Radiosynthesis and in vivo evaluation of 1-[¹⁸F]fluoroelacridar as a positron emission tomography tracer for P-glycoprotein and breast cancer resistance protein

Abstract

Aim of this study was to label the potent dual P-glycoprotein (Pgp) and breast cancer resistance protein (BCRP) inhibitor elacridar (1) with ¹⁸F to provide a positron emission tomography (PET) radiotracer to visualize Pgp and BCRP. A series of new 1- and 2-halogen- and nitro-substituted derivatives of 1 (4a-e) was synthesized as precursor molecules and reference compounds for radiolabelling and shown to display comparable in vitro potency to 1 in increasing rhodamine 123 accumulation in a cell line overexpressing human Pgp (MDCKII-MDR1). 1-[¹⁸F]fluoroelacridar ([¹⁸F]4b) was synthesized in a decay-corrected radiochemical yield of 1.7±0.9% by a 1-step no-carrier added nucleophilic aromatic ¹⁸F-substitution of 1-nitro precursor 4c. Small-animal PET imaging of [¹⁸F]4b was performed in naïve rats, before and after administration of unlabelled 1 (5 mg/kg, n=3), as well as in wild-type and Mdr1a/b^(−/−)Bcrp1^(−/−) mice (n=3). In PET experiments in rats, administration of unlabelled 1 increased brain activity uptake by a factor of 9.5 (p=0.0002, 2-tailed Student’s t-test), whereas blood activity levels remained unchanged. In Mdr1a/b^(−/−)Bcrp1^(−/−) mice, the mean brain-to-blood ratio of activity at 60 min after tracer injection was 7.6 times higher as compared to wild-type animals (p=0.0002). HPLC analysis of rat brain tissue extracts collected at 40 min after injection of [¹⁸F]4b revealed that 93±7% of total radioactivity in brain was in the form of unchanged [¹⁸F]4b. In conclusion, the in vivo behavior of [¹⁸F]4b was found to be similar to previously described [¹¹C]1 suggesting transport of [¹⁸F]4b by BCRP and/or Pgp at the rodent BBB. However, low radiochemical yields and a significant degree of in vivo defluorination will limit the utility of [¹⁸F]4b as a PET tracer.

1. Introduction

The adenosine triphosphate (ATP) binding cassette (ABC) transporters P-glycoprotein (Pgp^a, ABCB1) and breast cancer resistance protein (BCRP, ABCG2) act as gatekeepers at the level of the vascular endothelium of the blood-brain barrier (BBB) preventing brain uptake of a wide range of lipophilic molecules by active ATP-driven efflux transport ¹. Changes in ABC transporter function and expression are believed to be involved in several neurological disorders, such as therapy refractory epilepsy, Parkinson’s and Alzheimer’s disease ¹.

Visualization of ABC transporters with the non-invasive nuclear imaging method positron emission tomography (PET) might help to better understand disease related changes in transporter function/expression ². Two classes of radiotracers for visualization of Pgp have been described so far. Radiolabelled Pgp substrates such as (R)-[¹¹C]verapamil ³ or [¹¹C]-N-desmethyl-loperamide ⁴ and radiolabelled Pgp inhibitors such as [¹¹C]laniquidar ⁵, [¹¹C]elacridar ^6-7, [¹¹C]tariquidar ^8-9 and [¹¹C]MC18 ¹⁰ (Fig. 1). Radiolabelled Pgp inhibitors were initially developed to provide PET tracers which should bind to Pgp without being transported and therefore allow for visualization of Pgp expression levels as opposed to substrates which visualize Pgp function. However, the in vivo behavior of most radiolabelled Pgp inhibitors tested to date in rodents was found to be “substrate-like” in that brain activity uptake in baseline scans was equally low or even lower than for radiolabelled substrates and several times increased after administration of unlabelled inhibitor. Our recent experiments with [¹¹C]elacridar ([¹¹C]1, Fig. 1) in rats and transporter knockout and wild-type mice suggested that this radiotracer was efficiently transported by Pgp and/or BCRP at the rodent BBB ⁶. Given its favorable metabolic profile with radiometabolites not being taken up into brain tissue, [¹¹C]1 ^6-7 appeared as an interesting candidate probe to investigate Pgp and BCRP. Due to the short radioactive half-life of ¹¹C (20.4 min) the use of ¹¹C-labelled PET tracers is restricted to centers with an onsite cyclotron. In order to facilitate a broader applicability of this radiotracer we aimed at labelling 1 with the longer-lived radionuclide ¹⁸F (half-life: 109.8 min).

Figure 1.

Chemical structures of Pgp substrate and inhibitor PET tracers.

Here we report on the synthesis of a series of new halogen- and nitro-substituted derivatives of 1, which we found to possess comparable potency to 1 in inhibiting Pgp transport in vitro. Moreover, we report on the radiosynthesis and small-animal PET evaluation of 1-[¹⁸F]fluoroelacridar ([¹⁸F]4b), which we showed to display similar in vivo behavior as [¹¹C]1.

2. Results

2.1. Chemistry and radiolabelling

The para- and the meta position relative to the carboxylamide group in the acridone moiety of 1 were selected for no-carrier-added nucleophilic aromatic substitution with [¹⁸F]F⁻. 1-fluoroelacridar (4b) and 2-fluoroelacridar (4e) were synthesized as shown in Scheme 1 ^{6, 11}. As radiolabelling precursors, 1- and 2-chloro substituted derivatives 4a and 4d as well as 1-nitro derivative 4c were synthesized according to Scheme 1 ^{6, 11}. In small-scale experiments precursor molecules 4a, 4c and 4d were reacted with the K[¹⁸F]F-K₂₂₂ complex using different solvents (DMF, DMSO), different reaction temperatures (120-180°C) and different reaction times (5-60 min). Derivatives 4a and 4d failed to provide any ¹⁸F-incorporation under all conditions tested, whereas 4c gave the corresponding 1-[¹⁸F]fluoro-derivative [¹⁸F]4b in an incorporation yield of 10-20% based on radio-thin layer chromatography (radio-TLC) analysis of crude reaction mixture. The synthesis of [¹⁸F]4b was automated in a TRACERlab FX synthesis module. [¹⁸F]4b, ready for intravenous (i.v.) injection, was obtained in a decay-corrected radiochemical yield of 1.7±0.9% (n=7) based on [¹⁸F]F⁻ in a total synthesis time of 147 min. Radiochemical purity was greater than 98% and the specific activity at the end of synthesis was 8±4 GBq/μmol (n=7). The identity of [¹⁸F]4b was confirmed by high-performance liquid chromatography (HPLC) coinjection with unlabelled reference compound 4b.

Scheme 1.

Synthesis of nitro- and halogen-substitued derivatives of 1 (4a-e) and ¹⁸F-labelling of 4c to give [¹⁸F]4b. Reagents: (A) Cu⁰/K₂CO₃, EtOH, reflux; (B) POCl₃, CH₃CN, reflux; (C) TBTU, Et₃N, DMF, rt; (D) K[¹⁸F]F-K₂₂₂, DMSO, 175°C.

2.2. Assessment of Pgp inhibitory activity

Pgp inhibition of compounds 4a-e was determined relative to reference compound 1 using the rhodamine 123 (R123) accumulation assay in a cell line which overexpressed human Pgp (MDCKII-MDR1) (Table 1). Four compounds displayed comparable potency to 1 in inhibiting R123 efflux with half-maximum inhibitory concentrations (IC₅₀) in the 0.5-2 μM range (4b-e). For instance 1-fluoro-derivative 4b, which was also tested in vivo in ¹⁸F-labelled form (see below), had an only about 4 times higher IC₅₀ than 1. It is noteworthy that all derivatives were less effective than 1 in terms of increasing intracellular R123 concentration in MDCKII-MDR1 cells with maximal inhibitory effects relative to 1 <100% (Table 1).

Table 1.

Effects of compound 1 and the new modulators 4a-e on R123 efflux from MDCKII-MDR1 cell monolayers

compound	IC50 (μM)	logIC50a	Maximal inhibitory effect (%)b
1 (elacridar)	0.4	−0.36±0.06	100
4a	3.8	0.58±0.04	74±3
4b	1.7	0.24±0.04	81±4
4c	1.6	0.21±0.07	50±4
4d	1.4	0.14±0.08	88±5
4e	0.6	−0.21±0.08	69±1

^alogIC₅₀ values (fitted parameter±standard error) were estimated using GraphPad Prism software.

^bMaximal inhibitory effects (%) are expressed as inhibition caused by the highest concentration of the compound tested (4b, 4d, 4e: 10 μM; 4a, 4c: 100 μM) relative to the inhibitory effect caused by 3.2 μM 1 (100% inhibition).

2.3. Small-animal PET in rats and mice

[¹⁸F]4b was evaluated in two groups of rats. The first group underwent a 150 min PET scan, during which unlabelled 1 was i.v. administered at a dose of 5 mg/kg at 60 min after injection of [¹⁸F]4b. The second group of rats was scanned at 2 h after administration of 1. In the first group of rats, brain uptake of activity was low before unlabelled 1 was administered (Fig. 2A). Peak brain uptake was 0.46±0.12 (standardized uptake value, SUV) at 0.9-1.3 min after tracer injection. At 25 min after tracer injection, brain activity uptake had declined to 0.14±0.02 SUV. In response to administration of unlabelled 1 at 60 min after injection of [¹⁸F]4b there was a steep increase in brain activity concentration, reaching a SUV of 1.14±0.14 at 140 min after tracer injection (Fig. 2A), whereas blood activity levels remained unchanged (Fig. 2B). In group 2, which was scanned at 2 h after administration of unlabelled 1, brain activity uptake at 25 min after tracer injection was 9.5 times higher (p=0.0002) as compared to group 1 (Fig. 2A) with similar blood activity levels as in group 1 (Fig. 2B). In Fig. 3, representative PET summation images are shown for scans recorded before and after inhibitor administration. There was appreciable uptake of radioactivity in bone tissue which pointed to in vivo defluorination of [¹⁸F]4b (Fig. 3). In the first group of rats, activity uptake in humerus was 1.42±0.37 and 4.59±1.00 SUV at 1.8 and 140 min after injection of [¹⁸F]4b, respectively.

Figure 2.

TACs (mean SUV±SD) of [¹⁸F]4b in whole brain (A) and arterial blood (B) of rats (group 1: open squares, group 2: filled squares). In group 1, unlabelled 1 was administered as an i.v. bolus over 60 s at 60 min after injection of [¹⁸F]4b. The time point of administration of 1 is indicated by an arrow. Group 2 was scanned at 2 h after administration of 1. For comparison, whole-brain TACs (mean SUV±SD, n=3) previously obtained with [¹¹C]1 ⁶ are also shown in A (group 1: open circles, group 2: filled circles).

Figure 3.

Exemplary coronal (left), horizontal (center) and sagittal (right) PET summation images in rats, (A) before (group 1: 0-60 min) and (B) after administration of unlabelled 1 (5 mg/kg). The radiation scale is set from 0.1 to 2.0 SUV. Anatomical structures are indicated by arrows (br, brain; e, eye; h, humerus; sg, submandibular gland; pg, pituitary gland).

Radiometabolites of [¹⁸F]4b in plasma were assessed in a separate group of rats (n=3) by a solid-phase extraction (SPE) assay. At 10, 20, 30 and 40 min after injection of [¹⁸F]4b into rats, 6±4%, 6±3%, 12±5% and 13±4% of total plasma activity was recovered in SPE fractions 1 and 2 (corresponding to polar radiometabolites of [¹⁸F]4b) and 94±4%, 94±3%, 88±5% and 87±4% in fraction 3 (corresponding to unchanged [¹⁸F]4b and its lipophilic radiometabolites). Fraction 3 from the 40 min sample was further analyzed by HPLC, which showed that 72±27% of total radioactivity was in the form of unchanged [¹⁸F]4b, whereas the remainder of radioactivity eluted before [¹⁸F]4b on reversed-phase HPLC. [¹⁸F]4b was 98.0±1.7% (n=3) protein-bound in rat plasma. Brain extracts collected at 40 min after injection of [¹⁸F]4b into rats were also analyzed by HPLC showing that 93±7% of total radioactivity in brain was in the form of unchanged [¹⁸F]4b.

In Fig. 4, PET summation images and brain time-activity curves (TACs) of [¹⁸F]4b in wild-type and Mdr1a/b^(−/−)Bcrp1^(−/−) mice are shown. Brain TACs were several times higher in Mdr1a/b^(−/−)Bcrp1^(−/−) mice as compared to wild-type animals (Fig. 4B). Brain activity uptake at 60 min after tracer injection was normalized to blood activity levels measured after the PET scan. The mean brain-to-blood ratio of activity at 60 min after tracer injection was 7.6 times higher (p= 0.0002) in Mdr1a/b^(−/−)Bcrp1^(−/−) mice than in wild-type animals (brain-to-blood ratios, wild-type: 1.5±0.2; Mdr1a/b^(−/−)Bcrp1^(−/−): 11.6±1.4).

Figure 4.

(A) Sagittal PET summation images (0-60 min) of [¹⁸F]4b in wild-type and Mdr1a/b^(−/−)Bcrp1^(−/−) mice. The radiation scale is set from 0.1 to 2.0 SUV. Anatomical structures are indicated by arrows (br, brain). (B) TACs (mean SUV±SD, n=3 per mouse type) of [¹⁸F]4b in whole brain of wild-type (open squares), and Mdr1a/b^(−/−)Bcrp1^(−/−) mice (filled squares). For comparison TACs (mean SUV±SD, n=3) previously obtained with [¹¹C]1 ²⁴ are also shown (wild-type mice: open circles, Mdr1a/b^(−/−)Bcrp1^(−/−) mice: filled circles).

3. Discussion

The aim of this study was to label the potent dual Pgp/BCRP inhibitor 1 with the positron emitting radionuclide ¹⁸F in order to develop a longer-lived PET radiotracer, based on 1, than previously described [¹¹C]1 ^6-7 to investigate Pgp and BCRP in vivo. During the preparation of this manuscript a study by Kawamura et al. has appeared in the literature, which also reported the synthesis of an ¹⁸F-derivative of 1 (i.e. the corresponding 5-[¹⁸F]fluoroethoxy derivative), but this radiotracer was labelled in a different position than in our work ¹². We found that the acridone moiety of 1 gave straightforward access to the introduction of appropriate leaving groups (Cl, NO₂) in the 1- and 2-positions for ¹⁸F-substitution (Scheme 1). Derivatives 4a-e were prepared as reference compounds and radiolabelling precursors for ¹⁸F-fluorination of 1 (Scheme 1). As 4a-e are previously unknown close structural analogues of 1 they were tested for their Pgp inhibitory potencies in the R123 accumulation assay (Table 1). All compounds except for 1-chloro-derivative 4a were remarkably potent in the Pgp inhibition assay with IC₅₀ values in the 0.5-2.0 μM range, which was comparable to 1 (IC₅₀: 0.4 μM). As the primary imaging target was Pgp, BCRP modulation of derivatives 4a-e was not tested.

Radiolabelling experiments revealed that only the 1-position of the acridone moiety was moderately activated for ¹⁸F-substitution. 1-[¹⁸F]fluoroelacridar ([¹⁸F]4b) was synthesized in low radiochemical yield of 1.7±0.9% based on starting [¹⁸F]F⁻ employing rather drastic reaction conditions (60 min heating at 175°C) (Scheme 1). Surprisingly, the specific activity of [¹⁸F]4b at end of synthesis (8±4 GBq/μmol) was lower than expected for a typical no-carrier added nucleophilic aromatic ¹⁸F-substitution reaction (>37 GBq/μmol). The most likely explanation for the low specific activity of [¹⁸F]4b was that elimination of the NO₂ group in precursor molecule 4c had occurred under the employed labelling conditions giving rise to unlabelled 1, which we failed to separate from 4b on all tested HPLC systems. The obtained amounts of [¹⁸F]4b were sufficiently high to allow for conducting small-animal PET experiments.

The primary aim of the PET examination was to assess if in vivo [¹⁸F]4b behaved similarly to previously developed [¹¹C]1. Therefore we chose a similar study set-up as we had previously used for [¹¹C]1, which was studied in naïve rats by performing paired PET scans before and after administration of unlabelled 1 ⁶. However, due to the longer half-life of ¹⁸F it was not possible to perform paired PET scans in the same animals with [¹⁸F]4b and therefore separate groups of rats were examined. As previously observed for [¹¹C]1 ⁶ the in vivo behavior of [¹⁸F]4b was consistent with that of a transported substrate (Fig. 2A). Brain activity uptake of [¹⁸F]4b was low in the first group of rats, which was scanned before administration of unlabelled 1. In response to administration of 1 there was a steep increase in brain activity uptake (Fig. 2A). In the second group of rats, which was scanned after administration of 1, brain activity uptake was 9.5 times higher as compared to group 1. It seems unlikely that the rather low specific activity of [¹⁸F]4b should have significantly affected the PET results. Kawamura and co-workers have previously studied uptake of [¹¹C]1 into mouse brain in the presence of increasing amounts of unlabelled 1 (0.1-10 mg/kg, i.v.) and found that brain uptake of [¹¹C]1 started to increase at doses greater than 0.1 mg/kg with a half-maximum effect dose of 1.55 mg/kg ⁷. As the doses of unlabelled carrier administered in our in vivo experiments in rats and mice were <0.05 mg/kg the relatively low specific activity of [¹⁸F]4b should not have influenced its brain uptake. As a second approach to the rat studies, in which Pgp/BCRP was pharmacologically inhibited with unlabelled 1, we performed PET experiments with [¹⁸F]4b in wild-type and in transporter knockout mice, which lacked both Pgp and Bcrp1. Absence of Pgp/Bcrp1 at the BBB was reflected by a 7.6 times higher brain-to-blood ratio of activity as compared to wild-type mice. Overall, the in vivo behavior of [¹⁸F]4b was almost identical to that of [¹¹C]1 as reflected by quite similar TACs of both radiotracers in rats and mice (Fig. 2A, Fig. 4).

Our in vivo results obtained with [¹⁸F]4b are to a certain extent surprising as 1 has been previously characterized as a non-transported inhibitor of Pgp. It has been shown that 1 does not stimulate adenosine triphosphatase (ATPase) activity and has an efflux ratio that is <2 in cell monolayer transport assays, consistent with a lack of Pgp transport activity ^13-14. However, very recent data suggest that the structurally related Pgp inhibitor tariquidar (Fig. 1), which we have also labelled with ¹¹C and which was shown to behave similarly to [¹¹C]1 and [¹⁸F]4b in vivo ⁸, is an avid substrate of BCRP ¹⁵. Kannan and co-workers found that accumulation of ³H-labelled tariquidar (1 nM) was several fold-lower in a cell line overexpressing human BCRP relative to the parental cell line and increased to similar levels as in parental cells after co-incubation with the BCRP inhibitor fumitremorgin C, which was consistent with BCRP transport of tariquidar ¹⁵. On the other hand, when a cell line which overexpressed human Pgp, was incubated with [³H]tariquidar, cellular uptake was higher relative to parental cells, which pointed to binding of [³H]tariquidar to Pgp. In addition, recent PET results obtained with [¹¹C]tariquidar and [¹¹C]1 in Mdr1a/b^(−/−), Bcrp1^(−/−) and Mdr1a/b^(−/−)Bcrp1^(−/−) mice suggest that the two radiotracers are transported by Bcrp1 and possibly Pgp at the murine BBB ^7-8. Given the close structural similarity between 1 and tariquidar and the similar in vivo behavior of [¹¹C]1 and [¹⁸F]4b on one hand and [¹¹C]tariquidar on the other hand it seems very likely that the “substrate-like” behavior of [¹⁸F]4b was caused by BCRP and possibly Pgp transport. This is also in good agreement with the observation that pharmacological doses of 1, which are known to inhibit BCRP and Pgp transport at the BBB ¹⁶, significantly increased brain uptake of [¹⁸F]4b in rats. As no radiotracers have been described to date to measure BCRP transport in vivo it definitely appears worthwhile to further characterize PET tracers based on 1 and tariquidar with respect to such a future application. Moreover, dual Pgp/BCRP substrate radiotracers could help to better understand the interplay of these two transporters at the BBB. This class of radiotracers appears particularly attractive as it has been shown that they lack brain uptake of radiometabolites ^{7, 9} which has been found to be a considerable limitation of the Pgp substrate probes (R)-[¹¹C]verapamil ¹⁷ and [¹¹C]-N-desmethyl-loperamide ¹⁸. In line with previous results obtained with [¹¹C]1 ^6-7 and [¹¹C]tariquidar ^8-9, we could show that >90% of total activity in rat brain was due to unchanged [¹⁸F]4b. However, a considerable drawback of [¹⁸F]4b was the significant degree of in vivo defluorination observed during the time course of the PET experiments as reflected by the high amount of radioactivity in bone tissue which would hamper the quantification of [¹⁸F]4b uptake in organs with low uptake which are surrounded by bone structures.

4. Material and methods

4.1. General

Unless otherwise stated, all chemicals were purchased from Sigma-Aldrich Chemie GmbH (Schnelldorf, Germany), TCI Europe (Zwijndrecht, Belgium), Merck (Darmstadt, Germany) or Apollo Scientific Ltd (Bredbury, UK) at analytical grade and used without further purification. Isoflurane was obtained from Baxter Vertriebs GmbH (Vienna, Austria). The hydrochloride salt of 1 was obtained from Glaxo SmithKline (Research Triangle Park, NC, USA). For administration 1 hydrochloride was dissolved freshly on each experimental day in a 20% aqueous (aq.) EtOH solution and injected at a volume of 2 mL/kg. Aqueous [¹⁸F]F⁻ was produced in a General Electrics PETtrace cyclotron (General Electric Healthcare, Uppsala, Sweden) via the ¹⁸O(p,n)¹⁸F nuclear reaction by irradiation of a 2 mL water target containing 95.9% enriched [¹⁸O]H₂O (Hyox18, Rotem Industries, Beer Sheva, Israel) with 16.5 MeV proton beam. Typical irradiation times were 60 min with a beam current of 40 μA, which yielded a [¹⁸F]F⁻ amount of about 110 GBq at end of bombardment. ¹H- and ¹³C-NMR spectra were recorded on a Bruker Advance DPx200 (200 and 50 MHz). Chemical shifts are reported in δ units (ppm) relative to Me₄Si line as internal standard (s, bs, d, m, Cq for singlet, broad singlet, doublet, multiplet and quarternary carbon, respectively) and J values are reported in Hertz. Mass spectra (MS) were obtained with a Shimadzu (GC-17A; MS-QP5050A) spectrometer. Purity of the intermediates 2a-e and 3a-e was established by combustion analysis with a Perkin-Elmer 2400 CHN elemental analyzer and of the key compounds 4a-e by analytical HPLC confirming a purity >95%.

4.2. General synthesis procedure for compounds 2a-e

A suspension of 2-amino-3-methoxybenzoic acid (1.67 g, 10 mmol), 2-bromobenzoic acid derivative (11 mmol, 1.1 equivalents, eq.), K₂CO₃ (2.77 g, 20 mmol, 2 eq.) and copper powder (0.125 g, 2 mmol, 0.2 eq.) was stirred in EtOH (20 mL) and heated to reflux for 1.5 h. The suspension was cooled to room temperature and H₂O (20 mL) was added. The mixture was filtered on cellite to remove the copper. The filter bed was washed with H₂O and the resulting solution was acidified with concentrated HCl to a pH of 2-3. The resulting suspension was stirred for 1 h at 10°C, the solid was filtered and washed with H₂O. The product was recrystallized in EtOH/H₂O and dried under vacuum.

4.2.1. 2-((2-carboxy-5-chlorophenyl)amino)-3-methoxybenzoic acid (2a)

The compound was synthesized according to the general procedure using 2-bromo-4-chlorobenzoic acid (2.59 g, 11 mmol, 1.1 eq.) as 2-bromobenzoic acid derivative. Yield: 2.95 g (92%) white crystals. mp: 250-251 °C. MS m/z: 323 (23%, M⁺, ³⁷Cl), 321 (62%, M⁺, ³⁵Cl), 258 (100%), 230 (74%), 74 (58%). ¹H NMR (DMSO-d₆) δ 3.79 (s, 3H, OCH₃), 6.27 (s, 1H), 6.77 (d, 1H, J = 8.4 Hz), 7.27-7.44 (m, 2H), 7.45-7.55 (m, 1H), 7.85 (d, 1H, J = 8.4 Hz), 10.08 (s, 1H), 13.17 (bs, 2H). ¹³C NMR (DMSO-d₆) δ 55.7 (OCH₃), 111.4 (Cq), 113.6 (CH), 115.8 (CH), 116.7 (CH), 122.3 (CH), 125.4 (CH), 127.0 (Cq), 128.0 (Cq), 132.8 (CH), 138.1 (Cq), 148.1 (Cq), 153.6 (Cq), 167.7 (COOH), 168.6 (COOH).

4.2.2. 2-((2-carboxy-5-fluorophenyl)amino)-3-methoxybenzoic acid (2b)

The compound was synthesized according to the general procedure with 2-bromo-4-fluorobenzoic acid (2.41 g, 11 mmol, 1.1 eq.) as 2-bromobenzoic acid derivative. Yield: 3.00 g (98%) pale white crystals. mp: 206-209 °C. MS m/z: 305 (49%, M⁺), 242 (83%), 213 (100%), 184 (35%), 93 (47%). ¹H NMR (DMSO-d₆) δ 3.78 (s, 3H, OCH₃), 5.94-6.03 (m, 1H), 6.46-6.60 (m, 1H), 7.27-7.39 (m, 2H), 7.40-7.49 (m, 1H), 7.84-7.96 (m, 1H), 10.11 (s, 1H), 13.04 (bs, 2H). ¹³C NMR (DMSO-d₆) δ 55.7 (OCH₃), 100.1 (d, CH, J = 25.9 Hz), 104.0 (d, CH, J = 22.2 Hz), 109.2 (Cq), 115.7 (CH), 122.2 (CH), 125.4 (CH), 127.1 (Cq), 127.9 (Cq), 133.8 (d, CH, J = 11.5), 149.6 (d, Cq, J = 12.2), 153.6 (Cq), 165.4 (d, CF, J = 246.1 Hz), 167.7 (COOH), 168.6 (COOH).

4.2.3. 2-((2-carboxy-5-nitrophenyl)amino)-3-methoxybenzoic acid (2c)

The compound was synthesized according to the general procedure with 2-bromo-4-nitrobenzoic acid (2.71 g, 11 mmol, 1.1 eq.) as 2-bromobenzoic acid derivative. Yield: 3.19 g (96%) orange crystals. mp: 279-284 °C. MS m/z: 332 (67%, M⁺), 269 (67%), 89 (66%), 75 (100%), 51 (75%). ¹H NMR (DMSO-d₆) δ 3.78 (s, 3H, OCH₃), 7.02-7.07 (m, 1H), 7.28-7.47 (m, 2H), 7.48-7.57 (m, 2H), 8.08 (d, 1H, J = 8.8 Hz), 10.35 (bs, 1H), 13.43 (bs, 2H). ¹³C NMR (DMSO-d₆) δ 55.8 (OCH₃), 109.2 (CH), 110.6 (CH), 116.0 (CH), 117.6 (Cq), 122.6 (CH), 125.5 (CH), 126.3 (Cq), 127.8 (Cq), 132.6 (CH), 147.1 (Cq), 150.2 (Cq), 153.2 (Cq), 167.7 (COOH), 168.0 (COOH).

4.2.4. 2-((2-Carboxy-4-chlorophenyl)amino)-3-methoxybenzoic acid (2d)

The compound was synthesized according to the general procedure using 2-bromo-5-chlorobenzoic acid (2.57 g, 11 mmol) as 2-bromobenzoic acid derivative ¹⁹. Yield: 2.80 g (87%) ochre crystals. mp: 225-229°C. MS m/z: 323 (24%, M⁺, ³⁷Cl), 321 (63%, M⁺, ³⁵Cl), 260 (35%), 258 (100%), 230 (68%), 44 (85%). ¹H-NMR (DMSO-d₆): δ 3.75 (s, 3H, OCH₃), 6.35 (d, 1H, J = 9 Hz), 7.15-7.39 (m, 3H), 7.40-7.52 (m, 1H), 7.75-7.79 (m, 1H), 10.16 (bs, 1H), 12.97 (bs, 2H). ¹³C-NMR (DMSO-d₆): δ 55.6 (OCH₃), 113.9 (Cq), 115.6 (CH), 116.7 (CH), 120.0 (Cq), 122.2 (CH), 124.7 (CH), 126.4 (Cq), 128.6 (Cq), 129.7 (CH), 132.7 (CH), 145.7 (Cq), 153.4 (Cq), 167.8 (COOH), 168.0 (COOH).

4.2.5. 2-((2-Carboxy-4-fluorophenyl)amino)-3-methoxybenzoic acid (2e)

The compound was synthesized according to the general procedure with 2-bromo-5-fluorobenzoic acid (2.41 g, 11 mmol) as 2-bromobenzoic acid derivative ¹⁹. Yield: 2.84 g (93%) light yellow crystals. mp: 209-214°C. MS m/z: 305 (58%, M⁺), 242 (100%), 214 (95%), 83 (35%), 57 (53%). ¹H-NMR (DMSO-d₆): δ 3.74 (s, 3H, OCH₃), 6.30-6.41 (m, 1H), 7.10-7.35 (m, 3H), 7.42-7.59 (m, 2H), 9.93 (bs, 1H), 13.12 (bs, 2H). ¹³C-NMR (DMSO-d₆): δ 55.5 (OCH₃), 113.3 (d, Cq, J = 6.4 Hz), 115.6 (CH), 115.7 (d, CH, J = 22.8 Hz), 116.5 (d, CH, J = 7.3 Hz), 120.4 (d, CH, J = 22.8 Hz), 122.3 (CH), 124.1 (CH), 125.7 (Cq), 129.5 (Cq), 143.4 (d, Cq), J = 1.3 Hz), 153.3 (Cq), 153.6 (d, Cq, J = 232.5 Hz), 167.9 (COOH), 168.1 (d, COOH, J = 2.5 Hz).

4.3. General synthesis procedure for compounds 3a-e

Compounds 2a-e (5 mmol) were dissolved in CH₃CN (15 mL) and heated to reflux. Phosphorus(V)oxychloride (1 mL, 11 mmol, 2.2 eq.) was added over a period of 1 h. The solution was refluxed for further 2 h and then cooled to 10-15°C. H₂O (10 mL) was added and the mixture heated to reflux for additional 2.5 h. The suspension was cooled to 10°C and filtered. The solid was washed with H₂O and CH₃CN and then dried under vacuum.

4.3.1 1-Chloro-5-methoxyacridone-4-carboxylic acid (3a)

Cyclization of compound 2a (1.61 g, 5 mmol) yielding the title compound was carried out using the general approach. Yield: 1.12 g (73%) green solid. mp: 315-320 °C. MS m/z: 305 (25%, M⁺, ³⁷Cl), 303 (75%, M⁺, ³⁵Cl), 257 (66%), 151 (88%), 62 (100%). ¹H NMR (DMSO-d₆) δ 4.03 (s, 3H, OCH₃), 7.13-7.36 (m, 3H), 7.67 (d, 1H, J = 8.2 Hz), 8.24 (d, 1H, J = 8.2 Hz), 12.5 (s, 1H). ¹³C NMR (DMSO-d₆) δ 56.5 (OCH₃), 112.9 (CH), 113.5 (Cq), 116.9 (CH), 117.3 (Cq), 122.1 (CH), 122.1 (Cq), 123.1 (CH), 129.6 (Cq), 136.0 (CH), 139.6 (Cq), 142.8 (Cq), 147.2 (Cq), 168.8 (COOH), 175.4 (CO).

4.3.2. 1-Fluoro-5-methoxyacridone-4-carboxylic acid (3b)

Cyclization of compound 2b (1.53 g, 5 mmol) to 3b was carried out using the general approach. Yield: 1.16 g (81%) green/brown solid. mp: 275-278 °C. MS m/z: 287 (100%, M⁺), 269 (74%), 254 (67%), 241 (71%), 170 (37%). ¹H NMR (DMSO-d₆) δ 4.03 (s, 3H, OCH₃), 6.92-7.07 (m, 1H), 7.15-7.28 (m, 1H), 7.34 (d, 1H, J = 7.8 Hz), 7.67 (d, 1H, J = 7.8 Hz), 8.31-8.42 (m, 1H), 12.39 (s, 1H). ¹³C NMR (DMSO-d₆) δ 56.5 (OCH₃), 107.0 (d, CH, J = 21,6 Hz), 110.7 (d, Cq, J = 9.0 Hz), 110.7 (d, Cq, J = 4.2 Hz), 113.0 (CH), 116.6 (CH), 122.1 (Cq), 138.2 (d, CH, J = 13.2 Hz), 142.9 (d, Cq, J = 5.2 Hz), 147.3 (Cq), 165.0 (d, CF, J = 268.6 Hz), 168.6 (COOH), 174.8 (d, CO, J = 1.7 Hz).

4.3.3. 5-Methoxy-1-nitroacridone-4-carboxylic acid (3c)

Cyclization of compound 2c (1.66 g, 5 mmol) yielding the title compound was carried out using the general approach. Yield: 1.29 g (82%) brown solid. mp: 284-290 °C. MS m/z: 314 (31%, M⁺), 313 (34%), 121 (47%), 57 (100%), 51 (55%). ¹H NMR (DMSO-d₆) δ 4.02 (s, 3H, OCH₃), 7.19-7.41 (m, 2H), 7.51 (d, 1H, J = 8.0 Hz), 7.64 (d, 1H, J = 7.8 Hz), 8.50 (d, 1H, J = 8.0 Hz), 12.49 (s, 1H). ¹³C NMR (DMSO-d₆) δ 56.5 (OCH₃), 111.2 (Cq), 113.5 (CH), 114.1 (CH), 116.5 (CH), 117.3 (Cq), 121.0 (Cq), 122.7 (CH), 130.1 (Cq), 137.2 (CH), 141.1 (Cq), 147.4 (Cq), 151.6 (Cq), 167.9 (COOH), 173.4 (CO).

4.3.4. 2-Chloro-5-methoxyacridone-4-carboxylic acid (3d)

Cyclization of compound 2d (1.61 g, 5 mmol) to 3d was carried out using the general approach ¹⁹. Yield: 1.21 g (79%) dark green solid. mp: 309-314°C. MS m/z: 305 (35%, M⁺, ³⁷Cl), 303 (100%, M⁺, ³⁵Cl), 285 (85%), 270 (52%), 257 (76%). ¹H-NMR (d₆-DMSO): δ 4.05 (s, 3H, OCH₃), 7.19-7.43 (m, 2H), 7.69-7.77 (m, 1H), 8.25-8.38 (m, 2H), 12.07 (s, NH). ¹³C-NMR (d₆-DMSO): δ 56.5 (OCH₃), 113.2 (CH), 116.8 (CH), 117.3 (Cq), 120.7 (Cq), 122.1 (CH), 122.6 (Cq), 124.5 (Cq), 130.6 (Cq), 130.7 (CH), 135.7 (CH), 138.8 (Cq), 147.4 (Cq), 167.8 (COOH), 175.1 (CO).

4.3.5. 2-Fluoro-5-methoxyacridone-4-carboxylic acid (3e)

Cyclization of compound 2e (1.53 g, 5 mmol) yielding the title compound was carried out using the general approach ¹⁹. Yield: 1.22 g (85%) ochre solid. mp: >350°C. MS m/z: 287 (100%, M⁺), 269 (58%), 254 (63%), 241 (67%), 92 (51%). ¹H-NMR (d₆-DMSO): δ 4.05 (s, 3H, OCH₃), 7.18-7.41 (m, 2H), 7.68-7.77 (m, 1H), 8.08-8.23 (m, 2H), 12.04 (s, NH). ¹³C-NMR (d₆-DMSO): δ 56.5 (OCH₃), 112.9 (CH), 116.6 (d, CH, J = 22.5 Hz), 116.7 (CH), 117.3 (d, Cq, J = 6.5 Hz), 120.0 (Cq), 121.8 (CH), 122.5 (d, Cq, J = 6.0 Hz), 124.1 (d, CH, J = 25.7 Hz), 130.7 (Cq), 137.1 (Cq), 147.3 (Cq), 155.3 (d, CF, J = 239.0 Hz), 167.8 (d, COOH, J = 2.3 Hz), 175.4 (d, CO, J = 2.6 Hz).

4.4. General synthesis procedure for key compounds 4a-e

To a solution of 6,7-dimethoxy-2-(4-aminophenethyl)-1,2,3,4-tetrahydroisochinoline (0.312 g, 1 mmol), which has been prepared as described elsewhere ⁶, and compound 3a-e (1 mmol, 1 eq.) in DMF (2 mL) O-(benzotriazol-1-yl)-N,N,N’,N’-tetramethyluronium hexafluorophosphate (HBTU, 0.379 g, 1.05 mmol, 1.05 eq.) and Et₃N (0.29 mL, 2.1 mmol, 2.1 eq.) were added. The solution was stirred for 2 h at rt, and then a mixture of H₂O and isopropyl alcohol (1/1, v/v, 2 mL) was added. The mixture was stirred for additional 30 min. The obtained slurry was filtered, washed with MeOH and H₂O, and dried under vacuum.

4.4.1 1-Chloro-N-(4-(2-(6,7-dimethoxy-1,2,3,4-tetrahydroisoquinolin-2-yl)ethyl)phenyl)-5-methoxyacridone-4-carboxamide (1-chloroelacridar, 4a)

The compound was synthesized according to the general procedure with 3a (0.304 g, 1 mmol) as reactant. Yield: 0.335 g (56%) ochre solid. mp: 183-189 °C. HRMS (ESI/MS) calcd for C₃₄H₃₂O₅N₃ClH: 598.2109 found: 598.2102. ¹H NMR (DMSO-d₆) δ 2.69-2.91 (m, 8H, 4xCH₂), 3.58 (s, 2H, CH₂), 3.70 (s, 6H, 2xOCH₃), 4.02 (s, 3H, OCH₃), 6.61-6.69 (m, 2H), 7.17-7.43 (m, 5H), 7.64 (d, 2H, J = 8.0 Hz), 7.74 (d, 1H, J = 8.0 Hz), 8.31 (d, 1H, J = 8.0 Hz), 10.63 (s, 1H), 12.66 (s, 1H).

4.4.2. N-(4-(2-(6,7-dimethoxy-1,2,3,4-tetrahydroisoquinolin-2-yl)ethyl)phenyl)-1-fluoro-5-methoxyacridone-4-carboxamide (1-fluoroelacridar, 4b)

Starting from 3b (0.287 g, 1 mmol) the title compound was synthesized according to the general approach. Yield: 0.337 g (58%) brown solid. mp: 215-220 °C. HRMS (ESI/MS) calcd for C₃₄H₃₂O₅N₃FH: 582.2404 found: 582.2414. ¹H NMR (DMSO-d₆) δ 2.63-3.10 (m, 8H, 4xCH₂), 3.65 (s, 2H, CH₂), 3.69 (s, 6H, 2xOCH₃), 6.59-6.73 (m, 2H), 7.02-7.22 (m, 2H), 7.24-7.35 (m, 3H), 7.58-7.75 (m, 3H), 8.36-8.52 (m, 1H), 10.58 (s, 1H), 12.65 (s, 1H).

4.4.3. N-(4-(2-(6,7-dimethoxy-1,2,3,4-tetrahydroisoquinolin-2-yl)ethyl)phenyl)-5-methoxy-1-nitroacridone-4-carboxamide (1-nitroelacridar, 4c)

According to the general procedure the compound was synthesized using 3c (0.314 g, 1 mmol, 1eq) as reactant. Yield: 0.305 g (50%) brown solid. mp: 167-172 °C. HRMS (ESI/MS) calcd for C₃₄H₃₂O₇N₄H: 609.2349 found: 609.2355. ¹H NMR (DMSO-d₆) δ 2.70-3.04 (m, 8H, 4xCH₂), 3.67-3.83 (m, 8H, CH₂, 2xOCH₃), 4.05 (s, 3H, OCH₃), 6.65-6.75 (m, 2H), 7.25-7.44 (m, 4H), 7.56-7.84 (m, 4H), 8.52-8.63 (m, 1H), 10.81 (bs, 1H), 12.61 (bs, 1H).

4.4.4. 2-Chloro-N-(4-(2-(6,7-dimethoxy-1,2,3,4-tetrahydroisoquinolin-2-yl)ethyl)phenyl)-5-methoxyacridone-4-carboxamide (2-chloroelacridar, 4d)

The title compound was synthesized using the general approach starting from 3d (0.304 g, 1 mmol) as reactant. Yield: 0.353 g (59%) yellow solid. mp: 188-195 °C. HRMS (ESI/MS) calcd for C₃₄H₃₂O₅N₃ClH: 598.2109 found: 598.2105. ¹H NMR (CDCl₃) δ 2.72-3.02 (m, 8H, 4xCH₂), 3.67 (s, 2H, CH₂), 3.85 (s, 6H, 2xOCH₃), 4.06 (s, 3H, OCH₃), 6.56 (s, 1H), 6.61 (s, 1H), 7.08-7.32 (m, 4H), 7.63-7.76 (m, 2H), 7.95-8.13 (m, 1H), 8.31 (d, 1H, J = 2.2 Hz), 8.60 (d, 1H, J = 2.2 Hz), 9.92 (s, 1H), 12.25 (s, 1H)

4.4.5. N-(4-(2-(6,7-dimethoxy-1,2,3,4-tetrahydroisoquinolin-2-yl)ethyl)phenyl)-2-fluoro-5-methoxyacridone-4-carboxamide (2-fluoroelacridar, 4e)

Starting from 3e (0.287 g, 1 mmol) the title compound was synthesized according to the general procedure. Yield: 0.399 g (69%) dark-orange solid. mp: 152-157 °C. HRMS (ESI/MS) calcd for C₃₄H₃₂O₅N₃FH: 582.2404 found 582.2398. ¹H NMR (DMSO-d₆) δ 2.56-2.96 (m, 8H, 4xCH₂), 3.57 (s, 2H, CH₂), 3.70 (s, 6H, 2xOCH₃), 4.03 (s, 3H, OCH₃), 6.60-6.69 (m, 2H), 7.16-7.40 (m, 4H), 7.61-7.73 (m, 2H), 7.73-7.80 (m, 1H), 8.10-8.18 (m, 1H), 8.38-8.47 (m, 1H), 10.63 (s, 1H), 12.26 (s, 1H).

4.5. General procedure for small-scale [¹⁸F]fluorination of compounds 4a, 4c and 4d

Aqueous [¹⁸F]F⁻ ion from the cyclotron target was collected into a 3 mL Wheaton V-vial (Wheaton Science Products, Millville, USA) containing a solution of kryptofix 2.2.2 in CH₃CN (120 mg/mL, 100-200 μL; 12.0-24.0 mg, 32.7-65.4 μmol) and a solution of K₂CO₃ in H₂O (200 mg/mL, 20-40 μL; 4.0-8.0 mg, 28.9-57.8 μmol). The mixture was exposed to a stream of argon and concentrated to dryness at 130°C under repeated addition of CH₃CN (3x0.5 mL). To the dried K[¹⁸F]F-K₂₂₂ complex, the respective precursor (4a, 4c, 4d: 3-12 μmol) dissolved in dry DMSO or DMF (0.5 mL each) was added and the resulting solution stirred for 5-60 min at 120-180°C. An aliquot of the reaction mixture was analyzed by TLC. For TLC analysis Merck silica gel 60 F₂₅₄ plates were used with ethyl acetate/MeOH 2/1 (v/v) as mobile phase. TLC plates were analyzed using a phosphor imager (Cyclone, Packard Instruments, Meriden, CN, USA). Retardation factors (R_f): [¹⁸F]F⁻ : 0.0; [¹⁸F]4b: 0.6-0.7.

4.6. 1-[¹⁸F]fluoroelacridar ([¹⁸F]4b)

The synthesis of [¹⁸F]4b was automated in a TRACERlab FX synthesis module (General Electric Healthcare). After delivery of the irradiated [¹⁸O]H₂O to the synthesis module, [¹⁸F]F⁻ was trapped on a Chromafix 30-PS-HCO₃ ion exchange cartridge, which had been preactivated with EtOH (2 mL) and H₂O (2 mL). [¹⁸F]F⁻ was eluted to the synthesis reactor by rinsing the cartridge with a mixture of kryptofix 2.2.2. (24 mg, 63.8 μmol) in CH₃CN (0.9 mL) and K₂CO₃ (3.5 mg, 57.1 μmol) in H₂O (0.1 mL). The solvent was evaporated under vacuum for 5 min at 75°C, then CH₃CN (1 mL) was added and the mixture heated under vacuum for 5 min at 130°C to remove the remaining H₂O. To the dried K[¹⁸F]F-K₂₂₂ complex, radiolabelling precursor 4c (6 mg, 9.9 μmol) dissolved in dry DMSO (1 mL) was added and the reaction mixture heated for 60 min at 175°C. The reaction mixture was then cooled to 40°C, diluted with a mixture of CH₃CN and 0.1% aq. trifluoroacetic acid (TFA) (1/1, v/v, 3 mL) and injected into the built-in HPLC system. A Hamilton-PRP1 column (500x8 mm, 10 μm) was eluted at a flow rate of 4 mL/min with a gradient (0-20 min) from CH₃CN/0.1% aq. TFA 35/65 (v/v) to 45/55. The HPLC eluate was monitored in series for ultraviolet (UV) absorption at a wavelength of 257 nm and radioactivity. On this system radiolabelling precursor 4c and product [¹⁸F]4b eluted with retention times of 23 min and 19-20 min respectively. The product fraction was diluted with H₂O (100 mL) and passed over a C18 Sep-Pak Plus cartridge (Waters cooperation, Milford, MA), which had been preactivated with EtOH (5 mL) and H₂O (10 mL). The cartridge was then washed with H₂O (10 mL), followed by elution of [¹⁸F]4b with EtOH (3 mL). The EtOH was removed by heating at 90°C under a stream of argon and the product formulated in a mixture of 5 mg sodium ascorbate in 0.9% aq. NaCl solution/EtOH/polyethylene glycol 300 (50/15/35, v/v/v) for i.v. injection into animals. Radiochemical and chemical purity and specific activity of [¹⁸F]4b were determined by analytical radio-HPLC using a Hamilton-PRP1 column (290x4 mm, 10 μm) eluted at a flow rate of 1 mL/min using a gradient (0-25 min) from CH₃CN/0.1% aq. TFA 35/65 (v/v) to 45/55. UV detection was performed at a wavelength of 257 nm. The retention time of [¹⁸F]4b was 16-18 min on this HPLC system.

4.7. Cell line and culture condition

Madin-Darby canine kidney epithelial cells stably transfected with the human MDR1 gene (MDCKII-MDR1), which were obtained from the National Cancer Institute (Bethesda, MD, USA), were used for the fluorescent screening studies. Cells were cultured in Dulbecco’s Modified Eagles medium (DMEM) supplemented with 80 ng/mL colchicine to maintain high expression of Pgp as previously described ²⁰. Cells were seeded onto 24-well culture plates at a density of 65,000 cells/cm² and used for functional assays upon reaching confluency (typically 3 days).

4.8. Assessment of Pgp inhibitory activity

Pgp inhibition of derivatives 4a-e was determined as previously described ²¹. Briefly, confluent MDCKII-MDR1 monolayers were pretreated with pH 7.4 Tyrodes balanced salt solution (TBSS): 136 mM NaCl, 1.8 mM CaCl₂, 1 mM MgCl₂, 0.36 mM NaH₂PO₄, 5.56 mM D-glucose and 5 mM HEPES. Various concentrations of 4a-e (4b, 4d, 4e: 0.1-10 μM; 4a, 4c: 0.1-100 μM, n=3 per concentration) were added to the TBSS and the cells were incubated for 30 min at 37°C. Following the preincubation period, the TBSS was removed and replaced with fresh TBSS containing 3.2 μM R123 to assess drug efflux transporter activity. Cellular accumulation of the fluorescent dyes was monitored over a 60 min period, under control conditions and in the presence of various concentrations of 4a-e (see above). After this time, cells were washed three times with ice-cold phosphate buffered saline solution and the cells were solubilized in 1.0% Triton-X100 (0.5 mL). Aliquots (100 μL) of the solubilized cell solutions were removed for determination of intracellular R123 accumulation using a Synergy HT fluorescent plate reader. Protein content was determined using the Pierce BCA method and the data were expressed as the amount of fluorescent probe (nmol) per mg cell protein. Concentration response curves were fitted using GraphPad Prism 5 software (GraphPad Software, Inc., San Diego, CA, USA) and compared to maximal inhibition achieved using 1 at a concentration of 3.2 μM.

4.9. Animals

Adult female Sprague Dawley rats weighing 260-300 g were obtained from Harlan Netherlands (Horst, Netherlands). Female FVB (wild-type) and Mdr1a/b^(−/−)Bcrp1^(−/−) mice weighing 27-35 g were purchased from Taconic Inc. (Germantown, USA). Prior to each experiment, the animals were placed into an induction box and anesthetized with 2.5% isoflurane. When unconscious, the animals were taken from the box and kept under anesthesia with 1.5-2% isoflurane administered via a mask during the whole experiment. The animals were warmed throughout the whole experiment at around 38°C. Rats were implanted with catheters into the femoral artery (for blood sampling) and vein (for administration of [¹⁸F]4b and 1). In mice, a lateral tail vein was used for radiotracer administration. The study was approved by the local Animal Welfare Committee and all study procedures were performed in accordance with the Austrian Animal Experiments Act.

4.10. Small-animal PET imaging and PET data analysis

The anesthetized animals were positioned in the imaging chamber of the microPET Focus220 scanner (Siemens, Medical Solutions, Knoxville, USA). Animals were injected with [¹⁸F]4b (rats: 33±3 MBq (mean±standard deviation, SD) in a volume of about 0.3 mL corresponding to 9.6±6.5 nmol or 20±12 μg/kg body weight unlabelled 4b; mice: 6±1 MBq in a volume of about 0.1 mL corresponding to 2.0±1.1 nmol or 39±22 μg/kg unlabelled 4b) as an i.v. bolus over approximately 40 s. At the start of radiotracer injection, dynamic PET imaging was initiated.

Two groups of rats were scanned with [¹⁸F]4b (n=3 per group). One group underwent a 150 min PET scan with [¹⁸F]4b during which 1 (5 mg/kg) was administered i.v. over approximately 60 s at 60 min after injection of [¹⁸F]4b. The other group underwent a 60 min PET scan at 2 h after injection of 1. During experiments in rats, 5-μL arterial blood samples were withdrawn manually with pre-weighted micropipettes from the femoral artery (approximately every 5 s) during the first 3 min after radiotracer injection, followed by further 10-μL samples taken at 5, 10, 20, 30, 40, 50, 60, 70, 80 and 90 min (last 3 time points for group 1 only). Mice (n=3 per mouse type) underwent single 60-min PET scans with [¹⁸F]4b. At the end of the PET scan, venous blood was withdrawn from mice by retro-orbital puncture into pre-weighted micropipettes. Blood samples from rats and mice were weighted and counted for activity in a 1-detector Wallac gamma counter (Perkin Elmer Instruments, Wellesley, USA), which had been cross-calibrated with the PET camera. Blood activity data were corrected for radioactive decay and corrected for injected dose per gram body weight and expressed as SUV.

PET images were reconstructed by Fourier rebinning followed by 2-dimensional filtered back projection with a ramp filter. The standard data correction protocol (normalization, attenuation, decay correction and injection decay correction) was applied to the data. Whole brain and humerus (in rats) were manually outlined on multiple planes of the PET summation images using the image analysis software Amide and TACs, expressed as SUV, were calculated. For the mouse data, the SUV values measured with PET in brain tissue of individual animals during the last time frame min (50-60 min after radiotracer injection) were divided by the SUV value measured in whole blood at the end of the PET scan to obtain brain-to-blood ratios of activity. For all outcome parameters, differences between groups were tested with a 2-tailed Student’s t-test. The level of statistical significance was set to p<0.05.

4.11. Metabolism and plasma protein binding of [¹⁸F]4b

A group of 3 rats was injected with [¹⁸F]4b (74±23 MBq in a volume of about 0.3 mL corresponding to 17.2±3.9 nmol or 37±10 μg/kg body weight unlabelled 4b) without performing PET examination in order to assess metabolism of [¹⁸F]4b. At 10, 20 and 30 min after radiotracer injection 0.5 mL blood samples were collected into heparinized vials. At 40 min after radiotracer injection rats were sacrificed, their brains removed and a terminal blood sample (5 mL) collected. Plasma collected at all time points was centrifuged (3000×g, 5 min, 21°C) and analyzed for radiometabolites of [¹⁸F]4b using a previously described SPE assay ²². In brief, arterial plasma was spiked with a solution of unlabelled 4b in DMSO (1 mg/mL, 10 μL) and acidified with 5M aq. HCl (40 μL) and loaded on a Sep-Pak vac tC18 cartridge (Waters Corporation, Milford, USA), which had been pre-activated with MeOH (3 mL) and H₂O (5 mL). The cartridge was first washed with H₂O (5 mL) and then eluted with MeOH (4 mL). Radioactivity in all three fractions (plasma, H₂O, MeOH) was measured in the 1-detector Wallac gamma counter. For the 40 min blood sample, the MeOH fraction was 1:1 diluted with 0.1% aq. TFA and analyzed by HPLC (injected volume: 2 mL). A Chromolith Performance RP 18-e (100–4.6 mm) column was eluted at flow rate of 5 mL/min using a gradient (0-20 min) from CH₃CN/0.1% aq. TFA 20/80 (v/v) to 25/75. UV detection was performed at a wavelength of 257 nm. The retention time of [¹⁸F]4b was 13-15 min on this HPLC system. For validation of the SPE assay, [¹⁸F]4b dissolved in mouse plasma (0.5 mL) was subjected to the SPE procedure showing that all radioactivity was quantitatively recovered in the MeOH fraction.

Rat brain was washed twice with ice-cold H₂O and homogenized in 0.9% aq. NaCl solution (0.8 mL) using an IKA T10 basic Ultra-turrax (IKA Laboratory Equipment, Staufen, Germany). The brain homogenate was mixed with CH₃CN (1.5 mL) and centrifuged (3 min, 4°C,13,000×g). The supernatant was diluted 1:1 with 0.1% aq. TFA and injected into the same HPLC system used for analysis of radiometabolites in plasma (injected volume: 2 mL). Plasma protein binding of [¹⁸F]4b was determined by incubating fresh plasma samples obtained from naïve rats with [¹⁸F]4b during 30 min at 37°C, followed by ultrafiltration using Amicon Microcon YM-10 centrifugal filter devices (Millipore Corporation, USA) as described previously ²³.

5. Conclusions

In an attempt to label the potent dual Pgp/BCRP inhibitor 1 with ¹⁸F we synthesized a series of new 1- and 2- halogen- and nitro-substituted derivatives of 1, which were found to display comparable potency to 1 in inhibiting Pgp transport of R123 in MDCKII-MDR1 cells. The 1-[¹⁸F]fluoro derivative of 1 ([¹⁸F]4b) was synthesized in moderate radiochemical yield and characterized in rats and mice with small-animal PET. [¹⁸F]4b was found to display comparable in vivo behavior to previously described [¹¹C]1 in that brain activity uptake was low in baseline scans and several times increased after administration of unlabelled 1, suggesting that [¹⁸F]4b is efficiently transported by BCRP and/or Pgp and at the rodent BBB. However, the significant degree of in vivo defluorination observed with [¹⁸F]4b together with low radiochemical yields will in all likelihood limit its future utility as a PET tracer.