Indirubin Analogues Inhibit Trypanosoma brucei Glycogen Synthase Kinase 3 Short and T. brucei Growth

The protozoan parasite Trypanosoma brucei is the causative agent of human African trypanosomiasis (HAT). The disease is fatal if it remains untreated, whereas most drug treatments are inadequate due to high toxicity, difficulties in administration, and low central nervous system penetration.

ABSTRACT

The protozoan parasite Trypanosoma brucei is the causative agent of human African trypanosomiasis (HAT). The disease is fatal if it remains untreated, whereas most drug treatments are inadequate due to high toxicity, difficulties in administration, and low central nervous system penetration. T. brucei glycogen synthase kinase 3 short (TbGSK3s) is essential for parasite survival and thus represents a potential drug target that could be exploited for HAT treatment. Indirubins, effective leishmanicidals, provide a versatile scaffold for the development of potent GSK3 inhibitors. Herein, we report on the screening of 69 indirubin analogues against T. brucei bloodstream forms. Of these, 32 compounds had potent antitrypanosomal activity (half-maximal effective concentration = 0.050 to 3.2 μM) and good selectivity for the analogues over human HepG2 cells (range, 7.4- to over 641-fold). The majority of analogues were potent inhibitors of TbGSK3s, and correlation studies for an indirubin subset, namely, the 6-bromosubstituted 3′-oxime bearing an extra bulky substituent on the 3′ oxime [(6-BIO-3′-bulky)-substituted indirubins], revealed a positive correlation between kinase inhibition and antitrypanosomal activity. Insights into this indirubin-TbGSK3s interaction were provided by structure-activity relationship studies. Comparison between 6-BIO-3′-bulky-substituted indirubin-treated parasites and parasites silenced for TbGSK3s by RNA interference suggested that the above-described compounds may target TbGSK3s in vivo. To further understand the molecular basis of the growth arrest brought about by the inhibition or ablation of TbGSK3s, we investigated the intracellular localization of TbGSK3s. TbGSK3s was present in cytoskeletal structures, including the flagellum and basal body area. Overall, these results give insights into the mode of action of 6-BIO-3′-bulky-substituted indirubins that are promising hits for antitrypanosomal drug discovery.

INTRODUCTION

Protozoan parasites of the genus Trypanosoma are the causative agents of nagana in animals and sleeping sickness in humans (human African trypanosomiasis [HAT]), which are fatal diseases if they are left untreated. Over 50 million people living in 25 sub-Saharan African countries but also travelers visiting rural areas in these countries are at risk of contracting HAT (1, 2).

After a bite of an infected Glossina tsetse fly, trypanosomes reach the bloodstream and tissues of the mammalian host, where they multiply as extracellular parasites, causing phase 1 of HAT (with clinical signs of fever, headache, and swelling of lymph nodes). In phase 2, the parasites pass into the central nervous system (CNS) of the host, causing neurological damage, which eventually leads to coma and death (3). Until recently, infected people diagnosed in the second phase of the disease were left with only two drugs as an option for treatment, namely, melarsoprol and eflornithine (4). Both drugs have considerable limitations, such as high toxicity, difficulties in administration, and low CNS penetration. Combination therapy with nifurtimox and eflornithine is now the first-line treatment of second-stage HAT (5). Even with this new combination, however, there are already reports of parasites developing drug resistance in cell culture (6–9). Thus, for enforcing the armory against HAT, it is essential to continue the drug discovery efforts.

In the last 2 decades, significant emphasis was given to drug discovery based on molecular targets, an approach that facilitates rational drug optimization (10) and that can work in excellent complementarity with the traditional phenotypic approach. Recently, several ongoing efforts of targeted drug discovery in the search for novel antitrypanosomal agents have been reported (10), with new leads being tested in the clinic (11). Several molecular targets, including parasite kinases, have been prioritized (10).

In this context, protein kinases of the CMGC group have already been validated as potential targets that could be exploited for HAT treatment (10, 12). Among these is the Trypanosoma brucei short isoform of glycogen synthase kinase 3 (TbGSK3s) (13), one of the two homologues of GSK3 in T. brucei. This kinase is essential for the survival of the bloodstream form (BSF) of T. brucei (13) and has recently attracted attention as a target for the discovery of new antitrypanosomal agents (13–17).

As the inhibition of GSK3 is relevant to a wide range of diseases, a multitude of low-molecular-weight inhibitors has been developed (16, 18, 19). Among the GSK3 inhibitors, of special interest is the class of indirubins, a family of natural bis-indole derivatives known for over a century as minor constituents of plant-, animal-, and microorganism-derived indigo. Indirubins are potent ATP-competitive inhibitors of protein kinases in mammals (GSK3, cyclin-dependent kinases [CDKs], dual-specificity tyrosine-regulated kinases [DYRKs]) (20–23). Indirubin analogues have also been shown to be effective antiparasitics, showing good efficacy against T. cruzi parasites (24) and against Leishmania parasites (25–27), while members of this family are also potent Leishmania GSK3s (LGSK3s) inhibitors (25, 26).

Herein, we report on the screening of an in-house indirubin collection composed of 69 compounds and the identification of novel analogues for which antitrypanosomal activity is related to inhibition of TbGSK3s. We explain the TbGSK3s-indirubin interaction by the use of theoretical calculations and report on the biology of the cell cycle arrest and cell death induced by specific analogues. Moreover, we describe the intracellular localization of TbGSK3s to understand the role of TbGSK3s and the mechanism of action of TbGSK3s inhibitors.

RESULTS

In vitro screening of an in-house indirubin collection for derivatives with antitrypanosomal activity.

As indirubins are a family of compounds with potent inhibitory activity against other kinetoplastids, including Leishmania (25, 26) and T. cruzi (24), we evaluated their activity against the related parasite T. brucei brucei. First, a collection of 69 indirubins consisting of six subsets based on the substitution pattern (Fig. 1) was screened against the bloodstream form (BSF) of T. brucei BSF9013 parasites at 10 μM and 1 μM. A growth curve performed with the same seeding density used in the drug screening process ensured that the parasites were exponentially growing until 96 h, the final time point of T. brucei growth assessment (see Fig. S1 in the supplemental material). In the initial screening, 42 and 32 analogues for the thresholds of 10 μM and 1 μM, respectively, were found to cause at least 50% growth inhibition in T. brucei parasites (data not shown). The half-maximal effective concentration (EC₅₀) was subsequently calculated for the 32 promising compounds that displayed the most potent antitrypanosomal activity in the low-micromolar and nanomolar range (EC₅₀ = 0.050 to 3.2 μM) (thirteen 6-bromosubstituted 3′-oximes [6-BIO-3′] bearing an extra bulky substituent on the 3′ oxime [(6-BIO-3′-bulky)-substituted], two 7-substituted, one 6-bulky-substituted, nine 6-Br-substituted, three 5-substituted, and three 6-halogen-substituted analogues and one 3′-methyl-substituted indirubin) (Fig. 1).

FIG 1.

Inhibitory activities of indirubin analogues against T. brucei BSF9013 parasites and TbGSK3s and their cytotoxicity against cells of the HepG2 cell line. Inhibitory activity was evaluated using the alamarBlue assay, as described in the text. EC₅₀ values were determined from dose-response curves via linear interpolation. For the kinase assays, 7 μM ATP was used. The selectivity index (EC₅₀ for HepG2/EC₅₀ for BSF T. brucei) is also displayed. The experiment was repeated 3 independent times in triplicate. Standard deviations are presented.

Specifically, the analogues 6-BIO, compound 50, 5-methyl-6-BIO (5-Me-6-BIO), and compound 61, which have previously been reported to be antileishmanial agents (25), displayed higher activity against T. brucei parasites (EC₅₀ = 0.17 ± 0.1, 0.19 ± 0.1, 0.6 ± 0.2, and 0.92 ± 0.3 μM, respectively) (Fig. 1) than Leishmania donovani parasites (25, 26). Moreover, a specific group of indirubins, the 6-BIO-3′-bulky-substituted analogues, previously shown to have potent antileishmanial activity (EC₅₀ = 0.59 to 2.44 μM) with enhanced potency against the leishmanial GSK3s (LGSK3s) (26), displayed potent growth activity with EC₅₀ values in the nanomolar range (EC₅₀ = 0.05 to 0.6 μM) (Fig. 1). It is important to mention that among the active compounds, several pairs of bases-salts appeared to have similar in vitro activity including the base-salt pair 4 (EC₅₀ = 0.055 ± 0.015 μM) and 5 (EC₅₀ = 0.052 ± 0.002 μM) and the base-salt pair 8 (EC₅₀ = 0.540 ± 0.18 μM) and 9 (EC₅₀ = 0.650 ± 0.025 μM), as anticipated.

In order to test if the observed activity was related to kinase inhibition, we evaluated the antiparasitic activity of indirubins with an N1-methyl substitution (compounds 58 and 59), a modification known to inactivate the indirubin pharmacophore toward kinases (21). We established that the N1-methyl derivatives of 6-BIO and compound 50 (compounds 58 and 59, respectively), known to be inactive toward protein kinases (25), did not inhibit the growth of T. brucei parasites in culture. Finally, the activity of the indirubin collection that was previously evaluated for cytotoxicity against the murine macrophage cell line J774.1 (26) was also evaluated against the human cell line HepG2, a hepatic cell line widely used for determining the toxicity of chemicals and drugs (28) (Fig. 1). The analysis showed that 14 analogues had a selectivity toward T. brucei by 2 orders of magnitude, while compounds 10 and 16, bearing a piperazine and a pyrrolidine substitution respectively, had an EC₅₀ toward the parasite in the submicromolar range and a reduced toxicity toward hepatic cells (EC₅₀ > 50 μM) (cytotoxicity/antitrypanosomal activity, >10-fold). In addition to the information presented above, pentamidine was used as a control drug against T. brucei parasites and displayed an EC₅₀ value of 0.0037 ± 0.0004 μM, in line with previous reports (29).

Indirubin analogues are potent inhibitors of TbGSK3s.

Since indirubins are known inhibitors of Leishmania major LGSK3s, bearing 51.9% and 65.7% identity to Homo sapiens GSK3β and TbGSK3s, respectively, we examined the inhibitory activity of the 32 analogues that displayed strong antitrypanosomal activity (EC₅₀ < 3.2 μM) against TbGSK3s. Two indirubins with no antitrypanosomal activity (compounds 6 and 58) were also tested as negative controls for their inhibitory potential against TbGSK3s. To this end, recombinant TbGSK3s was generated in a baculovirus expression system, produced in Sf9 insect cells, and purified as described in Material and Methods (Fig. S2). The GS-1 peptide was used as the substrate in order to assess the kinase activity of TbGSK3s. The specific activity of TbGSK3s was found to be 800 U/mg, where 1 U represents the incorporation of 1 nM phosphate into 1 μM GS-1 per min at 30°C at a final ATP concentration of 1 μM. The K_m values were 5.8 ± 1.5 and 6.3 ± 1.3 μM for the substrate and ATP, respectively, in agreement with previously published values (13, 16). After constructing dose-response curves, we calculated the half-maximal inhibitory concentration (IC₅₀) values toward TbGSK3s (Fig. 1). TbGSK3s was inhibited by 29 out of 32 indirubin analogues (Fig. 1) with an IC₅₀ below 1 μM (IC₅₀ = 0.021 μM to 0.847 μM). The only exceptions were analogues 63 (IC₅₀ = 1.077 μM) and 15 (IC₅₀ = 2.277 μM), which presented a lower efficacy against TbGSK3s, and analogue 21, which displayed very weak activity against TbGSK3s (IC₅₀ > 3.33 μM). Few 6-Br-substituted analogues inhibited TbGSK3s in the submicromolar range (for example, compound 53, the IC₅₀ of which was 0.023 μM). However, all 6-bromosubstituted 3′-oximes (6-BIO) bearing an extra bulky substituent on the 3′ oxime, except compound 15, were found to inhibit TbGSK3s in the nanomolar range, with the most potent being the base-salt pair of pyrrolidine analogues (compounds 16 and 17), which displayed an IC₅₀ of 0.023 and 0.021 μM, respectively. The lower efficacy of compound 15 toward TbGSK3s could be due to different ionization constants (pK_a values), which mostly affect physicochemical properties, such as aqueous solubility, potentially influencing ionic interactions, hydrogen binding, and ligand interactions (30).

The above-described observations, in combination with the knowledge that indirubin analogues often target more than one kinase (25–27, 31), prompted us to investigate the connection between the antitrypanosomal activity of indirubins and the inhibitory activity against TbGSK3s by performing correlation studies based on the logarithms of the EC₅₀ and IC₅₀ values of the indirubins (Fig. 2). When all 34 indirubins (the 32 active analogues and the 2 inactive analogues) listed in Fig. 1 were included in the study, a moderate uphill (positive) correlation (Fig. 2A) (Pearson r = 0.4278, P = 0.0116) between TbGSK3s inhibition and antitrypanosomal activity was observed. To investigate whether specific backbone substitutions contributed to TbGSK3s inhibitory activity, the same correlation study was performed after dividing indirubins into specific groups based on their substituents: non-6-Br-substituted analogues (compounds 21, 32, 45, 62 to 65, and 67), 6-substituted indirubin analogues (6-BIO, 5-Me-6-BIO, and compounds 3 to 6, 8 to 17, 45, 50 to 57, and 63 to 65), 6-Br indirubin analogues (6-BIO, 5-Me-6-BIO, and compounds 3 to 6, 8 to 17, and 50 to 57), 6-BIO-3′-bulky indirubin analogues (compounds 3 to 7 and 10 to 17), and the 6-BIO-3′ pyrrolidine/piperazine-substituted indirubins (compounds 10 to 14, 16, and 17). Notably, for analogues that did not carry a bromine at position 6 of the indirubin scaffold, there was a moderate positive correlation between antitrypanosomal and anti-TbGSK3s activity (Pearson r = 0.4829, P = 0.2254) (Fig. 2B), in contrast to the findings for the 6-substituted and 6-Br-substituted analogues, which actually displayed a weak positive correlation (Pearson r = 0.3473 and P = 0.0702 [Fig. 2C] and Pearson r = 0.3883 and P = 0.0608 [Fig. 2D]). This suggests that the antitrypanosomal activity of the analogues that did not carry a substitution at position 6 not only is due to the inhibition of TbGSK3s but also could be due to the inhibition of additional kinases. It is also worth mentioning that a specific subgroup of indirubins, the 6-BIO-3′-bulky-substituted analogues (compounds 8 to 17), which previously showed enhanced activity against LGSK3s (26), displayed a higher positive correlation between antitrypanosomal and anti-TbGSK3s activity (Pearson r = 0.6193, P = 0.0138) (Fig. 2E). Moreover, when the correlation for indirubins which bore a bulky ring of pyrrolidine or piperazine at the 3′ oxime was evaluated and when analogue 15 of the latter subgroup was excluded, a stronger positive correlation was observed (Pearson r = 0.6645, P = 0.0723) (Fig. 2F). This suggests that the presence of polar, bulky substituents of the 3′ oxime of the indirubin scaffold likely offers an enhanced potential to inhibit TbGSK3s.

FIG 2.

Correlation between antitrypanosomal activity and TbGSK3s inhibition by different subgroups of indirubin analogues. Correlation for (A) all indirubin analogues (black), (B) non-6-Br-substituted indirubin analogues (green), (C) 6-substituted indirubin analogues (red), (D) 6-Br-indirubin analogues (dark purple), (E) 6-BIO 3’ bulky substituted indirubins (blue), and (F) 6-BIO 3’ bulky ring substituted indirubins pyrrolidine/piperazine (light purple) is shown. The graphs correlate the logarithmic EC₅₀ values (Log EC₅₀) against T. brucei BSF parasites and the logarithmic IC₅₀ values (Log IC₅₀) against TbGSK3s. The Pearson coefficient and P values are indicated. Dashed lines display the 95% confidence interval. n.s., not significant.

Binding mode of indirubin analogues in TbGSK3s.

In an effort to better understand the observed selectivity of indirubins against T. brucei and TbGSK3s over Leishmania and LGSK3s and the selectivity of specific indirubin scaffolds over others, in the absence of an experimentally determined structure, a homology model of TbGSK3s kinase was created. A set of active indirubin analogues was docked to the kinase active site following a stepwise protocol.

Rigid docking was performed by using the Glide SP algorithm (32, 33), selected top-ranked poses were redocked, and the relative free energies of binding were determined for the studied ligands by implementing the Embrace module of Macromodel software (Schrödinger Release 2018-1: Macromodel, 2018; Schrödinger, LLC, New York, NY), additionally accounting for protein flexibility. The binding mode of indirubins inside the TbGSK3s binding pocket was in perfect agreement with the usually observed type I, ATP-competitive inhibition mode determined for indirubins complexed with the human kinase homolog. Three hydrogen bonds were formed at the kinase hinge region between the indirubin ligands and backbone atoms of residues E102 and V104, thus stabilizing the protein-inhibitor complexes (Fig. 3A), while the bromine substituent of position 6 was in all cases oriented toward the side chain of gatekeeper M101 (Fig. 3C), contributing to selectivity over other kinases. When unsubstituted, the oxime at 3′ contributed to binding principally by interacting with the Gly-rich loop (derivatives 51, 52, 62, 64, and 65), in accordance with previously described models (22). In contrast, the polar side chains of 3′-bulky-substituted indirubins and especially those that carried ionizable amines (such as the piperazine and pyrrolidine analogues 10, 12, 14, and 16) were further stabilized inside the kinase active site by forming at least one additional hydrogen bond with residues of the active-site periphery (side chains of H156, N157, or D171) (Fig. 3B), in line with their corresponding binding modes to the human homologue (23). Results obtained for 7-substituted analogues were in good agreement with previously derived structure-activity relationship observations as well (34). Their poor capacity to adopt reasonable docking poses was in accordance with their diminished binding affinity for the parasite enzyme (IC₅₀ values of analogues 21 and 32, >3.33 and 0.847 ± 0.303 μM, respectively), whereas the steric clash between the 7-position halogen and the hinge region of TbGSK3s was the most likely explanation of this biological effect (34). To provide an additional level of validation for the constructed model, a correlation between experimental and predicted binding affinities was performed. The linear interaction energy (LIE) approach was implemented using a set of 20 indirubin analogues representative of the studied collection. A good correlation was obtained (r² = 0.64), showing that the model can explain a significant part of the determined structure-activity relationships between indirubin analogues and the parasite enzyme target (see Fig. S3 for more details).

FIG 3.

Proposed mode of indirubin analogue binding to the active site of TbGSK3s. (A) Inhibitors 6-BIO (yellow carbons) and 10 (orange carbons) are shown as ball-and-stick models inside the GSK3s binding pocket, which is represented as a molecular surface colored according to its electrostatic potential (red, negative; blue, positive). The protonated piperazine ring of inhibitor 10 forms a charge-assisted hydrogen bond with the side chain of D171 (salt bridge), thus offering additional stabilization to the kinase-inhibitor complex (inset B), while the indirubin core is anchored at the hinge region mainly through two hydrogen bonds with the backbone atoms of V104 (inset C). (D) Side chain conformation of the arginine residue defining the binding cavity outer boundary in GSK3, as derived by MM-GBSA (molecular mechanisms, the generalized Born model and solvent accessibility) flexible docking calculations of 5-Me-6-BIO in the T. brucei (cyan, residue R110) and L. major (plum, residue R109) homologs. The difference in the flexibility of arginine originates from its intramolecular interaction with either E106 in T. brucei or the shorter side chain of D105 in L. major (not shown), possibly explaining the preference of indirubins for the trypanosomal homolog, where induced-fit effects can be accommodated more readily.

Notably, flexible docking calculations of the 6-BIO-3′-bulky-substituted analogues revealed an interesting result which could possibly be used to rationalize, at least partially, the observed selectivity of indirubins toward the trypanosomal protein over its leishmanial homolog. A T-shaped π-π stacking interaction was formed between the aromatic system of the indirubin scaffold and the guanidine side chain of R110 in TbGSK3s (Fig. 3D). This interaction is rather typical in kinases that carry a conserved arginine residue at this position and has been observed in the corresponding protein-ligand complexes in previously presented models of human GSK3β and Aurora kinases (34). Although LGSK3s also carries an arginine at the corresponding position (R109), an already observed difference related to the intramolecular partner of this side chain could account for the observed higher affinity of indirubins toward TbGSK3s. Indeed, the arginine side chain interacts through a salt bridge with either D105 (in L. major) or E106 (in T. brucei), thus forming in both cases the outer boundary of the kinase active site. Due to the additional carbon atom of the glutamate side chain (T. brucei) compared to aspartate (L. major), the arginine-glutamate pair creates a more flexible element that can accommodate the extended bis-indole ring of indirubins more easily than its leishmanial counterpart can. This notion, already suggested previously (35), was consistently supported by results of docking calculations. In all studied complexes, the arginine side chain showed a considerable displacement of about 1.6 Å from its starting coordinates, thus indicating the pronounced flexibility of this structural element to accommodate optimal stacking interactions with indirubins. On the contrary, attempts to model the same interaction using the experimentally determined structure of LGSK3s (PDB accession number 3E3P) did not result in reasonable docking poses. This was well anticipated, as the poor suitability of the latter crystal structure to serve as a template for docking experiments arises from the fact that the Leishmania kinase structure was determined as an apoenzyme. Moreover, in the L. major crystal structure, the side chain of R109 is oriented in a way that rather hinders full access to the binding site. This particular conformation of R109 seems to be fairly stabilized by the already discussed salt bridge formed between the protonated arginine guanidinium group and the proximal carboxylate of D105. In conclusion, the hypothesis that the R109-D105 pair confers more limited flexibility on L. major binding pocket plasticity than its R110-E106 counterpart in T. brucei is in very good agreement with the biological activity data presented in Fig. 1.

The mode of trypanocidal action of 6-BIO-3′-bulky-substituted indirubin analogues is similar to that of TbGSK3s knockdowns.

To gain a better understanding of the mode of action of TbGSK3s inhibitors, we performed phenotypic comparisons between indirubin-treated parasites and parasites in which TbGSK3s gene expression was ablated by tetracycline (Tet)-inducible RNA interference (RNAi) (TbGSK3s^RNAi), as previously described (36).

To this end, the expression of TbGSK3s was ablated in BSF parasites (BSF9013, Lister 427) by tetracycline-induced RNA interference (RNAi) as previously reported (13). Western blot analysis using an affinity-purified polyclonal antibody raised against LGSK3s showed a unique band at the theoretical molecular mass of TbGSK3s (40 kDa) (Fig. S4A) whose intensity was significantly reduced upon expression of double-stranded TbGSK3s^RNAi (≈86% reduction of the kinase expression 48 h after RNAi induction) (Fig. S4B). The growth inhibition observed after TbGSK3s knockdown was consistent with published data (13). After 72 h of RNAi induction, the cell concentration was diminished by 80% (Fig. S4C). Ablation of TbGSK3s resulted in apoptosis-like phenotypes. More specifically, DNA nicks were observed by terminal deoxynucleotidyltransferase-mediated dUTP-biotin nick end labeling (TUNEL) (40.9% ± 1.2% of the parasites) (Fig. S4D). Moreover, propidium iodide (PI) and annexin V-fluorescein isothiocyanate (FITC) staining revealed after 24 h of TbGSK3s RNAi induction a proportion of early apoptosis-like parasites (8.8%) with an intact membrane and phosphatidylserine (PS) externalization on the surface of the cell. At later time points after TbGSK3s RNAi induction, late-apoptotic cells (with a loss of membrane integrity and PS externalization) accumulated (15.5% at 72 h) (Fig. S5).

Cell morphology studies revealed that TbGSK3s knockdown resulted in the formation of multinucleate cells with many kinetoplasts and flagella (up to 17% of the parasites after TbGSK3s RNAi induction) (Fig. S4EIII). Cell cycle analysis of TbGSK3s^RNAi cells confirmed the increase in the number of multinucleated cells at the early stage of 24 h (25.5%) up to 48 h (15.8%) upon RNAi induction. The same analysis showed that the parasite population with DNA from 4 nuclei (4N) increased to 34% 48 h after RNAi induction (Fig. 4A).

FIG 4.

Cell cycle analysis of indirubin-treated and TbGSK3s RNAi BSF T. brucei parasites by flow cytometry. The parasite DNA content at 0 h, 24 h, and 48 h after TbGSK3s^RNAi induction (A), 0 h, 24 h, and 48 h after compound 4 treatment at the EC₅₀ (EC₅₀ = 0.055 μM) (B), 0 h, 24 h, and 48 h after compound 10 treatment at the EC₅₀ (EC₅₀ = 0.078 μM) (C), 0 h, 24 h, and 48 h after compound 5-Me-6-BIO treatment at the EC₅₀ (EC₅₀ = 0.6 μM) (D), and 0 h, 24 h, and 48 h after compound 6-BIO treatment at the EC₅₀ (EC₅₀ = 0.17 μM) (E) is shown. The DNA content was measured by staining the cells with propidium iodide (PI), and the cell cycle status was analyzed by flow cytometry. Statistics on the DNA content and the initial cell densities (icd) are present in each graph. Cells were seeded at a density of 1 × 10⁵ cells/ml, and cells were counted at every time point before cell cycle analysis. At 24 h, the cell density was found to be between 4 × 10⁵ cells/ml and 4.5 × 10⁵ cells/ml, while at 48 h, the cell density was found to be between 3.3 × 10⁵ cells/ml and 4 × 10⁵ cells/ml.

The cell cycle was further analyzed by counting the numbers of nuclei (N) and kinetoplasts (K) per cell. As the cell cycle of T. brucei is characterized by nuclear and kinetoplast components, the number of N and K per cell is indicative of the phase of the cell cycle. Thus, 2N2K characterizes cells in cytokinesis, 2K1N characterizes cells in mitosis, and 1K1N characterizes cells in G₁ phase. Our results are consistent with data previously reported by Jones et al. (12) showing that ablation of TbGSK3s expression by RNAi at 48 h after tetracycline induction resulted in a small increase in the number of 1N2K and 2N2K cells and suggested that TbGSK3s may be involved in mitosis and/or cytokinesis (Table S1).

After evaluating the effects on cell death and cell cycle upon TbGSK3s knockdown, we compared the observed effects on parasites treated with 6-BIO-3′-bulky-substituted indirubin analogues (compounds 4 and 10). In this analysis, we also used 5-Me-6-BIO and 6-BIO, whose antitrypanosomal action correlates loosely with their inhibitory activity toward TbGSK3s.

Our results revealed that compounds 4 and 10 provoked a mode of cell death in T. brucei parasites similar to that provoked in parasites ablated in TbGSK3s. There was an increase of early apoptosis-like populations at 24 h and 48 h after indirubin treatment, followed by an increase of late-apoptotic and necrotic cells at 72 h (Fig. 5). The morphology of the parasites incubated with these compounds resembled the morphology of TbGSK3s^RNAi cells, as they became small and round, while at the same time DNA condensation and fragmentation were observed (data not shown). Incubation of T. brucei parasites with compounds 4 and 10 also resulted in a cell cycle profile similar to that of cells knocked down for TbGSK3s, including a 1.8- and 2.4-fold increase, respectively, of the 4N/2N ratio over that for untreated cells at 48 h (Fig. 4).

FIG 5.

Effect of indirubins on the cell death of BSF T. brucei parasites. The effect of treatment with compounds 4, 10, 5-Me-6-BIO, and 6-BIO at the EC₅₀ for 0 h (control), 24 h, 48 h, and 72 h on the cell death of BSF T. brucei parasites is shown. Cells were stained with annexin V-FITC [horizontal axis, first laser parameter (FL) FL1-H] and propidium iodide (PI; vertical axis, FL2-H) to assess apoptosis (annexin V staining) and membrane impermeability. The samples were then analyzed by flow cytometry. Untreated and DMSO-treated parasites were used as negative controls. Parasites treated with H₂O₂ for 20 min were used as apoptosis-like positive controls. Population percentages and the initial cell densities (icd) are presented in each graph. Graphical plots of the population fractions are also shown. Cells were seeded at a density of 1 × 10⁵ cells/ml, and cells were counted at every time point before cell cycle analysis. At 24 h, the cell density was found to be between 4 × 10⁵ cells/ml and 4.5 × 10⁵ cells/ml, while at 48 h, the cell density was found to be between 3.3 × 10⁵ cells/ml and 4 × 10⁵ cells/ml.

While incubation of parasites with the compounds 5-Me-6-BIO and 6-BIO resulted in small and rounded-up trypanosomes with apparent DNA condensation and fragmentation (Fig. 5), the assessment of parasite cell death and cell cycle suggested that these compounds induced a pattern of cell death different from that in TbGSK3s^RNAi cells and cells treated with compounds 4 and 10. Notably, at 48 h and 72 h, there was an increase in the population of necrotic cells for 5-Me-6-BIO, while 6-BIO inhibited growth rather than killed the parasites (Fig. 5). Cell cycle analysis revealed a rise of the population with an N of <1 after treatment with 5-Me-6-BIO (49.7% of cells after 72 h of treatment) and 6-BIO (6% of cells after 48 h of treatment) (Fig. 4).

TbGSK3s localization in T. brucei parasites highlights a role in the flagellum and the basal body.

As TbGSK3s localization might give us insights into the role of this kinase in the regulation of T. brucei cell division and reveal aspects of the inhibitory action of indirubins, we decided to investigate the localization of TbGSK3s. For immunolocalization studies, we used an affinity-purified anti-GSK3s polyclonal antibody raised against the LGSK3s homologue (25) which recognizes a unique band in a Western blot of total T. brucei protein extract at the theoretical molecular mass of TbGSK3s (40 kDa), as previously reported (Fig. S4A). Immunostaining of paraformaldehyde (PFA)-fixed T. brucei BSF parasites showed a promiscuous localization to the cytoplasm and flagellum and a less pronounced localization in the nucleus (data not shown). Methanol fixation, a method that results in the precipitation of proteins and the gradual removal of lipids, depending on the fixation time, showed that TbGSK3s localized throughout the parasite and was detected as granule-like structures in the cytoplasm, the nucleus, and the flagellum. Moreover, the antibody produced a bright signal near the kinetoplast (Fig. 6A). We further investigated the localization of TbGSK3s near the kinetoplast and flagellum in detergent-extracted parasites and by double staining with the axonemal marker Mab25 (37) and the Mab22 marker, which recognizes the fibers connecting the kinetoplast to the proximal region of mature and probasal bodies (38, 39). Detergent extraction prior to fixation removes membrane and cytosolic proteins, and only proteins firmly associated with the cytoskeleton remain present (40). Under these conditions, traces of TbGSK3s localized in the flagellum axoneme as individual spots (Fig. 6B). Staining for TbGSK3s was found to be distal from Mab22, suggesting that the protein is present in the basal and probasal body (Fig. 6C). As this dual localization is reminiscent of that of intraflagellar transport (IFT) proteins (37), we investigated if TbGSK3s behaves as an IFT protein. To this end we performed immunofluorescence analysis in the retrograde transport IFT140^RNAi mutant, which produces short flagella filled with IFT material (37). Double labeling of TbGSK3s and an IFT protein, IFT172 (present in the short bulbous flagellum), showed that there was no colocalization between IFT172 and TbGSK3s (data not shown), suggesting that this kinase does not behave as an IFT protein. The presence of TbGSK3s in the flagellum requires further investigation, and the possibility that this kinase is axonemal may indicate its involvement in the regulation of the microtubule-based skeleton of the flagellum.

FIG 6.

Localization of TbGSK3s in T. brucei parasites. Phase-contrast images (Phase), fluorescence microscopy images (black and white) showing nuclear and kinetoplast DNA stained with DAPI (DAPI), and images of TbGSK3s stained with anti-LGSK3s (TbGSK3s) of BSFs fixed with methanol (A), costained with Mab25 after detergent extraction (B), or costained with Mab22 in T. brucei after detergent extraction (C) are shown. Merged images of the blue fluorescence (DAPI) with the green fluorescence (TbGSK3s) and the red fluorescence (Mab25 or Mab22) are also shown. On the bottom right, inset represents a 2-fold magnification of a portion of the merged image. Arrows indicate staining in the flagellum and basal body areas. Bars, 8 μm.

DISCUSSION

Our previous knowledge of the inhibitory action of indirubin analogues against Leishmania species (25–27) and Trypanosoma cruzi (24) parasites, as well as against leishmanial GSK3s (26), inspired us to explore their action against T. brucei, another member of the kinetoplastid family. We aimed to identify new antitrypanosomal agents based on inhibition of TbGSK3s, whose presence is essential for parasite survival (13). It is worth mentioning that despite the fact that indirubin analogues and, specifically, 6-bromo-substituted derivatives serve as potent inhibitors of the mammalian GSK3β (22), inhibition of this kinase is tolerated in adult mammals (41, 42), and thus, trypanosomatid GSK3s could serve as an excellent molecular target.

Our results revealed that indirubin-based analogues are effective antitrypanosomal agents against BSF T. brucei parasites and have more potent antiparasitic action against T. brucei than against L. donovani or T. cruzi (24–26). This could be attributed to either differences in the biology of the parasite (membrane permeability, different GSK3σ requirements) or the ability to inhibit its molecular target(s) more potently, or both. In agreement with this, molecular docking calculations revealed that in TbGSK3s the residues that define the outer boundary of the active site (R110 and E106) form a flexible salt bridge that can accommodate the bis-indole indirubin system more easily than the corresponding, more rigid pair in LGSK3s (R109 and the shorter D105).

Indirubin analogues have been shown to inhibit more than one kinase in Leishmania (GSK3s [25, 26], Cdc2-related kinase 3 [CRK3] [25, 27], and casein kinase 1.2 [CK1.2] [31]), depending on the substitution pattern of the indirubin scaffold. We showed that the 6-BIO-3′-bulky-substituted analogues were the most selective inhibitors of TbGSK3s, in line with our theoretical calculations. In this context, addition of bulky polar substituents in position 3′ further stabilizes the TbGSK3s-inhibitor complex by creation of at least one additional hydrogen bond (23). Interestingly, the same scaffold displayed enhanced selectivity toward LGSK3s rather than CRK3 in Leishmania parasites (26) and thus can be considered a promising initial scaffold for the development of novel TbGSK3s inhibitors and antiparasitic drugs against trypanosomatids.

The inability of 7-substituted analogues to significantly inhibit TbGSK3s, despite their antiparasitic action, points to their interaction with different targets. In mammalian cells, 7-substituted indirubins are predicted to target Aurora kinases (34) and DYRKs (21). This raises the possibility that Aurora kinases and DYRKs could also be targets of 7-substituted indirubins in T. brucei parasites, and this may explain the poor correlation between their inhibitory activity against TbGSK3s and parasite growth. Finally, compound 32 (7-trifuoromethyl-3′-oxim-6′-carboxymethyl-indirubin) inhibited TbGSK3s (IC₅₀ = 0.847 μM), while it did not target the human GSK3β (>10 μM) and did not display any antileishmanial activity, according to previously published results (26). Compound 45, a potent TbGSK3s inhibitor, had rather moderate antitrypanosomal activity. The latter could be due to the different physicochemical properties of the compound, i.e., to the lower level of uptake by the cell.

RNAi-mediated ablation of TbGSK3s expression initially resulted in cytostasis (including a rise in 2N2K populations and >2N2K populations and a small rise in 1N2K parasites), which was followed by an apoptosis-like death after prolonged induction of RNAi. These results are similar to the rise in the 1N2K and 2N2K populations observed in a previous study (12) and indicate a possible deregulation of mitosis and cytokinesis (12). Parasites treated with 6-BIO-3′-bulky-substituted analogues 4 and 10 displayed a cell cycle and cell death phenotype similar to that of TbGSK3s^RNAi cells, suggesting that trypanosome cell death by these compounds could at least be mediated by inhibition of TbGSK3s.

In order to gain insights into the phenotypic changes brought about by TbGSK3s ablation and pharmacological inhibition of TbGSK3s, the localization of TbGSK3s was investigated. This study revealed that TbGSK3s is present throughout the cytoplasm but is also found in the flagellum and concentrated in the basal body area. This is consistent with a proteomic study that detected TbGSK3s in purified intact flagella of the procyclic stage of T. brucei (43). The protein was not found in flagella extracted with high salt (44), in agreement with the sharp signal reduction observed here upon detergent treatment. The orthologous kinase is also present in the flagellum (25) and in the basal body of Leishmania (unpublished data), suggesting that both T. brucei and Leishmania kinases might act on similar signaling networks. The dual presence of TbGSK3s in the flagellum and the basal body is reminiscent of the localization of intraflagellar transport proteins (37), but our results suggest that TbGSK3s is not directly involved in this process. Traces of TbGSK3s were detected in the axoneme, the microtubule-based skeleton of the flagellum, implying that TbGSK3s might be involved in signaling networks related to either flagellum morphogenesis or motility, or both. For example, an active tyrosine-phosphorylated form of GSK3 enriched in the axoneme of the regenerating flagellum of Chlamydomonas reinhardtii (45) and the more abundant active GSK3α form in highly motile bovine sperm than in less motile bovine sperm (46) indicate a role for GSK3 kinases in flagellum morphogenesis and motility (47–49). The basal body has a critical role in nucleating flagellar axonemal microtubules in flagellated eukaryotes, with consequences for organelle division and cell morphogenesis (50). As both the flagellum and basal body are important organelles in directing morphogenesis and cytokinesis (51, 52), elucidation of TbGSK3s substrates and interactors localized in the basal body and flagellum could give us insights into the mechanism by which TbGSK3s modulates (directly or indirectly) cell division. This knowledge could provide further insights into GSK3 signaling, important for drug discovery efforts.

In conclusion, this work reveals the potent inhibitory activity of indirubin analogues against TbGSK3s and T. brucei growth. As indirubins have previously been shown to inhibit T. cruzi (24) and Leishmania (25, 26) growth, they could justifiably emerge as potential leads for targeted drug discovery against trypanosomatid parasites.

MATERIALS AND METHODS

Cell culture.

A Trypanosoma brucei brucei bloodstream form (BSF9013, Lister 427) (53) was cultured in HMI-9 medium (54) (Gibco) (supplemented with 10% heat-inactivated fetal bovine serum [HI-FBS; Gibco], 10 mM HEPES [Gibco], and antibiotics [G418, 3 μg/ml]; AppliChem-BioChemica] [54]).

Chemical library.

The chemical library consisted of 69 indirubins (Fig. 1) that were synthesized as previously described (22, 23, 55, 56). The compounds were dissolved in dimethyl sulfoxide (DMSO; AppliChem-BioChemica) at 10 mM, and appropriate serial dilutions in DMSO were made. Indirubins were diluted in culture medium to give the desired final concentrations.

In vitro testing of the antitrypanosomal activity of indirubins.

The alamarBlue assay was applied in order to determine the antitrypanosomal activity of the indirubin analogues. T. brucei BSF9013 parasites (1 × 10⁶ cells/ml) were seeded into 96-well flat-bottom plates at a density of 5 × 10⁴ cells/ml in 200 μl HMI-9 medium containing increasing indirubin concentrations or the equivalent volume of the diluent, DMSO, each in quadruplicate. The final concentration of DMSO was always <1% (vol/vol) and did not affect the growth of the parasites. Following indirubin treatment for 72 h, alamarBlue (20 μl/well; Invitrogen) was added and the plates were incubated at 37°C in 5% CO₂ for a further 24 h. Colorimetric readings were performed at a test wavelength of 550 nm and a reference wavelength of 620 nm. Comparison of DMSO-treated controls with samples allowed the calculation of the concentration of indirubin necessary to reduce the growth rate of trypanosomes by 50% (EC₅₀ values). EC₅₀ values were determined from dose-response curves via linear interpolation. Experiments were performed in quadruplicate and repeated 3 independent times.

Induction of TbGSK3s downregulation by RNA interference.

The RNAi constructs (GSK1, GSK3/p2T7^TABlue) were constructed as previously described (57) and generously provided by Wesley C. Van Voorhis. The RNAi constructs (GSK1, GSK3/p2T7^TABlue) were linearized with the NotI restriction enzyme (New England Biolabs, Ipswich, MA) before they were electroporated into parasites (13, 57). Cell growth was monitored for 7 days. TbGSK3s^RNAi cells were cultured in HMI-9 medium supplemented with 2.5 μg/ml hygromycin and 2.5 μg/ml G418 in the absence or the presence of 1 μg/ml tetracycline (Sigma-Aldrich) for 24 h, 48 h, and 72 h.

Flow cytometry analysis.

BSF T. brucei parasites (5 × 10⁴ parasites/ml) were cultured at 37°C in 5% CO₂ for 48 h and 72 h in the presence of the tested indirubin analogues or DMSO (used as a control). The cell cycle analysis of the samples by fluorescence-activated cell sorting (FACS) was carried out as previously described for L. donovani parasites (25, 26). For determining the parasite population that displayed membrane integrity and PS externalization, FACS analysis was performed using the protocol for annexin V-FITC and propidium iodide (PI) staining (annexin V-FITC; Trevigen). The analysis was performed in a Becton, Dickinson FACSCalibur flow cytometer, and the data were analyzed using CellQuest software. All experiments were performed at least 3 times.

Cellular and nuclear morphology.

BSF T. brucei parasites posttreated for 24 h, 48 h, and 72 h with indirubins or DMSO or TbGSK3s^RNAi cells incubated with tetracycline were fixed in 4% (wt/vol) paraformaldehyde (PFA; Scharlau) and treated with 10 μg/ml RNase A (Sigma-Aldrich) and 50 μg/ml PI (BD Biosciences). Parasites were observed under a TCS SP Leica confocal fluorescence microscope. The results for at least 100 cells from three independent experiments were recorded for each condition.

In situ labeling of DNA fragments by TUNEL.

A cell death fluorescein detection kit (Roche Applied Science) was used according to the manufacturer’s instructions in order to detect in situ DNA strand breaks in T. brucei parasites incubated for 72 h with indirubins at their EC₅₀. Samples were analyzed under a Zeiss Axiophot fluorescence microscope using Ph2 Plan Neofluor 40× and Ph3 Plan 63× Apochromat lenses. The ratio of apoptosis (apoptotic to total cells) was determined by counting at least 400 cells per group in 3 independent experiments.

Gene cloning.

The TbGSK3s gene (Tb427.10.13780) was amplified by PCR from T. brucei bloodstream-form (BSF9013, Lister 427) genomic DNA using the following primers: sense primer 5′-GTGTGGATCCAATGTCGCTCAACCTTACCGA-3′ and antisense primer 5′-GAGACTCGAGCTTCTTCAGCAGATACTCCC-3′.

The amplified PCR product was cloned into the BamHI-XhoI site (New England Biolabs) of the Novagen vector pTriEx-1.1, to ensure the expression of the recombinant protein TbGSK3s fused to a C-terminal polyhistidine extension. The cloned gene was verified by sequencing to confirm that the sequence was identical to that in the database.

Generation of TbGSK3s recombinant baculovirus.

A baculovirus expression system was used to obtain recombinant TbGSK3s. The pTriEx-1.1-TbGSK3s plasmid was cotransfected with the BaculoGold DNA into Spodoptera frugiperda (Sf9) insect cells, according to the BD BaculoGold manual (BD Biosciences). Extracts of pTriEx-1.1 cotransfected with the BaculoGold DNA into Spodoptera frugiperda (Sf9) insect cells were used as a negative control for the expression of the kinase. The Sf9 insect cells were cultured in Sf900 II medium (Gibco) supplemented with 5% (vol/vol) fetal bovine serum and the antibiotic gentamicin at 50 μg/ml (Sigma-Aldrich) at 25°C.

The recombinant baculovirus was isolated using a plaque assay and used to again transfect Sf9 insect cells in order to express TbGSK3s. The transfected Sf9 insect cells were harvested and lysed, and TbGSK3s was detected in the supernatant of the lysate after sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) and Western blotting using a polyclonal IgG His probe antibody (1:500 dilution; stock solution, 200 μg/ml; Santa Cruz Biotechnology) and a polyclonal anti-rabbit GSK3β antibody (Abcam).

Production and purification of TbGSK3s.

Insect cells (10⁸) were infected with recombinant baculovirus at a multiplicity of infection (MOI) equal to 1. Specifically, after 7 days of viral infection, the cells and the medium were collected and centrifuged at 170 × g for 5 min. The Sf9 cell pellets containing TbGSK3s were suspended in lysis buffer (150 mM NaCl [AppliChem], 0.1% [wt/vol] SDS [Sigma-Aldrich], 1% [vol/vol] Nonidet P-40 [NP-40], 0.5% [vol/vol] sodium deoxycholate [Sigma-Aldrich], 0.02% [wt/vol] NaN₃ [Sigma-Aldrich], 100 μg/ml phenylmethylsulfonyl fluoride [PMSF; Roche], 1 μg/ml aprotinin [Sigma-Aldrich], 50 mM Tris-HCl, pH 8.0 [AppliChem], Roche proteinase inhibitor tablet), and cell lysis was carried out at 4°C for 30 min on a rotating shaker. After the incubation, the lysate was centrifuged at 15,000 × g at 4°C, the supernatant was transferred into a new tube, and the His-tagged protein was purified on Ni²⁺-nitrilotriacetate (Ni-NTA) resin according to the manufacturer’s instructions (Macherey-Nagel). Purified TbGSK3s (present in the elution with 250 mM imidazole) was stored in 10% (vol/vol) glycerol at −80°C for use in kinase assays. Under these conditions, the average yield of soluble TbGSK3s was 10 μg per 10⁸ insect cells, while the purity of the protein eluted from the Ni-NTA resin was above 95%, as judged by SDS-PAGE (see Fig. S1 in the supplemental material).

SDS-PAGE and immunoblotting.

Parasites or cell lysates (5 × 10⁶ cell equivalents) were resolved by 12% SDS–PAGE according to the method of Laemmli (58), transferred to nitrocellulose membranes (Hybond C; GE Healthcare), and probed with the appropriate primary antibodies and, after incubation, with peroxidase-conjugated secondary antibody. ECL Plus (enhanced chemiluminescence; GE Healthcare) or horseradish peroxidase (HRP)–3,3′-diaminobenzidine staining with nickel (Vector Laboratories) was used for detection. AlphaImager software (Alpha Innotech) was used to quantify the immunoblot bands.

Immunofluorescence.

T. brucei parasites were washed once with phosphate-buffered saline (PBS) and then (i) fixed in 4% (wt/vol) PFA for 20 min, (ii) fixed with ice-cold methanol for 5 min, or (iii) treated with 1% (vol/vol) NP-40 in PEM buffer (100 mM PIPES [piperazine-N,N′-bis(2-ethanesulfonic acid); Sigma-Aldrich], 2 mM EGTA [Sigma-Aldrich], 1 mM MgSO₄ [Merck]), for 10 s, followed by fixation with 4% (wt/vol) PFA for 20 min (40). In the case of PFA fixation, cells were permeabilized with 0.1% (vol/vol) Triton X-100 (Sigma-Aldrich) in PBS for 5 min and cells were blocked in PBS containing 1% (wt/vol) bovine serum albumin (BSA; Thermo Fisher Scientific). The samples were incubated directly for 1 h with anti-LGSK3s (1:250) (25) and mouse Mab22 (38) or Mab25 (59) antibodies (1:100) in 0.1% (wt/vol) BSA in PBS. The cells were washed three times for 5 min each time with 1 ml of PBS to remove excess antibodies. Alexa Fluor 546- and Alexa Fluor 488-conjugated anti-rabbit and anti-mouse immunoglobulin antibodies (Molecular Probes) were added at a final concentration of 2 μg/ml in 0.1% (wt/vol) BSA in PBS for 1 h at room temperature. The secondary antibody was removed by three washing steps with 1 ml PBS for 5 min, and the parasite DNA was stained for 1 min at room temperature with 1 μg/ml DAPI (4′,6-diamidino-2-phenylindole; Thermo Fisher Scientific). The samples were washed twice with PBS, and coverslips were mounted with Vectashield mounting medium (Vector Laboratories). Samples were observed under a TCS SP confocal microscope (Leica) using a 63× Apochromat lens and a DMI4000 (Leica) or a DMR (Leica) microscope. Images were acquired with a Horca 03G camera (Hamamatsu, Hamamatsu City, Japan) for the DMI4000 microscope or with a CoolSnap camera (Roper) for the DMR microscope.

Docking calculations.

The homology model of TbGSK3s (UniProt accession number Q388M1) was built using an alignment created by BLAST and the Prime module of the Maestro suite (Small-Molecule Drug Discovery Suite 2018-1, 2018; Schrödinger, LLC, New York, NY) on the human homolog template structure (PDB accession number 1UV5) with default settings. Rigid docking was performed with the Glide SP algorithm, and a scaling factor of 0.8 was utilized for both the protein and ligand atoms to account for possible induced-fit effects upon inhibitor binding. The best poses were redocked, and the free energies of binding were determined by using the MM-GBSA algorithm with the variable dielectric generalized Born model for water and protein flexibility in a radius of 6 Å around the binding pocket, as implemented in the Prime module of the Maestro suite.

Kinase assays.

Recombinant TbGSK3s expressed and purified from Sf9 insect cells was used in the kinase assays in the presence of the compounds in increasing concentrations following the instructions of the manufacturer of the kinase luminescent assay kit (Promega), as previously described (25, 26). For determining the kinase activity of TbGSK3s, GS-1 peptide (YRRAAVPPSPSLSRHSSPHQSpEDEEE) (proteinkinase.de) was used as a substrate. In order to determine the IC₅₀ values of the inhibitors, we used ATP and substrate concentrations at the calculated K_m values in a final volume of 30 μl. We ensured that the time of the reaction was in the linear range, and the kinase assays were performed for 30 min at 30°C. At this time point, different kinase concentrations of a freshly purified kinase were used, and dose-response curves were made to evaluate the reduction of luciferase activity. For the kinase reaction, a kinase concentration that reduced the luciferase activity within the lower readings of the linear range (typically, 3/4) was selected. The average concentration of kinase used in the different assays performed was 40 ng. After a 30-min incubation at 30°C in the presence of increasing concentrations of the inhibitors, the reaction was stopped by addition of cold Kinase-Glo reagent, and after 10 min the luminescence was measured. Sample with no inhibitor addition was used as a positive control. As a negative control, we used volumes of mock eluates of beads incubated with noninfected Sf9 extracts. The experiment was repeated 3 times in triplicate. IC₅₀ values (in micromolar) were determined from the dose-response curves.