Introduction

Bone metastasis occurs in almost all cancers, most frequently with prostate, lung, and breast cancers. Nearly 80 % of patients with advanced breast or prostate cancer would develop bone metastasis [1, 2]. These malignant tumor metastases have strong impact on the quality of living for those patients. Severe complications include bone pain, pathologic fracture, hypercalcemia, and spinal cord compression [3]. To detect the bone metastasis at early stage, novel bone imaging agents with high bone affinity are needed to improve the diagnostic accuracy of routine bone scanning [4].

Several 99mTc-labeled phosphonates have been developed for skeletal imaging, including pyrophosphates [5], polyphosphates [6, 7], and diphosphonates (DPs) [8, 9]. For example, the conjugates between methylenediphosphonate (99mTc-MDP) and hydroxy-methylenediphosphonate (99mTc-HMDP) and technetium-99m have been widely used both experimentally and clinically for detection of skeletal metastases and other focal bone lesions [8, 10, 11]. Technetium-99m has excellent physical characteristics (t 1/2 = 6.02 h, E γ = 142 keV) and can be easily available from a generator. It has become the most important nuclide for organ imaging in nuclear medicine. Phosphonate derivatives are good targeting ligands for technetium-99m conjugates to be used as bone imaging agents [8], since Fleisch [12] described that phosphonates had high affinity for bone mineral.

Diphosphonates are one class of important clinical drugs for bone disease [13] and are used in the therapy of early cancer treatment to prevent adverse effects of therapies on the bone health [14]. A number of DPs are commercially available, such as clodronate, etidronate, pamidronate, alendronate, ibandronate, tiludronate, incadronate, risedronate, and zoledronate (ZL). However, for these 99mTc-DPs, an interval of 2–6 h is needed between injection and bone imaging [15]. Shorting this interval would lessen the burden on patients in terms of examination time. To enable imaging at earlier time after injection, a radiopharmaceutical with higher affinity for bone, larger ratio of bone-to-soft tissue uptake, and more rapid clearance from blood is required [16].

The structure of the tether attached to diphosphates in DPs determines their biological activities. When the carbon chain of the tether increases, the binding affinity for bone increases significantly. For example, from pamidronate to alendronate where the carbon chain increases from 1 to 2, the binding affinity increased 10 times [17]. The dialkylation of the nitrogen with methyl and pentyl groups in pamidronate generated ibandronate. The potency of the latter was 50 times of the former [18]. ZL is the most potent third-generation DPs, and the tether has an imidazole ring [19]. We had evaluated a series of derivatives of ZL with different alkyl substituents in the imidazole ring and different number of methylene linker between the imidazole ring and geminal diphosphonate groups in our previous publications [20–27]. We found significant improvement on bone absorption and clearance from blood and soft tissues.

In this paper, we report our evaluation on a new series of ZL derivatives (Scheme 1). We examined their in vitro stability and in vivo biodistributions in mice.

Scheme 1
scheme 1

Syntheses of 99mTc-EIPrDP, 99mTc-EIBDP, and 99mTc-EIPeDP

Full size image

Experimental

Reagents, instruments and animals

All analytical chemical reagents were purchased from commercial sources and used without further purification. Na99mTcO4 and human serum were supplied by affiliated Jiangyuan Hospital of Jiangsu Institute of Nuclear Medicine. Melting points were measured on Yanaco MP-500 melting point apparatus (Shimadzu, Japan). Elemental analysis was carried out using an Elementar Vario EL III analyzer. Electron spray ion (ESI) mass spectra were determined using a Waters Platform ZMD4000 LC/MS (Waters, USA). Proton nuclear magnetic resonance (1H NMR) spectra were obtained on a Bruker DRX-500 spectrometer (Bruker, Germany), and the chemical shift values were referenced to the internal tetramethylsilane (TMS). The Xinhua chromatography paper (Shanghai, China) was used for Mini-Scan and Flow-count thin paper chromatography (TLC) (BIO-SCAN, USA). A Packard-multi-prias γ counter (Perkins Elmer, USA) were used for radioactive counting. The high performance liquid chromatography (HPLC) system was equipped with a Waters 1525 binary HPLC pump, a Waters 2487 dual wavelength absorbance detector, and a Perkin Elmer Radiomatic 610TR radioactivity detector, which were operated by Breeze and proFSA software. A reverse phase C18 (RP-C18) column (4.6 × 250 mm; 10 μm particle size) was used for HPLC analysis (Elite Analytical Instrument Company, Dalian, China).

Institute of Cancer Research (ICR) mice (weighing 18–24 g) were supplied by Shanghai SLAC Laboratory Animal Co. Ltd. (Shanghai, China). The animal experiment in this study was approved by the Animal Care and Ethics Committee of Jiangsu Institute of Nuclear Medicine.

Syntheses of EIPrDP, EIBDP, and EIPeDP

These diphosphonic acids were synthesized according to the procedure outlined in Scheme 1 [26–29].

Syntheses of compounds 2

Compound 1 (0.1 mol) was dissolved in CH2Cl2 (75 mL), followed by adding KOH (8.4 g, 0.15 mol), K2CO3 (13.8 g, 0.0835 mol) and N(n-Bu)4Br (0.7 g, 0.002 mol). The solution was treated with homolog of ethyl bromoacetate (0.1 mol) and heated to reflux for 7 h. To remove the inorganic salt, the solution was filtered and the filtrate was washed with brine solution. After drying, the organic phase was evaporated to give the ester of compound 2, which was used without further purification. HCl (1.2 mol/L, 100 mL) was added to the ester of compound 2 and the mixture was refluxed for 8 h. The solution was concentrated and the residue was recrystallized from isopropanol to give the product 2.

3-(2 - Ethyl-1H-imidazol-1-yl) propanoic acid ( 2a ): Yield, 58 %. mp 105–107 °C; ESI–MS, m/z (%): 167 (100, [M–H]+).

4-(2-Ethyl-1H-imidazol-1-yl) butanoic acid ( 2b ): Yield, 36 %. mp 145–147 °C; ESI–MS, m/z (%): 181(100, [M–H]+).

5-(2-Ethyl-1H-imidazol-1-yl) pentanoic acid ( 2c ): Yield, 60 %; ESI–MS, m/z (%): 195(100, [M–H]+).

Syntheses of compounds 3

Compound 2 (20 mmol) was dissolved in chlorobenzene (25 mL) and heated to 120 °C for 30 min, then phosphoric acid (85 %, 4.2 mL) was added. Phosphorus trichloride (7.6 mL) was added dropwise, and the reaction mixture was kept at 120 °C for 4 h. The chlorobenzene was decanted. The yellow residue was redissolved in HCl (9 mol/L, 20 mL) and heated to reflux for 5 h. Charcoal was added before filtration to decolor the solution. After filtration, the solvent was removed. Finally, the crude product was recrystallized from ethanol to give the white crystalline product 3.

1-Hydroxy-3-(2-ethyl-1H-imidazol-1-yl)propane-1,1-diyldiphosphonic acid ( 3a ): Yield, 43 %. mp 185–188 °C; 1H NMR (400 MHz, D2O): δ7.348 (d, J = 1.6 Hz, 1H, CH-ring), 7.235 (d, J = 1.6 Hz, 1H, CH-ring), 4.395–4.354 (m, 2H, N–CH2), 2. 956 (dd, J = 7.6 Hz, 2H, ring-CH2), 2.429–2.319 (m, 2H, HO–C–CH2), 1.305–1.267 (t, J = 7.6 Hz, 3H, –CH3); ESI–MS, m/z (%): 313(100, [M–H]+). Anal. calcd for C8H16N2O7P2 (%): C, 30.58; H, 5.13; N, 8.92; Found (%): C, 30.62; H, 5.23; N, 9.11.

1-Hydroxy-4-(2-ethyl-1H-imidazol-1-yl)butane-1,1-diyldiphosphonic acid ( 3b ): Yield, 31 %. mp 200–203 °C; 1H NMR (400 MHz, D2O): δ7.362 (d, J = 1.6 Hz, 1H, CH-ring), 7.262 (d, J = 1.6 Hz, 1H, CH-ring), 4.111 (t, J = 7.2 Hz, 2H, N–CH2), 2.990–2.933 (dd, J = 7.6 Hz, 2H, ring-CH2), 2.149–2.091 (m, 2H, HO–C–CH2), 1.954–1.915 (m, 2H, CH2–CH2–CH2), 1.309–1.271(t, J = 7.6 Hz, 3H,–CH3); ESI–MS, m/z (%): 327(100, [M–H]+). Anal. calcd for C9H18N2O7P2 (%): C, 32.94; H, 5.53; N, 8.54; Found (%): C, 33.02; H, 5.64; N, 8.61.

1-Hydroxy-5-(2-ethyl-1H-imidazol-1-yl)pentane-1,1-diyldiphosphonic acid ( 3c ): Yield, 29 %. mp 75–78 °C; 1H NMR (400 MHz, D2O): δ7.324 (d, J = 1.6 Hz, 1H, CH-ring), 7.251 (d, J = 1.6 Hz, 1H, CH-ring), 4.084 (t, J = 7.2 Hz,2H, N–CH2), 3.610–3.557 (dd, J = 7.2 Hz, 2H, ring-CH2), 2.963–2.906 (q, 2H, HO–C–CH2), 1.849–1.776 (m, 2H, CH2–CH2–COH), 1.306–1.269 (m, 2H, ring-CH2CH2), 1.129–1.094(t, J = 6.8 Hz, 3H, –CH3); ESI–MS, m/z (%): 342(100, [M–H]+). Anal. calcd for C10H20N2O7P2 (%): C, 35.10; H, 5.89; N, 8.19; Found (%): C,35.08; H, 6.02; N, 8.29.

Radiolabeling

To a 10 mL vial, 100 μL 3a–3c aqueous solution (0.25 g DPs dissolved in 5.0 mL 0.2 M sodium hydroxide solution), 100 μL freshly prepared solution of stannous chloride dihydrate (10 mg SnCl2·2H2O dissolved in 10.0 mL 0.5 M HCl solution), and 74.0 MBq freshly eluted Na99mTcO4 were added. After the pH value of the reaction system was adjusted to 6.0 by adding 0.2 M phosphate buffer solution (PBS) and diluted to 2 mL, the reaction mixture was kept at 70 °C for 30 min.

Quality control

The radiolabeling yield (RLY) of 4a–4c was determined by TLC. Strips of Xinhua No.1 paper chromatography of 13 cm long and 0.5 cm wide were marked at 1.5 cm from the bottom and was lined into sections 1 cm each, up to 10 cm. About 3 μL 99mTc-EIPrDP, 99mTc-EIBDP and 99mTc-EIPeDP solution was applied with a syringe at 1.5 cm from the bottom of each paper strip, and then these strips were developed in distilled water and acetone. After complete development, the strips were dried and assayed by Mini-Scan TLC. The RLY of 4a–4c were determined as follows: % RLY = 100 % − % free 99mTcO4 − % 99mTc-colloidal.

The radiochemical purity (RCP) of 4a–4c was determined by HPLC. The sample was filtered through a 0.22 μm Millipore filter carefully and 10 μL solution was injected into the HPLC column. The column was eluted with isocratic solvents of 70 % water and 30 % acetonitrile, and the flow rate was 1.0 mL/min. Radio-analysis of the labeled compound was performed using a Cd (Te) detector.

In vitro stability

The in vitro stability of 99mTc-EIPrDP, 99mTc-EIBDP and 99mTc-EIPeDP were studied in PBS (pH = 7.4), fetal bovine serum (FBS), or human serum (HS) for 1–6 h. Briefly, 200 μL (3.7 MBq) of 99mTc-EIPrDP, 99mTc-EIBDP and 99mTc-EIPeDP were pipetted into 200 μL of PBS, FBS, or HS, respectively. After incubation at 37 °C for 1–6 h, an aliquot of PBS solution were taken directly and radioactivity was analyzed with TLC. For the solution of FBS and HS, the aliquot were added to 100 μL of 50 % trifluoroacetic acid (TFA). After centrifugation, the upper solution was taken for TLC analysis.

Determination of octanol–water partition coefficient (Log P)

The octanol–water partition coefficient was determined for 4a–4c at pH = 7.4 by measuring the distribution of the radiolabeled compounds in n-octanol and PBS, respectively. A 100 μL sample of 4a–4c was diluted with 900 μL PBS. Then, it was mixed with 1 mL n-octanol and vortexed for 5 min. The mixture was further centrifuged at 4,000 rpm for 5 min to ensure complete separation of layers. 100 μL aliquots of organic and aqueous phases were collected with pipettor and the radioactivity was measured with a γ counter, respectively. The octanol–water partition coefficient (log P) was calculated using the formula log P = log (octanol CPM/PBS CPM) [30]. The reported value is the average obtained from three independent experiments.

Plasma protein binding assay

4a–4c (100 μL, 37 KBq) were mixed with the human plasma (100 μL) in the centrifuge tube. After the mixture was incubated at 37 °C for 2 h, the plasma protein was precipitated by adding 1 mL trichloroacetic acid (250 g/L) to the mixture. The supernatant and precipitate were separated by centrifugation at 3,000 rpm for 5 min. The radioactivities of both phases were measured separately. The above experimental procedure was repeated three times. The percentage of protein binding was determined by the following equation: Plasma protein binding % = (Precipitate CPM)/[Precipitate CPM + Plasma CPM]) × 100 %.

Kinetics of blood clearance

For pharmacokinetic study, 4a–4c (7.4 MBq, 0.2 mL aqueous solution) was administered to the mice via intravenous injection through the tail vein. A series of blood samples (20 μL) were collected in the microcap tubes by nicking the tail with a needle at 2, 5, 10, 15, 30, 60, 120, 180, and 240 min after injecting 4a–4c. The radioactivity of the blood samples was counted and expressed as %ID/g. Pharmacokinetic parameters were analyzed by the 3P97 software (Practical Pharmacokinetic Program from China’s pharmacological to learn mathematics professional committee), and the radioactivity can be expressed as a function of time with the following equation

$$ C \, = {\text{ Ae}}^{{ - \alpha {\text{t}}}} + {\text{ Be}}^{{ - \beta {\text{t}}}} . $$

In vivo distribution

Thirty mice with the body weight range of 18–24 g were randomly divided into six groups and injected via the tail vein with each radioactive agent of 0.2 mL aqueous solution and approximately 7.4 MBq of radioactivity. The mice were sacrificed by decapitation at 5, 15, 30, 60, 120 and 240 min post injection. Interested organs were collected and weighed, and 200 μL of blood were taken from carotid artery. The radioactivity of each sample was measured using a γ counter. The distribution of the radioactivity in different organs was calculated and expressed as  %ID/g = \( \frac{\text{organ CPM}}{{{\text{organ weight }} \times {\text{ injected CPM}}}} \) × 100 %.

Results and discussion

Chemistry and radiolabeling

EIPrDP, EIBDP, and EIPeDP were synthesized in four steps from the starting material 2-ethylimidazole (Scheme 1). Nucleophilic substitution of the bromoestes generated the imidazole-substituted esters. Hydrolysis of the esters under acidic conditions gave the carboxylic acid 2a–c. Nucleophilic addition of PCl3 to carboxylic acids under strong acidic conditions generated α-hydroxy phosphoric chlorides, which hydrolyzed in water to give diphosphates 3a–c.

99mTc-EIPrDP, 99mTc-EIBDP and 99mTc-EIPeDP were synthesized from the reaction of the diphosphates and 99mTcO4 − under reducing conditions (scheme 1). TLC analysis showed that Na99mTcO4 was completely reduced and 99mTc-colloidal amount was less than 2 % (results of 4a is shown in Fig. 1). All three tracers showed similar results. HPLC analysis revealed that free technetium (Na99mTcO4) eluted at 8.9 min, while 4a–4c eluted at 2.8 ± 0.1 min (Fig. 2). The nearly identical retention times of 4a–c indicated the structural analogy of these 99mTc-DPs [17]. Furthermore, the single peak in the HPLC-chromatogram clearly shows one complex, no residual Na99mTcO4, or other 99mTc impurities [28]. This labeling method meets clinical requirements for production of other 99mTc-labeled diphosphonates, such as 99mTc-MDP. These radiolabeled compounds were used immediately for both in vitro and in vivo studies.

Fig. 1
figure 1

TLC analyses of 4a: a developed in acetone (R f values for 99mTcO2·nH2O and 4a were 0–0.1, for Na99mTcO4, 0.9–1); b developed in H2O (R f value for 99mTcO2·nH2O was 0–0.1, for 4a was 0.8–0.9, for Na99mTcO4 was 0.9–1)

Full size image
Fig. 2
figure 2

HPLC analyses: a 4a rt = 2.75 min, 4b rt = 2.79 min, 4c rt = 2.85 min; b 99mTcO4− rt = 8.95 min)

Full size image

In vitro stability

More than 95 % of 99mTc-EIPrDP, 99mTc-EIBDP and 99mTc-EIPeDP remained intact in the PBS, FBS and HS after 6 h of incubation (Fig. 3). The results indicated that these complexes are stable enough for biodistribution and imaging studies.

Fig. 3
figure 3

In vitro stability of 4a–4c in PBS, FBS, and human serum after incubation at 1 h and 6 h at 37 °C, respectively. Data were expressed as the mean ± SD

Full size image

Log P value and plasma protein binding

The octanol–water partition coefficient (log P) for 99mTc-EIPrDP, 99mTc-EIBDP and 99mTc-EIPeDP were −1.70, −1.67, −1.65 in PBS at pH = 7.4 (Table 1). The lipophilicity of these complexes increases with the length of the carbon chain between the imidazolyl and geminal diphosphonates group. The log P values are important parameters for the biological distribution of these complexes [31].

Table 1 Values of octanol–water partition coefficient and plasma protein binding
Full size table

The percentages of plasma protein binding for 99mTc-EIPrDP, 99mTc-EIBDP, and 99mTc-EIPeDP, were 23.12, 22.74, and 23.24, respectively (Table 1). In general, the binding has significant influence on the bone uptake in the biodistribution [32]. However, almost no change in plasma protein binding was observed for this series of complexes, although the length of the carbon chain increased from 2 to 4. In contrast, the log P values and plasma protein binding observed here are larger than those for 99mTc-MIPrDP, 99mTc-MIBDP and 99mTc-MIPeDP [26], reflecting much stronger influence of substituents in the imidazole ring.

Kinetics of blood clearance

Pharmacokinetics of 4a–4c can be fitted with double exponential equations, C = 12.11e−0.110t + 4.36e−0.016t, C = 19.78e−0.101t + 3.27e−0.013t and C = 17.38e−0.057t + 0.794e−0.0046t, during 6 h post injection, respectively (Table 2; Fig. 4).

Fig. 4
figure 4

Pharmacokinetic curves of 99mTc-EIPrDP, 99mTc-EIBDP and 99mTc-EIPeDP in mice (n = 5, mean ± SD)

Full size image

The K12 and K21 of 99mTc-EIPeDP were 0.043 and 0.041 min−1 respectively, which were higher than those of 99mTc-EIBDP or 99mTc-EIPeDP. The values of CL were 0.95, 0.856 and 0.782 %ID/g min−1 and the AUC were 388, 432, 472 for 99mTc-EIPrDP, 99mTc-EIBDP and 99mTc-EIPeDP, respectively. The blood clearance of 99mTc-EIPrDP was faster than that of 99mTc-EIBDP or 99mTc-EIPeDP in the early times.

Table 2 Pharmacokinetic parameters of 4a–4c in mice
Full size table

Biodistribution studies

The complexes 4a–4c mainly accumulated in the bone, kidneys, and liver (Table 3). The uptake of 4a–4c in bone peaks from 60 min to 120 min post injection at 13.3 ± 1.23, 11.7 ± 0.28 and 8.69 ± 0.04 %ID/g for 99mTc-EIPrDP, 99mTc-EIBDP and 99mTc-EIPeDP, respectively. From previous work of our group, we found that the bone uptake increases along with the increase of carbon chain between the imidazolyl and geminal diphosphonate groups in ZL (from 99mTc-ZL to 99mTc-IPeDP) [25, 27]. The same trend was observed for 99mTc-labeled methyl-substituted DPs (from 99mTc-MIPrDP to 99mTc-MIPeDP) [26]. At constant carbon chain length, the bone uptake increases along with size of the alkyl substituents in the imidazole ring. The bone uptake at 5 and 60 min for 99mTc-IPrDP [25], 99mTc-MIPrDP [26], and 99mTc-EIPrDP (Table 3), were 3.19 and 4.69, 6.43 and 12.2, and 8.63 and 13.3 %ID/g, respectively. Here the trend of bone uptake has reversed, with 99mTc-EIPeDP at lowest, indicating that optimal number of the carbon chain exists for bone uptake of the radiotracers, with 99mTc-EIPrDP as the best so far.

Table 3 Biodistribution of 4a–4c in mice (mean ± SD, n = 5)
Full size table

No sign of toxicity through the study period was observed for 4a–4c. This is consistent with the general observation that ZL has low toxicity and can be used therapeutically at a high dose [33, 34].

4a–4c also showed predominant kidney and liver uptake (Table 3). After 2 h post injection, the kidneys uptakes of 4a–4c were 2.89 ± 0.11, 6.24 ± 0.23 and 3.25 ± 0.44 %ID/g, respectively; and the liver uptakes were 0.81 ± 0.02, 0.49 ± 0.03 and 1.50 ± 0.06 %ID/g, respectively. 99mTc-EIPeDP showed the highest liver uptake, suggesting that it not only cleared through the kidneys, but also was excreted by the liver. This kind of metabolic pathway increases the burden of liver.

Furthermore, 4a–4c cleared from the blood rapidly, with 5.59 ± 0.31, 5.15 ± 0.27 and 8.51 ± 0.17 % ID/g at 5 min post injection, and 0.10 ± 0.01, 0.25 ± 0.01 and 0.26 ± 0.01 %ID/g 2 h post injection (Table 3). 99mTc-EIPeDP has higher radioactivity level than 99mTc-EIPrDP or 99mTc-EIBDP. This is consistent with the previous observation that 99mTc-EIPeDP excreted more from liver than 99mTc-EIPrDP or 99mTc-EIBDP does.

The uptake ratios of bone to muscle, spleen and heart for 99mTc-EIPrDP were larger than those of 99mTc-EIBDP and 99mTc-EIPeDP, while the ratio of bone to liver is the best for 99mTc-EIBDP (Fig. 5). The results showed that 99mTc-EIPrDP had higher selective uptake in the skeletal system and lower background uptake in soft tissues, and held great potential as a better bone-imaging agent among these 99mTc-DPs. No advantage was observed for elongating the carbon chain between the ethyl-imidazolyl and geminal diphosphonate groups. 99mTc-EIPrDP is worthy of further investigation.

Fig. 5
figure 5

Uptake ratios of bone to soft tissues in mice at different time post injection of 4a–4c

Full size image

The uptake ratios of bone-to-soft tissue for 99mTc-EIPrDP are the best, compared with other 99mTc-DPs previously evaluated in our lab. At 60 min post injection of 99mTc-EIPrDP, the ratios of bone to muscle, liver, spleen were 111, 13, 95, respectively, while those for 99mTc-IPrDP and 99mTc-MIPrDP were 18, 0.14, 0.46 and 64, 13, 29, respectively [25–27].

The bone uptake of 99mTc-EIPrDP peaks the earliest at 60 min among 99mTc-IPrDP [25, 27], 99mTc-MIPrDP [26], 99mTc-EIBDP and 99mTc-EIPeDP. In the clinical applications, shorter interval between injection and bone imaging would reduce the total examination time for patients.

Conclusions

A series of novel complexes of technetium-99m and ethyl-substituted zoledronic acids (99mTc-EIPrDP, 99mTc-EIBDP and 99mTc-EIPeDP) were prepared, with high radiolabeling yield and radiochemical purity (96 ± 2 %). They were stable up to 6 h in PBS, FBS and HS. The biodistribution in mice showed that 99mTc-EIPrDP was best among the complexes examined. Moreover, 99mTc-EIPrDP shows higher bone accumulation in the skeletal system and lower background in soft tissues at 60 min, as compared with 99mTc-IPrDP and 99mTc-MIPrDP. We will continue to explore 99mTc-EIPrDP as a SPECT imaging agent in larger animals and animal models of bone metastases.