Design and synthesis of novel hydroxamic acid derivatives based on quisinostat as promising antimalarial agents with improved safety

In our previous work, the clinical phase II HDAC inhibitor quisinostat was identified as a promising antimalarial agent through a drug repurposing strategy, but its safety was a matter of concern. Herein, further medicinal chemistry methods were used to identify new chemical entities with greater effectiveness and safety than quisinostat. In total, 38 novel hydroxamic acid derivatives were designed and synthesized, and their in vitro antimalarial activities were systematically investigated. These compounds at nanomolar concentrations showed inhibitory effects on wild-type and drug-resistant Plasmodium falciparum strains in the erythrocyte stage. Among them, compound 30 , after oral administration, resulted in complete elimination of parasites in mice infected with Plasmodium yoelii , and also exhibited better safety and metabolic properties than observed in our previous work. Mechanistically, compound 30 upregulated plasmodium histone acetylation, according to western blotting, thus suggesting that it exerts antimalarial effects through inhibition of Plasmodium falciparum HDAC enzymes.


In brief
In this work, 38 novel hydroxamic acid derivatives were designed and synthesized based on clinical phase II antitumor HDAC inhibitor quisinostat. Among them, the optimal compound 30 retained significant antimalarial activity both in vitro and in vivo, while improving safety and metabolic properties.

Graphical abstract Highlights
• Compound 30 exhibited significant antimalarial activity against both wild-type and multidrug-resistant malarial parasites.
• Compound 30 displayed completely elimination of parasites in the murine P. yoelii model.
• Compound 30 considerably reduced cytotoxicity and ameliorated metabolic properties compared with lead compound quisinostat.
• Compound 30 was confirmed to kill P. falciparum parasites by inhibiting PfHDAC enzymes activity.

INTRODUCTION
Malaria is one of the oldest and deadliest infectious diseases and has been a public-health challenge worldwide. According to the World Malaria Report 2021, an estimated 241 million cases of malaria and 627,000 deaths due to malaria occurred in 2020, representing an increase of 14 million malaria cases and 69000 deaths over 2019 [1]. The rise in malaria cases and deaths has been partly associated with the disruption of malaria services due to the COVID-19 outbreak in 2019 [2,3]; however, global progress against malaria had already stalled before the outbreak. Although the World Health Organization has recommended widespread use of the RTS,S/AS01 vaccine to prevent malaria in young children in Africa, the vaccine does not provide 100% protection [4]. Therefore, chemotherapy drugs are important treatments. However, the emergence and spread of drug-resistant malaria has been the largest public health emergency in malaria control. Alarmingly frequent emergence of artemisinin-resistant strains has been observed in Southeast Asia and Africa in recent years [5][6][7][8][9], thus posing a challenge to the use of artemisinin-based combination therapies as first-line treatments [10]. Therefore, the development of new antimalarial drugs effective against resistant plasmodium is a major scientific and public health need that must urgently be addressed.
In our previous studies, the clinical phase II anticancer HDAC inhibitor quisinostat was identified as a promising antimalarial agent; however, this compound has high toxicity. The structural modification of quisinostat has been demonstrated to decrease its toxicity while retaining potent antimalarial activity; further experiments have confirmed that its derivatives have PfHDAC inhibitory activity [26][27][28]. The work herein was aimed at exploring a new structural framework to further improve the drug's safety and broaden the therapeutic window. Referring to successful previous studies, we retained the hydroxamic acid fragment as the zinc-binding group and replaced the 4-aminomethylpiperidine moiety of the linker with a 1-oxa-4,9-diazaspiro [5.5]undecane moiety to increase the rigidity of the compounds through conformational restriction. The N-methylindole fragment (CAP region) was then systematically modified to fully explore the structure-activity relationships (SARs) and further improve bioactivity (Figure 1). Finally, a total of 38 derivatives with new structural frameworks were obtained.

In vitro 72 h erythrocyte-stage parasitekilling assays
Highly synchronized ring-stage parasites (1% parasitemia and 2% hematocrit) were cultured in 96-well plates with a total volume of 200 μL of compound. The compounds were serially diluted from an initial concentration of 200 nM. After 72 hours of incubation, 100 μL of dissolution buffer (0.12 mg/mL saponin, 0.12% Triton X-100, 30 mM Tris-Cl and 7.5 mM EDTA) was added to dissolve the parasites. Each well was then stained with SYBR Green I (Invitrogen; supplied at 10,000× dilution) and incubated in the dark for 2 hours [30]. Under 485 nm excitation and 535 nm emission conditions, the fluorescence signal representing parasite DNA was monitored with a microplate reader (Bioteck, Synergy H1). The IC 50 was calculated in GraphPad 7. The results are shown as the mean ± SD from two independent experiments.

In vitro cytotoxicity assays
HepG2 and 293T cells were cultured in DMEM (HyClone) supplemented with 10% FBS (Gibco) and 1% penicillin/streptomycin (YEASEN) at 37°C under 5% CO 2 . Cells (70,000 cells/mL) were seeded in 96-well plates in a total volume of 100 μL and were cultured for 24 h. Tested compounds were added to the plate (two-fold dilutions ranging from 20 μM to 0.049 μM), and cell culture continued for 72 h. Then 10% Cell Counting Kit 8 (YEASEN) reagent was added to each well for the analysis of cell viability. The absorbance was monitored after 2 h at 450 nm with a microplate reader (Bioteck, Synergy H1). The IC 50 was calculated in GraphPad 7. The results are shown as the mean ± SD from two independent experiments.

Mouse liver microsome metabolism assays
This experiment was performed by Shanghai ChemPartner Co., Ltd. Liver microsomes (0.5 mg/mL) were purchased from Corning Corporation. Ketanserin was used as a positive control. First, the tested compounds were dissolved in DMSO (10 mM) and then diluted to 0.5 mM with acetonitrile. The compounds were then adjusted to a working concentration of 1.5 μM in liver microsomal buffer; 30 μL was mixed with 15 μL of 6 mM NADPH at 37°C. At 0, 5, 15, 30 and 45 minutes after incubation, 135 μL of acetonitrile was added to stop the reaction. The mixture was incubated on a shaker (IKA, MTS 2/4) at 600 rpm/min for 10 min, then centrifuged at 5594×g for 15 min (Thermo Multifuge×3R). The supernatant was diluted 1:1 with distilled water and analyzed by LC-MS/MS.

In vivo erythrocytic antimalarial assays
All animal experimental procedures followed the National Institutes of Health Guide for the Care and Use of Laboratory Animals under the supervision of the Animal Welfare Committee of Institute Pasteur of Shanghai, Chinese Academy of Sciences (IACUC issue No. A2018009). Each group contained five female BALB/c mice (6-8 weeks of age). At day 0, each mouse was i.p. inoculated with 10 5 P. yoelii parasites [31]. After 24 h, the solution of compounds (DMSO: 20 wt% aqueous 2-hydroxypropyl-β-cyclodextrin = 5:95 v/v) or solvent was delivered p.o. at the indicated dose once daily for five successive days. Piperaquine phosphate (PPQ) was used as a positive control and was delivered i.p. Parasitemia was counted from at least 5,000 red blood cells by smearing at days 6, 30 and 60. The graph was generated in GraphPad 7.

Western blot assays
First, the red blood cells in parasite samples were removed with 0.15% saponin, and the parasites were sonicated in 1% NP-40 lysis buffer. After centrifugation at 12,000×g for 15 min, the supernatant was resuspended in regular SDS loading buffer. Parasite proteins were separated with 10% SDS-PAGE and transferred to a PVDF membrane (Millipore). Antibodies to acetylated histone H3 (H3; Millipore; 06-599) or H3 (ABclonal; A2348) were used for immunodetection; the secondary antibody was conjugated to HRP (Jackson ImmunoResearch). Signal detection was performed with a chemiluminescent HRP substrate Immobilon-Western kit (Millipore).

Human HDAC inhibition assays
This experiment was performed by Shanghai ChemPartner Co., Ltd. Human HDACs 1-3, 6 and 8, and SIRT 2 were purchased from BPS. SAHA and suramin were used as positive controls. The tested compounds

Acta Materia Medica
were threefold diluted from the initial concentration of 10 μM and then were incubated with 15 μL enzyme/Tris buffer at room temperature for 15 min. Subsequently, 10 μL trypsin and Ac-peptide substrate/Tris buffer were added to start the reaction, and HDACs 1-3, 6 were incubated at room temperature. To start the reaction, 10 μL Ac-peptide substrate/Tris buffer was added and incubated at room temperature for 4 h. Subsequently, HDAC 8 and SIRT 2 were incubated with trypsin solution for 2 h. Enzymatic activity was measured with a Synergy MX instrument with excitation at 355 nm and emission at 460 nm. The IC 50 was calculated in GraphPad 7.

Pharmacokinetic assays
This experiment was performed by Shanghai ChemPartner Co., Ltd. Female BALB/c mice (6-8 weeks of age) were bred and acclimated for 1 week. The compounds were dissolved in 5% DMSO and diluted to 0.5 mg/ml with 20% aqueous 2-hydroxypropyl-β-cyclodextrin solution (stock concentration). Each mouse was intraperitoneally injected with a compound solution at a dose of 5 mg/kg. Blood samples were collected into tubes containing anticoagulant (K 2 -EDTA) at 0.25, 0.5, 1, 2, 4, 8 and 24 h after injection and centrifuged at 2,000×g at 4°C for 5 min to obtain plasma. LC-MS/MS of samples was performed on an ACQUITY UPLC HSS T3 1.8 μm column. The mobile phase was a mixture of phase A (0.1% formic acid in water) and phase B (0.1% formic acid in acetonitrile), and gradient mode was used with a flow rate of 0.6 mL/min at 60°C. Mass spectra were acquired on an API6500 triple quadrupole instrument equipped with an ESI source. Propranolol was used as the internal standard. Plasma concentrations of compounds were analyzed, and the pharmacokinetic parameters were calculated via WinNonlin.

In vitro inhibition activity, cytotoxicity and SARs of compounds 01−38
The in vitro inhibition activity of compounds 01-38 was systematically investigated against wild-type P. falciparum 3D7 and multi-drug-resistant P. falciparum Dd2. Dihydroartemisinin (DHA) was chosen as a positive control. The compounds' cytotoxicity in HepG2 and 293T cells was tested in parallel. The results are summarized in Table 1. Although the in vitro antimalarial activity of most of the compounds, compared with quisinostat, was poorer, the cytotoxicity was clearly lower, particularly for several compounds with favorable selectivity indexes greater than 1000. These results suggested that the narrow therapeutic window of quisinostat can be improved through structural modification. Herein, we concluded the following SARs of compounds 01-38: (I) replacing 4-aminomethylpiperidine with an 1-oxa-4,9-diazaspiro [5.5]undecane moiety slightly decreased the antimalarial activity but markedly decreased cytotoxicity (quisinostat vs 01); (II) aliphatic and monocyclic aromatic and biphenyl CAP groups were not conducive to improved antimalarial activity (02-11 and 15 vs 01); (III) naphthyl, anthryl and other double-aromatic rings are tolerated in the CAP region, except 2-benzothiophenyl. (12-14 and 16-26 vs 01); (IV) treatment of azaindolyl group as CAP region or removal of N-methyl of 3-indolyl decreased the cytotoxicity and improved the selectivity index (27-33 vs 01); (V) introducing a substituent on the 3-indolyl group enhanced the potency against Dd2 and the selectivity index in HepG2 cells (34-38 vs 01). Thus, valuable SARs and structure-toxicity relationships were obtained through the systematic analysis of these 38 hydroxamic acid derivatives; among them, compounds 26, 33, 34 and 38 were the most potent compounds against 3D7 and Dd2, and exhibited better cellular safety than quisinostat; moreover, compound 30 showed moderate antimalarial activity but the best safety properties, with a selectivity index > 2316 in two cell lines.

In vitro metabolic stability of compounds 26, 30, 33, 34 and 38
Compounds 26, 30, 33, 34 and 38 were selected to investigate their stability in mouse liver microsomes in vitro. Table 2, all tested compounds were more stable than quisinostat. Among the compounds, the half-life (t 1/2 ) of 30 was longest (> 150-fold longer than that of quisinostat and > 10-fold longer than that of compound 33). These results suggested that the structure of the CAP region strongly influences metabolic stability.

In vitro inhibition activity of compounds 30 and 33 against clinical multi-drug-resistant parasites
Compounds 30 and 33, two of the most stable compounds in mouse liver microsomes, were further evaluated for their antimalarial activity against clinical multi-drug-resistant lines (GB4, C2A, CP286 and 6218) via 72 h parasite-killing assays in vitro. Their IC 50 values against the multi-drug-resistant lines ( Table 3) were similar to those of the drug-sensitive line 3D7 (Table 1), thus implying that compounds 30 and 33 avoided cross resistance, possibly because of the differences in antimalarial mechanisms between these compounds and existing antimalarial drugs.

Compounds 30 and 33 upregulate histone acetylation in P. falciparum parasites
The mechanism of action of compounds 30 and 33 was explored by western blotting to detect the acetylation level of plasmodium histones (Figure 2). Plasmodium was pre-incubated with quisinostat and compounds 30 and 33 at 20-fold the IC 50 (3D7) concentration for 4 h, and plasmodium protein was extracted to assess acetylation of H3. Compared with those in the DMSO control group,

In vitro inhibition activity of compounds 30 and 33 against human HDACs
In addition, we investigated the enzymatic inhibitory activity of compounds 30 and 33 against human HDACs. Among the tested enzymes, HDACs 1-3 and HDAC 8 are class I HDACs, HDAC 6 is a class II HDAC, and SIRT 2 is an NAD + -dependent class III HDAC [32]. As shown in Table 4, the activity of compounds 30 and 33 against class I/II HDACs was markedly lower than that of the lead compound quisinostat, thus potentially explaining the lower cytotoxicity of the derivatives in this work.
Notably, for compound 30, the IC 50 value against HDAC1 was nearly 60 times lower than that of quisinostat.

In vivo asexual erythrocyte-stage antimalarial activity of compound 30
According to the in vitro experimental data, compound 30, because of its low cytotoxicity and the best metabolic stability in liver microsomes, was selected as a candidate compound to evaluate its efficacy in a mouse model of P. yoelii infection (Figure 3). Herein, BALB/c mice were randomly divided into five groups, and the tested drugs were administered for 5 days after the day of inoculation with 1×10 5 parasites. PPQ was chosen as a positive control and administered intraperitoneally, whereas quisinostat and compound 30 were administered orally. As shown in Figure 3A, all mice in the vehicle group died by day 7, thus indicating successful infection with P. yoelii. One mouse in the quisinostat-treated group died on day 6, presumably because of cumulative toxicity caused by continual administration, whereas compound 30 showed a good safety profile, and no mice died. We collected tail-vein blood from mice on day 6 (the first day after the end of drug administration), day 30 and day 60 and prepared blood smears to quantify parasitemia ( Figure 3B). Although compound 30 did not completely kill the parasites on day 6, it resulted in   clearly lower parasitemia rates (3.13% in the 120 mg/kg group and 0.12% in the 150 mg/kg group) than that in the vehicle group (95.23%), with killing rates above 95%. Notably, no parasites were observed in mice on days 30 and 60, thus indicating that all surviving mice had been cured. Overall, compound 30 was shown to be safe and effective in this model.

In vivo pharmacokinetic study of compound 30
The pharmacokinetic (PK) properties of compound 30 administered via intraperitoneal injection were further investigated in BALB/c mice ( Table 5). The maximum concentration (C max ) and area under the curve (AUC) of 30 were much higher than those of quisinostat. Higher plasma concentrations of the compound may result in more potent parasite killing in vivo. In addition, compound 30 displayed a slightly longer half-life (t 1/2 ) than quisinostat, and it was the derivative with the longest t 1/2 in the entire series.

DISCUSSION
In this work, we designed and synthesized a series of novel hydroxamic acid derivatives (01-38) based on the antitumor drug quisinostat as an antimalarial lead compound. The optimal compound, 30, showed potent inhibitory activity against P. falciparum parasites (IC 50 = 8.64 ± 0.44 nM against 3D7) and low cytotoxicity (IC 50 > 20 μM against HepG2 and 293T). Further evaluation indicated that 30 exhibited strong killing efficacy against several multi-drug-resistant clinical P. falciparum strains and therefore may contribute to solving the treatment failures caused by drug resistance. Moreover, the in vivo efficacy of compound 30 indicated complete elimination of parasites in a mouse model of P. yoelii infection and showed a safety profile superior to that of the lead compound quisinostat. In addition, 30 exhibited good metabolic stability in liver microsomes and favorable PK properties in mice. Preliminary mechanistic research indicated that 30 killed P. falciparum parasites by inhibiting PfHDAC enzyme activity. Together, these results suggest that 30 is a prospective prototype drug for malaria therapy that warrants further optimization. (19YF1412600).

CONFLICTS OF INTEREST
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.  a PK parameters in plasma after a single i.p. injection of 5 mg/kg compound in mice; C max : maximum plasma or hepatic concentration; T max : time to C max ; AUC last : area under the concentration−time curve from 0 to the last sampling time at which a quantifiable concentration was found; AUC inf : area under the concentration−time curve from 0 to infinity; t 1/2 : apparent elimination half-life; CL z /F: apparent clearance; V z /F: apparent volume of distribution. b Data previously reported [26].