Synthesis and Evaluation of [18F]FEtLos and [18F]AMBF3Los as Novel 18F-Labelled Losartan Derivatives for Molecular Imaging of Angiotensin II Type 1 Receptors.

Losartan is widely used in clinics to treat cardiovascular related diseases by selectively blocking the angiotensin II type 1 receptors (AT1Rs), which regulate the renin-angiotensin system (RAS). Therefore, monitoring the physiological and pathological biodistribution of AT1R using positron emission tomography (PET) might be a valuable tool to assess the functionality of RAS. Herein, we describe the synthesis and characterization of two novel losartan derivatives PET tracers, [18F]fluoroethyl-losartan ([18F]FEtLos) and [18F]ammoniomethyltrifluoroborate-losartan ([18F]AMBF3Los). [18F]FEtLos was radiolabeled by 18F-fluoroalkylation of losartan potassium using the prosthetic group 2-[18F]fluoroethyl tosylate; whereas [18F]AMBF3Los was prepared following an one-step 18F-19F isotopic exchange reaction, in an overall yield of 2.7 ± 0.9% and 11 ± 4%, respectively, with high radiochemical purity (>95%). Binding competition assays in AT1R-expressing membranes showed that AMBF3Los presented an almost equivalent binding affinity (Ki 7.9 nM) as the cold reference Losartan (Ki 1.5 nM), unlike FEtLos (Ki 2000 nM). In vitro and in vivo assays showed that [18F]AMBF3Los displayed a good binding affinity for AT1R-overexpressing CHO cells and was able to specifically bind to renal AT1R. Hence, our data demonstrate [18F]AMBF3Los as a new tool for PET imaging of AT1R with possible applications for the diagnosis of cardiovascular, inflammatory and cancer diseases.

non-specific uptake. Therefore, here we described the synthesis and 18 F-labeling of two novel losartan derivatives using less hydrophobic motifs such as: 2-fluoroethoxy and ammoniomethyltrifluoroborate (AMBF 3 3 Los, respectively ( Figure 2). The 2-[ 18 F]fluoroethyl tosylate is a 18 F-fluoroalkylating agent usually used to prepare PET tracers since it is easy to prepare with low volatility, high chemical stability and reactivity, and chemoselectivity [29,30]. Besides, the [ 18 F]fluoroethyl group is considered as a surrogate for a [ 11 C]methyl group because it can be coupled to the same functional groups [30]. Therefore, we designed the novel analog [ 18 F]fluoroethyl-losartan ([ 18 F]FEtLos; Figure 2) following a similar reaction scheme to the one reported for [ 11 C]methyl-losartan ( Figure 1).
On the other hand, the copper (I)-catalyzed Huisgen alkyne-azide 1,3-dipolar cycloaddition (CuAAC) is the most commonly used reaction for click chemistry, which has become a very attractive strategy in the design of PET tracers. This type of reaction shows an outstanding efficiency, regiospecificity and fast formation of the 1,4-disubstituted 1,2,3-triazoles [31] which are metabolically stable under physiological conditions [32,33]. In particular, the N-propargyl-N,N-dimethylammoniomethyl-trifluoroborate is a hydrophilic linker that afford AMBF3-conjugated biomolecules using a CuAAC reaction. The AMBF3 motif allows an easy one-step 18 F-labeling by 18 F- 19 F isotope exchange reaction in acidic aqueous media using only nanomole amounts of the precursor. AMBF3 labeled products are usually very stable in vivo, present high molar activity and radiochemical purity, and do not depend on High Pressure Liquid Chromatography (HPLC) for purification [34][35][36]. Therefore, [ 18 F]ammoniomethyltrifluoroborate-losartan ([ 18 F]AMBF3Los; Figure 2) was also synthesized and characterized in the present work following the AMBF3 approach.
In addition, we studied the AT1R binding affinity of the newly synthetized derivatives both in vitro, ex-vivo and in vivo to evaluate its usefulness and pharmacokinetics as PET tracers for the noninvasive imaging of AT1R. We found that [ 18 F]AMBF3Los presented the most suitable characteristics as an AT1R PET tracers: (1) its kit-like 18 F-labeling approach was very fast and efficient, and (2) 18 F- 19 F isotope exchange allowed the use of smaller amounts of the precursor than conventional radiolabeling methods.
The synthesis of the cold derivative FEtLos was carried out through an O-alkylation of the hydroxymethylene group at the C-5 position of losartan (1) with 2-fluoroethyl-tosylate (2) (Scheme 1A), readily prepared from the tosylation of 2-fluoroethanol using tosyl chloride diluted in pyridine [37]. Following the bimolecular nucleophilic substitution reaction (SN2), FEtLos was obtained with a yield of 26% while the sub-product 3 was obtained with a yield of 7% (FEtLos/3 ratio = 3.7). Structural characterization of FEtLos was achieved by 13 C NMR, 1 H NMR, 19 F NMR, ( 1 H, 13 C)-HMBC and HRMS analyses. In particular, the structure of FEtLos was confirmed based on the heteronuclear correlation over three bonds between the fluoroethyl methylene hydrogens (4.70 ppm) and the oxymethylene carbon (52.7 ppm) at the imidazole C-5 position ( Figure S2). The compound 3 was also structurally characterized ( Figure S3). On the other hand, the copper (I)-catalyzed Huisgen alkyne-azide 1,3-dipolar cycloaddition (CuAAC) is the most commonly used reaction for click chemistry, which has become a very attractive strategy in the design of PET tracers. This type of reaction shows an outstanding efficiency, regiospecificity and fast formation of the 1,4-disubstituted 1,2,3-triazoles [31] which are metabolically stable under physiological conditions [32,33]. In particular, the N-propargyl-N,N-dimethyl-ammoniomethyl-trifluoroborate is a hydrophilic linker that afford AMBF 3 -conjugated biomolecules using a CuAAC reaction. The AMBF 3 motif allows an easy one-step 18 F-labeling by 18 F-19 F isotope exchange reaction in acidic aqueous media using only nanomole amounts of the precursor. AMBF 3 labeled products are usually very stable in vivo, present high molar activity and radiochemical purity, and do not depend on High Pressure Liquid Chromatography (HPLC) for purification [34][35][36]. Therefore, [ 18 F]ammoniomethyltrifluoroborate-losartan ([ 18 F]AMBF 3 Los; Figure 2) was also synthesized and characterized in the present work following the AMBF 3 approach.
In addition, we studied the AT 1 R binding affinity of the newly synthetized derivatives both in vitro, ex-vivo and in vivo to evaluate its usefulness and pharmacokinetics as PET tracers for the noninvasive imaging of AT 1 R. We found that [ 18 F]AMBF 3 Los presented the most suitable characteristics as an AT 1 R PET tracers: (1) its kit-like 18 F-labeling approach was very fast and efficient, and (2) 18 F- 19 F isotope exchange allowed the use of smaller amounts of the precursor than conventional radiolabeling methods.
Following the bimolecular nucleophilic substitution reaction (S N 2), FEtLos was obtained with a yield of 26% while the sub-product 3 was obtained with a yield of 7% (FEtLos/3 ratio = 3.7). Structural characterization of FEtLos was achieved by 13 C NMR, 1 H NMR, 19 F NMR, ( 1 H, 13 C)-HMBC and HRMS analyses. In particular, the structure of FEtLos was confirmed based on the heteronuclear correlation over three bonds between the fluoroethyl methylene hydrogens (4.70 ppm) and the oxymethylene carbon (52.7 ppm) at the imidazole C-5 position ( Figure S2). The compound 3 was also structurally characterized ( Figure S3).
We next evaluated the in vitro cellular uptake of [ 18 F]AMBF3Los in CHO-AT1R cells, which overexpress the human AT1R ( Figure 3B), and CHO cells (control). As observed in Figure 3C, CHO-AT1R cells presented a higher uptake of [ 18 F]AMBF3Los in comparison to CHO cells. Incubation of 100 µM of losartan potassium blocked the CHO-AT1R uptake, indicating that the binding of [ 18 F]AMBF3Los was receptor-specific (p < 0.01). [ 18 F]FEtLos, on the other hand, did not show any specific AT1R binding for CHO-AT1R cells, which is in agreement with the low Ki value ( Figure S7).
We next evaluated the in vitro cellular uptake of [ 18 F]AMBF3Los in CHO-AT1R cells, which overexpress the human AT1R ( Figure 3B), and CHO cells (control). As observed in Figure 3C, CHO-AT1R cells presented a higher uptake of [ 18 F]AMBF3Los in comparison to CHO cells. Incubation of 100 µM of losartan potassium blocked the CHO-AT1R uptake, indicating that the binding of [ 18 F]AMBF3Los was receptor-specific (p < 0.01). [ 18 F]FEtLos, on the other hand, did not show any specific AT1R binding for CHO-AT1R cells, which is in agreement with the low Ki value ( Figure S7). [ 18 F]FEtLos was prepared by a 18 F-fluoroalkylation of losartan using the prosthetic group 2-[ 18 F]fluoroethyl tosylate (2*) (Scheme 3). Thus, 2* was manually prepared with a radiochemical yield of 30 ± 8% (decay-corrected from dried [ 18 F]F − ) and >99% chemical and radiochemical purity after HPLC-purification. The synthesis duration of 2* was of approximately 50 min. After 18 F-fluoroalkylation of losartan potassium, [ 18 F]FEtLos was purified by HPLC with a chemical and radiochemical purity >99% and its chemical identity was confirmed by HPLC co-injection of the cold compound (FEtLos) ( Figure S6). [ 18 F]FEtLos was prepared using a small amount of 2* and obtained with a radiochemical yield of 12 ± 5% (decay-corrected from 2*, approximately 100 min) and 1.4 ± 1.2 GBq (38 ± 33 mCi)/µmol molar activity (n = 10). The overall radiochemical yield was 2.7 ± 0.9% (decay-corrected from dried [ 18 F]F − ) with a total synthesis time of approximately three hours. The distribution coefficient of [ 18 F]FEtLos at pH 7.4 (log D 7.4 ) was 0.21 ± 0.09 (n = 3). The 18 F-labeled side-product 3* was also formed by N-18 F-fluoroalkylation of the losartan tetrazole moiety, confirmed by HPLC analyses of the crude reaction mixtures and compared with the HPLC profile of the cold compound 3.
On the other hand, [ 18 F]AMBF 3 Los was easily synthetized (35-55 min) by an 18 F-19 F isotopic exchange in an acidic aqueous medium, using only nanomole amounts of the AMBF 3 Los precursor and purified using a single Sep-Pak cartridge (Scheme 4). The manual radiosynthesis of [ 18 F]AMBF 3 Los, using 25 nmol of precursor and low activities of [ 18 F]F − , was achieved with a radiochemical yield of 11 ± 4%, a molar activity of 2.4 ± 0.5 GBq (0.06 ± 0.01 Ci)/µmol, and a radiochemical purity >97% (n = 20). When the radiosynthesis of [ 18 F]AMBF 3 Los was done using 100-fold higher activities of [ 18 F]F − and 4-fold higher mass of the precursor (100 nmol), a radiochemical yield of 10 ± 1% with a radiochemical purity >97% and molar activity of 108 ± 29 GBq (2.9 ± 0.8 Ci)/µmol was achieved. The chemical identity of [ 18 F]AMBF 3 Los was also confirmed by a single peak in the analytical HPLC profile after co-injection of the final formulation and the cold AMBF 3 Los, previously prepared ( Figure S6). [ 18 F]AMBF 3 Los displayed a log D 7.4 of −0.43 ± 0.02 (n = 3).

In Vivo Assays
A dynamic µPET study for healthy NOD.Cg-Prkdc scid Il2rg tm1Wjl /SzJ mice were next performed with [ 18 F]AMBF3Los ( Figure 4A,B). The µPET dynamic scan at baseline showed a high [ 18 F]AMBF3Los uptake in the mouse renal cortex at early time points (10, 15 and 20 min) that was reduced when losartan potassium (AT1R blocker) was co-injected with [ 18 F]AMBF3Los ( Figure 4A). To confirm this finding, a volume of interest (VOI) analysis was performed on the reconstructed images to generate the time-activity curves (TAC) for [ 18 F]AMBF3Los ( Figure 4B), which showed renal tracer uptake within the first few minutes, and the activity was slowly washed out. In contrast, the presence of losartan potassium (AT1R blocker) reduced renal uptake during the first 10 min.
To validate the µPET imaging studies, a biodistribution study of [ 18 F]AMBF3Los was performed at 60 min post injection (p.i.) as summarized in Table 1. No significant differences were observed compared to the µPET imaging quantification for the renal uptake. The high [ 18 F]AMBF3Los activity uptake in intestines suggests it may also be cleared through the hepatobiliary system.  3 Los uptake in CHO-AT 1 R or CHO cells seeded in 6-well plates after the tracer incubation for 60 min at 4 • C in the presence (white bars, n = 3) or absence (black bars, n = 3) of the AT 1 R blocker losartan potassium (100 µM/well). Graphs show the mean ± standard deviation (SD) of three independent experiments (n = 3). Data were analyzed by one unpaired t-test (multiple t tests); ** p < 0.01 (Holm-Sidak method).
We next evaluated the in vitro cellular uptake of [ 18 F]AMBF 3 Los in CHO-AT 1 R cells, which overexpress the human AT 1 R ( Figure 3B), and CHO cells (control). As observed in Figure 3C, CHO-AT 1 R cells presented a higher uptake of [ 18 F]AMBF 3 Los in comparison to CHO cells. Incubation of 100 µM of losartan potassium blocked the CHO-AT 1 R uptake, indicating that the binding of 3 Los was receptor-specific (p < 0.01). [ 18 F]FEtLos, on the other hand, did not show any specific AT 1 R binding for CHO-AT 1 R cells, which is in agreement with the low Ki value ( Figure S7).

In Vivo Assays
A dynamic µPET study for healthy NOD.Cg-Prkdc scid Il2rg tm1Wjl /SzJ mice were next performed with [ 18 F]AMBF 3 Los ( Figure 4A,B). The µPET dynamic scan at baseline showed a high [ 18 F]AMBF 3 Los uptake in the mouse renal cortex at early time points (10, 15 and 20 min) that was reduced when losartan potassium (AT 1 R blocker) was co-injected with [ 18 F]AMBF 3 Los ( Figure 4A). To confirm this finding, a volume of interest (VOI) analysis was performed on the reconstructed images to generate the time-activity curves (TAC) for [ 18 F]AMBF 3 Los ( Figure 4B), which showed renal tracer uptake within the first few minutes, and the activity was slowly washed out. In contrast, the presence of losartan potassium (AT 1 R blocker) reduced renal uptake during the first 10 min.
Molecules 2020, 25, x FOR PEER REVIEW 9 of 22 activity was approximately 5-times smaller in the kidney of the mouse previously blocked with losartan potassium in comparison to the unblocked one. Thus, these data suggest that [ 18 F]AMBF3Los specifically binds to renal AT1R. The renal AT1R binding specificity of [ 18 F]FEtLos was also evaluated by an ex vivo µPET/CT imaging of Balb/c Nude mice kidneys 10 min post injection (p.i.) of the tracer. As shown in Figure S8, [ 18 F]FEtLos showed a high uptake in the kidneys (baseline), which was blocked when losartan potassium (AT1R blocker) was co-injected with [ 18 F]FEtLos, despite the low affinity of the radiotracer. These positive results of the ex vivo renal AT1R binding of [ 18 F]FEtLos might be possible due to the high AT1R density in the kidneys, that allows visualizing specific uptake in spite of the low AT1R binding affinity of [ 18 F]FEtLos.  To validate the µPET imaging studies, a biodistribution study of [ 18 F]AMBF 3 Los was performed at 60 min post injection (p.i.) as summarized in Table 1. No significant differences were observed compared to the µPET imaging quantification for the renal uptake. The high [ 18 F]AMBF 3 Los activity uptake in intestines suggests it may also be cleared through the hepatobiliary system. Autoradiographies of the mouse kidneys were also taken after the 60-min dynamic µPET scan with [ 18 F]AMBF 3 Los showing that [ 18 F]AMBF 3 Los was mainly captured in the kidney cortex, and this uptake was blocked by the co-injection of losartan potassium ( Figure 4C). Indeed, the renal cortex activity was approximately 5-times smaller in the kidney of the mouse previously blocked with losartan potassium in comparison to the unblocked one. Thus, these data suggest that [ 18 F]AMBF 3 Los specifically binds to renal AT 1 R.
The renal AT 1 R binding specificity of [ 18 F]FEtLos was also evaluated by an ex vivo µPET/CT imaging of Balb/c Nude mice kidneys 10 min post injection (p.i.) of the tracer. As shown in Figure  S8, [ 18 F]FEtLos showed a high uptake in the kidneys (baseline), which was blocked when losartan potassium (AT 1 R blocker) was co-injected with [ 18 F]FEtLos, despite the low affinity of the radiotracer. These positive results of the ex vivo renal AT 1 R binding of [ 18 F]FEtLos might be possible due to the high AT 1 R density in the kidneys, that allows visualizing specific uptake in spite of the low AT 1 R binding affinity of [ 18 F]FEtLos.

Discussion
The renin-angiotensin system (RAS) plays a fundamental role in the control of the cardiovascular and renal systems. Recent evidences suggest that AT 1 R has been implicated in a few brain disorders and has been associated with cancer progression and prognosis. Therefore, the in vivo imaging of AT 1 R could provide a diagnostic and prognostic information for the management of several diseases.
Although a few PET AT 1 R radioligands have been evaluated for cardiovascular [39,40], renal [22][23][24][41][42][43][44] and brain [18,45] AT 1 R imaging, most of them were labeled with 11 C, which brings some disadvantages over 18 F in the clinical practice, such as the presence of an in-site cyclotron. Recently, it was reported the synthesis of [ 18 F]FPyKYNE-losartan (Figure 1), a losartan-derivative containing a hydrophobic motif that could favor a non-specific liver uptake [21]. Therefore, here we described the design, synthesis and characterization of two new AT 1  Both compounds were designed via the substitution of the hydroxyl group by fluoro ligands at the imidazole 5-position of losartan. Several substituents with different length such as a small methyl group [20] and a bulky side chain [46] have been introduced into the losartan pharmacophore at the same position. For instance, a few losartan derivatives were synthesized by adding nitric oxide (NO)-donor side chains to losartan at the imidazole 5-position ( Figure S9), displaying antagonist potency values similar to losartan [47]. Besides, the chelate-coupled losartan-Leucine-Diglycoloyl-Tetraethyleneglycol-Tetraamine ( Figure S10) showed superior AT 1 R affinity compared to parental losartan; radiolabeling of this compound with 99m Tc displayed an acceptable biodistribution profile in mouse model of post-myocardial infarction heart failure [46]. The increase for AT 1 R affinity of this losartan analog may be probably due to the higher possibility to form hydrogen bonds, which may increase their affinity to the receptor. In addition, the [ 18 F]FPyKYNE-losartan derivative also showed a high affinity to kidney AT 1 R in rats and pigs [44].
[ 18 F]FEtLos and the cold FEtLos were synthetized by both [ 18/19 F]fluoroethylation of losartan potassium and tetrazole-protected losartan through a S N 2 reaction. Many PET tracers have been synthesized using a two-step radiosynthesis because the direct [ 18 F]fluorination is not feasible most of the time. Direct fluorination usually requires high temperatures and non-physiological pH that can harsh the precursors [27]. Hence, 18 F-labeling using prosthetic groups, such as the 2-[ 18 F]fluoroethyl tosylate (2*) for 18 F-fluoroalkylation reactions [30], is commonly used in spite of the longer time of synthesis. Moreover, although the insertion of fluoroethylene group in the NH-tetrazole group is interesting, we decided to maintain the negatively-charged tetrazole moiety intact because of its crucial role for AT 1 R binding through hydrogen bridge interactions with basic amino acid residues of the receptor [48][49][50], its prolonged half-life and enhanced metabolic stability [51]. In addition, the tetrazolic ion is more stable than the alkoxide ion due to the spatial delocalization of the negative charge of the former, favoring reactive alkoxide formation and alkylation to produce FEtLos with a ratio of 3.7 with the corresponding NH-tetrazole isomer (3).
On the other hand, the cold AMBF 3 Los was synthetized via a copper (I)-catalyzed Huisgen alkyne-azide 1,3-dipolar cycloaddition, the most used click reaction; while the 18 F-labelled compound was obtained from an aqueous 18 F-19 F isotopic exchange reaction. The cold AMBF 3 Los was afforded in similar yield to other compounds using the same approach [38,52]. The one step 18 F-19 F isotopic exchange approach using organotrifluoroborates resembles a 'kit like" 18 F-labeling method and requires much lower amount of precursor (nanomoles) that the conventional radiolabeling methods (micromoles of precursor). This approach was previously used for radiolabeling peptides using high activities of [ 18 F]F − and very low amounts of precursor (75-100 nmol), within 30 min with a radiochemical purity >99%, a high molar activity (68-148 GBq/µmol), a radiochemical yield of 20-28%, with HPLC-free purification [34,38,53]. Small [ 18 F]AMBF 3 -conjugated molecules were also synthesized with a radiochemical yield of 14.8 ± 0.4% with a molar activity of 24.5 ± 5.2 GBq/µmol and radiochemical purity >99% [54]. Taking into account that the reaction medium is acid, a significant amount of [ 18 F]F − is lost by volatilization as [ 18 F]hydrogen fluoride during the labeling reaction under vacuum and heating. According to the literature, reaction pH is the crucial step of this radiolabeling approach, and pH 2-2.5 generally gives lower by-product (boronic acid) levels and better radiochemical yield [36]. The carrier [ 19 F]fluoride was used during the radiolabeling because it increases the [ 18 F]BF 3 formation while still providing a molar activity of 37 GBq/µmol [55].
Results of the in vitro receptor binding assays showed that only the derivative AMBF 3 Los displayed a good binding affinity to human AT 1 R compared to FEtLos. Therefore, the substitution of hydroxyl group by the 2-fluoroethoxy motif decreased the AT 1 R binding affinity due to the lack of hydrogen bonding or other intermolecular forces. The 1,2,3-triazole is a polar moiety and a good hydrogen-bond acceptor that may keep the receptor binding affinity of AMBF 3 Los, as reported for the [ 18 F]FPyKYNE-losartan derivative, which also contains a 1,2,3-triazole at the imidazole 5-position and displays high affinity for AT 1 R (dissociation constant, K D 49.4 nM) to rat kidney AT 1 R [44]. The structure-activity relationship at the imidazole 5-position revealed that it generally prefers small hydrogen-bonding substituents such as alcohols or carboxylic acids but will also tolerate a wide range of groups [48]. Hence, our results suggested that hydrogen-bonding substituents at the imidazole 5-position play an important role to AT 1 R binding affinity besides the tetrazole moiety.
Next, both in vitro and in vivo assays showed that [ 18 F]AMBF 3 Los specifically binds to the AT 1 receptors. Therefore, [ 18 F]AMBF 3 Los could be further evaluated as a specific PET radioligand to image the AT 1 receptors in disease, for instance, in cancer; whereas, the low binding affinity of [ 18 F]FEtLos limits its use as a AT 1 R PET tracer. In addition, the log D 7.4 measurements of [ 18 F]AMBF 3 Los (−0. 4) showed that this new derivative is more hydrophilic than the parental losartan with a log D 7.4 of 1.7 [56]. As such, we could observe that [ 18 F]AMBF 3 Los displayed a smaller liver uptake (% ID/g = 0.5) than the previously reported derivative [ 18 F]FPyKYNE-losartan (% ID/g = 66.7) in vivo at the same time-point (60 min) [44]. Therefore, [ 18 F]AMBF 3 Los exhibited a significant less non-specific liver uptake compared to the reported [ 18 F]FPyKYNE-losartan.
In particular, the PET imaging of AT 1 R in cancer has not been explored so far despite the upregulation of AT 1 R protein levels in human cells and tumor tissues of breast, prostate, gastric, bladder, ovarian and endometrial-derived cancers compared to non-cancerous tissues [57][58][59][60][61][62]. Accordingly, losartan has been shown to inhibit breast [63], gastric [62] and ovarian [64] cancer development and progression suggesting that AT 1 R could be a potential therapeutic target to complement current cancer treatment. Hence, [ 18 F]AMBF 3 Los as AT 1 R radioligand for PET imaging can be a very useful tool for assessing AT 1 R-positive tumors.

General Methods
All the chemicals and solvents were purchased with analytical grade from commercial sources and used without further purification. All the cold compounds were characterized by high resolution mass spectrometry (HRMS, Billerica, MA, USA) and nuclear magnetic resonance (NMR) experiments, recorded on a Bruker Daltonics micrOTOF-Q II ESI-Qq-TOF spectrometer and Bruker Avance DRX 300, DPX 400, DPX 500 equipments (Billerica, MA, USA). Chemical shifts (δ) were reported in parts per million (ppm) relative to an internal tetramethylsilane (TMS) standard. Coupling constants (J) were reported in Hertz (Hz

Quality Control
The chemical identity of [ 18 F]FEtLos and [ 18 F]AMBF 3 Los was determined by co-injection with their respective cold compounds onto HPLC using the analytical conditions previously described in each method of radiosynthesis. The radiochemical purity and molar activity of [ 18 F]FEtLos and [ 18 F]AMBF 3 Los were determined and the radiochemical yields were decay-corrected. The molar activity was always calculated at the end of each radiosynthesis using a calibration curve (nmol of the cold compound vs. area under the curve recorded from the HPLC analytical profile) (Figures S11 and S12).

Log D 7.4 Measurements
The log D 7.4 values were determined according to the literature [65]. Aliquots (2 µL) of [ 18 F]FEtLos and [ 18 F]AMBF 3 Los were added to polypropylene tubes containing 3 mL of octanol and 3 mL of PBS (phosphate-buffered saline). The samples were vortexed for 20 s (2 times) and centrifuged at 600× g for 15 min. Aliquots (0.1 mL) of each layer were counted in a gamma counter to determine the log D 7.4 as log (counts in octanol layer/counts in PBS layer).

Competition Binding Assays in Membranes Containing Human AT 1 R
The assays were performed using membranes expressing the human angiotensin II type 1 receptor from CHO-K1 cells, purchased from Perkin Elmer (Waltham, MA, USA). A serial dilution ranging from 10 −4 to 10 −12 M of Losartan potassium, FEtLos and AMBF 3 Los were used as cold ligands to displace the hot [ 125 I]-(Sar1,Ile8)-Angiotensin II (0.3 nM taking into account the activity of 125 I decay-corrected at the day of the assay and the specific activity of the hot compound (81.4 TBq (2200 Ci)/mmol)). The hot ligand [ 125 I]-(Sar1,Ile8)-Angiotensin II (81.4 TBq/mmol) was also purchased from Perkin Elmer (USA). The Multiscreen MSFCN6B50 96-well plates were used for this assay. The procedure was based on the recommended assay conditions from Perkin Elmer (Waltham, MA, USA) and procedures previously described for other AT 1 R ligands [22,23]. Briefly, the 96-well plate was preincubated for 1-2 h at 4 • C with 200 µL of 0.5% BSA (Bovine Serum Albumin) followed by washing (3 times) with an assay buffer. Next, 150 µL of the membranes, diluted (1:150) in the buffer assay (50 mM Tris-HCl pH 7.4, 5 mM MgCl 2 ) to a final concentration of 0.6 µg/well, were added to the wells followed by the addition of cold (25 µL) and hot ligand (25 µL). The plate was incubated for 60 min at 27 • C under stirring and at the end of the experiment all wells were washed (9 times) with ice-cold wash buffer (50 mM Tris-HCl pH 7.4) using vacuum. Afterwards, membranes were punched out and counted in a Perkin Elmer The AT1R mRNA abundance was measured in CHO-AT1R and CHO cells by RT-PCR. Total RNA from cells was isolated with 1 mL of Tri-Reagent (Sigma-Aldrich, St. Louis, MO, USA) according to the manufacturer's instructions. Complementary DNA was synthesized using the High capacity cDNA RT kit (Applied Biosystems, Foster City, CA, USA), according to the manufacturer's protocols. Quantitative PCR analysis was performed in triplicate using the Power SYBR ® Green PCR Master Mix (Applied Biosystems). Relative quantification was done using the ∆∆Ct method normalizing to β-actin gene expression. Primer human AT 1 R: forward 5 -3 , GCGTCAGTTTCAACCTC; reverse 5 -3 , TCCGGGACTCGTAATG. Primer hamster β-actin: Forward 5 -3 , GGCAGGCAAAGGTTACTCTG; reverse 5 -3 , TGGTGACAGGTGGACAAGAT.

Cell Binding Assays
The cell binding assays were performed based on previous reports [66] with modifications. CHO-AT1R and CHO cells (2 × 10 5 ) were plated in a 6-well plate overnight and then incubated for one hour at 4 • C with 10 µCi (370 kBq) of [ 18 F]AMBF 3 Los per well, in presence or absence of the AT 1 R blocker (losartan potassium (100 µM/well) in PBS). To saturate the AT 1 R, the cells of the blocking group were treated with losartan 30 min before adding the radiolabelled compound. At the end of the incubation period, the supernatant was aspirated and cells were washed six times with ice-cold PBS, and further removed with cell scraper and transferred to a gamma counter tube. The cellular activity was counted in a Cobra II gamma counter (Packard).

Animals
Experiments involving animals were approved by the Animal Ethics Committee of the University of British Columbia (Protocol number: A16-0128). Immunodeficient NOD.Cg-Prkdc scid II2rg tm1Wjl /SzJ (NSG) mice were obtained from an in-house breeding colony at the Animal Resource Centre at BC Cancer Research Facility. Mice were always divided into baseline and AT 1 R blocked groups to assess the AT 1 R binding specificity of [ 18 F]AMBF 3 Los with the co-injection of the AT 1 R blocker losartan (18 mg/kg, dissolved in PBS).

Imaging and Biodistribution Studies
The in vivo AT 1 R binding specificity of [ 18 F]AMBF 3 Los was evaluated in healthy mice by a dynamic PET scan. Healthy NSG mice were injected in the tail vein with 6 ± 1 MBq (170 ± 38 µCi) of [ 18 F]AMBF 3 Los and a sixty-minute dynamic PET scans was carried out on a µPET/CT scanner (Inveon, Siemens) following procedures reported previously [67]. Alternatively, [ 18 F]AMBF 3 Los was concomitantly injected with the AT 1 R blocker losartan (18 mg/kg) before the dynamic acquisition. Before and during the acquisition, mice were sedated with a continuous flow of isoflurane anesthesia (2% isoflurane in oxygen). The µPET/CT images were analyzed with PMOD v3.3 software (PMOD Technologies, Zürich, Switzerland). An elliptical volume-of-interest that enclosed the entire kidney was positioned manually on the PET images for the determination of the kidney volume. Then, 3-dimensional isocontours were drawn automatically. For each VOI, the percentage of the injected dose in the region of interest, divided by the injected dose per animal weight (% ID/g) could be calculated directly using the output parameters from the µPET.
After the sixty-minute dynamic PET scans, anesthetized mice were euthanized by carbon dioxide asphyxiation and blood, fat, testis, intestines, stomach, spleen, liver, pancreas, adrenal glands, kidneys, lungs, heart, muscle, bone and brain were harvested, rinsed in PBS, weighted and counted in a Perkin Elmer (Waltham, MA, USA) Wizard2 2480 automatic gamma counter to quantify the percent of injected dose per gram of tissue (% ID/g).

Autoradiography
Kidneys harvested from healthy NSG mice (after the biodistribution studies), were rinsed in PBS and frozen in isopentane/dry-ice bath. Ten µm-thick slides of the kidneys were obtained using a cryostat (Leica, Wetzlar, Germany), and thaw-mounted onto Superfrost Plus microscope slides (Fischerbrand from Thermo Fisher Scientific, Toronto, ON, Canada), following general procedures of the lab [67]. Afterwards, kidney slides were exposed to a phosphor screen overnight and imaged on a Typhoon FLA 9500 scanner (GE Healthcare, Chicago, IL, USA).

Statistical Analysis
Data was expressed as the mean ± standard deviation (SD). The statistical analysis was performed using GraphPad Prism 7.01 Software (San Diego, CA, USA). Data was analyzed by one unpaired t-test (multiple t tests). The outliers were removed before analyzing the data. A p-value < 0.05 was considered statistically significant.

Conclusions
Both FEtLos and AMBF 3 Los were successfully synthesized as novel derivatives of losartan. The radiolabelled derivative [ 18 F]FEtLos was synthetized high radiochemical purity and low molar activity due to the small amount of 18 F used, and resulted in a less lipophilic compound than losartan. Still, [ 18 F]FEtLos exhibited a low AT 1 R binding affinity. On the other hand, [ 18 F]AMBF 3 Los was synthetized with a high radiochemical purity and molar activity. Moreover, [ 18 F]AMBF 3 Los was slightly hydrophilic, and showed a high AT 1 R binding affinity and specificity. Our data suggests that [ 18 F]AMBF 3 Los might be a valuable PET imaging tracer for monitoring AT 1 R expression in several diseases.