*3.3. Hemichromes Accumulation and ROS Production*

The next element of our hypothesis was that enhanced accumulation of hemichromes in Syk inhibitor- and DHA-treated pRBCs should increase the oxidative stress within the pRBCs. To test this step in the hypothesis, we quantitated the level of free radicals generated within the RBCs by EPR using a spin trap, α-phenyl-*N*-*tert*-butylnitrone (PBN), to stabilize any generated free radicals. The six-lined spectra, characteristic of the PBN adduct formed after radical trapping in different samples, are presented in Figure 3A. Simulation of the spectra yielded the following parameters: *g* = 2.0059, a<sup>N</sup> = 14.9 G, and a<sup>H</sup> = 3.3 G, indicating that a hydroxyl radical was generated and transformed into a CH<sup>3</sup> radical after reaction with DMSO. Importantly, little or no signal was recordable in either control RBCs or healthy RBCs treated with DHA, P505-15, or both together, i.e., indicating that the combination of DHA + Syk inhibitor does not induce significant oxidative stress in healthy RBCs. In contrast, stimulation of HMC formation in healthy cells by treatment with phenylhydrazine (PHZ) was found to enable the DHA + Syk inhibitor combination to potently induce free radical formation, i.e., suggesting that HMCs must be present to catalyze the antimalarial-induced formation of free radicals. Consistent with these results, although only a low-intensity EPR signal was observed in untreated pRBCs (ring stage), this signal was significantly enhanced upon brief incubation (10 min) with 200 µM DHA (this high concentration was chosen due to the short time of incubation) and then further strengthened by addition of P505-15 (*p*-value = 0.028) (Figure 3A). Since very similar results were obtained when the ROS-sensitive fluorescent probe, CM-H2DCFDA (*p*-value = 0.022), was employed to quantitate oxidative stress (Figure 3B), we conclude that, in the presence of HMCs, induced by phenylhydrazine or parasite growth, intra-erythrocyte oxidative stress can be greatly increased by the addition of Syk inhibitors plus DHA. *Antioxidants* **2020**, *9*, x 10 of 22 using a spin trap, α-phenyl-*N*-*tert*-butylnitrone (PBN), to stabilize any generated free radicals. The six-lined spectra, characteristic of the PBN adduct formed after radical trapping in different samples, are presented in Figure 3A. Simulation of the spectra yielded the following parameters: *g* = 2.0059, a<sup>N</sup> = 14.9 G, and a<sup>H</sup> = 3.3 G, indicating that a hydroxyl radical was generated and transformed into a CH<sup>3</sup> radical after reaction with DMSO. Importantly, little or no signal was recordable in either control RBCs or healthy RBCs treated with DHA, P505-15, or both together, i.e., indicating that the combination of DHA + Syk inhibitor does not induce significant oxidative stress in healthy RBCs. In contrast, stimulation of HMC formation in healthy cells by treatment with phenylhydrazine (PHZ) was found to enable the DHA + Syk inhibitor combination to potently induce free radical formation, i.e., suggesting that HMCs must be present to catalyze the antimalarial-induced formation of free radicals. Consistent with these results, although only a low-intensity EPR signal was observed in untreated pRBCs (ring stage), this signal was significantly enhanced upon brief incubation (10 min) with 200 µM DHA (this high concentration was chosen due to the short time of incubation) and then further strengthened by addition of P505-15 (*p-*value = 0.028) (Figure 3A). Since very similar results were obtained when the ROS-sensitive fluorescent probe, CM-H2DCFDA (*p*-value = 0.022), was employed to quantitate oxidative stress (Figure 3B), we conclude that, in the presence of HMCs, induced by phenylhydrazine or parasite growth, intra-erythrocyte oxidative stress can be greatly increased by the addition of Syk inhibitors plus DHA.

**Figure 3.** (**A**) Intensity of electron paramagnetic resonance (EPR) spectra (arbitrary units) in the presence of *N*-*tert*-butyl-α-phenylnitrone (PBN) spin trapping agent in RBCs treated with phenylhydrazine (PHZ) (400 µΜ) with/without DHA (200 µΜ). RBCs and pRBCs treated with/out P505-15 (0.5 µΜ) for 24 h with/without 10-min incubation of DHA (200 µΜ). (**B**) The same experiments mentioned above (**A**) were performed using the fluorescent probe 2′,7′ dichlorodihydrofluorescein diacetate (CM-H2DCFDA), a cell-permeable indicator of reactive oxygen species (ROS). Data are the average ± SD of five independent experiments. Significant differences **Figure 3.** (**A**) Intensity of electron paramagnetic resonance (EPR) spectra (arbitrary units) in the presence of *N*-*tert*-butyl-α-phenylnitrone (PBN) spin trapping agent in RBCs treated with phenylhydrazine (PHZ) (400 <sup>µ</sup>M) with/without DHA (200 <sup>µ</sup>M). RBCs and pRBCs treated with/out P505-15 (0.5 <sup>µ</sup>M) for 24 hwith/without 10-min incubation of DHA (200 <sup>µ</sup>M). (**B**) The same experiments mentioned above (**A**) were performed using the fluorescent probe 20 ,70 -dichlorodihydrofluorescein diacetate (CM-H2DCFDA), a cell-permeable indicator of reactive oxygen species (ROS). Data are the average ± SD of five independent experiments. Significant differences between no-DHA RBCs and pRBCs at \*\* *p* < 0.001.

between no-DHA RBCs and pRBCs at \*\* *p* < 0.001. Figure 4A, chelation of the heme iron significantly decreased the intensity of the free radical signal (*p*-value = 0.019), indicating the requirement of accessible iron in the production of activated Figure 4A, chelation of the heme iron significantly decreased the intensity of the free radical signal (*p*-value = 0.019), indicating the requirement of accessible iron in the production of activated DHA and oxidative free radicals in the parasitized RBC.

DHA and oxidative free radicals in the parasitized RBC.

*Antioxidants* **2020**, *9*, x 11 of 22

**Figure 4.** (**A**) Intensity of EPR spectra (arbitrary units) in the presence of PBN spin trapping agent in pRBCs treated with/without P505-15 (0.5 µΜ) for 24 h with/without 10-min incubation with DHA (200 µΜ) and with different concentrations (100–400 µM) of Deferasirox (DFX). (**B**) The same experiments mentioned above (**A**) were performed using the probe CM-H2DCFDA. Data are the average ± SD of five independent experiments. Significant differences between untreated DHA-RBCs and pRBCs at \* *p* < 0.05, \*\* *p* < 0.001. **Figure 4.** (**A**) Intensity of EPR spectra (arbitrary units) in the presence of PBN spin trapping agent in pRBCs treated with/without P505-15 (0.5 µM) for 24 h with/without 10-min incubation with DHA (200 µM) and with different concentrations (100–400 µM) of Deferasirox (DFX). (**B**) The same experiments mentioned above (**A**) were performed using the probe CM-H2DCFDA. Data are the average ± SD of five independent experiments. Significant differences between untreated DHA-RBCs and pRBCs at \* *p* < 0.05, \*\* *p* < 0.001. **Figure 4.** (**A**) Intensity of EPR spectra (arbitrary units) in the presence of PBN spin trapping agent in pRBCs treated with/without P505-15 (0.5 µΜ) for 24 h with/without 10-min incubation with DHA (200 µΜ) and with different concentrations (100–400 µM) of Deferasirox (DFX). (**B**) The same experiments mentioned above (**A**) were performed using the probe CM-H2DCFDA. Data are the average ± SD of five independent experiments. Significant differences between untreated DHA-RBCs and pRBCs at \* *p* < 0.05, \*\* *p* < 0.001.

These data suggest that the availability of reactive iron in hemichrome-rich pRBCs but essentially absent from healthy RBCs is essential for the catalysis of activated DHA and ROS production. As evidenced in the previous paragraph, ROS scavengers decreased the production of ROS in pRBCs (Figure 5A,B) and MPs released from pRBCs (Figure 5C,D). These data suggest that the availability of reactive iron in hemichrome-rich pRBCs but essentially absent from healthy RBCs is essential for the catalysis of activated DHA and ROS production. As evidenced in the previous paragraph, ROS scavengers decreased the production of ROS in pRBCs (Figure 5A,B) and MPs released from pRBCs (Figure 5C,D). These data suggest that the availability of reactive iron in hemichrome-rich pRBCs but essentially absent from healthy RBCs is essential for the catalysis of activated DHA and ROS production. As evidenced in the previous paragraph, ROS scavengers decreased the production of ROS in pRBCs (Figure 5A,B) and MPs released from pRBCs (Figure 5C,D).

**Figure 5.** ROS production was performed using the probe CM-H2DCFDA in pRBCs pretreated with/without P505-15 (0.5 µΜ) and DHA (200 µΜ) and their MPs for 24 h at different concentrations **Figure 5.** ROS production was performed using the probe CM-H2DCFDA in pRBCs pretreated with/without P505-15 (0.5 µM) and DHA (200 µM) and their MPs for 24 h at different concentrations of *N*-acetyl-l-cysteine (NAC) (**A**,**C**) and GSH (Glutathione) (**B**,**D**). Significant differences between untreated DHA-RBCs/MPs and pRBCs/MPs at \* *p* < 0.05, \*\* *p* < 0.001.

with/without P505-15 (0.5 µΜ) and DHA (200 µΜ) and their MPs for 24 h at different concentrations

#### *3.4. Evaluation of Synergy between Syk Inhibitors and Artemisinins in Suppression of Parasitemia 3.4. Evaluation of Synergy between Syk Inhibitors and Artemisinins in Suppression of Parasitemia*

untreated DHA-RBCs/MPs and pRBCs/MPs at \* *p* < 0.05, \*\* *p* < 0.001.

*Antioxidants* **2020**, *9*, x 12 of 22

The previous paragraphs suggest that DHA and Syk inhibitors act synergistically to induce the formation of ROS in pRBCs. To determine whether this synergistic production of ROS might translate into synergistic suppression of parasitemia, we employed the combination index theorem of Chow–Talalay [55,56] to quantitate the synergy/antagonism between DHA and Syk inhibitors in eliminating *P. falciparum* from cultures of fresh human blood. As noted in Section 3, experimental data points in the required isobolograms that lie below the diagonal line indicate synergy, while those that lie on the diagonal demonstrate additivity, and those that reside above the line are interpreted to indicate antagonism. To construct these isobolograms, ring-stage pRBCs (12 h post infection) were treated with increasing concentrations of the desired drugs for 24 h, and the residual parasitemia was quantitated to determine the IC<sup>50</sup> value of each Syk inhibitor at the indicated concentration of artemisinins. As seen in Figure 6, regardless of the Syk inhibitor employed, all inhibitors synergized with DHA in suppressing parasitemia. The previous paragraphs suggest that DHA and Syk inhibitors act synergistically to induce the formation of ROS in pRBCs. To determine whether this synergistic production of ROS might translate into synergistic suppression of parasitemia, we employed the combination index theorem of Chow– Talalay [55,56] to quantitate the synergy/antagonism between DHA and Syk inhibitors in eliminating *P. falciparum* from cultures of fresh human blood. As noted in Section 3, experimental data points in the required isobolograms that lie below the diagonal line indicate synergy, while those that lie on the diagonal demonstrate additivity, and those that reside above the line are interpreted to indicate antagonism. To construct these isobolograms, ring-stage pRBCs (12 h post infection) were treated with increasing concentrations of the desired drugs for 24 h, and the residual parasitemia was quantitated to determine the IC<sup>50</sup> value of each Syk inhibitor at the indicated concentration of artemisinins. As seen in Figure 6, regardless of the Syk inhibitor employed, all inhibitors synergized with DHA in suppressing parasitemia.

**Figure 6.** Isobolograms showing the interactions between Syk Inhibitors (P505-15, R406, entospletinib, SYK II, piceatannol, and imatinib) and dihydroartemisinin (DHA), after 24 h of incubation in the *P*. *falciparum* Palo Alto strain. Synchronized *P. falciparum* cultures were treated for 24 h with different concentrations (from 0.05 to 2.5 µM) of different Syk inhibitors (P505-15, R406, entospletinib, SYK II, piceatannol, and imatinib) in combination with different concentrations of DHA (from 0.6 to 10 nM) at the ring stage. IC<sup>50</sup> concentrations of all the drug combinations were plotted in isobolograms to determine synergy, additivity, or antagonism. **Figure 6.** Isobolograms showing the interactions between Syk Inhibitors (P505-15, R406, entospletinib, SYK II, piceatannol, and imatinib) and dihydroartemisinin (DHA), after 24 h of incubation in the *P*. *falciparum* Palo Alto strain. Synchronized *P. falciparum* cultures were treated for 24 h with different concentrations (from 0.05 to 2.5 µM) of different Syk inhibitors (P505-15, R406, entospletinib, SYK II, piceatannol, and imatinib) in combination with different concentrations of DHA (from 0.6 to 10 nM) at the ring stage. IC<sup>50</sup> concentrations of all the drug combinations were plotted in isobolograms to determine synergy, additivity, or antagonism.

Moreover, as noted from their quantitative combination index values (Table 1), the strength of the observed synergies corresponded approximately with the potencies of the Syk inhibitors (i.e., Moreover, as noted from their quantitative combination index values (Table 1), the strength of the observed synergies corresponded approximately with the potencies of the Syk inhibitors (i.e., combination index (CI) values were 0.42 for P505-15 and R406, followed by entospletinib and SYK II at

0.51, and then piceatannol and imatinib at 0.56 and 0.73, respectively), and similar strong synergy was also observed when other forms of artemisinins were evaluated (Table S1, Supplementary Materials).


**Table 1.** Combination index value (CI) 24 and 48 h after treatment.

Combination index (CI) (Chow–Talalay) of different Syk inhibitors (P505-15, R406, entospletinib, SYK II, piceatannol, and imatinib) at different concentrations (50–500 nM) in combination with different artemisinin derivatives (dihydroartemisinin, artesunate, and artemether) after 24 and 48 h of incubation: additive effect (C = 1), synergism (CI < 1), and antagonism (CI > 1). <sup>1</sup> IC<sup>50</sup> on Syk catalytic subunit.

To add evidence to the role of oxidative stress mediated by HMCs in the mechanism leading to the synergistic interaction between Syk inhibitors and artemisinin, we conducted experiments growing *P. falciparum* in G6PD-deficient RBCs that, due to their reduced capability to produce NADPH (Nicotinamide adenine dinucleotide phosphate), are known to maximize the effects of oxidative stress, promoting the formation of HMCs both in vitro and in G6PD-deficient patients following ingestion of redox-active drugs or fava beans [30,62,63]. As previously described [34], parasites display the same growth rate in control and G6PD-deficient RBCs. In G6PD-deficient RBCs, we observed an increment in HMC content following DHA and Syk inhibitor treatment (Figure S2 Supplementary Materials). In addition, in G6PD-deficient RBCs, we measured a significant (*p* < 0.01) decrease in IC<sup>50</sup> for DHA (from 2.60 nM to 1.75 nM) and a variable increase in synergistic interaction between DHA and Syk inhibitors. The CI shifted from 0.42 to 0.36 for P505-15 and from 0.73 to 0.33 for imatinib. The change in CI was significant (*p* < 0.01) only for the less specific Syk inhibitor imatinib, possibly indicating that potent Syk inhibitors are capable of determining the maximal synergistic effect on DHA (Figure S3, Supplementary Materials).

It was also important to notice that treatments with DHA + Syk inhibitors did not cause any increase in HMCs in non-parasitized G6PD-deficient RBCs, confirming that this combination does not cause oxidative injury to RBCs. To further investigate this relevant issue, we measured hemolysis as the percentage of total Hb contained in RBCs released in the supernatant following 24 h of treatment with DHA (10 nM) + P505-15 (250 nM). In untreated control and G6PD-deficient RBCs, we could not observe a significant difference in hemolysis (1.21% ± 0.76% and 1.66% ± 0.53%, respectively); following 24 h of treatment, we observed a non-significant increase in hemolysis both in control and in G6PD-deficient RBCs (2.43% ± 1.76% and 3.06% + 1.53%, respectively).

Because chelation of iron by DFX was found to block the production of ROS by DHA + Syk inhibitor (Figure 6), we then predicted that DFX might similarly prevent the synergistic elimination of parasitemia by DHA + Syk inhibitor. As shown in Figure 7 and Table 2, this prediction was *Antioxidants*  indeed realized. **2020**, *9*, x 15 of 22

**Figure 7.** Isobolograms showing the interactions between Syk Inhibitors (P505-15 and R406) and dihydroartemisinin (DHA) and artesunate (AS) in combination with the iron chelator Deferasirox (DFX), after 24 h of incubation in the *P*. *falciparum* Palo Alto strain. Synchronized *P. falciparum* cultures were treated for 24 h with different concentrations (from 0.05 to 2.5 µM) of representative Syk inhibitors (P505-15 and R406) in combination with different concentrations of DHA and AS (from 0.6 to 10 nM) using a fixed concentration (50 µM) of the iron chelator Deferasirox (DFX) at the ring stage. **Figure 7.** Isobolograms showing the interactions between Syk Inhibitors (P505-15 and R406) and dihydroartemisinin (DHA) and artesunate (AS) in combination with the iron chelator Deferasirox (DFX), after 24 h of incubation in the *P*. *falciparum* Palo Alto strain. Synchronized *P. falciparum* cultures were treated for 24 h with different concentrations (from 0.05 to 2.5 µM) of representative Syk inhibitors (P505-15 and R406) in combination with different concentrations of DHA and AS (from 0.6 to 10 nM) using a fixed concentration (50 µM) of the iron chelator Deferasirox (DFX) at the ring stage.


**Table 2.** Combination index value (CI) 24 h after treatment. Combination index (CI) (Chow–Talalay) of different SYK inhibitors (P505-15 and R406) at different concentrations (50–500 nM) in combination with different artemisinin derivatives (dihydroartemisinin and artesunate) using a fixed concentration of the iron chelator Deferasirox (DFX) after 24 and 48 h of incubation.

Thus, regardless of whether dihydroartemisinin or artesunate was combined with P505-15 or R406 (Syk inhibitors), addition of DFX moved all data points back to the diagonal line, indicating the elimination of synergy between artemisinin and Syk inhibitor. These data argue that reactive iron and ROS production are critical to creation of the potent synergy between these two classes of antimalarial drugs.

Finally, to determine whether the combined action of Syk inhibitor plus artemisinin on *P. falciparum* survival might depend on the stage of parasite maturation, we compared the effects of the combination therapy with each monotherapy at different stages of parasite development. Figure 7 shows that, upon adding DHA + P505-15 (2 nM + 250 nM) at different stages of parasite development, the maximal activity was observed between 12 and 36 h post infection corresponding to ring and trophozoite stages. The data in Figure 8B reveal that none of the drugs have a measurable impact on pRBC morphology at 6 h post infection, after which the monotherapies (i.e., DHA or Syk inhibitor alone) display detectable but comparatively mild effects (few pycnotic cells (less than 2%)) on pRBC morphology.

In contrast, by 12 h post infection, the combination therapy is seen to exert a major influence on growth progression and pRBC morphology, essentially blocking the normal pathway of parasite maturation. While other explanations of the synergistic potency of the combination therapy can be envisioned, the data are very consistent with the ability of DHA to enhance the oxidative stress arising from hemichromes accumulated in pRBCs and with the capacity of Syk inhibitors to greatly augment this oxidative stress by preventing the discharge of these oxidative hemichromes in HMC-enriched microparticles.

*Antioxidants* **2020**, *9*, x 16 of 22

**Figure 8.** (**A**) Stage dependency of the efficacy (6, 12, 24, 36, 40, and 48 h) expressed as relative activity at one fixed dosage of DHA + P505-15 (2 nM + 250 nM). Relative activity of DHA + P505-15 (2 nM + 250 nM) added at 6, 12, 24, 36, 40, and 48 h of incubation. Values are expressed as percentage of DHA + P505-15 activity (treatment time at 24 h post infection) measured as % activity. Data are the average of five experiments ± SD. (**B**) Morphological changes in *P. falciparum* induced by Syk inhibitor (P505- 15) in combination with a fixed concentration (1.25 nM) of dihydroartemisinin after 6, 12, and 24 h of treatment. Representative images of selected damaged parasites at 6, 12, and 24 h of treatment in control and drug-treated cultures selected from Diff-Quik® fix-stained thin blood films. The micrographs were obtained using a Leica DM IRB microscope equipped with a 100× oil planar apochromatic objective with 1.32 numeric aperture and a DFC420C camera and DFC software version 3.3.1 (Leica Microsystems, Wetzlar, Germany). The scale bar in the figure is 7.5 µm. **Figure 8.** (**A**) Stage dependency of the efficacy (6, 12, 24, 36, 40, and 48 h) expressed as relative activity at one fixed dosage of DHA + P505-15 (2 nM + 250 nM). Relative activity of DHA + P505-15 (2 nM + 250 nM) added at 6, 12, 24, 36, 40, and 48 h of incubation. Values are expressed as percentage of DHA + P505-15 activity (treatment time at 24 h post infection) measured as % activity. Data are the average of five experiments ± SD. (**B**) Morphological changes in *P. falciparum* induced by Syk inhibitor (P505-15) in combination with a fixed concentration (1.25 nM) of dihydroartemisinin after 6, 12, and 24 h of treatment. Representative images of selected damaged parasites at 6, 12, and 24 h of treatment in control and drug-treated cultures selected from Diff-Quik® fix-stained thin blood films. The micrographs were obtained using a Leica DM IRB microscope equipped with a 100× oil planar apochromatic objective with 1.32 numeric aperture and a DFC420C camera and DFC software version 3.3.1 (Leica Microsystems, Wetzlar, Germany). The scale bar in the figure is 7.5 µm.

In contrast, by 12 h post infection, the combination therapy is seen to exert a major influence on

#### **4. Discussion**

Based on all the above data, we wish to propose a mechanism for the effects of artemisinins and Syk inhibitors on *P. falciparum* parasitemia. Following invasion of an erythrocyte, the parasite must digest hemoglobin in order to multiply ~18-fold during its 48-h intra-erythrocyte life cycle [19]. Parasite growth is accompanied by the production of ROS leading to hemoglobin oxidation and the formation of the hemichromes [13,64–66], which are known to be redox-active compounds [28,67,68]. In order to reduce this oxidative toxicity, the parasite promotes erythrocyte membrane weakening via activation of the phosphorylation of band 3 by Syk tyrosine kinase, which in turn enables the release of hemichrome-enriched MPs from the RBC membrane. In the absence of added drugs to enhance this oxidative stress, the parasite apparently copes with these oxidants from the pRBC. However, when a drug is added, it increases the pRBC's oxidative stress, and the parasite's ability to detoxify the added stress is exceeded as shown in Figures 3 and 4, leading eventually to parasite death. Syk kinase inhibitors prevent the discharge of HMCs in membrane-encapsulated microparticles [44] and force the accumulation/retention of redox-active HMC forms, enhancing the oxidative stress within the pRBC and, therefore, activating artemisinins. On the other hand, we observed that artemisinin compounds strongly enhance the generation of HMCs in pRBCs. This apparently cooperative behavior between Syk kinase inhibitors and artemisinins may be at the basis of their synergistic anti-plasmodial interaction.

A mechanism to explain the observed synergy between artemisinins and Syk inhibitors is also implied in the data of Figure 3A,B, where artemisinins are not observed to induce oxidant production in healthy RBCs. We explain this difference between healthy RBCs and pRBCs by noting that redox-active hemichromes and/or traces of free heme released from hemichromes or from the parasite food vacuole are required to catalyze the production of pharmaceutically active artemisinin [69]. In addition, this mechanism could be relevant to determine the high selectivity of artemisinin for parasitized RBCs and, consequently, their effectiveness. The lower effectiveness of artemisinins observed in patients with hemoglobinopathies could, therefore, determine the loss of selectivity due to artemisinin activation in non-parasitized RBCs containing traces of hemichromes and/or free heme.

Based on this pathway, the activation and/or the potency of artemisinins must depend on the abundance of HMCs in parasitized RBCs, which in turn will depend on the ability of the parasite to eliminate these HMCs in discharged microparticles. The release of MPs containing hemichromes was already described in oxidized erythrocytes [70,71], in parasitized RBCs [20,44], and in different hematological diseases characterized by hemichrome formation such as thalassemias and G6PD-deficient RBCs following oxidant treatment [39,66]. Supporting the proposed mechanism of action, iron chelators and ROS scavengers suppressed the anti-plasmodial activity of the Syk inhibitor + artemisinin treatment. On the contrary, G6PD-deficient RBCs enhanced the synergistic combination between Syk inhibitor + artemisinin combination and HMC accumulation. Importantly, no hemichrome formation or hemolysis was observed in non-parasitized G6PD-deficient parasites treated by Syk inhibitors and artemisinins, evidencing the need of the additional ROS generated by malaria parasites to trigger a virtuous cycle between the accumulation of redox-active HMCs and the activation of artemisinin. Those data are also suggestive of a lack of hemolytic effect exerted by Syk inhibitor + artemisinin combination in normal and G6PD-deficient RBCs, although more specific tests will be conducted to exclude the pro-hemolytic activity of this combination.

#### **5. Conclusions**

In conclusion, the presented data highlight the complex interactions occurring among ROS production due to parasite metabolism, redox-active HMCs, and artemisinin activation, documenting the role of Syk inhibition as a key element to synergistically improve the activity of artemisinins. Since some Syk kinase inhibitors such as R406 can be administered for long periods with minimal side effects [72,73], we propose that Syk kinase inhibitors could contribute measurably to the potencies of ACTs.

Moreover, classical tyrosine kinase was not identified in the *Plasmodium falciparum* genome [74]. The use of a drug targeting human Syk, an enzyme expressed in erythrocytes, displays the potential advantage of preventing the selection of mutant resistant parasites.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2076-3921/9/8/753/s1, Figure S1. Effect of RBC treatment with P505-15 (0.5 µM) and DHA (0.5 nM) on the accumulation of HMC after 12 h of incubation. Data are the average of five experiments ± SD; Figure S2 Effect of pRBC treatments with a representative Syk inhibitor (P505-15) (0.5 µM) and DHA (0.5 nM) on the accumulation of HMC after 3, 6, 12, and 24 h. Data are the average of five experiments ± SD. Significant differences to untreated pRBCs at \* *p* < 0.05; \*\* *p* < 0.001; Figure S3. Isobolograms showing the interactions between Syk Inhibitors (P505-15 and imatinib) and dihydroartemisinin (DHA), after 24 h of incubation in the P. falciparum Palo Alto strain treated with G6PD-d red cells; Table S1. Combination index (CI) (Chow-Talalay) of different SYK inhibitors.

**Author Contributions:** Conceptualization, K.R., A.P. and F.M.T.; Data curation, I.T., K.R., G.M., M.C.P., C.F., F.N., A.P. and F.M.T.; Formal analysis, I.T., A.P. and F.M.T.; Funding acquisition, K.R., A.P. and F.M.T.; Investigation, I.T., A.P. and F.M.T.; Methodology, I.T., K.R. and F.M.T.; Project administration, F.M.T.; Resources, A.P. and F.M.T.; Software, I.T. and P.V.; Supervision, A.P. and F.M.T.; Validation, G.M.; Visualization, C.F. and F.N.; Writing—original draft, F.M.T. and P.S.L.; Writing—review and editing, I.T., K.R., C.F., F.N., A.P. and F.M.T. All authors have read and agreed to the published version of the manuscript.

**Funding:** This study was funded by the Fondazione Banco di Sardegna: Prot. U1056.2014/AI.938MGB Prat. 2014.0040.

**Acknowledgments:** The authors would like to thank Pietro Fresu and Gioacchino Greco and MSc students Alessia Manca, Salvatore Duras, and Cristina D'Avino for their support. This study was supported by grants from the Fondazione di Sardegna Prot. U1056.2014/AI.938MGB Prat. 2014.0040. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

**Conflicts of Interest:** The authors declare no conflict of interest.
