Repositioning of a Diaminothiazole Series Confirmed to Target the Cyclin-Dependent Kinase CRK12 for Use in the Treatment of African Animal Trypanosomiasis

African animal trypanosomiasis or nagana, caused principally by infection of the protozoan parasites Trypanosoma congolense and Trypanosoma vivax, is a major problem in cattle and other livestocks in sub-Saharan Africa. Current treatments are threatened by the emergence of drug resistance and there is an urgent need for new, effective drugs. Here, we report the repositioning of a compound series initially developed for the treatment of human African trypanosomiasis. A medicinal chemistry program, focused on deriving more soluble analogues, led to development of a lead compound capable of curing cattle infected with both T. congolense and T. vivax via intravenous dosing. Further optimization has the potential to yield a single-dose intramuscular treatment for this disease. Comprehensive mode of action studies revealed that the molecular target of this promising compound and related analogues is the cyclin-dependent kinase CRK12.


In vitro drug sensitivity assays -Trypanosoma brucei brucei
Drug sensitivity assays were carried out with bloodstream-form (BSF) T. b. brucei 'single marker' S427 cells grown at 37°C with 5% CO2 in HMI-9T medium, 1 as previously described. 2

In vitro drug sensitivity assays -Trypanosoma congolense
EC50 values for test compounds against T. congolense (IL3000 strain) were conducted as previously described 3 with minor modifications. 4 Briefly, trypanosome cell densities were determined using a CASY cell counter (Schärfe) and diluted to a seeding density of 2 x 10 5 trypanosomes/mL. BSF trypanosomes were grown in HMI-9 (IMDM) media supplemented with 20% (v/v) bovine serum. 1 Test compounds were resuspended to a final concentration of 10 mg/mL in DMSO. Assays plates were incubated at 34 °C with 5 % CO2, for a total of 69 h, prior to addition of 10 µL of Resazurin (12.5 mg in 100 mL phosphate buffered saline, Aldrich) into each well, followed by a further 3 h incubation. Absorbance was read using a SpectraMax (Gemini XS) fluorescence reader at an excitation wavelength of 536 nm and an emission wavelength of 588 nm. EC50 values were determined using SOFTmax Pro 5.2 analysis software. All experiments were conducted with technical duplicates, and at least three biological replicates.

Ex vivo drug sensitivity assays -Trypanosoma vivax
Since T. vivax cannot be cultured axenically, BSF trypanosomes were harvested from an infected mouse with high parasitemia via cardiac puncture and used directly in ex vivo assays. EC50 values were determined against T, vivax (STIB 719/ ILRAD 560) as previously described 3 , with minor modifications. Briefly, cell densities were determined using a Neubauer chamber and then diluted to a seeding density of 4 x 10 5 trypanosomes / mL. BSF trypanosomes were cultured in HMI-9 (IMDM) media containing 20% (v/v) bovine serum. Assay plates were incubated at 37 °C with 5 % CO2, for a total of 45 h prior to addition 10 µL of Resazurin (12.5 mg in 100 mL phosphate buffered saline, Aldrich) into each well, followed by a further 3 h incubation. Absorbance was read using a SpectraMax (Gemini XS) fluorescence reader at an excitation wavelength of 536 nm and an emission wavelength of 588 nm. EC50 values were determined using SOFTmax Pro 5.2 analysis S-4 software. All experiments were conducted with technical duplicates and at least three biological replicates.

Cytotoxicity assays -Rat skeletal myoblasts (L6)
The cytotoxicity of test compounds was evaluated using a rat skeletal myoblast (L6) cell line, as previously described 3 , but with minor modifications. Cells were grown in RPMI media supplemented with 10% (v/v) fetal calf serum. L6 cells at a seeding density of 2 x 10 4 cells/mL were incubated in microtiter plates overnight at 37 ºC and 5 % CO2 until adherent. The following day, test compounds were added to the plates (3-fold serial dilutions), prior to incubation for 69 h. Following incubation, 10 µL of Resazurin (12.5 mg in 100 mL phosphate buffered saline, Aldrich) was added to each well and the plates were incubated for a further 3 h. The plates were read using a SpectraMax (Gemini XS) fluorescence reader at an excitation wavelength of 536 nm and an emission wavelength of 588 nm. EC50 values were determined using SOFTmax Pro 5.2 analysis software. All experiments were conducted as technical replicates, with at least three biological replicates.

Solubility
The aqueous solubility of test compounds was measured using laser nephelometry. Compounds were subject to serial dilution from 10 mM to 0.5 mM in DMSO. Aliquot were then mixed with MilliQ water to obtain an aqueous dilution plate with a final concentration range of 250 -12 µM. The final DMSO concentration in these plates was 2.5%. Triplicate aliquots were transferred to a flat-bottomed polystyrene plate which was immediately read on a NEPHELOstar nephelometer (BMG Lab Technologies). The laser scatter caused by insoluble particulates (relative nephelometry units, RNU) was plotted against compound concentration using a segmental regression fit, with the point of inflection being quoted as the compound's aqueous solubility (µM).
Solubility of compound 1 was measured by a different method. Here, compound 1 was dissolved in DMSO to give a 10 mM solution and solubility test samples prepared by adding a volume (5 µL) of this stock solution to phosphate buffered saline, pH 7.4 (195 µL, Sigma-Aldrich). This S-5 solution was then mixed for 24 h (rotary mixing, 900 rpm, 25°C) excluding light. After mixing, test samples were filtered to remove any undissolved material using a proprietary filter (Millipore Multiscreen HTS filter, 96-well format). Samples were drawn through the filter using a vacuum. The resulting filtrate was analysed for dissolved compound 1 using a truncated UHPLC methodology. A Shimadzu Nexera X2 UHPLC system was used, with a reversed-phase column and a simple formic acid gradient elution. The UHPLC parameters are shown below (Table S1): CLi(mL/min/g liver) = k x V x microsomal protein yield Where V (mL/mg protein) is the incubation volume/ mg protein added and microsomal protein yield is taken as 52.5 mg protein/ g of liver. Verapamil (0.5 µM) was used as a positive control to confirm acceptable assay performance. and 120 min. 100 µL of 80:20 water:acetonitrile was added to all samples and the analysis plate was centrifuged (10 min, 1,665 g, RT) prior to injection and analysis of samples by UPLC-MS/MS. The response (area ratio of test compound relative to internal standard) was plotted against time using an exponential decay model and rate of disappearance calculated. Hepatocyte CLint (mL/min/10 6 cells) was scaled to in vivo CLint (ml/min/g Liver) using the hepato-cellularity scaling factor of 120 x of test compound to internal standard) was plotted against time using an exponential decay model and rate of disappearance calculated. Hepatocyte CLint (mL/min/10 6 cells) was scaled to in vivo CLint (mL/min/g liver) using the hepato-cellularity scaling factor of 120 x 10 6 cells/g of liver.

Hepatocyte stability studies
S-8

Protein plasma binding
This was determined as previously described. 6

Maximum tolerated dose
Compound 1 as free base in 5% DMSO; 40% PEG400; 55% MilliQ water was dosed daily subcutaneously with doses of 50 mg/kg, 100 mg/kg and 300 mg/kg for 3 days. No tolerability issues were detected.

PK and formulation studies -male Sprague rats
Compound  Table 2). Studies conducted at Ridgeway

Formulation and tolerability studies for compound 2 (NorthWest Biopharm)
A study to identify alternative formulations was conducted. Compound 2, either as the hydrochloride salt or free base, was screened to identify a 50 mg/mL soluble formulation suitable for injection.

S-10
Different formulations investigated to reduce injection site tolerability (Table S2) The various formulations were then subjected to an injection site tolerability assessment (IST) in cattle to determine if any irritation at the site of injection was evident. Each formulation was administered to two female calves. The calves were cross-breeds (1 x Limousine cross 2 x Hereford cross, 3 x Aberdeen Angus cross). The three formulations of the IVP were supplied formulated ready for administration. Each formulation was administered by single intramuscular injection into the area of the semi-tendinosis muscle of the hind limb. and with ascending IM doses of 1 mL, 5 mL and 10 mL. Saline was used as a control. The site of administration for each formulation on each study day was specified in the table below (Table S3).

S-11
On the day of administration, prior to administration of test compounds to any animal and at 30 min and 1 h (± 5 min) and at 3 and 5 h (± 10 min) after administration, a whole blood sample was collected from that animal. This sample was used to determine creatinine kinase (CK) levels, a marker of muscle damage, and levels of compound 2 in the animal at each collection timepoint.
Approximately 5 mL of whole blood was collected from the jugular vein using an appropriate needle and syringe. The sample was divided between two 1.3 mL heparinised blood tubes and one 1.3 mL EDTA (K2) tube. Immediately after collection, the EDTA (K2) tube was placed into iced water.
Thereafter, 2 x 0.5 mL of this blood was accurately pipetted into two Eppendorf tubes each containing 1.0 mL of distilled water (also accurately pipetted and stored in a refrigerator). These tubes were pre-labelled for each animal and placed in a rack so that the samples, when frozen, remained at the bottom of the tubes. The blood and water were mixed well by pipetting gently up and down. The • Formulation 1: Following a 5 mL injection with formulation 1 (Compound 2 hydrochloride salt in Kolliphor EL, Glycerol Formal, 2-Pyrrolidone and 11% water suitable for injection), an 8 cm swelling appeared at the injection site. Thus, further study of this formulation was stopped prior to proceeding to 10 mL dosing.

S-12
• Formulation 2: Following a 5 mL injection with formulation 2 (compound 2 hydrochloride salt in Kolliphor EL, N-methyl pyrrolidone, 2-pyrrolidone, and 11% water suitable for injection), a 2-4 cm swelling appeared at the injection site, however, no pain or lameness observed. In contrast, no injection site changes were evident following an injection of 10 mL and there were no signs of associated pain or lameness.
• Formulation 3: Administration of a 5 mL injection of formulation 3 (compound 2 free base in Kolliphor EL, N-methyl pyrrolidone, 2-pyrrolidone and 11% water suitable for injection) resulted in a 1-2 cm swelling and a 10 mL injection caused 4 cm swelling.
It should be noted that swellings of <5 cm in these studies are not considered significant.
Creatine kinase levels were also markedly lower with formulations 2 and 3 when compared to those in animals treated with formulation 1.
During the study all calves remained clinically well with only local injection site reactions. Blood concentration at each dose volume for each formulation are summarised in Table S3. In parallel, all the formulations were assessed in rats (IM) and the results were comparable to those in cattle, observationally and with regard to creatine kinase levels. These observations offer confidence that rats can be used as a model animal in the drug development path going forward prior to studies in cattle.

Comparative cattle PK (Northwestern Biopharm)
The pharmacokinetics for compound 2 when dosed as free base in the best tolerated formulation: Kolliphor EL, N-methyl pyrrolidone, 2-pyrrolidone and 11% water at 5 mg/kg was assessed, IM compared to SC. Mean pharmacokinetic parameters are presented in Table S4. Bioanalysis of blood samples was conducted at the University of Dundee.

Salt screen
A screen of salt forms as alternatives to the hydrochloride was carried out at Drugabilis. Ten pharmaceutically acceptable counter ions were assessed and three were selected for further study

S-13
(citrate, fumarate and phosphate) (Figure 9). These alternative salt forms of compound 2 were prepared and formulated in the best vehicle determined by injection site tolerability analysis (0.3M sodium acetate buffer, Kolliphor EL, N-methyl pyrrolidone, 2-pyrrolidone, and 11% water suitable for injection). Rat IM pharmacokinetics was then carried out with each salt to determine the salt form to use for efficacy. None of these salt forms offered any improvement in rat IM pharmacokinetic profile over the free base in the tolerated vehicle ( Figure S7).

Mouse efficacy studies for T. brucei brucei.
Mouse efficacy studies for T. brucei brucei were carried out as previously detailed. 9

NMRI mouse efficacy studies for T. congolense and T. vivax.
Mouse models of infections were carried out using standardized protocols. Briefly, 1 mL samples of  and 50% deionized water). Mean blood concentration data presented in Figure S8.
Response to treatment was determined by monitoring parasitaemia twice a week until day 21.
After day 21, the mice were checked once a week until 63 d post-treatment. Mice surviving to the end of the experiment, and with blood smears clear of parasites, were considered cured.
One day after treatment, 80% of the animals in the treatment group were negative for parasites.
After 3 d, 100% of the animals in the treatment groups were negative for parasites. After 5 d, 40% of the animals were positive for parasites. After 6 d, 80% were positive and after 8 days, 100% were positive.

Suppression of parasitaemia -cattle
Friesian-Holstein calves (similar age and housing to those in 3.3) were infected on day 0. Animals were infected with fresh cattle blood from animals infected with T. congolense KON 2/133 (n = 4) or

Cattle efficacy for compound 2, intramuscular (CIRDES)
Felani-zebu cattle were housed in a fly-proof facility, acclimatised for 15 d prior to the study, cattle were treated with an anthelmintic, an ectoparasiticide and an antibiotic. Donor animals were infected with stabilates (frozen stocks) of either

Resistant line generation
Compound-resistant cell lines were generated by sub-culturing a clone of T. brucei in the continuous presence of 2. Starting at sub-lethal concentrations, drug concentrations in 3 independent cultures were increased in a step-wise manner, usually by 2-fold. When parasites were able to survive and grow in concentrations of drug equivalent to more than 10 times the established EC50 value, the resulting cell lines were cloned by limiting dilution in the presence of compound. Three clones (RES I-III) were selected for further biological study.

Whole genome sequencing and analysis
A standard alkaline lysis protocol was used to isolate genomic DNA from compound-resistant bloodstream T. brucei parasites (~5 x 10 8 ). Whole genome sequencing was performed using a HiSeq4000 next generation sequencing platform (Beijing Genomics Institute, Hong Kong).
Sequencing reads (100 bp) were aligned to the Trypanosoma brucei TREU927 genome (v39; tritrypDB) using Bowtie2 10 and Samtools 11 software. Samtools and BCFtools 12 were used to call single nucleotide polymorphisms (SNP) and indels compared with the wildtype starter clone, where S-17 overall quality score (QUAL) was >100. Artemis 13 was used to analyse chromosome and gene copy number variation (CNV), as well as visualisation of SNPs.

RIT-seq screening, Illumina sequencing and analysis
The RIT-seq library screening was performed as described previously. 14 The RNAi library was maintained with blasticidin (1 µg ml -1 ) and phleomycin (1 µg ml -1 ) at a minimum of 2 x 10 7 cells.
Following tetracycline (1 µg mL -1 ) induction for 24 h, compound 2 (4 nM), and compound 12 (98 nM), were added to cultures and supplemented with fresh compound and tetracycline as required. DNA was extracted from compound-selected cells and RNAi target fragments were amplified from compound-selected parasites by PCR using the Lib2f and Lib2r primers. 14 PCR products were fragmented and sequenced with an Illumina HiSeq platform at BGI (Beijing Genomics Institute).
Reads were mapped to the T. brucei 927 reference genome (v39; TriTrypDB [http://tritrypdb.org/]) using Bowtie 2 software, 10 with the following parameter: --very-sensitive-local. Following manipulation with SAMtools 11 the alignment files were searched with a custom script to identify reads with the following barcode: GCCTCGCGA. 14 The total and barcoded reads were then quantified using the Artemis genome browser. 15

S-18
with randomised codons at the appropriate positions shown in bold; synonymous edits that should promote insertion of randomised codons are underlined. Each template (40 µg) was delivered to the sgRNA strain by electroporation 24 h after Cas9-induction, using tetracycline at 1 µg/mL; tetracycline was removed at this point. Compound 2 selection (10 nM) was applied 6 h later. After 5 days of selection, drug-resistant cultures were subcloned and six clones were selected from each experiment. DNA was extracted from each subclone and a specific portion of the TbCRK12 gene encompassing the edited region was PCR-amplified and sequenced.

Molecular modelling
The sequence of TbCRK12 kinase domain (residues 323-685 of the Tb927.11.12310 entry in the TriTrypDB database) was used to identify homologues performing a Blast search on the Protein Data Bank proteins. The structure of the human CRK13 structure complexed with ADP (PDB ID 5EFQ), displaying a sequence identity of 35%, was selected as template for homology modelling. The sequence alignment between the Tb (target) and human (template) sequences was generate using the multisequence viewer tool in the Schrödinger modelling platform (Schrödinger 301 Release 2019-3: Schrödinger, LLC, New York, NY, 2020) and was manually curated to ensure that the highly conserved kinase catalytic residues were correctly aligned ( Figure S10). The homology model of the         S-35 S-36