gel is an autoradiograph and the bottom a Coomassie stain of the same gel. (C) TCMDC-135051 inhibition of recombinant PfCLK3 and the G449P mutant. (D) Maximal kinase activity of recombinant PfCLK3 compared to the activity of the G449P mutant. (E) Schematic of gene targeting strategy that would result in the expression of the G449P mutant (containing a triple HA tag) in place of wild-type PfCLK3. (F) The recombination event illustrated in (E) was identified in cloned G449P parasite cultures by PCR (A3 and A8). (G) Expression of the triple HA-tagged G449P mutant in genetically engineered parasite cultures (G449P) was determined by Western blotting. Left, gel probe of lysates with antibodies to HA; center, a loading control probed with antibodies to PfCDPK1; right, a Coomassie stain of the lysate preparations used in the Western blots. (H) Growth inhibition curves of TCMDC-135051 against parent 3D7 parasites and G449P parasites (A3 and A8). Data in (C), (D), and (H) are means ± SEM of at least three independent experiments. *P < 0.05 (t test).

Previous efforts to make inhibitor-insensitive versions of apicomplexan protein kinases have focused on the mutation of the gatekeeper residue, a key residue in the ATP binding pocket that can provide steric hindrance to ATP competitive protein kinase inhibitors (5, 29, 30). In contrast, our approach was based on a comparison of residues between two highly related kinases (PfCLK1 and PfCLK3) that showed differential sensitivity to an inhibitor. By swapping residues between the kinases, we introduced inhibitor insensitivity into our target kinase (e.g., PfCLK3) in a strategy that could be applied to other protein kinases.

Inhibition of PfCLK3 prevents trophozoite-to-schizont transition

To characterize the phenotypic response to PfCLK3 inhibition and to understand PfCLK3 function, we treated P. falciparum 3D7 parasites synchronized at ring stage (time point zero) with 1 μM TCMDC-135051. The parasites progressed to late ring stage (time point 20 hours) (Fig. 4A) but did not progress further to trophozoite stage (time point 30 and 40 hours), arresting with a condensed and shrunken appearance (Fig. 4A). Similar effects were observed if the parasites were treated at mid-ring stage (time point 10 hours). Treatment of the parasite at later time points (20 or 30 hours) blocked development of the parasite from the trophozoite to the schizont stage. The fact that the parasites at the schizont stage were not viable was further evidenced by the absence of ring-stage parasites when the culture was continued to the 50-hour time point (Fig. 4A). These data indicated that PfCLK3 inhibition prevented the transition of the parasites at early stages (ring to trophozoite) as well as late stages (trophozoite to schizont) of development and did not allow parasites to reach the next invasion cycle (Fig. 4A). These data further indicated that PfCLK3 inhibition resulted in rapid killing, with no evidence that the compound resulted in quiescence from which the parasite could recover after drug withdrawal. These features were confirmed in parasite reduction rate assays, which showed that treatment of parasites with 10 × EC50 of TCMDC-135051 completely killed the parasite in 48 hours; viable parasites could not be observed despite maintaining the parasite culture for 28 days after withdrawal of TCMDC-135051 (Fig. 4B).

Fig. 4 Inhibition of PfCLK3 prevents trophozoite-to-schizont transition, kills the parasite with rapid kinetics, and disrupts gene transcription.

(A) Smears of synchronized blood-stage P. falciparum cultures after treatment with TCMDC-135051 (2 μM) were taken at the indicated times after TCMDC-135051 administration. (B) The in vitro parasite reduction rate in the presence of 10 × EC50 of TCMDC-135051 was used to determine the onset of action and rate of killing. Data are means ± SEM. Previous results reported on standard antimalarials tested at 10 × EC50 using the same conditions are shown for comparison (44). (C and D) Illustration of the genes that are designated as significantly changing (moderate t test, n = 4) in transcription after treatment with TCMDC-135051 (1 μM, 60 min) of either (C) parent Dd2 parasites or (D) TM051C mutant parasites. Each line represents the log2 fold change in the probes used in the microarray. Numbers of genes represented by the probes are indicated. (E) Summary of the parasite processes associated with the genes where transcription is statistically significantly down-regulated after TCMDC-135051 treatment. (F) Assessment of intron-containing genes among genes that are up-regulated and down-regulated in Dd2 parasites after TCMDC-135051 treatment. (G) Assessment of intron-containing genes in the P. falciparum genome (data derived from Plasmodb).

Inhibition of PfCLK3 disrupts transcription

Because PfCLK3 has been proposed to regulate RNA processing (20) and is closely related to the human kinases PRPF4B and CLK2 that are involved in RNA splicing (19), we investigated changes in gene transcription in parent Dd2 parasites and the drug-resistant stain TM051C in response to exposure to TCMDC-135051. RNA isolated from trophozoite-stage parasites was extracted after treatment with 1 μM TCMDC-135051 for 60 min, during which the Dd2 and TM051C parasites maintained normal morphology. Genome-wide transcriptional patterns were determined using oligonucleotide microarray chips that probed 5752 P. falciparum genes (31). Under these conditions, 779 gene transcripts were significantly down-regulated in response to PfCLK3 inhibition in the Dd2 parasites and 155 genes were up-regulated (Fig. 4C and table S2). That the majority of these transcriptional changes were due to inhibition of PfCLK3 and not off-target events was supported by the fact that under the same conditions, only six genes were up-regulated and 88 down-regulated in the resistant TM051C parasite strain (Fig. 4D and table S3). By subtracting the transcriptional changes observed in the TM051C strain, defined here as “off-target,” from those observed with the Dd2 parent, the transcriptional changes due to “on-target” inhibition of PfCLK3 were defined (table S4). Among these “on-target” down-regulated genes were those involved in key parasite processes, such as egress and invasion, cytoadherence, parasite protein export, and involvement in sexual stages, as well as housekeeping functions including metabolism, RNA processing, lipid modification, and mitochondrial function (Fig. 4E and table S4). Of the 696 “on-target” genes identified as down-regulated by PfCLK3 inhibition (table S4), 425 matched those that have recently been determined to be essential for asexual P. falciparum survival (12) (table S4).

Gene ontology enrichment analysis was used to determine biological functions that were disproportionally down-regulated by PfCLK3 inhibition. In this analysis, genes associated with key biological functions, particularly protein modification, phospholipid biosynthesis, and lipid modification, were significantly overrepresented among those genes that were down-regulated (fig. S8 and table S5). We found that 93% of the “on-target” down-regulated genes contained introns (Fig. 4F and table S4), versus 52% of genes in the P. falciparum genome that are annotated as containing introns (32) (Fig. 4G). Hence, PfCLK3 inhibition significantly affected the transcription of genes that contained introns (P < 0.0001, Pearson χ2 test), further supporting its role in splicing.

In addition to the nearly 700 genes down-regulated in response to PfCLK3 inhibition, there were 154 genes that were significantly up-regulated (table S4). Among these were genes associated with RNA processing, such as splicing factor 1 (PF3D7_1321700) and pre-mRNA splicing factor SYF1 (PF3D7_1235900), indicating that at least some of the up-regulated genes may represent compensatory mechanisms. In support of this notion was the finding that PfCLK3 itself was within the up-regulated genes (table S4).

Cross-species and in vivo activity of TCMDC-135051

It might be predicted that the close similarity between orthologs of CLK3 in different malaria parasite species would result in TCMDC-135051 showing similar activities against CLK3 from different Plasmodium species. This indeed was the case, as in vitro kinase assays using recombinant PvCLK3 (P. vivax) and PbCLK3 (P. berghei) (fig. S9, A and B) showed that TCMDC-135051 had near-equipotent inhibition at these two orthologs, with pIC50 values of 7.47 ± 0.12 (IC50 = 0.033 μM) and 7.86 ± 0.10 (IC50 = 0.013 μM), respectively (fig. S10, A and B). Furthermore, in asexual blood-stage cultures of both P. knowlesi (an experimental model for P. vivax) and P. berghei [a rodent malaria model used for in vivo drug testing (33)], inhibition of CLK3 by TCMDC-135051 resulted in parasiticidal activity in both of these Plasmodium species (Fig. 5, A and B). The potent and efficacious effects of TCMDC-135051 in blood P. berghei cultures prompted an investigation of the in vivo activity of TCMDC-135051 in mice infected with P. berghei. Twice-daily intraperitoneal dosing of TCMDC-135051 into mice infected with P. berghei resulted in a dose-related reduction in parasitemia over a 5-day infection period, where the maximal dose (50 mg/kg) resulted in near-complete clearance of parasites from peripheral blood (Fig. 5C).

Fig. 5 Inhibition of PfCLK3 has parasiticidal activity at multiple parasite species, shows in vivo parasiticidal activity in P. berghei, blocks gametocyte development, and reduces transmission to the mosquito vector.

(A and B) Concentration effect curve of TCMDC-135051 on blood-stage P. knowlesi (A) and P. berghei (B) parasites. (C) TCMDC-135051 P. berghei in vivo growth inhibition curves and day 4 percentage suppression plots (inset). Error bars are SD from n = 4 mice groups. Statistical comparisons between mice treated with drug and vehicle are shown using one-way analysis of variance and Dunnett multiple-comparisons test. ****P < 0.0001. (D to I) Concentration effect of exposure of stage II to V P. falciparum (clone 3D7) gametocytes to TCMDC-135051 on gametocyte (GC) numbers in culture [(D) and (E)], exflagellation [(F) and (G)], and prevalence (number of mosquitos with oocyst infection per number of mosquitos dissected) of infection of Anopheles coluzzii mosquitos [(H) and (I)]. (D), (F), and (H) show means ± SEM of four independent experiments; (E), (G), and (I) show the predicted effects of drug concentrations according to the maximal GLMM, with the shaded area indicating 95% confidence intervals. From the GLMM analysis, the approximate EC50 values were calculated.

Activity of TCMDC-135051 at liver invasion and sporozoite development

TCMDC-135051 showed potent activity against P. berghei sporozoites in a liver invasion and development assay (34) in which the compound showed a pEC50 value of 6.17 ± 0.10 (EC50 = 0.40 μM) (fig. S11), although hepatocyte toxicity (fig. S11) was observed but only significantly at 10 μM (fig. S11).

Targeting PfCLK3 reduces transmission to the mosquito vector

The effects of PfCLK3 inhibition on sexual-stage parasites were tested in an assay developed using the P. falciparum Pf2004 parasite strain, which shows high levels of gametocyte production (35). TCMDC-135051 showed potent parasiticidal activity in asexual-stage Pf2004 (fig. S12A) [pEC50 in Pf2004 = 6.58 ± 0.01 (EC50 = 0.26 μM)] similar to that seen in asexual 3D7 and Dd2 parasites. In addition, TCMDC-135051 showed inhibitory activity between commitment of infected red blood cells to stage II gametocytes [pEC50 = 6.04 ± 0.11 (EC50 = 0.91 μM)] (fig. S12B).

These in vitro studies were followed by mosquito membrane feeding assays to test directly the impact of PfCLK3 inhibition on transmission of P. falciparum to the mosquito vector. In these experiments, stage II gametocytes (from 3D7 parasites) were exposed to TCMDC-135051 and allowed to develop to stage V in the continued presence of drug. These experiments showed a concentration-dependent decrease in stage V gametocyte number (Fig. 5D). When analyzed using a generalized linear mixed model (GLMM), this effect had a potency of EC50 = 0.8 μM (Fig. 5E). Furthermore, the inhibition of PfCLK3 showed a statistically significant decrease in exflagellation (Fig. 5, F and G; EC50 = 0.2 μM), which, combined with the effect on gametocyte number, contributed to a statistically significant reduction in transmission, as measured by the prevalence of oocysts in the gut of mosquitos in membrane feeding assays (Fig. 5, H and I).

These studies were further extended to test the effects of PfCLK3 inhibition on stage V gametocytes. Although exposure of stage V gametocytes to TCMDC-135051 for 24 hours did not affect gametocyte number (fig. S13, A and B), a small but significant reduction in exflagellation (~25% reduction, P < 2 × 10−16 as determined by GLMM) was observed at the highest concentration tested (fig. S13, C and D). A more pronounced effect was observed in membrane feeding experiments where mosquito transmission was significantly reduced by ~50% (fig. S13, E and F, P = 4.33 × 10−6). A reduction of mosquito infection prevalence of 50% is likely to have a major effect in field conditions where infection rates in mosquitos are usually <5%.


Our results identify PfCLK3 as a valid and druggable antimalarial target for both sexual and asexual stages of parasite development, including the liver stage. This suggests that targeting PfCLK3 might be a novel strategy for developing curative treatments for malaria by clearance of asexual blood-stage parasites and as a potential prophylactic by targeting the liver stage; moreover, the parasiticidal activity afforded by PfCLK3 inhibition at gametocytes would indicate that through this mechanism, transmission to the insect vector could also be affected. Because splicing of essential transcripts occurs at many stages of the parasite life cycle, it is attractive to hypothesize that inhibition of PfCLK3, which has been implicated in the phosphorylation of splicing factors necessary for the assembly and activity of the spliceosome (17, 18, 20), would have a wide-ranging impact on parasite viability. In support of this notion is the finding that PfCLK3 inhibition down-regulated more than 400 essential parasite transcripts. Interestingly, the majority of down-regulated transcripts are from genes that contain introns (91%), providing further evidence that PfCLK3 is involved in RNA splicing and that disruption of this essential process at multiple life-cycle stages is the likely mechanism by which inhibitors of PfCLK3 have parasiticidal activity.

The similarity of CLK3 orthologs in Plasmodium spp. suggests that inhibitors might also have activity across a number of Plasmodium species. This was confirmed here by almost equipotent inhibition of the kinase activity of PvCLK3, PbCLK3, and PfCLK3 by TCMDC-135051. This in vitro action was mirrored by ex vivo activity in P. berghei and P. falciparum but also in P. knowlesi (a model for P. vivax), indicating that inhibition of Plasmodium CLK might have cross-species activity. This, coupled to the reduction in transmission after PfCLK3 inhibition, points to PfCLK3 satisfying many of the criteria set by MMV for a suitable target for next-generation antimalarials—namely, a target that can deliver rapid, multistage parasite killing across multiple species with action as a transmission blocker (36).

One of the major barriers associated with the development of protein kinase inhibitors is the issue of selectivity, because the ATP binding pocket, to which the majority of protein kinase inhibitors bind, is very similar between protein kinases (37). Here, TCMDC-135051 showed surprising selectivity toward PfCLK3 even when compared to its paralog in P. falciparum PfCLK1 and its human ortholog (PRPF4B) and the closely related human kinase CLK2. Furthermore, our transcriptional studies revealed very few off-target events, and adaptive resistance and chemogenetic resistance were associated with single point mutations in PfCLK3; these findings indicate that the selectivity of TCMDC-135051 for PfCLK3 observed in vitro was maintained in the parasite. The fact that a hit from a library screen can show such selectivity against human and parasite kinases provides encouragement that PfCLK3-selective inhibitors can be generated that might provide therapeutic efficacy with low off-target toxicity.

Lipid kinases such as phosphatidylinositol 4-kinase are considered promising targets (38) in malaria, and there is abundant evidence that phosphorylation and phosphosignaling are crucial for the viability of both asexual and sexual stages of the malaria parasite (5, 8, 10, 11). Essential parasite protein kinase targets have been identified (8, 11), and academic and industrial laboratories have gained much experience in the design of protein kinase inhibitor drugs (14, 37). But despite these developments, the targeting of parasite protein kinases in antimalarial drug development is only in its infancy (6, 39). By focusing on an essential parasite kinase and taking advantage of high-throughput phenotypic screens of commercial and academic libraries (21, 40, 41) as a starting point to screen for inhibitors, we have identified a probe molecule that has not only established the validity of PfCLK3 as a target in malaria but also determined that this protein kinase is susceptible to selective pharmacological inhibition by small drug-like molecules. In this way, our study lends weight to the argument that targeting the essential parasite protein kinases identified through global genomic studies might be a valid therapeutic strategy in the development of molecules that meet many of the criteria set for the next generation of antimalarial drugs.

Methods summary

See supplementary materials for details.

High-throughput screening

Compounds were tested in single shot at 10 μM, or in dose response from 100 μM (11-point, 3-fold serial dilutions). Screening was performed in 1536-well plates, with final reaction and read-out volumes of 4 μl and 6 μl, respectively. The results from the high-throughput screening were further analyzed using Activity Base (ID Business Solutions Ltd., Surrey, UK). For each test compound, percent inhibition was plotted against compound concentration.

Evolution of compound-resistant lines and whole-genome sequencing

The P. falciparum Dd2 strain was cultured in triplicate in the presence of increasing concentrations of TCMDC-135051 to generate resistant mutants as described (42). After approximately 60 days of selection, parasites were cloned in 96-well plates by limiting dilution (43). The half-maximal (50%) inhibitory concentration was determined in dose-response format using a SYBR Green-I–based cell proliferation assay as described (41). To determine genetic variants that arose during selection, genomic DNA was sequenced on an Illumina Mi-seq and single-nucleotide variants were detected using the Genome Analysis Toolkit (GATK v1.6).

Generation of G449 mutant parasite

A fragment of the PfCLK3 gene containing part of exon2, exon 3, exon 4, and part of exon 5 (1143 bp), corresponding to 655 to 1797 bp in clk3 genomic sequence, was amplified using primer CLK3-HR1 and CLK3-HR2 and the amplified product named as PfCLK3. The homologous region (CLK3-HR) was cloned in pHH1-derived vector using restriction sites HpaI and BglII. The rest of the clk3 gene sequence downstream of CLK3-HR, corresponding to 1798 to 3152 bp of the PfClk3 genomic sequence, was modified by removing introns and the stop codon, and the coding sequence was optimized for E. coli codon usage to make it dissimilar to Pfclk3 genomic sequence. This fragment of gene, which we named as PfCLK3-codon optimized (CLK3-CO), was commercially synthesized and included BglII recognition site at 5′ and XhoI recognition site at 3′. CLK3-CO was cloned downstream of CLK3-HR region in the parent plasmid using BglII and XhoI restriction sites in such a way that the triple HA tag sequence in the parent plasmid remained in frame with the PfClk3 sequence. The BglII restriction site that was artificially introduced for cloning purpose was mutated back to original PfCLK3 coding sequence by site-directed mutagenesis using CLK3-BglII-KN1 and CLK3-BglII-KN2 primers. Site-directed mutagenesis was used again to mutate Gly in PfCLK3 at position 449 to Ala. The targeting vector generated was used for transfection of schizont-stage parasites.

Supplementary Materials

Materials and Methods

Figs. S1 to S13

Tables S1 to S5

References (4551)

References and Notes

Acknowledgments: We thank the Proteomics facility of LaCTAD (Laboratório Central de Tecnologias de Alto Desempenho em Ciências da Vida, UNICAMP, Campinas, Brazil); A. da Silva Santiago, A. M. Fala, and P. Zonzini Ramos for assistance with PRPF4B protein production; Aché Laboratórios Farmacêuticos for provision of compound A; E. Peat and D. Armstrong for the maintenance of the IBAHCM/Glasgow University mosquito insectaries; the Scottish National Blood Transfusion service for the provision of human blood and serum; N. Emami (Stockholm University) for assistance with serum supplies; P. Johnson (IBAHCM, University of Glasgow) for discussions on GLMM; and R. Tewari for providing the P. berghei cDNA library. Funding: Supported by an MRC Toxicology Unit program grant (A.B.T., M.M.A.), MRC Developmental Gap Fund (A.S.-A.), Lord Kelvin Adam Smith Fellowship (M.M.A.), GSK Open Lab Foundation Award (A.S.-A.), joint MRC Toxicology Unit and MRC Unit the Gambia PhD program (O.J.), and Daphne Jackson Fellowship (D.M.). A.P.W., M.M., M.M.A., K.C., N.V.S., and S.B.M. are supported by Wellcome Centre for Integrative Parasitology Core support award WT104111AIA. E.A.W. is supported by grants from the NIH (5R01AI090141 and R01AI103058) and by grants from the Bill & Melinda Gates Foundation (OPP1086217, OPP1141300) as well as by Medicines for Malaria Venture (MMV). Drug WR99210 for selection of transgenic parasites was a gift from Jacobus Pharmaceuticals. M.M. is supported through WT award 172862-01 and a Wolfson Merit award from the Royal Society. The Structural Genomics Consortium (SGC) is a registered charity (number 1097737) that receives funds from AbbVie, Bayer Pharma AG, Boehringer Ingelheim, the Canada Foundation for Innovation, the Eshelman Institute for Innovation, Genome Canada, the Innovative Medicines Initiative (European Union [EU]/European Federation of Pharmaceutical Industries and Associations [EFPIA]) (ULTRA-DD grant no. 115766), Janssen, Merck & Company, Merck KGaA, Novartis Pharma AG, the Ontario Ministry of Economic Development and Innovation, Pfizer, the São Paulo Research Foundation (FAPESP number 2013/50724-5), Takeda, and the Wellcome Trust (106169/ZZ14/Z). E.F.A. was supported by the Tres Cantos Lab Foundation (grant TC125). A.B.C. was supported by a Scottish Funding Council Global Challenges Research Fund award to L.C.R.-C. Author contributions: A.B.T. conceived the project, designed experiments, analyzed data, and was the primary author; M.M.A., A.S.-A., O.J., and L.C.R.-C. designed experiments, conducted experiments, analyzed data, and contributed to writing; E.L.F., A.M., K.M., A.B.C., D.S., N.M.B.B., S.B.M., Y.A.K., N.V.S., J.A., D.M., L.S., K.D., C.J., C.Z., M.J.V., M.J.L.-M., and M.L.L. conducted experiments; G.C. and K.C. conducted data analysis; P.H.C.G., J.M.E., D.C., D.C.N., A.P.W., A.G.J., E.F.A., M.M., E.A.W., and F.J.G. contributed to experimental design and to writing the manuscript. Competing interests: The authors declare no conflicts of interest. Data and materials availability: The GSK compounds were obtained under a materials transfer agreement from GSK. All other data are available in the manuscript or the supplementary materials. Some of the data in this manuscript have been deposited at

Stay Connected to Science

Navigate This Article