Structural Optimization of BIPPO Analogs as Potent Antimalarials

Malaria continues to pose a significant health threat, causing thousands of deaths each year. The limited availability of vaccines and medications, combined with the emergence of drug resistance, further complicates the fight against this disease. In this study, we aimed to enhance the antimalarial potency of the previously reported hit compound BIPPO (pIC50 5.9). Through systematic modification of pyrazolopyrimidinone analogs, we discovered the promising analog 30 (NPD-3547), which exhibited approximately one log unit higher in vitro potency (pIC50 6.8) against Plasmodium falciparum. Furthermore, we identified several other BIPPO analogs (23, 28, 29 and 47a) with potent antimalarial activity (pIC50 > 6.0) and favorable metabolic stability in mouse liver microsomes. These compounds can serve as new tools for further optimization towards the development of potential candidates for antimalarial studies.


Introduction
Malaria is a mosquito-transmitted parasitic disease caused by Plasmodium spp. Although the world has witnessed a decrease in reported malaria cases, the current situation is still worrying [1]. In 2019, there was an estimated number of 229 million infections with 409,000 deaths globally [2]. With a substantial amount (20%) of malaria research funding invested in vaccine discovery annually, there is so far only one vaccine (RTS,S), which was approved in 2015 [3,4]. However, due to its low efficacy, the WHO does not recommend its routine use in infants (6-12 weeks), who suffer greatly from malaria [5]. Another emerging malaria vaccine is R21, which is still under assessment for its safety and effectiveness at the WHO [6]. In humans, malaria is caused by five different species of Plasmodium [7]. Among them, P. falciparum and P. vivax are responsible for most infections; other cases are caused by P. ovale, P. malariae and P. knowlesi. Of all five different species, infections caused by P. falciparum lead to the highest number of deaths. Therefore, the available treatment mainly focuses on this species.
The current drug treatment of malaria relies on the combination therapy (ACT) of artemisinin with its analogs (dihydroartemisinin, artesunate or artemether) and a different class of antimalarial drug (e.g., amodiaquine or mefloquine) [8]. Due to the unique life cycle of malaria, recrudescence, relapse or reinfection may occur after some symptom-free periods [9]. Moreover, drug resistance has already become a problem in some Southwest Asian countries, although the WHO recommends ACT to delay the rise of resistance [10,11]. Above all, malaria causes a heavy life threat and economic burden in epidemic areas. Therefore, it is of high urgency to develop novel effective antimalarial treatments.

Design and Synthesis of BIPPO Analogs
Previously, structural modifications focused mainly on substituents on the benzyl group of BIPPO and its phenyl analogs [12]. To explore the structure-activity relationship of this scaffold in more detail, first a few structurally close BIPPO analogs with various R 1 substituents were synthesized with the route [13] shown in Scheme 1. The first step started from a commercially available 4-aminopyrazole 1 with a condensation reaction to form the amide intermediate 2, after which a ring closure reaction under basic conditions yields the desired products 3-32. Interestingly, during the synthesis of the R 1 analogs, instead of the originally designed 4-cyclohexanone analog, two alcohol diastereomers, 26 and 27, were obtained after the ring closure reaction, probably because the carbonyl group was reduced under basic conditions in the microwave reaction; reactions with similar conditions were reported previously [14,15]. The structures of these two diastereomers could be confirmed with selective proton decoupling 1 H NMR. Next, to explore the chemical space of R 2 and R 3 positions, a methyl group was introduced. Previously, Howard et al. reported the synthesis of R 2 methylated BIPPO analog (37a) from BIPPO with dimethyl sulfate (DMS). However, the analog with a methyl group at the R 3 position (37b) was not reported. Here, we report the synthesis of these two analogs with a different synthetic route (Scheme 2) and confirm their regiochemistry. Starting from the carboxylic acid intermediate 33 [13], the pyrazole methylation in the first step leads to the two regio-isomers 34a and 34b, which were separated by column chromatography and structurally identified with a 1D-NOESY NMR method. The subsequent amidation and reduction reactions resulted in intermediates 36a and 36b in high yields (92% and 96%), without purification of the intermediates 35a and 35b. Amide coupling and ring closure reactions under basic conditions, as shown in Scheme 1, yielded analogs with a methyl group at R 2 (37a, 38a) and R 3 (37b, 38b) positions in moderate to good yields (37-70%).
reduced under basic conditions in the microwave reaction; reactions with simi tions were reported previously [14,15]. The structures of these two diastereom be confirmed with selective proton decoupling 1 H NMR. Next, to explore the space of R 2 and R 3 positions, a methyl group was introduced. Previously, How reported the synthesis of R 2 methylated BIPPO analog (37a) from BIPPO with sulfate (DMS). However, the analog with a methyl group at the R 3 position (37b reported. Here, we report the synthesis of these two analogs with a different route (Scheme 2) and confirm their regiochemistry. Starting from the carboxyl termediate 33 [13], the pyrazole methylation in the first step leads to the two regi 34a and 34b, which were separated by column chromatography and structura fied with a 1D-NOESY NMR method. The subsequent amidation and reduction resulted in intermediates 36a and 36b in high yields (92% and 96%), without pu of the intermediates 35a and 35b. Amide coupling and ring closure reactions un conditions, as shown in Scheme 1, yielded analogs with a methyl group at R 2 and R 3 (37b, 38b) positions in moderate to good yields (37-70%).
The last modification focused on the R 4 position, where a tert-butyl group clopentyl group were introduced instead of the isopropyl group in 3. Thes (46a/b, 47a/b) could be obtained with the previously reported route for 3 (Schem During the chemical characterization of the compounds, it was observed carbon signals from the non-N-substituted pyrazoles were not visible in 13 C NM By using 13 C NMR combined with 2D NMR (HSQC and HMBC), the compou unambiguously characterized [12,16]. substituents were synthesized with the route [13] shown in Scheme 1. The first step started from a commercially available 4-aminopyrazole 1 with a condensation reaction to form the amide intermediate 2, after which a ring closure reaction under basic conditions yields the desired products 3-32. Interestingly, during the synthesis of the R 1 analogs, instead of the originally designed 4-cyclohexanone analog, two alcohol diastereomers, 26 and 27, were obtained after the ring closure reaction, probably because the carbonyl group was reduced under basic conditions in the microwave reaction; reactions with similar conditions were reported previously [14,15]. The structures of these two diastereomers could be confirmed with selective proton decoupling 1 H NMR. Next, to explore the chemical space of R 2 and R 3 positions, a methyl group was introduced. Previously, Howard et al. reported the synthesis of R 2 methylated BIPPO analog (37a) from BIPPO with dimethyl sulfate (DMS). However, the analog with a methyl group at the R 3 position (37b) was not reported. Here, we report the synthesis of these two analogs with a different synthetic route (Scheme 2) and confirm their regiochemistry. Starting from the carboxylic acid intermediate 33 [13], the pyrazole methylation in the first step leads to the two regio-isomers 34a and 34b, which were separated by column chromatography and structurally identified with a 1D-NOESY NMR method. The subsequent amidation and reduction reactions resulted in intermediates 36a and 36b in high yields (92% and 96%), without purification of the intermediates 35a and 35b. Amide coupling and ring closure reactions under basic conditions, as shown in Scheme 1, yielded analogs with a methyl group at R 2 (37a, 38a) and R 3 (37b, 38b) positions in moderate to good yields (37-70%).
The last modification focused on the R 4 position, where a tert-butyl group and a cyclopentyl group were introduced instead of the isopropyl group in 3. These analogs (46a/b, 47a/b) could be obtained with the previously reported route for 3 (Scheme 3) [13].
During the chemical characterization of the compounds, it was observed that some carbon signals from the non-N-substituted pyrazoles were not visible in 13 C NMR spectra. By using 13 C NMR combined with 2D NMR (HSQC and HMBC), the compounds were unambiguously characterized [12,16]. The last modification focused on the R 4 position, where a tert-butyl group and a cyclopentyl group were introduced instead of the isopropyl group in 3. These analogs (46a/b, 47a/b) could be obtained with the previously reported route for 3 (Scheme 3) [13].

Antimalarial Activities of BIPPO Analogs
Previously, Howard et al. reported (sub)micromolar activity against P. falciparum for a series of BIPPO analogs with benzyl substituents [12]. To further understand the structure-activity relationship (SAR) of this interesting scaffold, close analogs of 3 with substituents at R 1 -R 4 positions were tested against P. falciparum and human MRC-5 cells as the control for non-specific toxicity (Table 1).
To improve the chemical diversity and solubility of 3, instead of the benzyl group, a pyridylmethyl group (4) was introduced at the R 1 position, which led to a sixfold decreased antimalarial potency compared to 3. Analogs 5, 6 and 12-17 were designed to understand the influence of linker flexibility, linker length and chemical space around the linker. Except for 13, 14 and 17 with an equal potency compared to 3, the antimalaria potencies of other analogs (5, 6, 12 and 15-16) were 5-to 25-fold lower than 3. All analogs with aromatic substituents (6-11) directly attached at the R 1 position exhibited significantly decreased activity compared with 3.
For the R 2 , R 3 and R 4 analogs, decreased antimalarial activities were observed for the N-methyl analogs 37a (pIC50 of 5.0) and 37b (pIC50 of 5.3) compared with 3. At the R 4 position, the isopropyl substituent of 3 was replaced with a tert-butyl (46a) or cyclopentyl groups (46b). Since both analogs show similar antimalarial potencies compared with 3, no further modifications were made at the R 4 position, as the larger substituents also led to a decreased aqueous solubility (e.g., cLogS of -4.0 for 46b compared to -2.9 for 3).
Following our initial screening of close analogs of 3, the large activity differences against P. falciparum following variation at the R 1 position (especially between 3 and 6, benzyl group versus phenyl group) and equal potency of 13 (α-methyl) suggested R 1 as a promising position for follow-up modifications. To further explore the R 1 position of this scaffold, a series of BIPPO analogs with various substituents at the R 1 position was syn- During the chemical characterization of the compounds, it was observed that some carbon signals from the non-N-substituted pyrazoles were not visible in 13 C NMR spectra. By using 13 C NMR combined with 2D NMR (HSQC and HMBC), the compounds were unambiguously characterized [12,16].

Antimalarial Activities of BIPPO Analogs
Previously, Howard et al. reported (sub)micromolar activity against P. falciparum for a series of BIPPO analogs with benzyl substituents [12]. To further understand the structure-activity relationship (SAR) of this interesting scaffold, close analogs of 3 with substituents at R 1 -R 4 positions were tested against P. falciparum and human MRC-5 cells as the control for non-specific toxicity (Table 1).
To improve the chemical diversity and solubility of 3, instead of the benzyl group, a pyridylmethyl group (4) was introduced at the R 1 position, which led to a sixfold decreased antimalarial potency compared to 3. Analogs 5, 6 and 12-17 were designed to understand the influence of linker flexibility, linker length and chemical space around the linker. Except for 13, 14 and 17 with an equal potency compared to 3, the antimalaria potencies of other analogs (5, 6, 12 and 15-16) were 5-to 25-fold lower than 3. All analogs with aromatic substituents (6-11) directly attached at the R 1 position exhibited significantly decreased activity compared with 3.
For the R 2 , R 3 and R 4 analogs, decreased antimalarial activities were observed for the N-methyl analogs 37a (pIC 50 of 5.0) and 37b (pIC 50 of 5.3) compared with 3. At the R 4 position, the isopropyl substituent of 3 was replaced with a tert-butyl (46a) or cyclopentyl groups (46b). Since both analogs show similar antimalarial potencies compared with 3, no further modifications were made at the R 4 position, as the larger substituents also led to a decreased aqueous solubility (e.g., cLogS of -4.0 for 46b compared to -2.9 for 3).
Following our initial screening of close analogs of 3, the large activity differences against P. falciparum following variation at the R 1 position (especially between 3 and 6, benzyl group versus phenyl group) and equal potency of 13 (α-methyl) suggested R 1 as a promising position for follow-up modifications. To further explore the R 1 position of this scaffold, a series of BIPPO analogs with various substituents at the R 1 position was synthesized and tested against P. falciparum and MCR-5 cells. From a series of analogs without a substituent at R 1 or with alkyl R 1 substituents with increasing sizes (18)(19)(20)(21)(22)(23), it appeared that the antimalarial potency increased with the size of the R 1 substituent. Analogs 19-21 with relatively small acyclic aliphatic substituents and 18 show lower or equal potency compared with 3, while 22, 23 and 28-30 with bulkier substituents exhibit improved activity (Table 2). Notably, the introduction of a cyclohexyl group (23) and an adamantanyl group (30) leads to a five-and eightfold potency increase, respectively, compared with 3. To improve the solubility of 23, three heterocyclic substituted analogs (24, 25, 32) and two analogs (26, 27) with a hydroxyl group were synthesized. Unfortunately, these modifications all resulted in less active analogs compared to 23 ( Table 2), indicating that heteroatoms and polar groups are not tolerated at this position. activity (Table 2). Notably, the introduction of a cyclohexyl group (23) and an adamantanyl group (30) leads to a five-and eightfold potency increase, respectively, compared with 3. To improve the solubility of 23, three heterocyclic substituted analogs (24, 25, 32) and two analogs (26, 27) with a hydroxyl group were synthesized. Unfortunately, these modifications all resulted in less active analogs compared to 23 ( Table 2), indicating that heteroatoms and polar groups are not tolerated at this position. From all our efforts to improve the potency of 3 at the R 1 position, 30 with an adamantanyl group turned out to be the most potent compound (pIC50 6.8) against P. falciparum without showing noticeable toxicity for human MCR-5 cells. Taking 30 as a starting point, further modifications focused on R 2-4 positions based on the previous synthetic routes (Schemes 1 and 2). Unfortunately, no potency improvement was achieved within this series (Table 3). Biological analysis of the R 2-4 analogs of 30 resulted in a similar SAR as for the related analogs of 3 ( Table 3). The introduction of a methyl group at the R 2 position (38a) led to a more than 400-fold potency decrease. Analog 38b with a methyl group at the R 3 position was more than 15-fold less active compared with 30. The introduction of a tert-butyl group (47a) and a cyclopentyl group (47b) at the R 4 position led to a 4-and 2.5-fold potency decrease, respectively. From all our efforts to improve the potency of 3 at the R 1 position, 30 with an adamantanyl group turned out to be the most potent compound (pIC 50 6.8) against P. falciparum without showing noticeable toxicity for human MCR-5 cells. Taking 30 as a starting point, further modifications focused on R 2-4 positions based on the previous synthetic routes (Schemes 1 and 2). Unfortunately, no potency improvement was achieved within this series (Table 3). Biological analysis of the R 2-4 analogs of 30 resulted in a similar SAR as for the related analogs of 3 ( Table 3). The introduction of a methyl group at the R 2 position (38a) led to a more than 400-fold potency decrease. Analog 38b with a methyl group at the R 3 position was more than 15-fold less active compared with 30. The introduction of a tert-butyl group (47a) and a cyclopentyl group (47b) at the R 4 position led to a 4-and 2.5-fold potency decrease, respectively.

Metabolic Stability Test
The modifications of 3 resulted ultimately in an increase in the pIC50 of 3 from 5.9 to 6.8 by introducing an adamantanyl group (30) at the R 1 position. Thus, compounds 23, 28-30, 47a and 47b with high potencies (pIC50 > 6) were evaluated for their in vitro metabolic stability in human and mouse liver microsomes (S9 fraction) with diclofenac as a reference compound. As summarized in Figure 2, substituents in the R 4 position affect metabolic stability. Analogs 23, 28 and 29 with an isopropyl group at the R 4 position showed similar metabolic stability. They exhibited good stability with human liver microsomes, with more than 50% of the parent compounds remaining after one hour incubation in conditions with Phase I and Phase II metabolism. Their stability against mouse liver microsomes is lower, especially for their Phase I metabolism; only 38% and 26% of the parent compounds are left for 23 and 29 after one hour of incubation. Analog 28 with a difluorocyclohexyl group at the R 1 position exhibited slightly improved metabolic stability with mouse liver microsomes; 68% of the parent compound was observed after one hour of incubation. For the adamantanyl analogs 30, 47a and 47b, metabolic stability differs drastically, with Phase I identified as the main metabolic pathway. Analog 47b was metabolized for >95% within 30 min with mouse liver microsomes, while 90% of 47a (tert-butyl analog) was left after one hour of incubation. In general, all compounds showed better metabolic stability than diclofenac in the mouse Phase II and both human Phase I/II assays. The analogs 23, 29, 30 and 47b were metabolized faster than diclofenac in the mouse Phase I assay.

Metabolic Stability Test
The modifications of 3 resulted ultimately in an increase in the pIC 50 of 3 from 5.9 to 6.8 by introducing an adamantanyl group (30) at the R 1 position. Thus, compounds 23, 28-30, 47a and 47b with high potencies (pIC 50 > 6) were evaluated for their in vitro metabolic stability in human and mouse liver microsomes (S9 fraction) with diclofenac as a reference compound. As summarized in Figure 2, substituents in the R 4 position affect metabolic stability. Analogs 23, 28 and 29 with an isopropyl group at the R 4 position showed similar metabolic stability. They exhibited good stability with human liver microsomes, with more than 50% of the parent compounds remaining after one hour incubation in conditions with Phase I and Phase II metabolism. Their stability against mouse liver microsomes is lower, especially for their Phase I metabolism; only 38% and 26% of the parent compounds are left for 23 and 29 after one hour of incubation. Analog 28 with a difluoro-cyclohexyl group at the R 1 position exhibited slightly improved metabolic stability with mouse liver microsomes; 68% of the parent compound was observed after one hour of incubation. For the adamantanyl analogs 30, 47a and 47b, metabolic stability differs drastically, with Phase I identified as the main metabolic pathway. Analog 47b was metabolized for >95% within 30 min with mouse liver microsomes, while 90% of 47a (tert-butyl analog) was left after one hour of incubation. In general, all compounds showed better metabolic stability than diclofenac in the mouse Phase II and both human Phase I/II assays. The analogs 23, 29, 30 and 47b were metabolized faster than diclofenac in the mouse Phase I assay.

Conclusions
Based on its potency as an antimalarial and its drug-like properties, we took BIPPO (3) as a starting point for a hit optimization program. Systematic modification identified the R 1 position in the structure of BIPPO as a key position to improve antimalarial potency. The introduction of aliphatic substituents at this position yielded the adamantanyl analog  Table S1.

Conclusions
Based on its potency as an antimalarial and its drug-like properties, we took BIPPO (3) as a starting point for a hit optimization program. Systematic modification identified the R 1 position in the structure of BIPPO as a key position to improve antimalarial potency. The introduction of aliphatic substituents at this position yielded the adamantanyl analog 30, which is around one log unit more potent than the parent compound BIPPO against asexual blood-stage Plasmodium. The metabolic stability assay indicates that BIPPO analogs 23, 28, 29 and 47a have a sufficient stability profile for in vivo studies. In summary, this systematic modification of BIPPO yields a series of analogs with high antimalarial potency against the blood-stage form of P. falciparum. Together with their good drug-like properties and in vitro metabolic stability, they can serve as tool compounds for further hit-to-lead optimization towards candidates for advanced antimalarial studies.

Chemistry
All starting materials were obtained from commercial suppliers and used without purification. Synthesis of 1, 3, 6, 18, 19, 36b, 37a and 43b was reported previously [12,13,[16][17][18][19][20]. Anhydrous THF, DCM and DMF were obtained by passing through an activated alumina column prior to use. All reactions were carried out under a nitrogen atmosphere unless mentioned otherwise. TLC analyses were performed using Merck F 254 aluminum-backed silica plates and visualized with 254 nm UV light. Flash column chromatography was executed using Biotage Isolera equipment. All HRMS spectra were recorded on a Bruker microTOF mass spectrometer using ESI in positive-ion mode. All NMR spectra were recorded on either a Bruker Avance 300, 500 or 600 spectrometer. The peak multiplicities are defined as follows: s, singlet; d, doublet; t, triplet; q, quartet; p, pentet; dd, doublet of doublets; dt, doublet of triplets; td, triplet of doublets; br, broad; m, multiplet; and app, apparent. The spectra were referenced to the internal solvent peak as follows: CDCl 3 (δ = 7.26 ppm in 1 H NMR, δ = 77.16 ppm in 13  The eluent program used is as follows: flow rate: 1.0 mL/min, start with 95% A in a linear gradient to 10% A over 4.5 min, hold 1.5 min at 10% A, in 0.5 min in a linear gradient to 95% A, hold 1.5 min at 95% A, total run time: 8.0 min. Compound purities were calculated as the percentage peak area of the analyzed compound by UV detection at 254 nm. Note: not all 13 C signals are visible in spectrum due to tautomerism of non-N-substituted pyrazoles; 2D NMR (HSQC and HMBC) spectra were measured to assign 13 C signals if applicable.
The general method for the synthesis of final compounds: An amine (1.0 eq) and the corresponding acid (1.0 eq), PyBrop (1.1 eq) and TEA (2.0 eq) were combined in DCE and heated using microwave irradiation at 120 • C for 20 min. The reaction mixture was purified using column chromatography to obtain the amide intermediates. Then, the amide intermediate was combined with KO t Bu (2.0 eq) in i PrOH and heated using microwave irradiation at 130 • C for 30 min. The reaction mixture was concentrated in vacuo and purified using column chromatography to obtain the final product.

Antimalarial Screening
The assay for antimalarial activity was carried out as described in detail in Pereira et al. [21].

Metabolic Stability
The assay for metabolic stability in human and mouse liver microsomal fractions (S9) was performed as described [22].