Reagents and chemicals
Synthesis of the reference compound, 8-(4-methoxyphenyl)-6-(3,6,6-trimethyl-4-oxo-4,5,6,7-tetrahydro-1H-indol-1-yl)-3,4-dihydroisoquinolin-1(2H)-one (4) and precursor compound, 8-(4-hydroxyphenyl)-6-(3,6,6-trimethyl-4-oxo-4,5,6,7-tetrahydro-1H-indol-1-yl)-3,4-dihydroisoquinolin-1(2H)-one (9), was performed as described by Ernst et al. (Ernst et al. 2014b) and a detailed synthetic protocol and characterization is provided in Supplementary Material (Figure S1-S11). Hsp90 inhibitors, Onalespib and PU-H71 were purchased from commercial suppliers Bio-Connect Life Sciences, Selleckchem or MedChem Express and used without further purification. SNX-0723 ((S)-2-fluoro-6-((tetrahydrofuran-3-yl)amino)-4-(3,6,6-trimethyl-4-oxo-4,5,6,7-tetrahydro-1H-indol-1-yl)benzamide) was synthesized as previously reported (Huang et al. 2009). Chemical structures and corresponding total polar surface area (tPSA) and LogP values were drawn/calculated using ChemDraw Ultra 15.0 (Perkin Elmer).
Radiosynthesis
Carbon-11 was produced by proton irradiation of a N2 + H2 (5%) gas mixture in a Cyclone 18/9 cyclotron (IBA Louvain-la-Neuve, Belgium) and obtained as [11C]CH4 by a 14N(p,α)11C nuclear reaction. A home-built gas phase recirculation module was used for the conversion of [11C]CH4 to [11C]CH3I. The synthesis of [11C]SNX-ab was performed by bubbling of [11C]CH3I with a helium flow through a solution of the corresponding phenol SNX-ab precursor (250–300 µg) and Cs2CO3 (2.5-3 mg) dissolved in anhydrous dimethylformamide (DMF) (200–250 µL) at 100°C for 3 min. After cooling down, the crude mixture was diluted with 1.3 mL of water and the reaction mixture was purified by HPLC on a RP-C18 column (XBridge C18 column, 5 µm, 4.6 mm × 150 mm; Waters, Milford, USA) eluted with 60/30 Na2HPO4 0.01 M pH 9.3/ACN at a flow rate of 1.5 mL/min. The preparative HPLC collected fraction of [11C]SNX-ab is diluted with water and passed over a single use C18 Sep-Pak cartridge (activated with 5 ml ethanol 100% and 10 mL water). The Sep-Pak is then rinsed with 10 mL of water. [11C]SNX-ab is eluted from the Sep-Pak with approximately 1.1 mL ethanol into the final product sterile vial through a single-use sterile syringe filter (Millex-GV filter 0.22 µm, ø 13 mm Millipore, Billerica, MA). Next, a volume of approximately 11.9 mL NaCl 0.9% water for injection is rinsed over the Sep-Pak and filtered via the same filter into the final sterile vial to obtain a final EtOH concentration < 10%. A sample (100 µL) is taken for quality control (QC) analysis by HPLC on a RP-C18 column (XBridge C18 column, 3.5 µm, 3.0 mm × 100 mm; Waters, Milford, USA) eluted with 64/36 NaOAc 0.05 M pH 5.8/ACN at a flow rate of 0.6 mL/min to assess chemical and radiochemical purity. The column effluent was passed through a UV detector (254 nm) and a NaI(Tl) scintillation radioactivity detector as described in the HPLC analysis section. The identity of the tracer was confirmed by co-injection with authentic reference compound, SNX-ab, on the same HPLC system.
In vitro autoradiography
Mice were anesthetized with 2.5% isoflurane in O2 at a flow rate of 1 L/min after which they were sacrificed by decapitation. Tumour or brain tissue was excised, rinsed with saline to remove blood and rapidly frozen in cooled 2-methylbutane (-40°C). Cryotome sectioning (Shandon cryotome FSE; Thermo Fisher, Waltham, MA) was performed to obtain 20 µm-sections, which were fixed on adhesive microscope slides (Superfrost Plus; Thermo Fisher Scientific) and stored at -20°C.
The frozen tissue slices were air-dried and submerged in tris(hydroxymethyl)aminomethane hydrochloride (tris.HCl) 50 mM pH 7.4 for 10 min at room temperature. Next, the slices were air dried and pre-incubated with 200–300 µL tris.HCl 50 mM pH 7.4 + 0.3% BSA supplemented with dimethylsulfoxide (DMSO) (10%) (= control) or 10–100 µM of either SNX-ab, Onalespib, PU-H71 or SNX-0723 dissolved in DMSO (10%) for 10 min to assess binding specificity. The incubation solutions were removed and the sections were dipped in tris.HCl 50 mM pH 7.4 + 0.3% BSA at 4°C. The slices were again air-dried and subsequently incubated with [11C]SNX-ab (74 kBq/mL in 200–300 µL tris.HCl 50 mM pH 7.4 + 0.3% BSA) for 10 min at room temperature. The slices were washed twice for 5 min in tris.HCl 50 mM pH 7.4 + 0.3% BSA at 4°C with a final dip in water at 4°C. In a follow-up experiment, 6.5% EtOH was added to the washing steps to reduce possible non-specific binding to the glass microscope plates. The air-dried slices were exposed to a phosphor storage screen (super-resolution screen; Perkin Elmer, Waltham, MA) overnight. The autoradiograms were obtained by reading the screens using a Cyclone Plus system (Perkin Elmer).
Autoradiography images were analysed using Optiquant software (Perkin Elmer) and results are expressed as digital light units per square mm (DLU/mm2). Percentage block versus control was calculated as (1 - (average DLU/mm2 in the presence of 100 µM blocker)) / (average DLU/mm2 tracer only) x 100% on 3–4 tissue sections within the same experiment and expressed as mean ± SD.
In vitro cell binding studies
U87 cells were seeded at 125,000 cells/well density in the 6-well plates and incubated at 37°C for 48 h in the presence of 5% CO2. Cells were gently washed once with PBS prior to the incubation with 200 µM of either SNX-ab, Onalespib, PU-H71 or SNX-0723 blocking agents dissolved in the cell culture medium (total 1% DMSO) for 60 min at 37°C and 5% CO2 to assess binding specificity. Cell culture medium with 1% DMSO was used as a baseline control. All conditions were tested in triplicates. The cell culture media was removed and the cells were gently washed once with PBS prior to the incubation with 250 kBq/mL of [11C]SNX-ab in the cell culture medium (total 1% ethanol) for 30 min at 37°C. After incubation, cellular uptake was terminated by washing the cells three times with 1 mL of ice-cold PBS. Surface-bound tracer was removed by two consecutive incubations with 1.5 mL of glycine-HCl (50 mM, pH 2.8) for 5 min at room temperature. Cells were washed with 1 mL of ice-cold PBS and lysed by adding 400 µL of the lysis buffer (reagent A100, Chemometec, Allerod, Denmark) and gently mixing by pipetting. The lysis buffer was collected and quenched with 400 µL of the neutralization buffer (reagent B, Chemometec, Allerod, Denmark) by first washing the wells and then adding the neutralization buffer to the lysed fraction. The PBS wash, glycine-HCl washes and lysed fractions were collected separately and the radioactivity in each fraction was counted with an automated gamma counter as described above. The number of cells per well was counted using an automated counting device with nucleocasettes (NucleoCounter® NC-200™, Chemometec). Results were expressed as percentage of the applied radioactivity bound to 1×106 cells normalized to the baseline control conditions and plotted with Graphpad Prism 8.4.0 (Graphpad Software) as a mean ± SD.
Ex vivo plasma radio metabolite studies
Healthy C57BL/6 mice were anesthetized using 2.5% isoflurane in O2 at a flow rate of 1 L/min and intravenously (i.v.) injected with ~ 5 MBq of [11C]SNX-ab via a tail vein. The mice were subsequently sacrificed by decapitation at 10 min post tracer injection (n = 3) and the blood was collected in K2EDTA-containing tubes (BD vacutainer, BD, Franklin Lakes, NJ, USA) and stored on ice. The plasma was separated by centrifugation of the blood for 5 min at 2330 × g. The isolated plasma was weighed, counted in a gamma counter and spiked with authentic reference compound (10 µL of 1 mg/mL solution in DMSO). The plasma samples were analysed by RP-HPLC on a Chromolith RP-C18 column (3 mm × 100 mm, Merck, Darmstadt, Germany) eluted with gradient mixtures of CH3CN (A) and NaOAc 0.05 M pH 5.5 (B) (Table S1). After passing through an in-line UV detector at 254 nm coupled to a 3-inch NaI(Tl) scintillation detector connected to a single channel analyser, the HPLC eluent was collected in 1 mL fractions of which the radioactivity was measured in an automated gamma counter as described above.
Ex vivo biodistribution studies
Healthy C57BL/6 mice were anesthetized using 2.5% isoflurane in O2 at a flow rate of 1 L/min and i.v. injected with ~ 5.5 MBq of [11C]SNX via a tail vein. The mice were subsequently sacrificed by decapitation at 10 or 60 min post tracer injection (n = 3 per time point). All organs of interest, including blood, were collected in tared tubes and weighed. The radioactivity in each organ was counted in an automated gamma counter as described above. For the calculation of the total radioactivity in blood, muscle and bone, the masses were assumed to be respectively 7%, 40% and 12% of the total body mass (Burns et al. 2007; Horti et al. 2006; Vermeulen et al. 2019). The plasma was separated from the blood cell fraction by centrifugation of the blood for 5 min at 2330 × g. The bone marrow was isolated from bone (femur) by several washing and centrifugation (2330 × g) steps using PBS(Liu and Quan 2015; Lwin et al. 2016) (Liu and Quan 2015). Data were expressed as percentage of injected dose (%ID) and standardized uptake value (SUV) and plotted with Graphpad Prism 8.4.0 (Graphpad Software) as mean ± SD.%ID was calculated as (counts per min (cpm) in organ / total cpm recovered) × 100%. SUV was calculated as (radioactivity in cpm in organ / weight of organ in g) / (total cpm recovered / total body weight in g).
In vivo µPET imaging studies
µPET imaging studies were performed on healthy C57BL/6 mice. The mice were anesthetized using 2.5% isoflurane in O2 at a flow rate of 1 L/min before the start of the scan and kept under anesthesia during the entire scan period. Immediately after i.v. injection with ~ 4 MBq of [11C]SNX-ab via a tail vein, the mice were scanned dynamically for 90 min using Molecubes (β-CUBE), followed by a CT scan using Molecubes (X-CUBE), allowing full body field of view PET/CT imaging. To assess Hsp90-specific binding, mice (n = 2 per pre-treatment) were injected intraperitoneally (i.p.) 20 min before tracer injection with either 100–200 µl vehicle or SNX-ab (1 mg/kg) dissolved in an aqueous solution of 5% DMSO and 95% (2-hydroxypropyl)-β-cyclodextrin (40%) in H2O. Pre-treatment solutions were sterile filtered through a 0.22 µm membrane filter (Millex-GV, Millipore).
PET data were histogrammed into 17 frames (4 × 15 s, 1 × 38 s, 3 × 60 s, 1 × 180 s, 1 x 450 s, 7 × 600 s) and reconstructed using 30 iterations of the manufacturer’s MLEM algorithm with corrections for randoms, scatter, attenuation and decay into a 192 x 192 image matrix containing 0.4 mm voxels. CT data were reconstructed using a regularized statistical (iterative) image reconstruction algorithm using non-negative least squares, using an isotropic 200 µm voxel size and scaled to Hounsfield Units (HUs) after calibration against a standard air/water phantom. A manual region of interest (ROI’s) was drawn over the whole brain using PMOD software (v3.3, PMOD Technologies, Zürich, Switzerland) after normalizing for injected dose and bodyweight to generate tissue time-activity curves (TACs) scaled to Standard Uptake Value (SUV). TACs were plotted as SUV mean ± SD using Graphpad Prism 8.4.0 (Graphpad Software).