Preparation and Biological Evaluation of a [55Co]-2-Acetylpyridine Thiosemicarbazone

Due to the anti-proliferative properties of cobalt-thiosemicarbazone complexes, the production of [55Co](III)-bis-(2-acetylpyridine thiosemicarbazone) ([55Co](III)[APTS]2) was investigated. Co-55 (T1/2=17.53 h) was produced by 150 μA irradiation of a natural nickel target by 15 MeV protons. The 55Co was separated from the irradiated target material using a two-step method with a radiochemical yield of >95% followed by radionuclidic and chemical purity control. [55Co](III)chloride was mixed with 2-acetylpyridine thiosemicarbazone for 30 min at room temperature to yield [55Co](III)[APTS]2 (radiochemical purity > 98% shown by RTLC/HPLC). A specific activity of about 10–20 Ci/mmol was obtained. The final solution was diluted in normal saline to 5% ethanolic solution for biological evaluation. The stability of the final product was checked in the absence and presence of human serum at 37°C to 24 h. The partition co-efficient of the final complex at the pH of 7 was 1.00±0.08. A significant tumor accumulation (%ID/g; 3.5%) was observed in tumoral tissue 21 h post injection in fibrosarcoma-bearing mice by biodistribution studies. Co-incidence imaging also demonstrated tumor uptake from 21–35 h however at 35 h tumor uptake is more specific and significant.


Introduction
Cobalt thiosemicarbazone complexes exhibit interesting biological properties specially anti-proliferative effects. For instance, furane containing cobalt complexes have demonstrated potent cytotoxicity against the growth of leukemias, lymphomas, human lung, colon, ovary and uterine carcinoma cell cultures [1]. Some pyridoxal thiosemicarbazone cobalt (III) complexes have shown antileukemic activity toward human cell lines U937 and CEM [2]. In another study, the 9,10-phenanthrenequinone cobalt complexes have exhibited antiproliferative activity in the human breast cancer cell-line, T47D [3].
Cobalt offers a selection of radionuclides suitable for imaging as well as tracing techniques [4]. The most commonly used cobalt radionuclide for tracing in long-period studies is 57 Co emitting photons at energy range of 100-200 keV suitable for SPECT detector systems. 55Co seems a potential radionuclide for positron emission tomography (PET), but emitting high energy photons is a major drawback in imposing unwanted dose to patients (Table 1).

Tab. 1.
Physical characteristics of Co-55 and Co-57 [4] Radionuclide Cobalt-APTS is a well studied complex with characterized structure using spectroscopic and crystallographic methods as well as demonstrating anti-proliferative activity against human cancer cell lines [5].
In continuation of our works on the development of possible radiometal-based imaging agents [6], and vast research works on the pyridine-based metal complexes [2,[7][8][9], (due to their resemblance to pyridoxal metabolites [10]), it was interesting to develop a possible PET imaging agent by incorporating 55 Co into APTS ligand to prepare a radiolabeled complex, i.e. [ 55 Co](III)[APTS] 2 ( Figure 1). Preliminary coincidence imaging and postmortem biodistribution studies in wild-type and fibrosarcoma-bearing animals were also performed.

Cobalt-55 production
Cobalt-55 was produced by bombardment of a 30 μm thick natural nickel target using a 150 μA current of 15-8 MeV protons. The production yield was 270.2 μCi/μAh at the end of bombardment. The radiochemical separation was based on a two step no-carrier-added method resulting in a yield of 95%. The radionuclidic purity was higher than 99.3%. The remaining activity was attributed to 57 Co. No nickel ions were detected at a detection limit of 2 ppm.

Radiolabeling of APTS with [ 55 Co]
It has been shown that pyridoxal thiosemicarbazone cobalt complexes form [Co(HL) 2 ] 2 X in their crystal structure and when dissolved in solution, [Co(HL) 2 ] + forms an octahedral cationic complexes. A more polar complex is formed after incorporation of cobalt cation which is lipophillic and carries a positive charge. Uncomplexed 55 Co in the form of 55 Co 3+ elutes at an R f = 0.8. The radiochemical yields were higher than 98% in each case (n=9) (Figure 2), while the APTS complex moved with an R f of 0.1 ( Figure 3). In HPLC chromatograms the Co 3+ cation is eluted at 1.2 minutes as a fast washing component ( Figure 4), while the free unlabeled ligand elutes at 3.22 minutes ( Figure 5) and the labeled complex elutes at 6.28 minutes ( Figure 6).
The final radiolabeled complex diluted in normal saline was passed through a 0.22 micron filter (Millipore) to sterilize the product, due to possible thermal instability.

Electrophoretic studies
A major component of complex solution migrated to cathode in EDTA 2solution demonstrating the total cationic property of the complex. As shown in Fig.1, considering the application of refluxing HNO 3 to dissolve the Ni target, the most probable cobalt cation would be Co 3+ . Thus, coordination of two nitrogen atoms (2 unpaired electrons in each) in and the two thiol-groups (in anionic form) would result in a positive charge for the complex.

Serum Stability Studies
Incubation of [ 55 Co](III)[APTS] 2 in freshly prepared human serum for 24 h at 37°C showed no significant loss of 55 Co from the complex during the course of the studies after RTLC study of the cut-off filter flow-through, and the radiochemical purity of complex remained at 98% for 24 h under physiologic conditions.

Partition co-efficient of [ 55 Co](III)[APTS] 2
As expected from the RTLC behavior, the lipophilicity of the [ 55 Co](III)[APTS] 2 compound was high as determined by the octanol/water partition coefficient (P) for the 55 Co-complex and was found to depend on the pH of the preparation. At a pH of 7 (final formulation) the lipophilicity was 1.00±0.08. The water solubility of the tracer is a bit changed when the pH is out of 5.5-7 range.

Biodistribution studies in tumor bearing mice
A few hours post-injection, the radioactivity content increased in the kidneys and liver and this pattern remained constant out to 21 hours ( Figure 7).

Bio-distibution of [ 55 Co](III)[APTS] 2 (100μCi IV) in fibrosarcoma-bearing mice 21 h post-injection
Major part uptake of radioactivity accumulated was observed in the reticulloendothelial system including liver and spleen. No significant radioactivity in the tumor at other time points was observed (data not shown) while the best time range for the tracer uptake in tumor showed to be 20-35 h. Intestines exhibited a significant uptake which could be attributed to liver excretion of the tracer or metabolites. A significant tumor/muscle uptake ratio (80.1) was obtained.  Animal experiments were carried out in compliance with the United Kingdom Biological Council's Guidelines on the Use of Living Animals in Scientific Investigations, 2 nd edn. Mice weighing 150-200 g were purchased from Razi Institute of Iran. Images were taken using a dual-head SPECT system (SMV, France, Sopha DST-XL).

Production of 55 Co
The procedure used for the targetry and bombardment for the production of 55 Co was similar to that of 57 Co already reported [11], with the exception of a 30 µm layer of nickel as the target and 15 MeV protons for target irradiation (15-8 MeV on the target). The radiochemical separation process including recovery of the cobalt from the nickel and copper were carried out immediately after the target bombardment. The recovery of cobalt and copper from the nickel was carried out in the same manner as explained above for the radiochemical separation of 57 Co [11]. The only differences were the use of 20 ml of refluxing warm 7 M HNO 3 in the first step and 20 ml of 4 M HCl was used for the recovery of radiocobalt and radiocopper ions.

Radionuclidic purity
Radionuclidic purity of the products was assayed by gamma spectroscopy of the final samples using an HPGe detector coupled with a Canberra™ multi-channel analyzer. The peaks were observed for 1 h for 55 Co.

Chemical purity
The production was based on the irradiation of natural nickel electroplated on to a goldlayered backing. Therefore, the presence of nickel cation was detected using visible colorimetric assays. The most important photometric reagents for determining nickel are dioximes, which provide specific and fairly sensitive methods. Dimethylglyoxime reacts with nickel ions in a neutral or ammonia medium forming a pink, flocculent precipitate. Even at 2 ppm of standard nickel concentration, the colored Ni-dimethylglyoxime complex is visible to the naked eye [13]. The amount of gold cation was monitored in the final solution using color formation with acidic rhodamine B reagent, based on a previously reported colorimetric method [14].

Preparation of [ 55 Co](III)bis-(2-acetylpyridine thiosemicarbazone) [ Bis{N'-[1-(pyridin-2-yl-kN)ethylidene]carbamohydrazonothioatok 2 N',S}( 55 Co)cobalt(1+) ]
The acidic solution of [ 55 Co]CoCl 3 (3 mCi in 2.5-3 ml) was transferred to a 5 ml-vial and heated to dryness using a flow of N 2 gas at 50-55°C. Fifty micro µl of 2-acetylpyridine thiosemicarbazone in methanol (1 mg/ml, 240 nmol) was added to the cobalt residue and vortexed at 25°C for 3-5 min and then left at room temperature for 30 min. The resulting mixture was diluted by the addition of normal saline (4.5 ml) and rapidly checked by RTLC and HPLC for radiochemical purity. The final solution was then passed through a 0.22 μm filter and pH adjusted to 5.5-7. The same procedure was used for the preparation of

Radiochemical purity of [ 55 Co](III)[APTS] 2
Radio thin layer chromatography RTLC was performed using an in-house made radiochromatogram scanner coupled to an HPGe detector. A step motor was installed to permit counting of 0.4 cm segments each 30 seconds through a slot in a shielded chamber. RTLC was performed using absolute ethanol (developing solution) on polymer-backed silica gel. The radiochemical yields were determined by comparison of the activities in the un-complexed 55 Co and the [ 55 Co](III)[APTS] 2 major radio peak.

High performance liquid chromatography
HPLC was performed on the final preparation using HPLC grade EtOH as the eluent with at a flow rate of 1.3 ml/min (pressure: 120-140 kgF/cm2) for 40 min using a Si Kromasil 100, 5 μm (250×46 mm).

Paper electrophoresis
Ionic charge of the cobalt-55 complex in final solution was checked using cellulose acetate paper electrophoresis (Gellman) in 0.05N EDTA at 200V for 10 min. A major component of complex solution migrated to cathode in EDTA 2solution.

Stability of [ 55 Co](III)[APTS] 2 complex in the final product
A sample of [ 55 Co](III)[APTS] 2 (5 mCi, 2mL) was kept at room temperature for 24 hours and checked by RTLC at various time intervals. A sample (100 μl) was taken from the shaken mixture and diluted 3-4 times with normal saline and passed through a 5KDa cutoff filter (Waters) followed by determination of free cobalt cation content (Rf. 0.8) to [ 55 Co](III)[APTS] 2 (Rf. 0.0) by RTLC using ethanol as eluent.

Serum stability studies
In order to perform serum stability studies, 500 μl of freshly prepared human serum was added to 976 μCi of

Determination of partition coefficient
The partition coefficient of the [ 55 Co](III)[APTS] 2 was measured following 1 min of vortexing a mixture compromised of 1 ml 1-octanol and 1 ml of isotonic acetate-buffered saline (pH=7) with approximately 100 μCi of the radiolabeled metal complex (100-150 μl) at 37°C. Following further incubation for 5 min, the octanol and aqueous phases were sampled and counted in a dose calibrator. A 500 µl sample of the octanol phase from this partitioning was repartitioned two to three times with fresh buffer to ensure that traces of hydrophilic 55 Co impurities did not alter the calculated P values. The reported (log P) values are the average of the second and third extractions from three to four independent measurements, (log P) values and represent the mean of five measurements.

Induction of fibrosarcoma tumors in mice
Tumor induction performed by the use of poly aromatic hydrocarbon injection in rodents as reported previously [16]. For tumor model preparation, 10µl of 3-methylcholanthrene solution in extra-virgin olive oil (4 mg/ml) was injected SC to the dorsal area of the mice. After 14-16 weeks the tumor weighed 0.2-0.4 g and was not grossly necrotic. Tumor tissues of some random animals were sent for pathological tests and were diagnosed as fibrosarcoma.

Biodistribution studies
A volume (0.1 ml) of the final [ 55 Co](III)[APTS] 2 solution containing 100 µCi activity (≤ 2 µg APTS in 100 µl) was injected via the dorsal tail vein. The total amount of activity injected into each rat was determined by counting the 1-ml syringe before and after injection in a dose calibrator with fixed geometry. The animals were sacrificed by CO 2 asphyxiation at selected time intervals (2, 4, 12, 21, 35 h) post injection, tissues (including liver, stomach, muscle, intestine, spleen, heart, kidney, skin, lung, sternum and the tumor) were weighed and their specific activities determined by counting on an HPGe detector to obtain the counts as a percentage of the injected dose per gram of tissues.

Imaging of [ 55 Co](III)[APTS] 2 in tumor bearing mice
Fibrosarcoma-bearing mice were used for tumor imaging when tumors reached a size of 0.5-1 cm 3 , 14-16 weeks after its induction. Images were taken 2-48 hours post injection including (21 and 35 hours) after administration of the radiopharmaceutical in the coincidence mode by a dual-head SPECT system. The mouse-to-high energy septa distance was 12 cm. Images were performed of tumor bearing mice. The useful field of view (UFOV) was 540 mm×400 mm. The spatial resolution in the coincidence mode was 10 mm FWHM at the CFOV. Sixty four projections were acquired for 30 seconds per view with a 64×64 matrix.

Conclusion
The methods described herein for the production and radiochemical separation of 55 Co were simple and cost effective, resulting in high production yields and acceptable levels of contamination. Total labeling and formulation of [ 55 Co](III)[APTS] 2 took about 40 minutes, with a yield of greater than 99% and a specific activity of ≈ 10-20 Ci/mmol. No significant amounts of labeled by-products were observed by HPLC analysis of the final preparations. The radio-labeled complex was stable in aqueous solution, and in human serum at 37ºC, for at least 24 h and no significant amount of other radioactive species were detected by RTLC/HPLC. The biodistribution of the ligand was checked in tumor bearing mice out to 48 h showing a significant tumor uptake (3%) after 21 h. Co-incidence imaging of the tumor-bearing mice (2-48 h post injection) receiving [ 55 Co](III)[APTS] 2 showed the accumulation of the tracer in the tumoral tissue 20-40 h, while the best time point seem to be 35 h due to background reduction. [ 55 Co](III)[APTS] 2 is a potential tracer, with suitable half life and good chemical stability for tumor diagnostic applications, while further animal studies are required to establish the tumor uptake and best time points in tumoral models.