Discovery of Inhibitors of Trypanosoma brucei by Phenotypic Screening of a Focused Protein Kinase Library

A screen of a focused kinase inhibitor library against Trypanosoma brucei rhodesiense led to the identification of seven series, totaling 121 compounds, which showed >50 % inhibition at 5 μm. Screening of these hits in a T. b. brucei proliferation assay highlighted three compounds with a 1H-imidazo[4,5-b]pyrazin-2(3H)-one scaffold that showed sub-micromolar activity and excellent selectivity against the MRC5 cell line. Subsequent rounds of optimisation led to the identification of compounds that exhibited good in vitro drug metabolism and pharmacokinetics (DMPK) properties, although in general this series suffered from poor solubility. A scaffold-hopping exercise led to the identification of a 1H-pyrazolo[3,4-b]pyridine scaffold, which retained potency. A number of examples were assessed in a T. b. brucei growth assay, which could differentiate static and cidal action. Compounds from the 1H-imidazo[4,5-b]pyrazin-2(3H)-one series were found to be either static or growth-slowing and not cidal. Compounds with the 1H-pyrazolo[3,4-b]pyridine scaffold were found to be cidal and showed an unusual biphasic nature in this assay, suggesting they act by at least two mechanisms.


Introduction
HumanA frican trypanosomiasis (HAT), or sleeping sickness, as it is more commonly known,isanoften fatal infection endemic in sub-Saharan Africa. Current treatments for HATa re inadequate duet ot oxicity and inappropriate means of administration for remote rural settings,w here the disease is predominantly found. Fortunately,t he number of cases of HATappears to be decreasing;h owever,t here is an urgent need for new effective treatments forH AT,t oi mprovet reatment and to facilitate the eventual eradication of this disease. HATi sc aused by the protozoan parasites Trypanosomab rucei gambiense and T. b. rhodesiense.The parasites have acomplex life cycle, involving the humanh ost andt he tsetse fly,w hich is the vector.I nitially there is ap eripheralinfection in which the parasite is pre-dominantly found in the blood. However, the parasites eventually invade the central nervouss ystem, which can lead to coma and death. [1,2] Protein kinases are am ajor drug target for human diseases, particularly oncology. [3] Analysis of the T. brucei genome indicates that there are 156 protein kinases in the trypanosome kinome. [4][5][6] Whilst parasite protein kinasesh ave been relatively poorlyi nvestigated, there appear to be significant differences from human kinases.Inp articular: ·A relatively large number of protein kinases associated with cell-cycle control,w hich may be ac onsequenceo ft he complex life cycleofthe parasite; ·A lack of tyrosine kinases( except dual-specificity kinases); ·A number of unique kinases.
In addition, an umber of protein kinases have been genetically validated as potentiald rug targets in T. brucei using RNA interference (RNAi)o rg ene knockouts. [4,5,[7][8][9] Given that protein kinases are generally known to be druggable, they represent promising drug targets in trypanosomatids. The Drug Discovery Unit (DDU) in Dundee has recently carriedo ut screensa nd chemicalo ptimisation programmes against severalp arasite protein kinases.A lthough we have prepared compounds with significant (low-nanomolar) activity against the proteins, this did not always translate into potent cellular activity. [10] Ac omplementary methodt ot arget-based drug discoveryi s phenotypicd rug discovery, [11] whichh as several advantages. Firstly,a ll possible drug discovery targets are present in their natural environment, allowing an unbiaseda nd more physio-As creen of af ocused kinase inhibitor library against Trypanosoma brucei rhodesiense led to the identification of seven series, totaling1 21 compounds, which showed > 50 %i nhibition at 5 mm.S creeningo ft hese hits in a T. b. brucei proliferation assay highlighted three compounds with a1 H-imidazo [4,5-b]pyrazin-2(3H)-one scaffold that showeds ub-micromolar activity and excellent selectivity against the MRC5 cell line. Subsequentr ounds of optimisation led to the identification of compounds that exhibited good in vitro drug metabolism and pharmacokinetics (DMPK) properties, although in general this series suffered from poor solubility.Ascaffold-hopping exercise led to the identification of a1 H-pyrazolo [3,4-b]pyridine scaffold, which retained potency.An umber of examples were assessed in a T. b. brucei growth assay,w hich could differentiate static and cidala ction.C ompounds from the 1H-imidazo [4,5b]pyrazin-2(3H)-one series were found to be either static or growth-slowinga nd not cidal. Compounds with the 1H-pyrazolo [3,4-b]pyridine scaffoldw ere found to be cidal and showed an unusualb iphasic nature in this assay,s uggesting they act by at least two mechanisms. logicallyr elevant screening platform;t his may give rise to compounds that inhibitm ore than one target. Indeed it has been found in the oncology field that compounds that inhibit more than one protein kinase are often required for activity. Secondly, as the primary screening platform is af unctional efficacy screen, the relationship between target and phenotype does not need to be established. Finally,c ompounds must be able to penetrate cells and have as ufficientf ree fraction in the assay to elicit their response, eliminating compounds with inappropriate properties. [12][13][14] We therefore decided to conduct ap henotypic screen of af ocused kinase compound library against whole parasites.As imilar exercise was recently reported by Diaz et al.,i nw hich ap henotypic screen of ak inase-targeted library from GlaxoSmithKline (GSK) was reported and gave rise to an umber of actives. [15] There is also ar ecent report of al arge screen against kinetoplastids with 1.8 million compounds from GSK. [16] The ideal target product profile to treat HATr equires ac ompound that can treat both stage 1 (peripheral) and stage 2( CNS) infection; [8] thus the compound should have blood-brain barrier(BBB) permeability.

Results and Discussion
The focused screen The Dundee focusedp rotein kinase library, [17] which at that point contained3 885 compounds, was assayed by the Swiss Tropical and Public Health Institute (STPH)a gainst T. b. rhodesiense at 1a nd 5 mm.F rom this original triage,s even series, totaling 121 compounds, were identified which showed > 50 % inhibition of parasite growth at 5 mm.T hese were progressed into EC 50 determination in a T. b. brucei proliferation assay and assessed in aM RC5 proliferationa ssay to provide an early indicator of toxicityt om ammalian cells. From this, seven compounds showedE C 50 values < 1 mm against T. b. brucei.T he most active hitc ompound was a1 H-imidazo [4,5-b]pyrazin-2(3H)-one( 1), which had an EC 50 value of 80 nm against T. b. brucei and was nontoxic to the mammalian MRC5 cell line (EC 50 > 50 mm). Twoa dditional 1H-imidazopyrazinone analogues had EC 50 values of < 1 mm against T. b. brucei (Table 1).
The physicochemical properties were calculated in StarDrop (www.optibrium.com). It has been proposed that for ac ompoundt oh ave BBB permeability,i ts hould have at opological polar surfacea rea (tPSA) of < 90 2 and am olecular weight (M r )less than 450 Da. [18] Wager's research group have proposed more sophisticatedm odels. [19][20][21] The M r and tPSA values were all within the preferred range for ap otentialC NS-penetrant compound. The calculated logP values were also in the range of CNS-penetrant compounds. [19] Based on the initial data we decidedt op rogress the project into hits-to-leads development.

Hit validation
Medicinal chemistry was initiated with the aim of validating and profiling the hit series and increasing the potencyo ft he compounds. The desired 1H-imidazopyrazinones were synthesised in four steps from 2-aminopyrazine 4. Brominationw ith NBS gave the versatile scaffold 2,5-dibromo-4-aminopyrazine 5. Regioselective displacement of the 5-bromine group with amines could be conducted using microwaveh eating to give intermediates 6a-f (Scheme 1). [13][14][15]17, and 18 were prepared by Suzukir eaction witht he appropriate diamine precursor followed by cyclisation with CDI. Cyclisation of the appropriate diamine (6a or b)w ith CDI gave 7a or 7b,w hich were then reactedu nder Suzukic onditions to give compounds 12 and 19-30. Treatment of 7a under hydrogenation conditions afforded compound 16.
Enantiomers 1 and 8 were equipotent, whilst removalo ft he benzylic methylg roup (compound 9) resultedi naf ivefold decrease in activity (Table 2). Replacemento ft he benzyl group with methylcyclopropane (compound 10)h ad little effect on the activity, whilst further truncation of the amine to am ethyl group (in 11)r esultedi na1 00-fold dropi na ctivity relative to methyl cyclopropane( in 10). Replacement www.chemmedchem.org of the methylbenzyl group of 1 with am ethoxypropyl group (in 13)r esulted in as evenfold loss in activity.I nterestingly,r eplacemento ft he methylbenzyl group of 1 with cyclohexane (compound 12)r esulted in at hreefold improvementi na ctivity. Compounds showed good selectivity for the parasiteo ver MRC5 cells. The 4-fluorophenyl group of 1 could be replaced with ap henyl ring (in 14)w ith no loss of activity ( Table 3). The larger 4-trifluoromethyloxy analogue 15 was1 0-fold less active than 1,b ut was comparable with the hitc ompound 2,f urther demonstrating that changes to the R 1 group were tolerated. Replacement of the phenylr ing of 14 with ah ydrogen atom in 16 resulted in a > 300-fold loss of activity relative to 14. Substitution of the phenyl ring with a3-fluoro group was tolerated in 17,w hile the larger and more polar 3-methanesulfonamide group (in 18)r esulted in a3 0-fold loss of activity.

Kinase panel and DMPK data
These data validated the series, as there was ad efinite structure activity relationship, and several examples demonstrated sub-100nm EC 50 values in the T. b. brucei proliferation assay. The compounds also showed excellents electivity over human MRC5 cells. Consequently,i tw as decidedt op rofile the com-poundsf urtherf or potential inhibitiono fh uman kinases and to studyt heir DMPK properties to ensure that there were no major issues which may impact further development.
The DDU kinase-focused compound set contains lead-like scaffolds that are designed to target protein kinases;they have kinase hinge binding motifs. Four of the 1H-imidazopyrazinones were screened against ah uman/mammalian kinase panel at the Universityo fD undee (www.kinase-screen.mrc.ac.uk) at ac oncentration of 10 mm.A tt he time of testingt he panel contained 79 mammalian kinases selected to provide ab road coverage of the mammalian kinome (~500 kinases in the human genome). [3] None of the compounds were active against any of the kinases at 10 mm.A lthough the resultsd id not provide any clues as to the molecular target in the parasite, they did show the 1H-imidazopyrazinones to have promising selectivity over mammalian kinases.
Twok ey compounds, 1 and 12,s howedl ow in vitro mouse hepatic microsomal intrinsic clearance (CL int < 0.5 and 0.7 mL min À1 g À1 ,r espectively), consistentw ith good metabolic stability. Compound 1,w ith promising in vitro anti-parasitic activity,l ow toxicityt oh uman cells, and promising in vitro metabolic stability, was progressedi nto in vivo pharmacokinetic (PK) determination. After as ingle oral dose at 5mgkg À1 to female NMRIm ice, the compound showed al ow C max value and took approximately 2h to reach C max ,b oth of which are suggestive of solubility-limited absorption (Table 4). Furthermore, 1 showedh igh protein binding (0.3 %u nbound). Conse-  quently,t he observed free bloodc oncentration was likely to be too low for the compound to be efficacious in vivoe ven at high dose, so further development was required to improve free oral exposure for 1.Common strategies to address the key issues of poor solubility and/or high protein binding include decreasing the lipophilicity (logP), addition of solubilising groups,a nd reducing the planarity of the molecule to minimise crystal stacking. [22][23][24][25] Hit to leads The R 2 position Ar ound of optimisation for the R 2 position was commenced. During this process R 1 was retaineda sacyclohexylm oiety,a s this substituent gave the greatest activity and, due to its nonplanar structure, shoulda lso increase the solubility of the compounds.Avariety of substituents were investigated at the para, meta,a nd ortho positions of the phenyl group at the R 2 position. The substituents were designed either to disrupt the packing and/or to reduce lipophilicity.C ompounds were prepared using the route described in Scheme 1( Ta ble5).
For the para-substitutedc ompounds,t he electron-donating methoxy group in 20 showed an eightfold drop in activity.T he introduction of polar groups was poorly tolerated, with hydroxymethyl compound 21 being 10-foldl ess active, while amide 22 and dimethylaminomethyl compound 23 were3 0and 50-foldl ess active, respectively.S ubstituents at the para positiong enerally caused ad rop in activity,r elative to those in the ortho or meta positions. Encouragingly,t he hydroxymethyl substituents retained activity in the ortho and meta positions (28 and 24,respectively) and gave rise to good solubility.
Furtherw ork was undertaken in which the phenyl ring of 19 was replaced with ah eterocycle or saturated ring system, which should increase solubility. For the introduction of amines we used aB uchwald reactiono ni ntermediate 7b (Scheme 2). For the introduction of aromatic heterocyclesw e used the chemistry describeda bove. Both 4-pyridyl 29 and 3pyridyl 30 compounds were equipotent to phenyl 19,d emonstratingatolerance for ah ydrogen bond acceptora tthe 3and 4-positions (Table6). Both piperidine 31 and the more polar morpholine 32 were tolerated, but the basic piperazine 33 was 50-fold less active than the phenylcompound 19.

The R 1 position
The initial rounds of medicinalc hemistry demonstratedt hat changes to the R 1 group were tolerated and that the cyclohexyl ring of 12 was slightly more active than the initial a-methylbenzylg roup of 1.T oc ompletet he exploration of the R 1 group, an optimisation array was conducted, maintaining ap henylr ing at the R 2 position forr eference. 3-Bromo-5-phenylpyrazine-2-amine 35 was prepared from 3-amino-6-bromopyrazine-2-carboxylic acid 34 in two steps (Scheme 3). Displacement of the 3-bromo group followed by cyclisation with CDIgave compounds 36-42.T runcation of 1 to an isopropyl group (in 36)r esulted in a1 0-foldd rop in activity while rigidificationb yr eplacemento fc yclohexyl with an aromatic ring (in 37)r esultedi nacomplete loss of activity (Table 7). The majority of changes were designed to introduce polar solubilising groups.A lkyl ethers 38 and 39 showedo nly two-and fivefold losses of activity, respectively.A lkoxy compound 40 was~10-foldl ess active than the cyclohexyl, whereas the basic compounds 41 and 42 showedp oor antitrypanosomal activity. Both 28 and 38 showed decreased protein binding relative to 1 (7.2 and 15 %u nbound, respectively, versus 0.3 %u nbound) and improved solubility (> 250 and 55 mm,r espectively,c ompared with Scheme2.Reagents and conditions:a )Pd 2 dba 3 (5 mol %), amine (2 equiv), rac-binap (10 mol %), tBuOK (3 equiv), 1,4-dioxane, 160 8C( microwave), 20 min,5 -24 %. www.chemmedchem.org < 12 mm). However,b oth compounds were less metabolically stabile when incubated with mouse hepatic microsomes (CL int = 2.4 and 1.6 mL min À1 g À1 ,r espectively,v ersus < 0.5 mL min À1 g À1 )b ut stillp rogressable. Given these results, both 28 and 38 were progressed into in vivo PK studies. Compound 28 showed significantly higher levels of exposure relative to 38.C ompound 28 also exhibited an improvement in the time to reach C max compared to 1 (15 minversus 2h)w hen dosed orally (Table 8), presumably aresult of its improved solubility.H owever, 28 was cleared quite rapidly after both oral and i.p. administration, likely due to the higher metabolic instabilityo ft he compound;i tw ould be unlikely to deliver sufficient duration of efficacious free drug concentration at at olerated dose. We decided to investigate alternative scaffoldsw ith the aim of improving metabolic stability and solubility.
Scaffold hopping:the core group Initial work was conducted to investigate the importance of the pyrazine nitrogen atoms (by removal of both of them, Scheme4). Thus, 1-bromo-3-fluoro-4-nitrophenyl 43 was reacted with a-methylbenzyla mine to give intermediate 44 (Scheme 4). Thenitro group wasreduced, then 45 was cyclised with CDI to give (R)-6-bromo-1- Finally ac ross-coupling reaction gave the desired compound 47.R emoval of both ring nitrogen atoms of 1 (EC 50 :0 .08 mm)t og ive 47 (EC 50 :1 0mm)r esulted in a1 00-fold loss of activity,i ndicating that one or both of the pyrazine nitrogen atomsa re important.
It was hypothesised that the "urea" moiety may be ac ause of the poor solubility,p ossibly through its ability to cause antiparallel hydrogen bonded dimers. However,i ft his was to be replaced, certainf eatures would need to be retained:t he carbonyl group could act as ah ydrogenb ond acceptor, and the NH group as ah ydrogen bond donor.A sN 1w as at ertiary nitrogen it was hypothesised that does not make any important interactions and could be replaced by ac arbon atom. This would then allow the carbonyl to be replaced by an itrogen, which could also act as ah ydrogen bond acceptor, giving an aza-indazole scaffold:aknown kinase scaffold.
The Scheme5. Reagents and conditions:a)NaH,MeI, DMF,RT, 3h,3 8%. Scheme6.Reagents and conditions:a)1)cyclohexanone, AcOH, CH 2 Cl 2 ;2)Na(OAc) 3  Removal of one the nitrogen atoms from the pyraziner ing was predicted to lead to improvements in metabolic stability, protein binding,a nd solubility (Table 11). However,r eplacement of the core scaffoldw ith the aza-indazole had an egative effect on the in vitro DMPK properties of the compounds (compare 51 and 60; 55 and 63).
Compound 56 was selected for in vivo PK studies, as it had the most favourablei nv itro profile (potency, metabolic stability,p rotein binding, and solubility). However, 56 had lower exposure levels than 28 and was also cleared rapidly,i nl ine with its higher CL int value ( Table 12).
Given our experience from other programmes, in order for ac ompound to show efficacy in am ouse model of HAT, it is important that the compound is cidal, rather than static,i ni ts activity against T. brucei. Compounds that are consistent with this are 60 and 63.H owever,g iven their metabolic instability and poor potency( minimum cidal concentrations of 17 mm), it is highlyu nlikely that either would deliver efficacy in am ouse modelo fH AT.F urther work is neededt oi mprovet he potency and pharmacokinetic properties of the molecules.

Conclusions
Through our phenotypics creening approach witharelatively small library,w eh ave developed as eries of compounds that exhibit good levels of potency in the T. b. brucei growth inhibition assay,a nd that are nontoxic to MRC5 mammalian cells. Despitet he fact that the screening library was af ocusedp rotein kinase library,i tw as possible to obtain compounds with selectivity over the mammalian protein kinasesa ssayed. These compounds possess reasonable physicochemical properties and show oral exposure, but have modest half-lives. By changing the substituents R 1 and R 2 it wasp ossible to optimise the compounds for potencya nd to alter key physicochemicala nd DMPK properties, in the absence of knowledge of the molecular target. It was also possible to scaffold-hop, although it is possible that the new scaffold works on adifferent molecular target. The ultimate aim would be to devise ac ompound that has oral bioavailability and the capacity to penetrate the BBB. The compounds prepared in this study generally have physicochemical properties consistent with BBB penetration as reported by Wager, [19] in terms of M r ,t PSA, clogP,a nd number of hydrogen bond donors. Although of course for compounds taken forward, BBB permeability would have to be determined experimentally.
Key compounds in as tatic-cidal assay showeda n unusualb iphasic nature, giving two points of inflection on the dose-response curve. This suggests at least two modes of action (or,l ess likely due to the use of ac lonal line, differentials usceptibility of cell sub-populations). Compounds from the 1H-imidazo [4,5-b]pyrazin-2(3H)-one scaffold appeared to be either static or growth-slowing, whilst compounds from the 1H-pyrazolo [3,4-b]pyridine series were cidal. It is possible that compounds from the 1H-imidazo [4,5-b]pyrazin-2(3H)-one scaffold will be cidal at higher concentrations, and that the key difference between the two scaffolds is not ad ifferencei n mode-of-action butad ifferencei np otencya gainst the target(s) represented by the second phase of the biphasic curves.
Aw ay forwardw ould be to identifyt he molecular target(s)o ft he compounds, specifically which target(s) causes the cytocidala ctivity,a nd to focus efforts on potent inhibition of this target. Subsequent progression of the series will requirei dentification of compounds that are cytoci-  dal. It will also be necessary to optimise the key properties of the molecule (microsomal stability, solubility,a nd protein binding). Although it has provedp ossible to optimise the compound in terms of potencya nd to some extenta tl east the pharmacokinetic parameters, we have reached ab iological issue which has proved difficult to resolve in the absence of knowledge of the cidal target. In this case, ac ombination of phenotypica nd target-based approaches would probably be optimal.

Experimental Section
Chemistry Compounds 1, 2, 3 and 13 were purchased from BioFocus. Chemicals and solvents were purchased from Aldrich Chemical Co., Fluka, ABCR, VWR, Acros, Fisher Chemicals, and Alfa Aesar and were used as received unless otherwise stated. Air-and moisturesensitive reactions were carried out under an inert atmosphere of argon in oven-dried glassware. Analytical thin-layer chromatography (TLC) was performed on pre-coated TLC plates (layer 0.20 mm silica gel 60 with fluorescent indicator UV254, Merck). Developed plates were air dried and analyzed under aU Vl amp (l 254/ 365 nm). Flash column chromatography was performed using prepacked silica gel cartridges (230-400 mesh, 40-63 mm, SiliCycle) (unless otherwise stated) using aT eledyne ISCO Combiflash Companion or Combiflash Retrieve. 1 Ha nd 13 CNMR spectra were recorded on aB ruker AvanceII5 00 spectrometer ( 1 Ha t5 00.1 MHz, 13 Ca t1 25.8 MHz), or aB ruker DPX300 spectrometer ( 1 Ha t 300.1 MHz). Chemical shifts (d)a re expressed in ppm recorded using the residual solvent as internal reference in all cases. Signal splitting patterns are described as singlet (s), doublet (d), triplet (t), quartet (q), pentet (p), multiplet (m), broad (br), or ac ombination thereof. Coupling constants (J)a re quoted to the nearest 0.1 Hz. LC-MS analyses were performed with either an Agilent HPLC 1100 series instrument connected to aB ruker Daltonics MicrOTOFo ra n Agilent Technologies 1200 series HPLC connected to an Agilent Te chnologies 6130 quadrupole spectrometer,w here both instruments were connected to an Agilent diode array detector.L C-MS chromatographic separations were conducted with aW aters Xbridge C 18  linear gradient from 80:20 to 5:95 over 3.5 min and then held for 1.5 min;f low rate:0 .5 mL min À1 .A ll tested compounds had am easured purity of ! 95 %( by TIC [total ion current] and UV) as determined by this analytical LC-MS system. High-resolution electrospray MS measurements were performed on aB ruker Daltonics Mi-crOTOFm ass spectrometer.M icrowave-assisted chemistry was performed using aB iotage initiator microwave synthesiser.

Biological assays
Medium-throughput screen for African trypanosomes and cytotoxicity:M easurement of inhibition of the proliferation of L-6 cells (mouse muscle fibroblasts) and T. brucei bloodstream-stage cells was performed as previously described. [35,36] Trypanosome and MRC5 proliferation assay:M easurement of inhibition of the proliferation of MRC5 cells (human lung fibroblasts) and T. brucei bloodstream-stage cells was performed using am odification of ac ell viability assay previously described. [35,36] Compounds (50 mm to 0.5 nm)w ere incubated with 210 3 cells per well in 0.2 mL of the appropriate culture medium (MEM with 10 %f oetal bovine serum for MRC5 cells) in clear 96-well plates. Plates were incubated at 37 8Ci nt he presence of 5% CO 2 for 69 h. Resazurin was then added to af inal concentration of 50 mm,a nd plates were incubated as above for af urther 4hbefore being read on aB ioTek Flx800 fluorescence plate reader.
Determination of cidality against T. brucei bloodstream-stage cells was performed using as tatic-cidal (SC) assay as previously reported. [26] Briefly,b loodstream-form trypanosomes were seeded into 384-well plates at 410 5 mL À1 (50 mLp er well), followed by immediate addition of resazurin (50 mm final) to one of the plates, and all plates were incubated at 37 8C, 5% CO 2 .A fter 4h the t 0 plate was read using aP erkinElmer Victor 3p late reader (l ex = 528 nm, l em = 590 nm). The second plate was processed in the same manner 20 hl ater,a nd at 44 ht he last plate was processed. The minimum cidal concentration (MCC) is defined as the lowest concentration of drug that results in ad ecrease of resorufin signal over time. For dose-response curves from this assay either amonophasic or ab iphasic equation was used, depending on which one provided the best fit. For monophasic fits, the following four-parametric equation [Eq. (1)] was used: in which A = percent inhibition at bottom, B = percent inhibition at top, C = EC 50 , D = slope, x = inhibitor concentration, and y = percent inhibition. Equation (2) was used for biphasic fits:  (3): in which V [mL (mg protein) À1 ]i st he incubation volume per mg protein added, and microsomal protein yield is taken as 52.5 mg protein per gram of liver.V erapamil (0.5 mm)w as used as ap ositive control to confirm acceptable assay performance.
Plasma protein binding experiments:I nb rief, a9 6-well equilibrium dialysis apparatus was used to determine the free fraction in plasma for each compound (HT Dialysis LLC, Gales Ferry,C T, USA). Membranes (12-14 kDa cutoff) were conditioned in deionised H 2 O for 60 min, followed by conditioning in 80:20 deionised H 2 O/EtOH for 20 min, and then rinsed in isotonic buffer before use. Female CD1 mouse plasma was removed from the freezer and allowed to thaw on the day of experiment. Thawed plasma was then centrifuged (Allegra X12-R, Beckman Coulter,USA), spiked with test compound (10 mgg À1 ), and 150 mLa liquots (n = 6r eplicate determinations) loaded into the 96-well equilibrium dialysis plate. Dialysis against isotonic buffer (150 mL) was carried out for 5hin at emperature-controlled incubator at~37 8C( Barworld Scientific Ltd.,U K) using an orbital microplate shaker at 125 rpm (Barworld Scientific). At the end of the incubation period, aliquots of plasma or buffer were transferred to micronic tubes (Micronic B.V., the Netherlands), and the composition in each tube was balanced with control fluid such that the volume of buffer to plasma is the same. Sample extraction was performed by the addition of 400 mLMeCN containing an appropriate internal standard. Samples were allowed to mix for 1min and then centrifuged at 3000 rpm in 96-well blocks for 15 min (Allegra X12-R, Beckman Coulter,USA). All samples were analysed by UPLC-MS/MS on aQ uattro Premier XE Mass Spectrometer (Waters Corp.). The unbound fraction was determined as the ratio of the peak area in buffer to that in plasma.
Solubility experiments:T he kinetic aqueous solubility of the test compounds was measured using laser nephelometry.C ompounds were subject to serial dilution from 10 mm to 0.5 mm in DMSO. An aliquot was then mixed with Milli-Q H 2 Otoobtain an aqueous dilution plate with af inal concentration range of 250-12 mm,w ith af inal DMSO concentration of 2.5 %. Triplicate aliquots were transferred to af lat-bottomed polystyrene plate which was immediately read on the NEPHELOstar (BMG Lab Te chnologies). The amount of laser scatter caused by insoluble particulates (relative nephelometry units, RNU) was plotted against compound concentration using as egmental regression fit, with the point of inflection being quoted as the compounds' aqueous solubility (in mm).
In vivo pharmacokinetics in mice:C ompounds 1 and 28 were dosed orally by gavage as as olution at 5mgf ree base per kg or 20 mg free base per kg, respectively (dose volume:5or 10 mL kg À1 ,d Ethics: All regulated procedures on living animals were carried out under the authority of ap roject license issued by the Home Office under the Animals (Scientific Procedures) Act 1986, as amended in 2012 (and in compliance with EU Directive EU/2010/63). License applications will have been approved by the University'sE thical Review Committee (ERC) before submission to the Home Office. The ERC has ag eneral remit to develop and oversee policy on all aspects of the use of animals on University premises and is as ubcommittee of the University Court, its highest governing body.