(C) PLOS One [1]. This unaltered content originally appeared in journals.plosone.org.
Licensed under Creative Commons Attribution (CC BY) license.
url:https://journals.plos.org/plosone/s/licenses-and-copyright

------------



The 3-phosphoinositide–dependent protein kinase 1 is an essential upstream activator of protein kinase A in malaria parasites

['Eva Hitz', 'Department Of Medical Parasitology', 'Infection Biology', 'Swiss Tropical', 'Public Health Institute', 'Basel', 'University Of Basel', 'Natalie Wiedemar', 'Armin Passecker', 'Beatriz A. S. Graça']

Date: 2022-01 

Cyclic adenosine monophosphate (cAMP)-dependent protein kinase A (PKA) signalling is essential for the proliferation of Plasmodium falciparum malaria blood stage parasites. The mechanisms regulating the activity of the catalytic subunit PfPKAc, however, are only partially understood, and PfPKAc function has not been investigated in gametocytes, the sexual blood stage forms that are essential for malaria transmission. By studying a conditional PfPKAc knockdown (cKD) mutant, we confirm the essential role for PfPKAc in erythrocyte invasion by merozoites and show that PfPKAc is involved in regulating gametocyte deformability. We furthermore demonstrate that overexpression of PfPKAc is lethal and kills parasites at the early phase of schizogony. Strikingly, whole genome sequencing (WGS) of parasite mutants selected to tolerate increased PfPKAc expression levels identified missense mutations exclusively in the gene encoding the parasite orthologue of 3-phosphoinositide–dependent protein kinase-1 (PfPDK1). Using targeted mutagenesis, we demonstrate that PfPDK1 is required to activate PfPKAc and that T189 in the PfPKAc activation loop is the crucial target residue in this process. In summary, our results corroborate the importance of tight regulation of PfPKA signalling for parasite survival and imply that PfPDK1 acts as a crucial upstream regulator in this pathway and potential new drug target.

Funding: This work was supported by funding from the Swiss National Science Foundation ( https://www.snf.ch/en ) to TSV (grant number BSCGI0_157729) and PM (grant number 310030_156264), and from the Rudolf Geigy Foundation ( https://en.geigystiftung.ch/ ) to EH. BASG received a PhD fellowship from the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement number 860875 ( https://ec.europa.eu/programmes/horizon2020/en/home ). IV received funding from the Medical Research Council UK ( https://mrc.ukri.org/ ) (grant number MR/N009274/1) and the EPA Cephalosporin Fund ( https://register-of-charities.charitycommission.gov.uk/charity-details/?regid=309698&subid=0 ) (grant number CF 329). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Here, we used reverse genetics approaches to study the function of PfPKAc in asexual blood stage parasites, sexual commitment, and gametocytogenesis. Our results confirm the essential role for PfPKAc in merozoite invasion and show that while PfPKAc plays no obvious role in the control of sexual commitment or gametocyte maturation, it contributes to the regulation of gametocyte-infected erythrocyte deformability. We further demonstrate that overexpression of PfPKAc is lethal in asexual blood stage parasites. Intriguingly, whole genome sequencing (WGS) of parasites selected to tolerate PfPKAc overexpression identified mutations exclusively in the gene encoding the P. falciparum orthologue of phosphoinositide-dependent protein kinase-1 (PfPDK1). Using targeted mutagenesis, we show that the T189 residue is crucial for PfPKAc activity and that activation of PfPKAc is likely dependent on PfPDK1-mediated phosphorylation.

While several studies demonstrated the importance of cAMP in activating PfPKAc, the role of PfPKAc phosphorylation in regulating PfPKAc activity remains elusive. High-throughput phosphoproteomic approaches identified several phosphorylated residues in PfPKAc, including T189 that corresponds to the PDK1 target residue T197 in the activation segment of mammalian PKAc [ 31 – 34 ]. However, if and to what extent phosphorylation of T189 is important for PfPKAc activation in P. falciparum and whether T189 phosphorylation is deposited via autophosphorylation or by another kinase, is unknown. Furthermore, besides the well-established role for PfPKAc in parasite invasion, other possible functions of PfPKA in asexual and sexual development are only poorly understood.

In P. falciparum, PfPKA kinase consists of only one catalytic (PfPKAc) and one regulatory (PfPKAr) subunit [ 25 , 26 ], and cAMP levels in blood stage parasites are regulated by PfPDEβ [ 27 ], which hydrolyses both cAMP and cGMP and PfACβ that synthesises cAMP [ 8 ]. Analysis of conditional loss-of-function mutants showed that PfPKAc is essential for the process of merozoite invasion, where it is required for the phosphorylation and timely shedding of the invasion ligand AMA1 from the merozoite surface [ 5 , 6 , 8 , 9 ]. Likewise, depletion of cAMP levels through conditional disruption of pfacβ phenocopied the invasion defect observed for the pfpkac mutant [ 8 ]. Interestingly, a conditional pfpdeβ null mutant, which displays increased cAMP levels and PfPKAc hyperactivation, also showed a severe merozoite invasion defect that was linked to elevated phosphorylation and premature shedding of AMA1 [ 27 ]. These studies highlighted that tight regulation of PfPKAc activity is crucial for successful merozoite invasion and parasite proliferation. In addition, PfPKA seems to have additional functions in blood stage parasites. Global phosphoproteomic studies of pfpdeβ, pfacβ, and pfpkac conditional knockout (KO) cell lines identified 39 proteins as high confidence targets of cAMP/PfPKA-dependent phosphorylation [ 8 , 27 ]. These proteins include invasion factors (e.g., AMA1 and coronin) and several proteins with predicted roles in other processes (e.g., chromatin organisation and protein transport) or with unknown functions [ 8 , 27 ]. In addition, cAMP/PfPKA-dependent signalling has been implicated in the regulation of ion channel conductance and new permeability pathways (NPPs) in asexual blood stage parasites [ 28 ] and gametocytes [ 29 ] as shown through the use of pharmacological approaches (PKA/PDE inhibitors, exogenous 8-Bromo-cAMP) and transgenic cell lines (deletion of PfPDEδ, overexpression of PfPKAr) [ 28 , 29 ]. Similar experiments identified a putative role for cAMP/PfPKA-dependent signalling in regulating gametocyte-infected erythrocyte deformability [ 30 ].

Protein kinase A (PKA) was discovered in the 1970s and is one of the best characterised eukaryotic protein kinases [ 10 ]. In its inactive state, the PKA holoenzyme is a tetramer consisting of 2 regulatory subunits (PKAr) and 2 catalytic subunits (PKAc) [ 11 ]. Upon binding of cAMP to PKAr, the PKAc subunits are released. PKAc release thus depends on cAMP levels, which are regulated by adenylyl cyclases (ACs) and phosphodiesterases (PDEs) that synthesise and hydrolyse cAMP, respectively [ 11 ]. Furthermore, phosphorylation of PKAc is essential for its activity. In vitro, PKAc was shown to be active upon release from PKAr due to autophosphorylation [ 12 , 13 ]. However, research in budding yeast and human cells demonstrated that the 3-phosphoinositide–dependent protein kinase-1 (PDK1) phosphorylates and activates PKAc in vivo [ 14 – 18 ]. PDK1 has originally been identified as the kinase responsible for activating PKB/Akt in response to growth factor–induced phosphoinositide 3-kinase (PI3K) signalling in human cells [ 17 , 19 – 21 ]. Subsequent studies have shown that PDK1 also activates a large number of other AGC type kinases including PKA, PKG, and PKC [ 17 , 21 ]. AGC kinases dock with PDK1 via their so-called PDK1-interacting fragment (PIF), a hydrophobic motif that binds to the PIF-binding pocket in the N-terminal region of the PDK1 kinase domain, and this interaction allows PDK1 to activate its substrates by activation loop phosphorylation [ 17 , 21 , 22 ]. In case of human PKAc, the PDK1-dependent phosphorylation of T197 in the activation loop plays a crucial role in controlling PKAc structure, activity, and function [ 15 , 18 , 23 , 24 ].

Erythrocyte invasion by merozoites is a highly regulated multistep process starting with the initial attachment of the merozoite to the RBC surface, followed by parasite reorientation and formation of a so-called tight junction [ 4 ]. The tight junction is the intimate contact area between the merozoite and RBC membranes that moves along the merozoite surface during the actin–myosin motor-driven invasion process [ 4 ]. Alongside the secreted rhoptry neck proteins, the micronemal transmembrane protein apical membrane antigen 1 (AMA1) is an integral component of the tight junction [ 4 ]. The cytoplasmic domain of AMA1 bears an essential role in merozoites during RBC invasion [ 5 – 7 ]. In particular, the phosphorylation of residues in the AMA1 cytoplasmic tail (S610 and T613) is essential for AMA1 function in RBC invasion [ 5 – 7 ]. Recent research has shown that the P. falciparum cyclic adenosine monophosphate (cAMP)-dependent protein kinase A (PfPKA) is responsible for AMA1 phosphorylation at S610 and hence essential for successful erythrocyte invasion [ 5 , 6 , 8 , 9 ].

Malaria is caused by protozoan parasites of the genus Plasmodium. Infections with Plasmodium falciparum are responsible for the vast majority of severe and fatal malaria cases. People get infected through female Anopheles mosquitoes that inject sporozoites into the skin tissue during their blood meal. After reaching the liver, sporozoites infect and multiply inside hepatocytes, generating thousands of merozoites that are released into the blood stream. Merozoites invade red blood cells (RBCs) and develop intracellularly through the ring stage into a trophozoite and finally a schizont stage parasite, which undergoes 4 to 5 rounds of nuclear division followed by cytokinesis to produce up to 32 new merozoites. Upon rupture of the infected red blood cell (iRBC), the released merozoites infect new erythrocytes to initiate another intraerythrocytic developmental cycle (IDC). The consecutive rounds of RBC invasion and intraerythrocytic parasite proliferation are responsible for all malaria-related pathology and deaths. Importantly, during each round of replication, a small subset of trophozoites commits to sexual development, and their ring stage progeny differentiates over the course of 10 to 12 days and 5 distinct morphological stages (stages I to V) into mature male or female gametocytes. Sexual commitment occurs in response to environmental triggers that activate expression of the transcription factor PfAP2-G, the master regulator of sexual conversion [ 1 – 3 ]. When taken up by an Anopheles mosquito, mature stage V gametocytes develop into gametes and undergo fertilisation. The resulting zygote develops into an ookinete that migrates through the midgut wall and transforms into an oocyst, generating thousands of infectious sporozoites ready to be injected into another human host.

Many PDK1-dependent AGC kinases (e.g., AKT/PKB, RSK, PKC, S6K, and SGK) are downstream effectors in the PI3K/protein kinase B (AKT) or mitogen-activated protein kinase (MAPK) growth factor signalling pathways and are aberrantly activated in various types of cancer in humans [ 60 ]. In addition, PDK1 expression itself is augmented in many tumours [ 60 ]. For these reasons, human PDK1 is pursued as a potential drug target for cancer therapy, and a large number of inhibitors targeting human PDK1 have been developed and patented over the past 15 years [ 61 – 63 ]. For instance, BX-795 and BX-912, 2 related aminopyrimidine compounds, inhibit recombinant human PDK1 activity with half-maximal inhibitory concentrations (IC 50 ) of 11 nM and 26 nM, respectively [ 64 ], and the aminopyrimidine-aminoindazole GSK2334470 has similar in vitro potency against PDK1 (15 nM IC 50 ) [ 65 , 66 ]. While BX-795 was shown to also inhibit several other human kinases in vitro [ 67 ], GSK2334470 displayed high specificity for PDK1 over a large panel of other recombinant human kinases [ 65 ]. In cell-based assays, all 3 compounds inhibited the PDK1-dependent phosphorylation of several AGC kinase substrates at submicromolar concentrations [ 64 – 66 ]. Here, we tested these 3 commercially available ATP-competitive inhibitors of human PDK1 for their potential to kill P. falciparum asexual blood stage parasites using a [ 3 H]-hypoxanthine incorporation assay [ 68 ]. We found that BX-795, BX-912, and GSK2334470 all inhibited parasite proliferation with IC 50 values of 1.83 μM (± 0.23 SD), 1.31 μM (± 0.24 SD), and 1.83 μM (± 0.11 SD), respectively ( S14 Fig ). In an attempt to test whether the lethal effect of these molecules is due to the specific inhibition of PfPDK1, we repeated the dose response assays on NF54/PDK1 cKD parasites cultured in the presence (control) or absence of Shield-1 (PfPDK1 depleted). However, we observed no reduction in IC 50 values for PfPDK1 depleted (−Shield-1) compared to NF54/PDK1 cKD control parasites (+Shield-1) or compared to NF54 WT parasites cultured in the presence or absence of Shield-1 ( S14 Fig ), suggesting that all 3 inhibitors are not specific for PfPDK1 but likely target additional/other essential parasite kinases. To investigate the effects these compounds have on intraerythrocytic parasite development, we assessed the morphology of drug-treated parasites by visual inspection of Giemsa-stained thin blood smears. Analysis of NF54 WT early ring stage parasites treated with each of the 3 PDK1 inhibitors over a period of 60 hours revealed that parasites were unable to progress beyond the ring stage and became pyknotic thereafter, in contrast to untreated control parasites that progressed through schizogony and gave rise to ring stage progeny as expected ( S14 Fig ). Furthermore, additional results obtained from separate 18-hour treatments of early ring stages, early trophozoites or early schizonts suggest that all 3 inhibitors are active against all intraerythrocytic stages, as in all cases the morphology of drug-treated parasites was reminiscent of dying or pyknotic forms and viable trophozoites, schizonts, and ring stage progeny were not observed ( S14 Fig ). Hence, based on the promising activities of these molecules against blood stage parasites, attempts to identify their target(s) as well as the screening of extended PDK1 inhibitor libraries and experimental validation of PfPDK1 as a drug target would be worthwhile activities to be pursued in future research.

(A) Expression of PfPKAcT189V-GFP under OE-inducing (–GlcN) and control conditions (+GlcN) by live cell fluorescence imaging and western blot analysis. Synchronous parasites (0 to 8 hpi) were split (±GlcN) 40 hours before collection of the samples. Representative fluorescence images are shown. Parasite DNA was stained with Hoechst. Scale bar = 5 μm. GlcN, glucosamine. For western blot analysis, lysates derived from equal numbers of parasites were loaded per lane. MW PfPKAc-GFP = 67.3 kDa, MW PfGAPDH = 36.6 kDa. The full size western blot is shown in S13 Fig . (B) Increase in parasitaemia (left) and parasite multiplication rates (right) of NF54/PKAcT189V cOE parasites over 3 generations under PfPKAcT189V-GFP OE-inducing (–GlcN) and control conditions (+GlcN). Synchronous parasites (0 to 6 hpi) were split (±GlcN) 18 hours before the first measurement in generation 1. Open squares represent data points for individual replicates and the means and SD (error bars) of 3 biological replicates are shown. Differences in multiplication rates have been compared using a paired 2-tailed Student t test (statistical significance cutoff: p < 0.05). The raw data are available in the source data file ( S2 Data ). DIC, differential interference contrast.

Previous research in human cell lines and yeast identified a specific threonine residue in the PKAc activation loop (T197 in mammals) as the target of PDK1-dependent phosphorylation [ 15 , 18 , 23 ]. In PfPKAc, T189 likely represents the activation loop phosphorylation site corresponding to T197 in human PKAc. Hence, we tested whether the T189 residue is indeed important for PfPKAc activity. To achieve this, we employed the same approach as already used to obtain NF54/PKAc cOE parasites ( S5 Fig ) to generate the NF54/PKAcT189V cOE line that conditionally overexpresses a mutated version of PfPKAc in which T189 has been substituted with a nonphosphorylatable valine residue [ 58 , 59 ] ( S13 Fig ). Correct insertion of the PfPKAcT189V cOE cassette into the glp3 locus was verified by PCR on gDNA ( S13 Fig ). Live cell fluorescence microscopy and western blot analysis confirmed that PfPKAcT189V-GFP OE was efficiently induced upon removal of GlcN (Figs 6A and S13 ). Strikingly, in contrast to the lethal effect provoked by OE of PfPKAc-GFP, OE of the PfPKAcT189V-GFP mutant (–GlcN) had no effect on intraerythrocytic parasite development, multiplication, and survival when compared to the control population (+GlcN) ( Fig 6B ). Hence, these results suggest that PfPKAc activity is strictly dependent on phosphorylation of T189 in the activation loop segment. In combination with the findings obtained through the mutational analysis of PfPDK1, they further imply that PfPDK1 is directly or indirectly responsible for T189 phosphorylation and thus activation of PfPKAc.

(A, B) Expression of PfPKAc-GFP under OE-inducing (–GlcN) and control conditions (+GlcN) in NF54/PKAc cOE S1/PDK1_wt parasites (A) and NF54/PKAc cOE M1/PDK1_mut parasites (B) by live cell fluorescence imaging and western blot analysis. Synchronous parasites (0 to 8 hpi) were split (±GlcN) 40 hours before collection of the samples. Representative fluorescence images are shown. Parasite DNA was stained with Hoechst. Scale bar = 5 μm. For western blot analysis, lysates derived from an equal number of parasites were loaded per lane. MW PfPKAc-GFP = 67.3 kDa, MW PfGAPDH = 36.6 kDa. The full size western blots are shown in S11 Fig . (C, D) Increase in parasitaemia of NF54/PKAc cOE S1/PDK1_wt parasites (C) and NF54/PKAc cOE M1/PDK1_mut parasites (D) over 3 generations under PfPKAc-GFP OE-inducing (–GlcN) and control conditions (+GlcN). Synchronous parasites (0 to 6 hpi) were split (±GlcN) 18 hours before the first measurement in generation 1. Open squares represent data points for individual replicates and the means and SD (error bars) of 3 biological replicates are shown. The raw data are available in the source data file ( S2 Data ). DIC, differential interference contrast; GlcN, glucosamine.

To test if PfPDK1 is indeed involved in regulating PfPKAc activity, we used CRISPR/Cas-9–based targeted mutagenesis to change the PfPDK1 M51R mutation back to the WT sequence in the NF54/PKAc cOE S1 survivor line (NF54/PKAc cOE S1/PDK1_wt). Sanger sequencing verified the successful reversion of the PfPDK1 M51R mutation ( S11 Fig ), and live cell fluorescence imaging and western blot analysis confirmed that GlcN removal still triggered efficient PfPKAc-GFP OE in NF54/PKAc cOE S1/PDK1_wt parasites (Figs 5A and S11 ). Strikingly, PfPKAc-GFP OE (–GlcN) led to a complete block in parasite replication, showing that reverting the PfPDK1 M51R mutation completely restored the PfPKAc OE-sensitive phenotype (Figs 5C and S11 ). To complement these experiments, we also attempted to introduce the PfPDK1 M51R mutation into NF54 WT parasites and the NF54/PKAc cOE clone M1. Two independent attempts to mutate PfPDK1 in NF54 WT parasites failed, suggesting that fully functional PfPDK1 is strictly required for parasite viability under normal PfPKAc expression levels. By contrast, we readily succeeded in introducing the PfPDK1 M51R mutation into the NF54/PKAc cOE clone M1 (NF54/PKAc cOE M1/PDK1_mut) ( S11 Fig ). Western blot analysis and live cell fluorescence imaging confirmed that GlcN removal still triggered efficient PfPKAc-GFP OE (Figs 5B and S11 ). Parasite multiplication assays revealed that the PfPDK1 M51R point mutation rendered NF54/PKAc cOE M1/PDK1_mut parasites tolerant to PfPKAc-GFP OE (–GlcN) (Figs 5D and S11 ). However, in contrast to NF54/PKAc cOE S1 parasites, which carry the same M51R PfPDK1 mutation, the multiplication rates of NF54/PKAc cOE M1/PDK1_mut parasites overexpressing PfPKAc-GFP (–GlcN) reached only 60% to 80% compared to the matching control (+GlcN) (Figs 5D and S11 ). This observation suggested that an additional selection step had taken place in NF54/PKAc cOE S1 parasites that conferred full tolerance to PfPKAc-GFP OE (see Fig 3G ). Indeed, based on the WGS data the parental NF54/PKAc cOE M1 clone carries 8 copies of the PfPKAc cOE transgene cassette, whereas the NF54/PKAc cOE S1 survivor population carries only 2 copies ( S12 Fig ). We therefore believe that due to the negative impact of continuous PfPKAc-GFP OE on parasite viability, NF54/PfPKAc cOE S1 parasites carrying fewer PfPKAc cOE cassettes and hence lower overall PfPKAc expression levels had a comparative advantage during the selection process for PfPKAc-GFP OE tolerance.

Taken together, we show that PfPDK1 is expressed in the parasite nucleus and cytosol throughout asexual development, reaching peak expression in schizonts. While PfPDK1 is likely essential, the results obtained with the PfPDK1 cKD mutant suggest that residual PfPDK1 expression levels are sufficient to sustain parasite viability. Furthermore, PfPDK1 seems to play no major role in regulating sexual commitment, gametocytogenesis or male gametogenesis, but it is again conceivable that residual PfPDK1 expression in the cKD mutant may have been sufficient to maintain these processes.

(A) Expression of PfPDK1-GFPDD in ring (18 to 24 hpi), trophozoite (24 to 30 hpi), and schizont (42 to 48 hpi) stages cultured under protein-stabilising (+Shield-1) conditions by live cell fluorescence imaging. Representative fluorescence images are shown. Parasite DNA was stained with Hoechst. Scale bar = 5 μm. (B) Expression of PfPDK1-GFPDD under protein-depleting (–Shield-1) and control conditions (+Shield-1) by live cell fluorescence imaging and western blot analysis. Synchronous parasites (0 to 8 hpi) were split (±Shield-1) 40 hours before collection of the samples. Representative fluorescence images are shown. Parasite DNA was stained with Hoechst. Scale bar = 5 μm. For western blot analysis, lysates derived from an equal numbers of parasites were loaded per lane. MW PfPDK1-GFPDD = 101.1 kDa, MW PfGAPDH = 36.6 kDa. The full size western blot is shown in S9 Fig . (C) Increase in parasitaemia (left) and parasite multiplication rates (right) under PfPDK1-GFPDD-depleting (–Shield-1) and control (+Shield-1) conditions. Synchronous parasites (0 to 6 hpi) were split (±Shield-1) 18 hours before the first measurement in generation 1. Open squares represent data points for individual replicates and the means and SD (error bars) of 3 biological replicates are shown. Differences in multiplication rates have been compared using a paired 2-tailed Student t test (statistical significance cutoff: p < 0.05). The raw data are available in the source data file ( S2 Data ). cKD, conditional PfPKAc knockdown; DIC, differential interference contrast.

To gain further insight into the function of PfPDK1, we attempted to generate a pfpdk1 KO line by gene disruption using a CRISPR/Cas-9–based single plasmid approach [ 53 ]. However, consistent with the results obtained in previous kinome- or genome-wide KO screens in P. falciparum and Plasmodium berghei [ 31 , 54 , 55 ], we failed to obtain a pfpdk1 null mutant, indicating that the gene is essential in asexual parasites. We therefore engineered the PfPDK1 conditional knockdown line NF54/PDK1 cKD by tagging the endogenous pfpdk1 gene with gfp fused to the dd sequence in NF54 WT parasites ( S8 Fig ). Correct editing of the pfpdk1 gene and absence of the WT locus was confirmed by PCR on gDNA. Donor plasmid integration downstream of the pfpdk1 gene was detected in a subset of parasites ( S8 Fig ), but this is not expected to compromise pfpdk1-gfpdd expression since the 551 bp 3′ homology region (HR) used for homology-directed repair includes the native terminator as based on published RNA sequencing (RNA-seq) data [ 56 ]. Live cell fluorescence imaging and western blot analysis revealed that PfPDK1-GFPDD is expressed in the cytosol and nucleus throughout the IDC with highest expression in schizonts, consistent with published gene expression data [ 56 , 57 ] (Figs 4A and S9 ). Furthermore, PfPDK1-GFPDD expression was efficiently reduced after Shield-1 removal compared to parasites grown in the presence of Shield-1 (Figs 4B and S9 ). Surprisingly, PfPDK1-GFPDD-depleted parasites (–Shield-1) showed slightly higher multiplication rates compared to the control (+Shield-1) ( Fig 4C ). This increased multiplication rate cannot be attributed to the removal of Shield-1 itself, since NF54 WT parasites cultured in the presence or absence of Shield-1 multiplied equally ( S9 Fig ). We also tested whether PfPDK1-GFPDD plays a role in gametocytogenesis but did not detect any differences in SCRs, gametocyte morphology or ExRs when comparing PfPDK1-GFPDD-depleted (–Shield-1) with control parasites (+Shield-1) ( S10 Fig ).

(A) Top: Schematic of the Pf3D7_1121900/pfpdk1 gene. Asterisks indicate the approximate localisation of the 5 missense mutations identified in the 6 different NF54/PKAc cOE survivors. Bottom: Summary of the sequence information obtained from WGS of gDNA of the 2 NF54/PKAc cOE clones (M1, M2) and the 6 independently grown survivors (S1–S6). Missense mutations and their positions within the pfpdk1 coding sequence as well as the corresponding amino acid substitutions in the PfPDK1 protein sequence are shown. (B) Predicted PfPDK1 structure shown in orthogonal (top left versus top right) or opposing (top left versus lower left) views. PfPDK1 was modelled on the crystallographic structure of human PDK1 (PDB ID 1UU9) [ 52 ]. PfPDK1 segments with no correspondence in human PDK1 (amino acids 1 to 27, 188 to 307, and 423 to 525) were omitted from modelling. The ATP substrate (sticks), mutated amino acids (substitution in parenthesis; red spheres), and residues forming the PIF-binding pocket [ 22 ] (green sticks) are indicated. The PIF-binding pocket is shown in surface representation in the lower right view. c., cDNA; cOE, conditional overexpression; PIF, PDK1-interacting fragment; WGS, whole genome sequencing.

The above findings suggested that genetic mutations in the NF54/PKAc cOE S1 survivor population might cause their tolerance to elevated PfPKAc-GFP expression levels. To address this hypothesis, we performed WGS of the 2 parental NF54/PKAc cOE clones M1 and M2 and the 6 independently grown survivor populations (NF54/PKAc cOE S1-S6), 3 each originating from the M1 and M2 clones, respectively. Intriguingly, we found that all 6 NF54/PKAc cOE survivors, but not the parental M1 and M2 clones, carried missense mutations in the gene encoding a putative serine/threonine protein kinase (Pf3D7_1121900) ( Fig 3A , S1 Data ). No other mutations were identified in the NF54/PKAc cOE survivor populations, consistent with the Sanger sequencing results, neither the endogenous nor the ectopic pfpkac gene carried mutations in any of the 6 survivors. The Plasmodium vivax orthologue of Pf3D7_1121900 (PVX_091715) is annotated as putative 3-phosphoinositide dependent protein kinase-1 (PDK1) ( www.plasmodb.org ), and a multiple sequence alignment suggested that Pf3D7_1121900 is indeed an orthologue of PDK1, a kinase widely conserved among eukaryotes and known as a master regulator of AGC kinases including PKA [ 17 , 50 ] ( S7 Fig ). However, similar to PDK1 enzymes from most fungi, nonvascular plants, and other alveolates, Pf3D7_1121900 and its P. vivax orthologue lack the carboxyl-terminal phospholipid-binding pleckstrin homology (PH) domain that is found in PDK1s from animals and vascular plants and important to localise PDK1 to the plasma membrane for PKB activation in response to the PI3K-dependent production of phosphatidylinositol bis-/trisphosphates [ 17 , 50 , 51 ]. Construction of a homology model of the Pf3D7_1121900 protein kinase domain based on the human PDK1 crystallographic structure (PDB ID 1UU9) [ 52 ] allowed us to visualise the location of the amino acids mutated in the 6 NF54/PKAc cOE survivors ( Fig 3B ). None of these mutations affected the putative PIF-binding pocket. Rather, all mutated amino acids were located at or in the periphery of the predicted ATP-binding cleft, although only one mutation (N45S) may influence ATP coordination directly (Figs 3B and S7 ). We hence surmise that all identified mutations alter the catalytic activity but not the protein interaction preferences of PfPDK1.

(A) Expression of PfPKAc-GFP in NF54/PKAc cOE M1 parasites under OE-inducing (–GlcN) and control conditions (+GlcN) as assessed by live cell fluorescence imaging and western blot analysis. Synchronous parasites (0 to 8 hpi) were split (±GlcN) 40 hours before sample collection. Representative fluorescence images are shown. Parasite DNA was stained with Hoechst. Scale bar = 5 μm. For western blot analysis, lysates derived from equal numbers of parasites were loaded per lane. MW PfPKAc-GFP = 67.3 kDa, MW PfGAPDH = 36.6 kDa. The full size western blot is shown in S6 Fig . (B) Representative images from Giemsa-stained thin blood smears showing NF54/PKAc cOE M1 and M2 parasites under PfPKAc-GFP OE-inducing (–GlcN) and control conditions (+GlcN). Synchronous parasites (0 to 8 hpi) were split (±GlcN) 40 hours before the images were captured. Scale bar = 5 μm. (C) Number of nuclei per schizont in NF54/PKAc cOE M1 parasites under PfPKAc-GFP OE-inducing (–GlcN) and control conditions (+GlcN). Each open circle represents one parasite. Data from 3 biological replicate experiments are shown, and 100 parasites were counted in each experiment. The boxplots show data distribution (median, upper, and lower quartile and whiskers). The raw data are available in the source data file ( S2 Data ). (D) Increase in parasitaemia in NF54/PKAc cOE M1 parasites over 3 generations under PfPKAc-GFP OE-inducing (–GlcN) and control (+GlcN) conditions. Synchronous parasites (0 to 6 hpi) were split (±GlcN) 18 hours before the first measurement in generation 1. Open squares represent data points for individual replicates and the means and SD (error bars) of 3 biological replicates are shown. The raw data are available in the source data file ( S2 Data ). (E) Expression of PfPKAc-GFP in NF54/PKAc cOE S1 survivor parasites under OE-inducing (–GlcN) and control conditions (+GlcN) as assessed by live cell fluorescence imaging and western blot analysis. Parasites were cultured and samples prepared as described in panel A. Scale bar = 5 μm. MW PfPKAc-GFP = 67.3 kDa, MW PfGAPDH = 36.6 kDa. The full size western blot is shown in S6 Fig . (F) Number of nuclei per schizont in NF54/PKAc cOE S1 survivor parasites under PfPKAc-GFP OE-inducing (–GlcN) and control conditions (+GlcN). Each open circle represents one parasite. Data from 3 biological replicate experiments are shown, and 100 parasites were counted in each experiment. The boxplots show data distribution (median, upper, and lower quartile and whiskers). The raw data are available in the source data file ( S2 Data ). (G) Increase in parasitaemia in NF54/PKAc cOE S1 survivor parasites over 3 generations under PfPKAc-GFP OE-inducing (–GlcN) and control (+GlcN) conditions. Parasites were cultured as described in panel D. Open squares represent data points for individual replicates and the means and SD (error bars) of 3 biological replicates are shown. The raw data are available in the source data file ( S2 Data ). cOE, conditional overexpression; DIC, differential interference contrast; GlcN, glucosamine.

The merozoite invasion and post-invasion developmental defects observed for pfpdeβ KO parasites had been linked to increased cAMP levels and PfPKAc hyperactivity [ 27 ]. To further study the consequences of PfPKAc overexpression, we generated a PfPKAc conditional overexpression (cOE) line using CRISPR/Cas-9–based gene editing. To this end, we inserted an ectopic pfpkac-gfp transgene cassette into the nonessential glp3 (cg6, Pf3D7_0709200) locus in NF54 WT parasites (NF54/PKAc cOE) ( S5 Fig ). Here, the constitutive calmodulin (PF3D7_1434200) promoter and a glmS ribozyme element [ 48 ] control expression of the pfpkac-gfp gene. Since the initial transgenic NF54/PKAc cOE population still contained some parasites carrying the WT glp3 locus, we obtained clonal lines by limiting dilution cloning [ 49 ]. In 2 clones (NF54/PKAc cOE M1 and M2), correct integration of the inducible PfPKAc-GFP OE expression cassette and absence of WT parasites was confirmed by PCR on gDNA ( S5 Fig ). Live cell fluorescence imaging and western blot analysis confirmed the efficient induction of PfPKAc-GFP OE upon GlcN removal in both clones (Figs 2A and S6 ). Interestingly, PfPKAc-GFP OE (–GlcN) resulted in a complete block in parasite development half way through the IDC ( Fig 2B ). To study this growth defect in more detail, we quantified the number of nuclei per parasite in late schizont stages (40 to 46 hpi), 40 hours after triggering PfPKAc-GFP OE (–GlcN) in young ring stage parasites (0 to 6 hpi). This experiment revealed that NF54/PKAc cOE M1 parasites overexpressing PfPKAc-GFP did not develop beyond the late trophozoite/early schizont stage as most parasites contained only 1 or 2 nuclei as opposed to the control population (+GlcN) that progressed normally through several rounds of nuclear division ( Fig 2C ). To test whether these parasites were still able to produce progeny, we performed parasite multiplication assays. NF54/PKAc cOE M1 ring stage parasites were split at 0 to 6 hpi, cultured separately in the presence or absence of GlcN and parasite multiplication was quantified over 3 generations. PfPKAc-GFP OE (–GlcN) completely failed to multiply as no increase in parasitaemia was observed, in contrast to the control population (+GlcN) that multiplied normally (Figs 2D and S6 ). Notably, however, viable PfPKAc-GFP OE parasites emerged approximately 2 weeks after maintaining NF54/PKAc cOE M1 parasites constantly in culture medium lacking GlcN. We termed these parasites “PfPKAc OE survivors” (NF54/PKAc cOE S1). Importantly, NF54/PKAc cOE S1 parasites still overexpressed PfPKAc-GFP in absence of GlcN (Figs 2E and S6 ). Furthermore, Sanger sequencing confirmed that neither the endogenous nor the ectopic pfpkac genes in the NF54/PKAc cOE S1 survivor population carried any mutations. Quantifying the number of nuclei per schizont and parasite multiplication assays revealed that NF54/PKAc cOE S1 parasites were completely tolerant to PfPKAc-GFP OE and developed and multiplied identically irrespective of whether PfPKAc-GFP OE was induced (–GlcN) or not (+GlcN) (Figs 2F, 2G and S6 ).

In summary, our results demonstrate that PfPKAc plays no major role in the regulation of sexual commitment, gametocyte maturation, or male gametogenesis but that gametocyte-iRBC rigidity is at least partially regulated by PfPKAc. Furthermore, we discovered that the presence of 2.5 mM GlcN in the culture medium affects both sexual commitment and ExRs, which needs to be taken into account when studying these processes in conditional mutants employing the glmS riboswitch system.

Lastly, we tested if PfPKAc activity is involved in regulating gametocyte rigidity. Immature gametocytes display high cellular rigidity and sequester in the bone marrow and the spleen, whereas stage V gametocytes are more deformable and can reenter the bloodstream to be taken up by feeding mosquitoes [ 45 – 47 ]. Experiments employing PKA and PDE inhibitors, a transgenic cell line overexpressing PfPKAr, or treatment with exogenous 8-bromo-cAMP to increase cellular cAMP levels provided evidence for a potential role for PfPKAc in maintaining the rigidity of immature gametocyte-infected erythrocytes [ 30 ]. To test if PfPKAc is indeed involved in controlling this process, we measured the deformability status of immature stage III and mature stage V gametocytes using microsphiltration experiments [ 46 ]. Microsphiltration exploits the fact that differences in cellular rigidity correlate with cell retention rates in a microsphere-based artificial spleen system [ 46 , 47 ]. We observed a slight but significant decrease in the retention rates of PfPKAc-GFPDD-depleted gametocytes (–Shield-1) compared to the control (+Shield-1), both in immature stage III (day 6) (76.0% ± 5.7 SD versus 82.2% ± 8.0 SD) and mature stage V (day 11) gametocytes (13.9% ± 16.8 SD versus 27.6% ± 14.5 SD) ( Fig 1E ). By contrast, NF54 WT gametocytes cultured in the presence or absence of Shield-1 showed no difference in retention rates ( S4 Fig ).

Next, we tested if PfPKAc-GFPDD depletion affects gametocyte morphology or male gametocyte exflagellation. To conduct these experiments, we induced sexual commitment by culturing parasites in serum-free minimal fatty acid medium (–SerM) [ 3 ]. After reinvasion, parasites were split and cultured separately under PfPKAc-GFPDD-depleting (–Shield-1/+GlcN) or PfPKAc-GFPDD-stabilising conditions (+Shield-1/–GlcN). For the first 6 days of gametocyte maturation, the growth medium was supplemented with 50 mM N-acetyl-D-glucosamine (GlcNac) to eliminate asexual parasites [ 44 ]. Despite efficient depletion of PfPKAc-GFPDD expression (Figs 1D and S4 ), we could not detect morphological abnormalities in PfPKAc-GFPDD-depleted gametocytes in any of the 5 developmental stages (I to V) based on visual inspection of Giemsa-stained thin blood smears ( S4 Fig ). While the exflagellation rates (ExRs) of PfPKAc-GFPDD-depleted male stage V gametocytes (–Shield-1/+GlcN) were significantly reduced by more than 50% compared to the control (+Shield-1/–GlcN), this effect was again caused by the presence of GlcN in the culture medium ( S4 Fig ). When PfPKAc-GFPDD expression was depleted by Shield-1 removal only (–Shield-1/–GlcN), no significant differences in ExRs were observed even though PfPKAc-GFPDD expression was efficiently reduced ( S4 Fig ). In line with this result, GlcN treatment also led to a significant reduction in ExRs of NF54 WT gametocytes (+GlcN) to less than 50% compared to the control (–GlcN) ( S4 Fig ).

To investigate whether PfPKAc activity regulates sexual commitment, we split ring stage parasites (0 to 6 hpi) and cultured them separately under PfPKAc-GFPDD-depleting (–Shield-1/+GlcN) and PfPKAc-GFPDD-stabilising conditions (+Shield-1/–GlcN). Sexually committed parasites were identified based on PfAP2-G-mScarlet positivity in late schizonts (40 to 46 hpi) using high content imaging. We observed a significant increase in SCRs in PfPKAc-GFPDD-depleted (–Shield-1/+GlcN) compared to control parasites (+Shield-1/–GlcN) (1.56-fold ± 0.13 SD) ( S3 Fig ). We can exclude that these differences are caused by the Shield-1 compound as we have previously shown that Shield-1 treatment has no effect on SCRs [ 43 ]. However, treatment with GlcN also caused a similar increase in SCRs in NF54/AP2-G-mScarlet control parasites (1.44-fold ± 0.14 SD) ( S3 Fig ). We therefore conclude that PfPKAc plays no major role in regulating sexual commitment and that the increased SCRs observed in PfPKAc-GFPDD-depleted parasites are caused by the presence of 2.5 mM GlcN in the culture medium.

(A) Expression of PfPKAc-GFPDD in late schizonts under protein- and RNA-depleting (–Shield-1/+GlcN) and control conditions (+Shield-1/–GlcN) as assessed by live cell fluorescence imaging and western blot analysis. Synchronous parasites (0 to 8 hpi) were split (±Shield-1/±GlcN) 40 hours before sample collection. Representative fluorescence images are shown. Parasite DNA was stained with Hoechst. Scale bar = 5 μm. For western blot analysis, parasite lysates derived from equal numbers of parasites were loaded per lane. MW PfPKAc-GFPDD = 79.8 kDa, MW PfGAPDH = 36.6 kDa. The full size western blot is shown in S1 Fig . (B) Increase in parasitaemia over 3 generations under PfPKAc-GFPDD-depleting (–Shield-1/+GlcN) and control conditions (+Shield-1/–GlcN). Open squares represent data points for individual replicates and the means and SD (error bars) of 3 biological replicates are shown. The raw data are available in the source data file ( S2 Data ). (C) Representative images from Giemsa-stained thin blood smears showing the progeny of parasites cultured under PfPKAc-GFPDD-depleting (–Shield-1/+GlcN) or control conditions (+Shield-1/–GlcN) conditions at 0 to 6 hpi (generation 2). Scale bar = 5 μm. (D) Expression of PfPKAc-GFPDD in stage V gametocytes (day 11) under protein- and RNA-depleting (–Shield-1/+GlcN) and control conditions (+Shield-1/–GlcN) as assessed by live cell fluorescence imaging and western blot analysis. Representative fluorescence images are shown. Parasite DNA was stained with Hoechst. Scale bar = 5 μm. For western blot analysis, parasite lysates derived from equal numbers of parasites were loaded per lane. MW PfPKAc-GFPDD = 79.8 kDa, MW PfGAPDH = 36.6 kDa. The full size western blot is shown in S4 Fig . (E) Retention rates of stage III (day 6) and stage V (day 11) gametocytes cultured under PfPKAc-GFPDD-depleting (–Shield-1) (orange) and control conditions (+Shield-1) (blue). Coloured squares represent data points for individual replicates and the means and SD (error bars) of 2 biological replicates performed in 6 technical replicates each are shown. Differences in retention rates have been compared using an unpaired 2-tailed Student t test (statistical significance cutoff: p < 0.05). The raw data are available in the source data file ( S2 Data ). cKD, conditional PfPKAc knockdown; DIC, differential interference contrast; GlcN, glucosamine; hpi, hours postinvasion; PKA, protein kinase A; S, Shield-1.

To study PfPKAc function, we generated a conditional PfPKAc knockdown (cKD) line using a selection-linked integration (SLI)-based gene editing approach [ 35 ]. Successful engineering tags the pfpkac gene with the fluorescent marker gene gfp fused to the fkbp destabilisation domain (dd) sequence [ 36 , 37 ], followed by a sequence encoding the 2A skip peptide [ 38 , 39 ], the blasticidin S deaminase (bsd) marker gene, and, finally, the glmS ribozyme element [ 40 ] ( S1 Fig ). The FKBP/DD system allows for protein destabilisation in the absence of the small molecule ligand Shield-1 [ 36 , 37 ], whereas the glmS ribozyme in the 3′ untranslated region of the mRNA causes transcript degradation in the presence of glucosamine (GlcN) [ 40 ]. To be able to easily quantify sexual commitment rates (SCRs), we modified the pfpkac gene in NF54/AP2-G-mScarlet parasites [ 41 ], which allows identifying sexually committed cells by visualising expression of fluorophore-tagged PfAP2-G [ 3 , 42 ]. After drug selection of transgenic NF54/AP2-G-mScarlet/PKAc cKD parasites, correct editing of pfpkac and absence of the wild-type (WT) locus was confirmed by PCR on genomic DNA (gDNA) ( S1 Fig ). Live cell fluorescence imaging demonstrated that PfPKAc-GFPDD was expressed in the cytosol and nucleus in mid schizonts before it localised to the periphery of developing merozoites in late schizonts as also described elsewhere [ 8 , 9 ] ( S1 Fig ). PfPKAc-GFPDD expression was efficiently depleted in parasites cultured under–Shield-1/+GlcN conditions compared to the matching control (+Shield-1/–GlcN) ( Fig 1A ). To assess the growth phenotype of PfPKAc-GFPDD-depleted parasites, we split ring stage parasite cultures (0 to 6 hours postinvasion, hpi), maintained them separately under PfPKAc-GFPDD-depleting (–Shield-1/+GlcN) and PfPKAc-GFPDD-stabilising conditions (+Shield-1/–GlcN) and quantified parasitaemia over 3 generations using flow cytometry (Figs 1B and S2 ). As expected, PfPKAc-GFPDD-depleted parasites were unable to proliferate because the merozoites failed to invade new RBCs ( Fig 1C ). Hence, by combining 2 inducible expression systems, we engineered a PfPKAc cKD mutant that allows efficient depletion of PfPKAc-GFPDD expression and phenocopies the lethal invasion defect previously observed with conditional PfPKAc KO mutants [ 89 ].

Discussion

PfPKA is essential for the proliferation of asexual blood stage parasites, and the phosphorylation of the invasion ligand AMA1 is one of its crucial functions [5,6,8,9]. In addition, recent research described important roles for the PDE PfPDEβ and the AC PfACβ in regulating cAMP levels and hence PfPKA activity [8,27]. Here, we studied the function and regulation of the catalytic PfPKA subunit PfPKAc to obtain further insight into PfPKA-dependent signalling in P. falciparum blood stage parasites.

Using the NF54/AP2-G-mScarlet/PKAc cKD parasite line, we confirmed the previously described essential role for PfPKAc in merozoite invasion [8,9]. We also demonstrated that PfPKAc plays no major role in regulating sexual commitment. The dispensability of PfPKAc in the sexual commitment pathway seems rather surprising since early studies performed over 3 decades ago claimed a potential involvement of cAMP signalling in regulating sexual commitment [69,70]. However, these studies only indirectly suggested an involvement of cAMP/PfPKA-signalling in this process. For instance, Kaushal and colleagues determined the effect of high exogenous cAMP concentrations (1 mM) on sexual commitment and reported that under static culture conditions (high parasitaemia without addition of uninfected red blood cells [uRBCs]) nearly all parasites developed into gametocytes [69]. Rather than reflecting the true induction of sexual commitment by cAMP signalling, we suspect that observations may have been related to the selective killing of asexual stages by high cAMP concentrations (as indeed reported in their study) and/or the stimulation of high SCRs due to LysoPC depletion from the growth medium at high parasitaemia [3]. Our results further suggest that PfPKAc is not required for the morphological maturation of gametocytes and for male gametogenesis. However, even though our cKD system allowed for efficient depletion of PfPKAc expression, we cannot exclude the possibility that residual PfPKAc expression levels still supported normal sexual development. At this point, we would also like to reiterate that our experiments conducted with the NF54/AP2-G-mScarlet/PKAc cKD line showed that GlcN (2.5 mM), but not Shield-1 (675 nM), acts as a confounding factor when studying sexual commitment and male gametocyte exflagellation. We therefore advise to use the FKBP/DD-Shield-1 [36,37] or DiCre/rapamycin [71–73] conditional expression systems when studying these processes.

The cellular rigidity of immature P. falciparum gametocytes is linked to a dynamic reorganisation of the RBC spectrin and actin networks [45] as well as the presence of parasite-encoded STEVOR proteins at the iRBC membrane [47,74]. By contrast, the increased deformability gained by stage V gametocytes is accompanied by the reversal of these cytoskeletal rearrangements [45] and dissociation of STEVOR from the iRBC membrane [47,74]. Interestingly, results obtained from experiments using pharmacological agents to increase cellular cAMP levels or to inhibit PKA activity demonstrated that gametocyte rigidity is positively regulated by cAMP/PKA-dependent signalling [30,74]. While potential PKA substrates involved in this process are largely unknown, the PKA-dependent phosphorylation of the cytoplasmic tail of STEVOR (S324) is important to maintain cellular rigidity of immature gametocytes, and dephosphorylation of this residue is linked to the increased deformability of stage V gametocytes [74]. Given that PfPKA is not known to be exported into the iRBC cytosol, however, PKA-dependent phosphorylation events in the RBC compartment are likely exerted by human rather than parasite PKA. Notably though, overexpression of the regulatory subunit PfPKAr, which is expected to lower PfPKAc activity, caused increased deformability of stage III gametocytes [30]. Consistent with these data, we demonstrated that PfPKAc depletion caused a significant, yet only moderate, increase in stage III and V gametocyte deformability. While these results provide direct evidence for a role of PfPKAc-dependent phosphorylation in regulating the biomechanical properties of gametocyte-iRBCs, they also suggest that cAMP signalling through PfPKA is not the only driver of this process. We envisage that PfPKAc activity may regulate the expression, trafficking, or function of parasite-encoded proteins destined for export into the iRBC or of proteins of the inner membrane complex and/or microtubular and actin networks underneath the parasite plasma membrane that play important roles in determining cellular shape during gametocytogenesis [75–77]. Comparative phosphoproteomic analyses of the conditional PfPKAc mutants generated here and elsewhere [8,9] may be a promising approach to test this hypothesis and identify the actual substrates involved.

Three recent studies employing DiCre-inducible KO parasites for PfPKAc [8,9], the PDE PfPDEβ [27], and the AC PfACβ [8] highlighted the importance for tight regulation of PfPKAc activity in asexual blood stage parasites. In these studies, induction of the corresponding gene KOs in ring stage parasites caused no immediate defects in intraerythrocytic parasite development but resulted in a complete or severe block in RBC invasion by newly released merozoites due to prevention of PfPKAc activity (PfPKAc and PfACβ KOs) [8,9] or PfPKAc hyperactivation (PfPDEβ KO) [27], respectively. These findings are consistent with the specific expression pattern of all 3 cAMP signalling components in late schizont stages [78]. Interestingly, however, Flueck and colleagues showed that some PfPDEβ KO merozoites successfully invaded RBCs but were then unable to develop into ring stage parasites [27], providing compelling evidence for a lethal effect of PfPKAc hyperactivity also on early intraerythrocytic parasite development. Similarly, we showed that the OE of PfPKAc through a constitutively active heterologous promoter blocked parasite progression through schizogony. We believe this detrimental effect is due to the incapacity of endogenous PfPKAr to complex and inactivate the excess of PfPKAc enzymes, resulting in illegitimate activity of free PfPKAc and hence untimely phosphorylation of substrates prior to the intrinsic PfPKA activity window in late schizonts. While we did not engage in further explorations towards identifying the molecular mechanisms underlying the lethal consequences of PfPKAc overexpression, we discovered another kinase that is likely required for PfPKAc activation. We identified this function by selecting for “PfPKAc OE survivor” parasites able to tolerate PfPKAc OE. All 6 independently selected NF54/PKAc cOE survivor populations carried mutations in the same gene encoding a putative serine/threonine kinase (Pf3D7_1121900). Bioinformatic analyses and structural modelling suggested this kinase is an orthologue of the eukaryotic phosphoinositide-dependent protein kinase 1 (PDK1), hence termed PfPDK1.

Interestingly, all PfPDK1 mutations identified in the various NF54/PKAc cOE survivors are positioned proximal to the ATP-binding cleft and do not coincide with the PIF-binding pocket, suggesting that these mutations impair the catalytic efficiency of PfPDK1 rather than its capacity to interact with substrates. It therefore seems that the most straightforward manner for the parasite to overcome the lethal effect of PfPKAc OE was to acquire mutations reducing PfPKAc activation through PfPDK1-mediated phosphorylation. We confirmed this scenario by (1) reverting the PfPDK1 M51R mutation in the NF54/PKAc cOE S1 survivor, which rendered these parasites again sensitive to PfPKAc OE; and (2) by introducing the PfPDK1 M51R mutation into the PfPKAc OE-sensitive clone M1, which rendered these parasites resistant to PfPKAc OE. Notably, we could also show that OE of the PfPKAcT189V mutant form of PfPKAc, which carries a nonphosphorylatable valine residue instead of the target threonine in the activation loop, had no negative effect on intraerythrocytic parasite development and multiplication. Together, these striking results imply that in addition to the cAMP-mediated release of PfPKAc from the regulatory subunit PfPKAr, phosphorylation of the T189 residue is essential for PfPKAc activity and provide compelling evidence that PfPDK1 is the kinase that targets this residue. While this hypothesis is entirely consistent with the evolutionary conserved role for PDK1 in activating PfPKAc in other eukaryotes [14–18], further experiments will be required to confirm that PfPDK1 indeed interacts with and activates endogenous PfPKAc via T189 phosphorylation in vivo.

Conditional depletion of PfPDK1 did not result in any obvious multiplication or developmental defects in asexual and sexual blood stage parasites, showing that largely diminished PfPDK1 protein levels are still sufficient to sustain parasite viability and proliferation. However, several lines of evidence strongly argue for an essential role for PfPDK1 in asexual parasites and that at least one of its vital functions is to activate PfPKAc. First, previous studies [31,54] and our own attempts failed to obtain PfPDK1 null mutants via gene disruption approaches. Second, none of the different mutations identified in the 6 NF54/PKAc cOE survivors introduced a nonsense loss-of-function mutation into the pfpdk1 open reading frame. This observation again supports the notion that PfPDK1 function is vital and that the PfPDK1 mutant enzymes retain residual kinase activity. Third, we were only able to introduce the M51R PfPDK1 mutation into parasites overexpressing PfPKAc but not into WT parasites, suggesting that parasites expressing functionally compromised PfPDK1 mutants can only survive if impaired PfPDK1-dependent PfPKAc activation is compensated for by elevated PfPKAc expression levels. Given that PDK1 is widely conserved in eukaryotes and required to activate AGC kinases at large [17], it is conceivable that PfPDK1 may also regulate the activity of PfPKG or PfPKB, the only other 2 known members of the AGC family in P. falciparum, which are both essential in blood stage parasites [31,79]. Furthermore, the manifestation of the PfPKAc OE phenotype in late trophozoites/early schizonts implies that PfPDK1 is active and phosphorylates other substrates already at this stage, hours prior to the expression window of endogenous PfPKAc in late schizonts. Importantly, however, the fact that the NF54/PKAc cOE survivor parasites expressing functionally impaired PfPDK1 mutant enzymes are fully viable suggests that the PfPDK1-dependent phosphorylation of other substrates is either not essential or can still be executed at functionally relevant baseline levels by mutated PfPDK1.

In summary, we provide unprecedented functional insight into the cAMP/PfPKA signalling pathway in the malaria parasite P. falciparum. Our results complement earlier studies highlighting the importance of tight regulation of PfPKA activity for parasite survival, showing that diminished as well as augmented PfPKAc expression levels are lethal for asexual blood stage parasites. In addition to the well-established roles for the regulatory subunit PfPKAr, the AC PfACβ and the PDE PfPDEβ in regulating PfPKAc activity via cAMP levels, we provide compelling evidence that PfPDK1 is required to activate PfPKAc, most likely through activation loop phosphorylation at T189. In light of the essential role for PfPDK1 in this and possibly other parasite AGC kinase-dependent signalling pathways, and the promising anti-parasite activity of PDK1 kinase inhibitors, PfPDK1 represents an attractive candidate for further functional and structural studies and to be explored as a possible new antimalarial drug target.

[END]

[1] Url: https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3001483

(C) Plos One. "Accelerating the publication of peer-reviewed science."
Licensed under Creative Commons Attribution (CC BY 4.0)
URL: https://creativecommons.org/licenses/by/4.0/


via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/