A Plasmodium falciparum protein tyrosine phosphatase inhibitor identified from the ChEMBL‐NTD database blocks parasite growth

Post‐translational modifications, especially reversible phosphorylation, are among the most common mechanisms that regulate protein function and biological processes in Plasmodium species. Of the Plasmodium phosphatases, phosphatase of regenerating liver (PfPRL) is secreted and is an essential phosphatase. Here, we expressed PfPRL in a heterologous expression system, and then purified and characterized its phosphatase activity. We found that Novartis_003209, a previously identified inhibitor, inhibited the PfPRL phosphatase activity of recombinant PfPRL and blocked in vitro parasite growth in a dose‐dependent manner. Further, in silico docking analysis of Novartis_003209 with all four P. falciparum tyrosine phosphatases (PTP) demonstrated that Novartis_003209 is a Plasmodium PTP inhibitor. Overall, our results identify a scaffold as a potential starting point to design a PTP‐specific inhibitor.

Post-translational modifications, especially reversible phosphorylation, are among the most common mechanisms that regulate protein function and biological processes in Plasmodium species. Of the Plasmodium phosphatases, phosphatase of regenerating liver (PfPRL) is secreted and is an essential phosphatase. Here, we expressed PfPRL in a heterologous expression system, and then purified and characterized its phosphatase activity. We found that Novartis_003209, a previously identified inhibitor, inhibited the PfPRL phosphatase activity of recombinant PfPRL and blocked in vitro parasite growth in a dose-dependent manner. Further, in silico docking analysis of Novartis_003209 with all four P. falciparum tyrosine phosphatases (PTP) demonstrated that Novartis_003209 is a Plasmodium PTP inhibitor. Overall, our results identify a scaffold as a potential starting point to design a PTP-specific inhibitor.
Malaria, a parasitic disease caused by Plasmodium, is still one of the leading causes of death worldwide [1]. Although malaria cases have dropped significantly in the past decade, it is believed that they may rise again as a result of the emergence of resistance to current antimalarials [1][2][3]. Therefore, there is an urgent need to identify newer targets and develop a new pipeline of antimalarials. Reversible phosphorylation mediated by kinases and phosphatases is one of the most common mechanisms by which protein functions are regulated in eukaryotes and Plasmodium [4][5][6][7][8]. Phosphorylation balance is crucial for different parasites' developmental stages [5,8]. Several genome-wide and proteome-wide studies have identified 100 kinases and 67 phosphatases in the Plasmodium falciparum genome, and many of these regulatory proteins are essential for the parasite life cycle [5,[8][9][10][11][12]. Therefore, these essential kinases/phosphatases are being considered important targets for the development of new antimalarials.
Protein tyrosine phosphatases (PTPs) are essential signaling enzymes that, together with protein tyrosine kinases, regulate diverse cellular processes such as cell motility, division, proliferation, and survival [13][14][15][16]. As the name suggests, PTPs regulate tyrosine phosphorylation of proteins and have a distinct active site signature motif, HCX5R [15,17]. Many human diseases, including cancer, diabetes/obesity, autoimmune disorders, and infectious diseases, have been linked to aberrant tyrosine phosphorylation [18,19]. Therefore, many of these tyrosine phosphatases are being studied as therapeutic targets. For example, CDC25 has emerged as an important target for cancer and autoimmune diseases, PTP1b for obesity and type II diabetes and SHP-2 for rheumatoid arthritis [16,18,20,21]. We earlier illustrated the P. falciparum phosphatome and identified 67 phosphatases grouped into 13 superfamilies [10]. Thirty-three of these phosphatases do not have any human homologs, and six of these phosphatases are Plasmodium-specific phosphatases [10]. Among these phosphatases, phosphatases of regenerating liver (PfPRL) family homolog is an essential PTP and colocalizes with AMA-1, a membrane-associated protein linked to red blood cell (RBC) invasion of Plasmodium merozoites [22]. We have previously explored PfPRL as a drug target by in silico virtual screening of ChEMBL-NTD library and identified a list of compounds. Among these compounds, Novar-tis_003209 was docked with PfPRL with the least binding energy and showed stable interaction with PfPRL [23].
In the present study, we cloned and expressed PfPRL protein and established its phosphatase activity assay. Further, we analyzed the inhibitory potential of Novartis_003209 against the purified recombinant PfPRL protein and performed comparative molecular docking analysis with the other Plasmodium PTPs catalytic site. The structure-activity relationship discussed in the present study may pave the way for the rational design of new inhibitor(s) targeting PTPs of the human malaria parasite P. falciparum.

Recombinant PfPRL expression and protein purification
For PfPRL PTP expression, the PfPRL gene was amplified using Q5 polymerase, a forward primer (ATGAAC TTGTGTCCA) with a BamHI site, and a reverse primer (CATAAAATGACATTT) incorporating SalI site to express PfPRL protein from amino acids 11N to 218 M. The PCR amplified product was inserted into the multiple cloning sites (MCS) sites of the pJET vector. Positive clones were sequenced, and the cloned fragment was inserted into the BamHI-SalI sites of the pQE-30 vector (6X-HIS tag is located upstream of BamHI-SalI (MCS) cloning sites). The positive clones for PfPRL in the pQE-30 vector were transformed in Escherichia coli m15 strain cells for protein expression. Recombinant N terminus HIStagged-PfPRL was expressed for 8 h at 25°C with 0.25 mM IPTG. Cells were harvested, and the cell pellet was resuspended in lysis buffer (50 mM Tris pH 7.5, 300 mM NaCl, 0.01% Triton, and 10% glycerol). Cells were lysed using lysis buffer by sonication; cell lysate was centrifuged at 14,300 g for 20 min. The supernatant solution containing the soluble protein was affinity-purified to near purity using a Ni-NTA + column. The purity of the protein was analyzed on SDS/PAGE and by western blot analysis.

Recombinant PfPRL phosphatase activity assay
Phosphatase activity of nearly homogenous purified PfPRL protein was analyzed using 3-O-methylfluorescein phosphate (OMFP) hydrolysis in a reaction mixture containing 50 mM sodium acetate (pH 5.5), 100 mM NaCl, 10 mM DTT, 20% glycerol, and (4 µM-1 mM) OMFP at 37°C. The reaction was initiated by the addition of 4 µM of Histagged-PfPRL, and the release of 3-O-methylfluorescein (OMF) was monitored at 485/535 nm in Victor 1420 multilabel counter (Perkin Elmer, USA). His-HDP (heme detoxification protein, a protein used in process of hemozoin formation [24,25] and does not have phosphatase domain or show phosphatase activity) was used as a negative control. The compound Novartis_003209, identified from in silico study, was analyzed in an inhibition assay. The inhibitory potential of Novartis_003209 was tested against purified recombinant PfPRL in the presence of 50 mM sodium acetate (pH 5.5), 100 mM NaCl, 10 mM DTT, 20% glycerol, and 50 µM OMFP at 37°C. Novartis_003209 was procured from MolPort. Purity and other details for the received compounds may be retrieved from the MolProt site (https://www.molport.com/) with catalog numbers STK766939 (MolProt, Riga, Latvia). GRAPHPAD PRISM 9 (GraphPad Software, Inc, USA) software was used to plot phosphatase activity and inhibition assay graphs.

Parasite growth inhibition assay
The effect of phosphatase inhibitors on parasite growth was evaluated on the 3D7 strains of P. falciparum. The parasite culture was synchronized using 5% sorbitol, and the assay was started at the schizont stage with hematocrit and parasitemia of synchronized schizont stage culture adjusted to 2% and 1%, respectively. Novartis_003209 (0.07-10 lM) was added to the parasite culture in separate 96-well plates, and parasitemia was estimated after an incubation period of 24 h using flow cytometry. Briefly, cells from the samples were collected and washed with phosphate-buffered saline (PBS), followed by staining with ethidium bromide (10 lgÁmL À1 ) for 20 min at 37°C in dark. The cells were subsequently washed twice with PBS and analyzed on FACSCalibur (Becton Dickinson, USA), using CELLQUEST software (Becton Dickinson, USA). Fluorescence signal (FL2) was detected with the 590 nm bandpass filter using an excitation laser of 488 nm, collecting 100 000 cells per sample. Uninfected RBCs stained in a similar manner were used as a control. Following data acquisition, each sample was analyzed for percentage parasitemia by determining the proportion of FL2-positive cells using CellQuest.
Comparative molecular docking of Novartis_003209 to other Plasmodium PTPs NCBI BLASTp (https://blast.ncbi.nlm.nih.gov/Blast.cgi) and PlasmodDB (release 49) [26] were used to identify P. falciparum proteins showed similarity with PfPRL. Three-dimensional structure for none of 4 Plasmodium tyrosine phosphatases has been resolved yet, so homology and threading-based techniques were used to generate a 3D structure for the Plasmodium PTPs [27][28][29]. The best-predicted model was selected for molecular dynamic simulation for further optimization, validation, and downstream analysis. GROMACS (ver 4.6.3) with the CHARMM27 force field was used to perform molecular dynamic simulation in an aqueous environment [30,31]. The methods and parameters for the MD simulations were followed as mentioned in our previous study [23]. The final simulated protein structure's quality was verified using the Ramachandran plot [32]. Ligand preparation and docking were performed using raccoon and AutoDockTools4 [33,34]. Autodock4 uses Lamarckian Genetic Algorithm and Empirical Binding Free Energy Function to calculate ligand's binding energy to the target protein [34]. PYMOL (https://pymol.org/2/) and LigPlus were used to visualize and generate images [35].

Results
PfPRL expression, purification, and its activity analysis We have earlier reported an in silico drug screening analysis against PfPRL and identified nine potential hits. Among these hits, Novartis_003209 showed the lowest free binding energy [23]. To validate the inhibitory activity of Novartis_003209, we cloned and expressed PfPRL protein (amino acid 11N-218 M) in an E. coli heterologous expression system (Fig. 1A,B). SDS/PAGE and western blot analysis using anti-HIS antibody showed expression of recombinant PfPRL with the apparent size of~25 kDa (Fig. 1A,B). Solubility studies revealed that recombinant protein was expressed in both soluble and pellet fractions in E. coli expression system (Fig. 1C). We purified the recombinant PfPRL protein on the Ni-NTA + column. And the purified protein was analyzed on SDS/PAGE and western blot using an anti-His antibody. As shown in Fig. 1A, PfPRL was purified with more than 85 % purity.
The presence of Novartis_003209 blocks parasite growth during blood stages of P. falciparum life cycle Phosphatase of regenerating liver has been shown to play a crucial role in parasite growth and survival as the attempts to generate its knockout were unsuccessful [5,36]. Further, PfPRL has been shown to be an extracellular secreted protein at asexual blood stages of P. falciparum [37]. Therefore, to investigate the role of PfPRL at asexual blood stages, we tested the effect of Novartis_003209 on parasite growth in human RBCs in an in vitro P. falciparum culture. The mature schizont stage parasites (3D7 strain) at 2% hematocrit and 1% parasitemia were treated with the phosphatase inhibitor Novartis_003209 (0.07-10 lM) in a 96-well cell culture plate. The parasitemia was estimated 24 h postinfection in control and treated samples. A dosedependent decrease in the parasitemia was observed in Novartis_003209-treated samples with a maximum of 90% reduced parasitemia at 5 µM concentration with the IC 50 value of 0.273 µM (95% Cl profile likelihood, 0.2314 to 0.3189 µM; Fig. 1F), indicating a potential role of PfPRL in parasite growth during blood stages. Interestingly, growth inhibition assay results showed that Novartis_003209 blocked 50% parasite growth at a significantly lower concentration than what it showed for blocking the in vitro recombinant phosphatase activity. Thus, these results advocated that Novartis_003209 may have other potential targets, including PfPRL phosphatase. In silico docking revealed Novartis_003209 to have an affinity for the PTP catalytic site As the Novartis_003209 showed a lower IC 50 value in an in vitro parasite growth assay than its effect on recombinant PfPRL activity, we speculated that Novartis_003209 might be targeting other P. falciparum proteins. Therefore, we searched for any P. falciparum proteins with a minimum of 30% sequence coverage and ≥ 30% sequence identity with PfPRL. In addition to this, we also included PTPs encoded by P. falciparum, as PTPs share a conserved catalytic domain, and Novartis_003209 may be targeting the conserved PTP domain. The BLASTp search with PfPRL sequence did not show any similar P. falciparum proteins. The three other PTPs with conserved HCX5R motif: PF3D7_1127000, PF3D7_0309000, and PF3D7_1455100, showed domain similarity with PfPRL. To validate the hypothesis that Novar-tis_003209 may be targeting the other three PTPs, we performed multiple sequence alignment (MSA) of the four P. falciparum PTPs which showed the presence of conserved HCX5R motif (Fig. S2). The MSA analysis revealed that histidine (H) is replaced with cysteine (C) in the signature PTP motif region of PF3D7_1455100 (Fig. S2). As no crystal structure was available and attempts to generate full homology-based models failed, we generated partial 3D coordinates of PF3D7_0309000 domain (312-432 amino acid (aa); PDB template ID: 4KI9) and PF3D7_1455100 domain (2-143; PDB template ID: 1ZZW), using SWISSMO-DEL. A single template did not predict 3D coordinates for the complete PF3D7_1127000 PTP domain (122-276 aa); hence, a 3D model was generated using the Phyre2 server. Molecular dynamic simulation of the three predicted PTPs showed stable protein conformation, further validated by Ramachandran plot analysis (Figs S3-S5). Next, molecular docking analysis was performed in triplicate using AutoDockTools to check the binding affinity of Novartis_003209 with the 3 PTPs. The comparative binding energy analysis revealed the affinity of Novartis_003209 for the 3 PTPs, highest for the PF3D7_0309000 (binding energy: 7.7 kcal/mol) but comparatively lower than PfPRL (difference of 1.5 kcal/mol) and lowest for the PF3D7_1127000 (binding energy:~6.45 kcal/mol; Table 1). The molecular docking analysis showed the binding of Novartis_003209 in the catalytic pocket for all the four PTPs, involving at least one residue from the conserved catalytic PTP motif HCX5R within 4 A (HIS153 for PfPRL, CYS379 for PF3D7_0309000, CYS141 for PF3D7_1127000, and CYS89 for PF3D7_1455100). These observations show that Novartis_003209 potentially targets other Plasmodium PTPs in vivo and the binding is PTP specific rather than PfPRL (Figs 2 and S6). Comparative molecular docking studies of Novartis_003209 with human PRL-3 (PDB ID: 1V3A) showed binding of À6.01 kcal/mol, showing Novartis_003209 has a higher affinity for PfPRL compared with human phosphatase of regenerating liver (PRL) protein [23]. Further, previous bioassay studies, including cytotoxicity against human hepatocellular carcinoma cell line (Huh7), show that Novartis_003209 is either inactive or acts at > 100 µM concentration, suggesting that Novartis_003209 acts on human proteins and cells at a comparatively very high concentration as compared to the human malaria parasite (https://pubchem.ncbi.nlm.nih.gov/compound/ 2928525). Together, the study has identified Novar-tis_003209 as an inhibitor that binds to PfPRL and other Plasmodium tyrosine phosphatases to inhibit the phosphatase activity and block the parasite growth.

Discussion
Reversible phosphorylation is a fundamental regulatory cellular process that governs several cellular functions required for cell survival, such as growth, division, adhesion, and motility. This process is mediated by kinases and phosphatases, to which a delicate balance is required for the successful completion of these cellular processes. Dysregulation of protein phosphorylation or alterations in phosphorylation patterns is a primary cause of many human diseases such as cancers, diabetes, autoimmune disorders, and neurological disorders [13,14,[18][19][20]. Kinases and phosphatases are fast emerging as novel drugable targets for many diseases, including diseases caused by various pathogens such as Plasmodium [13,14,[18][19][20][21]38]. We have carried out genome-wise phosphatome analysis of P. falciparum genome previously and identified four tyrosine phosphatases in the genome [10]. Subsequently, we targeted one of the Plasmodium tyrosine phosphatases, PfPRL, a secretory and an essential phosphatase for in silico inhibitor screening using the ChemBL-NTB database [23]. Here, we characterize one of the inhibitors from the ChemBL-NTB database, Novartis_003209, for its potential inhibitor activity as well as for the antiparasitic effects.
A previous study has expressed PfPRL as a fusion protein with GST and showed that the recombinant PfPRL functions possess phosphatase activity and are localized in the same compartment as PfAMA-1, suggesting its role in merozoite invasion or egress [22]. In the present study, we expressed the PfPRL gene in an E. coli expression vector, pQE-30, that adds a small tag 6X-HIS at the N-terminal of the PfPRL protein.
Next, we characterized the phosphatase activity of the recombinant PfPRL protein by measuring the hydrolysis of OMFP and analyzed the effect of Novar-tis_003209 on the hydrolysis of OMFP as Novartis_003209 had been earlier shown to bind PfPRL by in silico docking studies. Novartis_003209 inhibited the phosphatase activity of PfPRL in a dosedependent manner with an IC 50 value of < 10 µM. These results are in line with a previous report that has shown a block in OMFP inhibition by PfPRL in the presence of various general phosphatase inhibitors such as sodium orthovanadate and phosphate inhibitor set [22]. Next, we measured the effect of Novar-tis_003209 in an in vitro parasite growth inhibition assay by adding different concentrations of inhibitor at the mature schizont stage of the P. falciparum parasites. A dose-dependent decrease in parasitemia was observed in the Novartis_003209-treated samples, with an IC 50 of 0.273 µM.
The difference in IC 50 values observed for Novar-tis_003209 inhibition for the enzyme activity vs. on the parasite growth prompted us to check whether Novar-tis_003209 may be acting on other Plasmodium proteins, especially other Plasmodium tyrosine phosphatases or observed phenotype is solely due to changes in the PfPRL phosphatase activity as PfPRL is an essential phosphatase for parasite survival and small changes in its activity may substantially dysregulate critical parasite processes. To validate the hypothesis that Novartis_003209 may be binding to other parasite proteins, we searched for P. falciparum proteins showing similarity with PfPRL. Interestingly, none of the Plasmodium proteins showed similarity > 30% to PfPRL. Next, we performed comparative protein-ligand docking studies for Novartis_003209 with all the four Plasmodium PTPs: PF3D7_1113100 (PfPRL, from previous study [23]), PF3D7_0309000 Plasmodium falciparum dual-specificity protein phosphatase (PfYVH1), PF3D7_1127000, and PF3D7_1455100 and compared these binding affinities with the human PRL. Among these PTPs, PfPRL and PfYVH1 have been functionally characterized and deemed essential for the parasite survival and growth [5,22,39]. Although not much work has been performed regarding the other two Plasmodium PTPs, however, genome-wide knockout studies have shown that the two other PTPs are redundant. The comparative docking studies showed Novartis_003209 binds the catalytic sites of all the four Plasmodium PTPs, with PF3D7_0309000 showing binding energy comparable to PfPRL. Together, the data advocate that Novartis_003209 is a Plasmodium tyrosine phosphatase-specific inhibitor. This also explains the lower IC 50 for in vitro parasite culture and targets all the Plasmodium PTPs. Further, Novartis_003209 is inactive on human cells and proteins (https://pubchem. ncbi.nlm.nih.gov/compound/2928525), suggesting that Novartis_3209 could be further explored as a potential antimalarial [23].
In summary, present study supports our previous in silico findings, where we identified nine potential inhibitors targeting P. falciparum tyrosine phosphatase, PfPRL. Here, we show that Novartis_003209, inhibitor with the lowest free binding energy against PfPRL in the in silico screening, blocks PfPRL phosphatase activity and inhibits parasite growth. Additionally, the molecular docking simulations showed the binding affinity of Novartis_003209 to all the Plasmodium PTPs, suggesting that it could be a general Plasmodium tyrosine phosphatase inhibitor. It will be interesting to develop diverse compounds based on the structural scaffold of Novartis_003209 to make it target specific and study the efficacy of diverse yet related molecules to develop new and efficacious antimalarial.

Supporting information
Additional supporting information may be found online in the Supporting Information section at the end of the article. A, suggesting correct and stable protein fold prediction for PF3D7_0309000. Additionally, during 10ns molecular dynamics run, the protein structure did not break, confirming stable predicted 3D structure. (G) Ramachandran plot analysis to check the quality of 3D model. Fig. S4. Molecular dynamics simulation showing stable conformations of the predicted 3D structures of PF3D7_1127000. (A) Number of energy minimization steps required to achieve maximum force less than 1000 kJ mol-1 nm-1. (B) Fluctuations in temperature at constant volume (isothermal-isochoric process) show that the system reaches the target temperature (300K) quickly and remain stable over the remainder of the equilibration. (C-D) Fluctuations in pressure and density at a constant temperature. (E-F) Root mean square deviation (RMSD) calculation using protein backbone structure and radius of gyration fluctuations during the 10 ns production simulation. Each analysis shows fluctuations within 2 A, suggesting correct and stable protein fold prediction for PF3D7_1127000. Additionally, during 10ns molecular dynamics run, the protein structure did not break, confirming stable predicted 3D structure. (G) Ramachandran plot analysis to check the quality of 3D model. Fig. S5. Molecular dynamics simulation showing stable conformations of the predicted 3D structures of PF3D7_1455100. (A) Number of energy minimization steps required to achieve maximum force less than 1000 kJ mol-1 nm-1. (B) Fluctuations in temperature at constant volume (isothermal-isochoric process) show that the system reaches the target temperature (300K) quickly and remain stable over the remainder of the equilibration. (C-D) Fluctuations in pressure and density at a constant temperature. (E-F) Root mean square deviation (RMSD) calculation using protein backbone structure and radius of gyration fluctuations during the 10 ns production simulation. Each analysis shows fluctuations within 2 A, suggesting correct and stable protein fold prediction for PF3D7_1455100. Additionally, during 10ns molecular dynamics run, the protein structure did not break, confirming stable predicted 3D structure. (G) Ramachandran plot analysis to check the quality of 3D model.