Structure–Activity Relationship Studies of Pyrrolone Antimalarial Agents

Previously reported pyrrolones, such as TDR32570, exhibited potential as antimalarial agents; however, while these compounds have potent antimalarial activity, they suffer from poor aqueous solubility and metabolic instability. Here, further structure–activity relationship studies are described that aimed to solve the developability issues associated with this series of compounds. In particular, further modifications to the lead pyrrolone, involving replacement of a phenyl ring with a piperidine and removal of a potentially metabolically labile ester by a scaffold hop, gave rise to derivatives with improved in vitro antimalarial activities against Plasmodium falciparum K1, a chloroquine-and pyrimethamine-resistant parasite strain, with some derivatives exhibiting good selectivity for parasite over mammalian (L6) cells. Three representative compounds were selected for evaluation in a rodent model of malaria infection, and the best compound showed improved ability to decrease parasitaemia and a slight increase in survival.


Introduction
Malaria is a serious disease endemic in tropical and subtropical regions of the world. It is a major threat to public health in more than 100 countries and is responsible for > 200 million clinical cases each year and probably about 1 million deaths, the majority of whom are young children and pregnant women. [1][2][3][4] In addition, malaria is responsible for a huge economic impact in endemic countries. [2] With increasing globalisation and global warming, there is a risk of malaria spreading to new areas. [3] No vaccine is currently available for malaria, and the resistance of the protozoa to clinically used chemotherapeutic agents is increasingly common. Therefore, an urgent need exists to develop new classes of antimalarial drugs that operate by novel mechanisms of action.
We have previously reported on the discovery and structure-activity relationship (SAR) of a series of pyrrolones, which have potent antimalarial activity; [5] these were initially discovered through a phenotypic screening process conducted by the World Health Organisation (WHO) Programme for Research and Training in Tropical Diseases. The prototypical compound TDR32750 (Figure 1 a) showed potent antimalarial activity against a panel of strains of Plasmodium falciparum and significant depression of parasitaemia (> 99 %) in the Plasmodium berghei mouse model of malaria when dosed intraperitoneally. Unfortunately, the compound has low oral bioavailability; good oral bioavailability is a requirement for most target product profiles for malaria. While the precise reason for the low bioavailability is unknown, it is most likely due to a combination of metabolic instability (as evidenced by studies in hepatic microsomes) and low aqueous solubility. [5] The ester linkage is also susceptible to hydrolysis in blood, which contributes to the overall systemic clearance, as previously reported. Our previous SAR studies around TDR32750 [5] encompassed changes to the A-, B-and C-rings (Figure 1 a), which led to the conclusion that the A-ring was reasonably tolerant of changes, unlike the B-ring. Modifications to the C-ring were not investigated apart from replacement of the ester with a variety of amides, which led to compounds with decreased activity.
In order to improve the physicochemical properties of the molecules and the prospects for further development of the compound series, our aim was to decrease the clog D value to less than three. Here, we report on changes to the A-ring and C-ring, which have led to compounds with enhanced antimalarial activity and improved physicochemical properties. In particular, our work focused around some piperidine analogues of Previously reported pyrrolones, such as TDR32570, exhibited potential as antimalarial agents; however, while these compounds have potent antimalarial activity, they suffer from poor aqueous solubility and metabolic instability. Here, further structure-activity relationship studies are described that aimed to solve the developability issues associated with this series of compounds. In particular, further modifications to the lead pyrrolone, involving replacement of a phenyl ring with a piperidine and removal of a potentially metabolically labile ester by a scaf-fold hop, gave rise to derivatives with improved in vitro antimalarial activities against Plasmodium falciparum K1, a chloroquine-and pyrimethamine-resistant parasite strain, with some derivatives exhibiting good selectivity for parasite over mammalian (L6) cells. Three representative compounds were selected for evaluation in a rodent model of malaria infection, and the best compound showed improved ability to decrease parasitaemia and a slight increase in survival.
the A-ring in our lead molecule (Figure 1 b). The piperidine moiety has an amino group, which when unconjugated is basic and should improve the solubility properties.
The lead compound, TDR32750, contains an ester in the Cring; we were concerned that this might be a metabolic liability due to hydrolysis by esterases, since previous work has confirmed that there is gradual degradation of the ester in vivo. [5] However, this is probably not the only mode of metabolism, as we have also shown that the compound undergoes cytochrome P450 (CYP450)-mediated degradation in hepatic microsomes. [5] Nevertheless, we thought it prudent to try and replace the ester to improve the clearance properties. Previously, we reported efforts at replacing the ester with an amide, but this led to loss of activity; possibly due to the amide causing a conformational change compared to the ester. [5] In this paper, we report the effect of cyclising the ester to form a fused pyridone, which might lock the compound into a more favourable conformation (Figure 1 b).

In vitro activity
Compounds were evaluated for activity against P. falciparum K1, a chloroquine-and pyrimethamine-resistant parasite strain, and counter-screened against mammalian L6 cells ( Table 1). Many of the compounds showed potent antimalarial activity, with eight compounds showing EC 50 values of less than 10 nm Figure 1. a) Prototypical compound TDR32750, structure and in vitro activities. [5] b) Generic structure of compounds prepared in this study.  Table 1. and two compounds (29 and 44) showing sub-nanomolar potencies. Most of the compounds showed good selectivity for parasite over mammalian cells.

A-Ring modifications
The following observations were made regarding the in vitro antimalarial activity for compounds with variations around the A-ring: * In general, simple alkyl substituents (18-23) led to compounds that exhibit a large decrease in activity compared with TDR32750. This also applied to a methylene cyclopropyl derivative (39). Interestingly, methylene cyclohexyl derivative 40 showed only a slight drop in activity compared with TDR32750, although adding heteroatoms to the ring to improve the physicochemical properties of the compounds was counter-productive for activity (41 and 42). Similarly, cyclohexyl analogues linked via an ethylene linker (37 and 38) showed decreased activity. * Where the substituent was an aromatic ring directly attached to the piperidine ring (30), the compound retained reasonably potent activity (EC 50 = 0.024 mm); however, this compound would have decreased basicity at the piperidine ring. * Compounds with a phenyl ring separated from the piperidine by a methylene linker in general showed good potency (17, 26-29, 35, 36, 43, 44, 51). * Replacement of the phenyl ring by a heteroaromatic ring (to improve physicochemical properties) was generally detrimental to activity (24, 25, 45-47, 49, 52, 53). The marked exception to this was compound 43, a pyridine, albeit with an electron-withdrawing trifluoromethyl group, which is likely to significantly decrease the basicity of the pyridine nitrogen. The oxazole (48) and N-methylpyrrole (50) derivatives also retained activity. * Increasing the spacer length between the piperidine and the aromatic ring also led to a drop in activity (31-34).
* The phenyl ring was tolerant to different substitutions, although the morpholine analogue (54) had greatly decreased activity.

C-Ring modifications
Modifications to the C-ring showed interesting activities in vitro against P. falciparum K1 ( Table 2). The most active compounds displayed sub-micromolar activities; compounds 56, 57, 58 and 63 were found to be the most active. The following observations were made regarding the SAR of the C-ring-modified analogues: * Compound 56 with a free NH group on the pyrrolo[3,2-c]pyridine ring showed an EC 50 value of 0.047 mm. Small alkyl substituents on the "pyridone" nitrogen were tolerated (57, R = Me: EC 50 = 14 nm; 58, R= Et: EC 50 = 21 nm), whilst a propyl substituent (59) led to a tenfold drop in activity. * Basic substituents did not appear to be tolerated, with compounds (60 and 61) exhibiting approximately a fivefold decrease in EC 50 value relative to the unsubstituted analogue 56. * The para-trifluoromethylbenzyl (63) and para-pyridyl (64) analogues appeared to substantially retain potency compared with TDR32750, although the straight benzyl derivative (62) showed around a fivefold loss in activity.

In vitro drug metabolism and pharmacokinetics (DMPK)
For key compounds, metabolic stability when incubated with mouse liver microsomes was measured, and for some compounds solubility was also assessed. For the A-ring variants, the solubility increased, particularly at low pH, probably due to the basic nature of the piperidine ring. In those derivatives with an aromatic or benzyl substituent, the basicity of the amine is decreased, which probably explains the lower solubility at pH 6.5. Disappointingly, the in vitro intrinsic clearance (CL int ) values were all rather high-the preferred value is < 50 mL min À1 mg À1 protein, although the aim is for a value of less than 20 mL min À1 mg À1 . The most stable compounds were 17, 20, 31, 35, 40, 48, and 50, with derivatives 20 and 31 being relatively inactive. Compound 48 is particularly stable, presumably due to the relatively low clog D value and the metabolic stability of the isoxazole ring.
For C-ring variants, the most active compounds (56-58) were investigated for microsomal stability. Compound 56 showed moderate stability, whilst the N-alkyl congeners were less stable, possibly due to CYP450-mediated metabolism of the alkyl groups.

In vivo studies
Three compounds were selected for further study in vivo. Compound 43 was selected from the A-ring-modified series, as  it represents a compound with good potency against the malaria parasite and very good selectivity compared with mammalian cells, albeit with a slightly increased microsomal instability. Compounds 56 and 57 were chosen as C-ring-modified derivatives, and also for their potency, selectivity and clog D value. Compounds 57 and 58 have very similar properties; compound 57 was chosen as it has marginally better properties. Multidose oral efficacy studies of compounds 43, 56, and 57 against green fluorescent protein (GFP)-transfected P. berghei ANKA-infected mice were conducted according to the standard protocol (Peters), with test compounds administered either per oral (po) or via intraperitoneal (ip) injection (Table 3).
Both via po and ip administration, compound 57 was the most efficacious in terms of decrease of parasitaemia. At a dose of 50 mg kg À1 ip, compound 57 decreased the level of parasitaemia by 99.9 %, and there was a significant increase in survival time. When dosed orally, the decrease in parasitaemia was less at 50.5 %. Compound 57 shows limited microsomal stability, so the decrease in oral activity of the compound is likely to be due to poorer exposure as a consequence of first-pass metabolism, but further work is required to confirm this hypothesis. Compound 43 failed to show significant activity in vivo, despite showing good potency in vitro. This might be due to poor pharmacokinetics, but this needs to be further investigated.

Conclusions
Modifications to our prototypical lead compound TDR32750 have been shown to yield compounds that retain in vitro activity. These changes include removal of the ester functionality that might be a point of metabolism and addition of a basic centre that should be helpful in increasing the solubility. Although these changes have not yet provided compounds with good oral in vivo activity, the results do suggest that there is further scope for the optimisation of the metabolic and physicochemical properties of this series, which could potentially lead to orally active antimalarial compounds.

Experimental Section
Profiling software: StarDrop version 5.3 with the P450 metabolism plug-in module (http://www.optibrium.com) was used to predict the sites of metabolism of the compounds. The same software was used to calculate the cLog D values for all compounds.
Chemistry: Chemicals and solvents were purchased from Sigma-Aldrich or Fluka and were used as received unless otherwise stated. Air-and moisture-sensitive reactions were carried out under an inert atmosphere of argon in oven-dried glassware. Analytical thin-layer chromatography (TLC) was performed on precoated TLC plates (layer 0.20 mm silica gel 60 with fluorescent indicator UV254; Merck). Developed plates were air-dried and analysed under a UV lamp (254/365 nm). Flash column chromatography was performed using prepacked silica gel cartridges (230-400 mesh, 40-63 mm; SiliCycle) using a Teledyne ISCO Combiflash Companion or Combiflash Retrieve. Microwave irradiation was conducted using a Biotage Initiator unit. The machine consists of a continuous focused microwave power delivery system with operator-selectable  Technologies 1200 series HPLC connected to an Agilent Technologies 6130 quadrupole spectrometer, where both instruments were connected to an Agilent diode array detector. Liquid chromatography-mass spectrometry (LCMS chromatographic separations were conducted with a Waters X bridge C18 column (50 mm 2.1 mm, 3.5 mm particle size), with a mobile phase of water/acetonitrile + 0.1 % HCOOH, or water/acetonitrile + 0.1 % NH 3 , using a linear gradient from 80:20 to 5:95 over 3.5 min and then held for 1.5 min, at a flow rate of 0.5 mL min À1 . All compound samples evaluated in biological assays had a measured purity of ! 95 % (by total ion current (TIC) and UV) as determined using this analytical LCMS system. High-resolution mass spectrometry (HRMS) using electrospray ionisation was performed on a Bruker Daltonics MicrOTOF mass spectrometer.