Dichloromethane, acetate, sodium thiosulfate, chloroform, acetonitrile, potassium dihydrogen phosphate, dipotassium hydrogen phosphate, ethylene diamide tetra acetic acid (EDTA), and glucose-6-phosphate (G6P) were purchased from Nacalai Tesque (Kyoto, Japan). β-Nicotinamide-adenine dinucleotide phosphate (β-NADP⁺) and isoflurane were purchased from Wako Pure Chemical Industries (Osaka, Japan), and glucose-6-phosphate dehydrogenase (G6PD) was purchased from Oriental Yeast (Osaka, Japan). tert-Butyl hydroperoxide (t-BuOOH) was purchased from Tokyo Chemical Industry (Tokyo, Japan). ¹²⁵I-NaI was purchased from American Radiolabeled Chemicals (St. Louis, United States). ¹²³I-NaI was purchased from FUJIFILM Toyama Chemical (Tokyo, Japan). Mequitazine was provided by Sumitomo Chemical (Tokyo, Japan). All other reagents and chemicals were of analytical or high-performance liquid chromatography (HPLC) grade.

Mequitazine {10-[(3RS)-1-azabicyclo [2.2.2] oct-3-ylmethyl]-10H-phenothiazine} is a long-acting and selective histamine H₁–receptor antagonist used clinically. To develop a SPECT imaging probe using mequitazine, 4-(tributylstannyl) benzyl bromide was synthesized as a precursor. Mequitazine was labeled with radioiodine in two steps (Figure 1). In the first step, 1.0 μL of 4-(tributylstannyl) benzyl bromide was diluted by adding 50 μL of dichloromethane. This solution was mixed with 20 μL of acetic acid (70%), 33 μL of t-BuOOH, and 3.7 MBq of ¹²⁵I-NaI. The reaction mixture was then incubated at room temperature for 5 min. The oxidation reaction was stopped by the addition of saturated sodium chloride solution and heated to reflux under N₂. The resultant product was ¹²⁵I-4-iodobenzyl bromide. In the second step, ¹²⁵I-4-iodobenzyl bromide was diluted by adding 1 mL of acetonitrile. This solution was mixed with 32 mg of mequitazine and incubated at room temperature for 5 min. The resultant product was ¹²⁵I-4-iodobenzylmequitazine (this new SPECT imaging probe was designated ¹²⁵I-BMQ). Separation and purification of ¹²⁵I-BMQ was carried out on an HPLC system consisting of a pump (model L-7100, Hitachi, Japan), UV detector (Chromaster 5410, Hitachi, Japan), and γ-ray detector (model RLC-701, Hitachi-Aloka Medical, Japan) equipped with a 5C₁₈ AR-II-column (4.6 mm × 250 mm; 5 μm, Nacalai Tesque, Kyoto, Japan). PowerChrom (ver. 2.3.3, eDAQ, Japan) was used for data processing. The mobile phase consisted of acetonitrile (solvent A) and 20 mM phosphate buffer (pH 2.3) (solvent B). Separation was carried out as follows: 35% solvent A; 65% solvent B, 0–2 min; 35%–50% linear gradient, 2–9.5 min; 50%–80% linear gradient, 9.5–16 min; flow rate 1.0 mL/min. The elution of ¹²⁵I-BMQ was monitored at 250 nm UV. For UV chromatograms, mequitazine and ¹²⁷I-BMQ (cold-labeling) were mixed for comparison of radioisotope (RI) chromatograms (hot ¹²⁵I-labeling). The structure of ¹²⁷I-BMQ was confirmed by proton nuclear magnetic resonance spectroscopy (¹H-NMR, JEOL JNM-ECS400; Jeol Resonance Inc., Tokyo, Japan). ¹H-NMR data were as follows: (400 MHz, CDCl₃/TMS) δ: 7.66 (d, J = 8.2 Hz, 2H), 7.34 (d, J = 8.7 Hz, 2H), 7.17–7.15 (m, 4H), 6.95 (t, J = 7.8 Hz, 4H), 5.00 (d, J = 12.8 Hz, 1H), 4.96 (d, J = 12.8 Hz, 1H), 4.13 (dd, J = 14.2, 9.6 Hz, 1H), 3.98 (dd, J = 14.2, 6.0 Hz, 1H), 3.77–3.72 (m, 4H), 3.57–3.46 (m, 2H), 2.66 (br s, 1H), 2.28–2.27 (m, 1H), 1.96–1.81 (m, 4H). ¹³C-NMR data were as follows: (100 MHz, CDCl₃) δ: 144.56, 138.31, 134.87, 127.79, 127.71, 126.34, 126.11, 123.26, 116.10, 97.49, 65.96, 57.27, 54.13, 53.92, 48.29, 31.98, 24.86, 21.87, 19.54.

Animal studies were approved by the Animal Care Committee at Kanazawa University (AP-173851) and conducted in accordance with international standards for animal welfare and institutional guidelines. For preparation of pooled mouse liver microsomes (MLMs), three 6-week-old male ddY mice (Japan SLC, Tokyo, Japan) were euthanized under anesthesia using isoflurane. The liver of each mouse was removed and weighed. After adding 3 mL of Krebs-Ringer’s phosphate buffer (pH 7.4) per gram of liver, the livers were pooled and homogenized using an ultrasonic homogenizer (SONIFIER250, Branson, MO, United States). The homogenate was centrifuged for 20 min at 9,000 g, and the supernatant was centrifuged for 60 min at 100,500 g. The resulting precipitate was suspended with Krebs-Ringer phosphate buffer and centrifuged for 60 min at 10,500 g. The resulting precipitate was obtained as microsomes, and the protein content was determined according to the bicinchoninic acid method (Smith et al., 1985). The pooled MLMs were stored at −80°C.

To analyze CYP-mediated metabolism of ¹²⁵I-BMQ, nicotinamide adenine dinucleotide phosphate (NADPH) was used as an energy source for CYP. The NADPH-generating system created a CYP-activated state for ¹²⁵I-BMQ metabolism. Metabolism of ¹²⁵I-BMQ via CYP was carried out using the NADPH-generating system (0.5 mM β-NADP⁺, 5 mM MgCl₂, 5 mM G6P, 1 U/mL G6PD), 100 mM sodium potassium phosphate buffer (pH 7.4), 50 μM EDTA disodium salt, and 1,000 μg protein/20 μL pooled MLMs in a final volume of 250 μL (NADPH [+]). The same mixture without the NADPH-generating system was used as control (NADPH [−]). Samples were incubated at 37°C for 60 min with gentle shaking, and the reaction was stopped by adding 100 µL of ethanol. The samples were then centrifuged at 18,000 g for 5 min, and the supernatant of each sample was analyzed by thin layer chromatography (TLC). The supernatant was spotted onto a silica gel TLC plate 60 F₂₅₄ (Merck, Darmstadt, Germany), which was developed using methanol/acetic acid in a 100:1 ratio. After development and complete drying, the TLC plates were cut into 21 fractions, and the radioactivity associated with each fraction was measured using a γ-ray counter (AccuFLEX γ 7000, Aloka, Tokyo, Japan). The fractionation ratio of ¹²⁵I-BMQ, ¹²⁵I-BMQ, and other metabolites was calculated by dividing the radioactivity count for each fraction by the total radioactivity count.

In order to analyze the inhibition of specific ¹²⁵I-BMQ-metabolizing CYP isoenzymes, specific inhibitors of the human CYP isoenzymes employed in this study were used, including α-naphthoflavone (inhibitor of CYP1A1 and 1A2) (Kim et al., 2005; Dinger et al., 2014; Valicherla et al., 2019), sulfaphenazole (inhibitor of CYP2C9) (Kim et al., 2005; Turpeinen et al., 2005; Wang et al., 2010; Dinger et al., 2014; Valicherla et al., 2019), fluconazole (inhibitor of CYP2C19) (Dinger et al., 2014; Valicherla et al., 2019), paroxetine and quinidine (inhibitors of CYP2D6) (Kim et al., 2005; Turpeinen et al., 2005; Wang et al., 2010; Dinger et al., 2014; Valicherla et al., 2019), 4-methylpyrazole (inhibitor of CYP2E1) (Wang et al., 2010; Valicherla et al., 2019), and ketoconazole (inhibitor of CYP3A4) (Kim et al., 2005; Turpeinen et al., 2005; Valicherla et al., 2019). These specific inhibitors were used to identify the CYPs involved in the biotransformation of ¹²⁵I-BMQ. The inhibitors and mequitazine (self-inhibitor) were dissolved at a concentration of 10 μM (0.8% [v/v] final concentration DMSO).

A mixture consisting of 50 µL of NADPH-generating system (NADPH [+]), 50 µM disodium EDTA salt in 100 mM sodium potassium phosphate buffer (pH 7.4), 1,000 µg protein/20 µL pooled MLMs, and 25 µL of inhibitor in a final volume of 250 µL was used to examine the inhibition of ¹²⁵I-BMQ metabolism. Samples were incubated at 37°C for 15 min with gentle shaking, and the reaction was stopped by addition of 100 µL of ethanol and centrifugation at 18,000 g for 5 min. The final supernatant was analyzed by TLC using methanol/acetic acid at a 100:1 ratio. The percentages of ¹²⁵I-BMQ metabolites in each sample were then calculated (n = 3).

¹²⁵I-BMQ was prepared to a specific radioactivity of 185 kBq/mL by adding saline. A total of 20 fasting 6-week-old male ddY mice were injected with ¹²⁵I-BMQ into the tail vein (18.5 kBq/100 µL/mouse), and after 2, 10, 30, 60, and 120 min, four mice each were euthanized under isoflurane, and blood, brain, thyroid, lung, heart, liver, gallbladder, stomach, pancreas, spleen, intestine, kidney, and urine tissue were collected. The tissue was weighted and the radioactivity was measured using a γ-ray counter to calculate the injected dose percent (%ID) or injected dose percent per gram of tissue (%ID/g).

¹²⁵I-BMQ was prepared to a specific radioactivity of 10 MBq/mL by adding saline and injected via tail vein into three mice (2.0 MBq/200 µL/mouse). After 30 min, each of the three mice was euthanized under isoflurane, and the gallbladder was removed. Radioactive products in the bile were analyzed by TLC using chloroform/acetic acid/H₂O at a ratio of 60:40:1. Bile was spotted directly onto the TLC plate.

¹²³I-BMQ was labeled and purified according to the same method used for ¹²⁵I-BMQ. Whole-body SPECT imaging of normal mice was performed using a U-SPECT-II/CT system (MILabs, Utrecht, Netherlands). Two normal mice were injected with 14.5 MBq of ¹²³I-BMQ via the tail vein. Whole-body SPECT images were acquired under 2.0% isoflurane anesthesia every 10 min from 5 to 115 min after injection. The images were reconstructed using the filter-backed projection method, with 16 subsets and six iterations. The voxel size was set to 0.8 mm × 0.8 mm × 0.8 mm. Neither attenuation nor scatter correction was performed. Post-reconstruction smoothing filtering was applied using a Gaussian smooth 3D filter at 1.2 mm. Images were displayed using PMOD (ver. 3.7). On SPECT images, volumes of interest were drawn for the liver, intestines (duodenum, small and large intestines), and kidney. Respective %ID values were calculated over time.

For whole-body SPECT imaging in CYP2D-inhibited mice, two mice were injected intraperitoneally with 30 mg/kg of paroxetine (MacLeod et al., 2017). At 60 min after paroxetine injection, 14.5 MBq of ¹²³I-BMQ was injected via the tail vein. Whole-body SPECT images were acquired under 2.0% isoflurane anesthesia every 10 min from 5 to 115 min after ¹²³I-BMQ injection. SPECT images of CYP2D-inhibited mice were analyzed according to the same method used for control mice.

All results represent the mean of at least three experiments, and data are expressed as the mean ± standard deviation. Data were analyzed using the F-test and Student’s t-test, and p < 0.01 or <0.05 was considered statistically significant.

To identify ¹²⁵I-BMQ, ¹²⁷I-BMQ (cold-label) was used as a reference standard. Figure 2 shows HPLC chromatograms of mequitazine, ¹²⁷I-BMQ (cold-label), and ¹²⁵I-BMQ. The UV chromatogram confirmed the separation of mequitazine (retention time: 8 min) and ¹²⁷I-BMQ (retention time: 13.5 min), and the RI chromatogram showed that the main peak (retention time: 13.5 min) had the same retention time as ¹²⁷I-BMQ (In the HPLC system used in this study, the RI detector is located after the UV detector. The slight difference in retention times of the ^125/127I-BMQ peaks is due to the longer tube between the RI and UV detectors.) This confirmed that two-step ¹²⁵I labeling of mequitazine provided ¹²⁵I-BMQ with a labeling index of approximately 88%–91% and radiochemical purity >98%.

In this study, ¹²⁵I-BMQ and its metabolites were separated by TLC. The rate of flow (Rf) values in this analytical condition ranged from 0.15–0.30 for ¹²⁵I-BMQ and 0.65–0.75 for ¹²⁵I-NaI. In the metabolic reaction without NADPH (NADPH [−]), ¹²⁵I⁻ was not detected and only ¹²⁵I-BMQ was detected. On the other hand, in the NADPH-mediated metabolic reaction (NADPH [+]), as in NADPH [−], ¹²⁵I⁻ was not detected, ¹²⁵I-BMQ was detected, and other radioactive metabolites (Rf value: 0.05–0.10) were detected (Figure 3). With NADPH [+] conditions, the percentage of radioactive metabolite significantly increased from approximately 3%–25% with reaction time, and the percentage of ¹²⁵I-BMQ significantly decreased from approximately 93%–76% with reaction time. These results showed that the radioactive metabolite was an NADPH-mediated metabolite of ¹²⁵I-BMQ.

Figure 4 shows the percentages of radioactive metabolite produced from ¹²⁵I-BMQ. Under the condition with NADPH [−], paroxetine, quinidine, and mequitazine, the percentage of metabolites were 2.64%, 12.5%, 4.47%, and 1.34%, respectively significantly lower than under NADPH [+]. Under the condition with 4-methylpyrazole and ketoconazole, the percentage of ¹²⁵I-BMQ was slightly lower, but not significantly.

Table 1 shows the biodistribution of ¹²⁵I-BMQ in normal mice. The injected ¹²⁵I-BMQ was rapidly distributed to tissues throughout the body. In the thyroid and stomach, radioactivity was consistently low. Radioactivity in the liver increased immediately after injection and then decreased gradually. In the gallbladder, intestine, and kidney, radioactivity increased gradually after injection.

The percentage of radioactive metabolites of ¹²⁵I-BMQ in mouse bile was analyzed by in vitro TLC. In this analytical conditions, the Rf values of ¹²⁵I⁻ and ¹²⁵I-BMQ in bile were 0.45–0.55 and 0.65–0.75, respectively. In the bile of mice injected with ¹²⁵I-BMQ, no ¹²⁵I-BMQ was detected, only ¹²⁵I⁻ and radioactive metabolite of ¹²⁵I-BMQ (Rf values; 0.00–0.35). The percentage of metabolites of ¹²⁵I-BMQ was >90% (Figure 5).

Figure 6 shows SPECT images of ¹²³I-BMQ in normal mice and CYP2D-inhibited mice. In normal mice, radioactivity was found in the liver and intestine (Figure 6A, frame 1). Radioactivity in the liver gradually decreased and that in the intestine gradually increased. Radioactivity excreted in the small intestine translocated to the large intestine (Figure 6A, frame 10). In CYP2D-inhibited mice, radioactivity was detected in the liver but not clearly in the intestine (Figure 6B, frame 1). The radioactivity in the intestine gradually increased, but remained on the duodenal side and did not translocated to the large intestine side (Figure 6B, frame 10).

Figure 7 shows the time-activity curves of ¹²³I-BMQ in the liver and intestines in SPECT images of normal and CYP2D-inhibited mice. CYP2D-inhibited mice accumulated more in the liver than control mice. Accumulation in the intestines was low in the early stages of ¹²³I-BMQ injection, but gradually accumulated to the same level as in control mice.