Development of Novel Radiogallium-Labeled Bone Imaging Agents Using Oligo-Aspartic Acid Peptides as Carriers

68Ga (T 1/2 = 68 min, a generator-produced nuclide) has great potential as a radionuclide for clinical positron emission tomography (PET). Because poly-glutamic and poly-aspartic acids have high affinity for hydroxyapatite, to develop new bone targeting 68Ga-labeled bone imaging agents for PET, we used 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid (DOTA) as a chelating site and conjugated aspartic acid peptides of varying lengths. Subsequently, we compared Ga complexes, Ga-DOTA-(Asp)n (n = 2, 5, 8, 11, or 14) with easy-to-handle 67Ga, with the previously described 67Ga-DOTA complex conjugated bisphosphonate, 67Ga-DOTA-Bn-SCN-HBP. After synthesizing DOTA-(Asp)n by a Fmoc-based solid-phase method, complexes were formed with 67Ga, resulting in 67Ga-DOTA-(Asp)n with a radiochemical purity of over 95% after HPLC purification. In hydroxyapatite binding assays, the binding rate of 67Ga-DOTA-(Asp)n increased with the increase in the length of the conjugated aspartate peptide. Moreover, in biodistribution experiments, 67Ga-DOTA-(Asp)8, 67Ga-DOTA-(Asp)11, and 67Ga-DOTA-(Asp)14 showed high accumulation in bone (10.5±1.5, 15.1±2.6, and 12.8±1.7% ID/g, respectively) but were barely observed in other tissues at 60 min after injection. Although bone accumulation of 67Ga-DOTA-(Asp)n was lower than that of 67Ga-DOTA-Bn-SCN-HBP, blood clearance of 67Ga-DOTA-(Asp)n was more rapid. Accordingly, the bone/blood ratios of 67Ga-DOTA-(Asp)11 and 67Ga-DOTA-(Asp)14 were comparable with those of 67Ga-DOTA-Bn-SCN-HBP. In conclusion, these data provide useful insights into the drug design of 68Ga-PET tracers for the diagnosis of bone disorders, such as bone metastases.


Introduction
Bone contains abundant proliferation factors, and is therefore a convenient environment for tumors to metastasize and grow. Indeed, malignant tumors frequently metastasize to the bone [1]. With the development of therapeutic methods and drugs, early diagnoses of bone metastases must be more important. Significant advances in imaging technologies such as X-ray computed tomography (CT) and magnetic resonance imaging (MRI) have been made during the last a few decades; however, because of its high sensitivity, nuclear medicine bone scanning is the optimal test for detecting bone metastases. Over the last thirty years, 99m Tcbisphosphonate complexes such as methylenediphosphonate ( 99m Tc-MDP) and hydroxymethylenediphosphonate ( 99m Tc-HMDP) have been widely used as radiopharmaceuticals in bone scintigraphy for disorders such as metastatic bone cancer, Paget's disease, and osteoporotic fractures [2][3][4][5]. The accumulation of 99m Tc-bisphoshonate complexes in bone must be derived from the binding of phosphonate groups in bisphosphonate to calcium (Ca 2+ ) in hydroxyapatite crystals in bone, but the mechanism of high uptake to lesion sites has not been completely elucidated. One of factors should be the increased vascularity and regional blood flow caused from disease. However, it has been reported that regional bone blood flow alone does not account for the increased uptake of 99m Tc-bisphoshonate complexes [6]. Other factors should be involved in their binding and interaction with bone. It is generally assumed that 99m Tc-bisphoshonate complexes accumulate at sites of active bone metabolism, especially, at osteoblastic lesions [7,8]. Newly formed bone has a much larger surface area than stable bone does. That is, the crystalline structure of hydroxyapatite in newly formed bone is amorphous and has a greater surface area than that in normal bone [9]. In the cases of 99m Tc-bisphoshonate complexes, the phosphonate groups coordinate with not only Ca 2+ but also 99m Tc [10], which might decrease the inherent accumulation of bisphosphonate (MDP or HMDP) in bone. Incidentally, 99m Tc-bisphoshonate complexes cannot be isolated as well-defined single chemical species, but as mixtures of short-and long-chain oligomers, may reduce the efficacy of radiopharmaceuticals. Biological behaviors of these tracers are also affected by the degree of ionization and by variable oligomer constitutions of preparations [11]. To overcome the shortcomings of 99m Tc-bisphoshonate complexes, we and other groups have designed and developed 99m Tc-mononuclear complex-conjugated bisphosphonate compounds [12][13][14][15], in which phosphonate groups are not coordinated with 99m Tc. As expected, some of these compounds showed superior biodistribution compared with previous compounds. Of note, this drug concept is applicable to both 99m Tc-complex radiopharmaceuticals and other radiometals [16][17][18][19][20][21][22][23][24][25][26].
Sodium fluoride labeled with 18 F ( 18 F-NaF) for bone imaging was initially reported by Blau et al. in 1962 [27], and subsequently was approved by FDA in 1972. 18 F-NaF accumulates in bone because fluoride anions are isomorphously exchanged with the hydroxyl group in hydroxyapatite (Ca 10 (PO 4 ) 6 (OH) 2 ) and fluoroapatite (Ca 10 (PO 4 ) 6 F 2 ) is formed. After the development of 99m Tc-labeled bone scintigraphy agents, such as 99m Tc-MDP, 18 F-NaF was replaced by them because the physical characteristics of 99m Tc were more convenient for imaging with conventional gamma cameras in those days. However, in the last two decades, positron emission tomography (PET) and PET/CT have evolved significantly and become widespread. The changes caused the reemergence of 18 F-NaF and bone imaging agents for PET are desired because current PET have higher spatial resolution and greater sensitivity than conventional gamma cameras. Actually, it was reported that 18 F-NaF PET imaging was significantly more sensitive than 99m Tc-MDP planar and 99m Tc-MDP single photon emission computed tomography (SPECT) imaging [28]. However, most positron emitters, such as 18 F, need high cost cyclotron facilities, and it limits the availability for PET.
Meanwhile, the radionuclide 68 Ga has great potential for clinical PET and could become an attractive alternative to 18 F because of its radiophysical properties, particularly as a generatorproduced nuclide with a half-life (T 1/2 ) of 68 min [29]. Namely, it does not require an on-site cyclotron and can be eluted on demand. Indeed, in principle, the long half-life of the parent nuclide 68 Ge (T 1/2 = 270.8 days) provides a generator with a long life span. Therefore, the appearance of 68 Ga-labeled compounds for bone imaging has been desired and some compounds have been reported in recent years [30][31][32][33][34].
Several noncollagenous bone proteins have repeating sequences of acidic amino acids (Asp or Glu) in their structures, offering potential hydroxyapatite-binding sites. For example, osteopontin and bone sialoprotein, 2 major noncollagenous bone matrix proteins, have repeating Asp and Glu rich sequences, respectively [35][36][37]. Reportedly, poly-glutamic and poly-aspartic acids have high affinity for hydroxyapatite and could be used to deliver drugs to bone tissues [38][39][40].

Synthesis of DOTA-(Asp) n
The protected peptidyl resin was manually constructed by an Fmoc-based solid-phase methodology using Fmoc-Asp(OtBu)-Wang resin and Fmoc-Asp(OtBu). The peptide chain was constructed in cycles of (I) 15 minutes of deprotection with 20% piperidine in dimethylformamide (DMF) and (II) 2 hours of coupling with 3 equivalents of Fmoc-Asp(OtBu), 1,3-diisopropylcarbodiimide (DIPCDI) and 1-hydroxybenzotriazole hydrate (HOBt) in DMF. The coupling reaction was then repeated after Kaiser test was positive for the resin [41]. After construction of the peptide chain on the resin, the Fmoc protecting group was removed using 20% piperidine in DMF, and a mixture containing 2 equivalents of DOTA-tris, DIPCDI, and HOBt in DMF was added and allowed to react for 2 hours, as described above. To cleave peptides from the resin and deprotect, 0.5 mL of thioanisole and 5 mL of trifluoroacetic acid (TFA) were added to the fully protected peptide resin at 0uC and stirred at room temperature for 2 hours. After resin removal by filtration, ether was added to the filtrate at 0uC to precipitate crude peptide. The crude products were purified by reversed-phase (RP)-HPLC performed with a Hydrosphere 5C18 column (106150 mm; YMC, Kyoto, Japan) at a flow rate of 4 mL/min with an isocratic mobile phase of water containing 0.1% TFA [in the case of DOTA-(Asp) 2 ] or with a Cosmosil 5C 18 -AR 300 column (106150 mm; Nacalai Tesque, Kyoto, Japan) at a flow rate of 4 mL/min with a 0-20% methanol

Hydroxyapatite-binding Assays
Hydroxyapatite-binding assays were performed according to previously described procedures with slight modifications [33]. In brief, hydroxyapatite beads (Bio-Gel; Bio-Rad, Hercules, CA, USA) were suspended in Tris/HCl-buffered saline (50 mM, pH 7.4) at 2.5 mg/mL, 10 mg/mL, and 25 mg/mL. For the solutions of 67 Ga-DOTA-(Asp) n (n = 2, 5, 8, 11, or 14), the ligand concentrations were adjusted to 19.5 mM by adding DOTA-(Asp) n . Two hundred microliters of each 67 Ga-DOTA-(Asp) n solution was added to 200 mL of the hydroxyapatite suspension, and the samples were gently shaken for 1 hour at room temperature. After centrifugation at 10,000 g for 5 minutes, the radioactivity of the supernatants was measured using an auto well gamma counter (ARC-7010B, Hitachi Aloka Medical, Ltd., Tokyo, Japan). Control experiments were performed using the same procedure without hydroxyapatite beads, which showed less than 0.1% adsorption of radioactivity to vials. The ratios of binding were determined as follows: The effect of bisphosphonate on the binding of 67 Ga-DOTA-(Asp) 14 or 67 Ga-DOTA-Bn-SCN-HBP to hydroxyapatite beads was also examined. In these experiments, 100 mL of 67 Ga-DOTA-(Asp) 14 or 67 Ga-DOTA-Bn-SCN-HBP solutions containing varying concentrations of the bisphosphonate compound alendronate were incubated with 100 mL of suspensions containing 1 mg of hydroxyapatite beads. After centrifugation, the radioactivity of the supernatant was measured, and hydroxyapatite-binding ratios were calculated as described above.

Biodistribution Experiments
Experiments with animals were conducted in strict accordance with the Guidelines for the Care and Use of Laboratory Animals of Kanazawa University. The animal experimental protocols used were approved by the Committee on Animal Experimentation of Kanazawa University (Permit Number: AP-132633). Biodistribution experiments were performed after an intravenous administration of each diluted tracer solution (37 kBq/100 mL) to 6-weekold male ddY mice (27-32 g, Japan SLC, Inc., Hamamatsu, Japan). To investigate the effect of an excess amount of bisphosphonate on biodistribution, alendronate (20 mg/kg) was intravenously administered to mice 1 minute before the intravenous injection of 67 Ga-DOTA-(Asp) 14 or 67 Ga-DOTA-Bn-SCN-HBP. Four to six mice each were sacrificed by decapitation at 10, 60, and 180 minutes post-injection. Tissues of interest were removed and weighed. Complete left femurs were isolated as representative bone samples, radioactivity was determined using an auto well gamma counter, and counts were corrected for background radiation and physical decay during counting.

Protein-binding Assay
Serum protein binding ratios of 67 Ga-DOTA-(Asp) n (n = 2, 5, 8, 11, or 14) and 67 Ga-DOTA-Bn-SCN-HBP were evaluated by an ultrafiltration method. In these experiments, 6-week-old male ddY mice received intravenous boluses of radiotracer. After 3 minutes, the mice were anesthetized with ether, and blood was collected by heart puncture. Serum samples were prepared and applied to an Amicon Ultra-0.5 Centrifugal Filter Unit with Ultracel-30 membrane (Millipore). The units were centrifuged at 14,000 g for 20 minutes at room temperature. The radioactivity counts of the initials and filtrates were determined using an auto well gamma counter. The protein-binding ratios were then calculated as follows: Protein-binding ratio % ð Þ~100{ radioactivity of filtrate ð Þ = radioactivity of initial solution ð Þ |100

Statistical Analysis
Data are expressed as means 6 standard deviations where appropriate. In biodistribution experiments using alendronate as a blocking agent, differences were identified using unpaired Students' t test and were considered significant when p,0.05.

Protein-binding Assay
Proportions of 67 Ga-DOTA-(Asp) n (n = 2, 5, 8, 11, or 14) and 67 Ga-DOTA-Bn-SCN-HBP bound to serum protein ( Figure 5) shows that the binding of 67 Ga-DOTA-(Asp) n compounds to serum proteins decreased with an increase in the length of the aspartic acid chain. The serum protein binding rate of 67 Ga-DOTA-Bn-SCN-HBP was greater than that of 67 Ga-DOTA-(Asp) n compounds.

Discussion
Previously, we introduced the concept of radiometal complexconjugated bisphosphonate compounds for the development of bone-seeking radiopharmaceuticals [42,43]. Moreover, in recent years, superior activities of newly developed radiogallium complex-conjugated bisphosphonate compounds have been reported by us and other groups [30][31][32][33][34]. In these drug compounds, the bisphosphonate structure has high affinity for hydroxyapatite, which is a specific component of bone tissues, leading to targeting of bone tissues. In a previous study, it was reported that the in vitro binding profile of Fmoc-(Asp) n (n = 2, 4, 6, 8, or 10) to hydroxyapatite increased with the increase in the length of the peptide [44]. Here a similar strategy was applied using aspartic acid peptides as the carrier to bone tissues instead of bisphosphonate. Indeed, we have demonstrated that the binding of 67 Ga-DOTA-(Asp) n (n = 5, 8, 11, or 14) to hydroxyapatite beads increased with increased length of the aspartic acid peptide. This   result is consistent with the previous study and was reflected by bone accumulation of 67 Ga-DOTA-(Asp) n in biodistribution experiments. Moreover, these biodistribution experiments showed greater bone accumulation with increasing length of the peptide conjugates from 67 Ga-DOTA-(Asp) 2 to 67 Ga-DOTA-(Asp) 11 . The longer compounds 67 Ga-DOTA-(Asp) 11 and 67 Ga-DOTA-(Asp) 14 accumulated equally in bone and showed superior biodistribution characteristics as that of bone imaging radiopharmaceuticals, with high accumulation in bone and rapid clearance from other tissues. Despite lower bone accumulation than the bisphosphonate 67 Ga-DOTA-Bz-SCN-HBP, the bone/blood ratios of radioactivity after injection of 67 Ga-DOTA-(Asp) 11 and 67 Ga-DOTA-(Asp) 14 , which are an index as bone imaging, were comparable or higher ( Figure 6), presumably due to more rapid blood clearance than 67 Ga-DOTA-Bz-SCN-HBP. This may reflect the lower serum protein binding ratios of 67 Ga-DOTA-(Asp) 11 and 67 Ga-DOTA-(Asp) 14 compared to that of 67 Ga-DOTA-Bn-SCN-HBP ( Figure 5). We assumed that the high accumulation of radioactivity in the bone after injection of these compounds was due to hydroxyapatite binding of bisphosphonate or aspartic acid structures in bone tissues. To estimate the hydroxyapatite binding of these compounds, alendronate inhibition experiments were performed in vitro and in vivo. In these hydroxyapatite binding assays, 67 Ga-DOTA-(Asp) 14 and 67 Ga-DOTA-Bn-SCN-HBP binding was inhibited by alendronate, confirming that the mechanism by which 67 Ga-DOTA-(Asp) 14 and 67 Ga-DOTA-Bn-SCN-HBP accumulate in bone involves coordination of their functional groups to the Ca 2+ in hydroxyapatite crystals [45]. However, 67 Ga-DOTA-Bn-SCN-HBP binding was inhibited by lower concentrations of alendronate compared with 67 Ga-DOTA-(Asp) 14 . Moreover, in biodistribution experiments, the inhibition of radioactive bone accumulation by alendronate was greater after injection of 67 Ga-DOTA-Bn-SCN-HBP than that of 67 Ga-DOTA-(Asp) 14 . Although the precise mechanisms remain unclear, the binding patterns of these compounds to hydroxyapatite may differ. Wang et al. reported that alendronate and (D-Asp) 8 , which were used as bone-targeting moieties on conjugated fluorescein isothiocyanate   (FITC)-labeled N-(2-hydroxypropyl)methacrylamide (HPMA) copolymers (P-ALN-FITC and P-D-Asp 8 -FITC). In the study, P-D-Asp 8 -FITC preferentially bound bone resorption surfaces, whereas P-ALN-FITC appeared to bind both formation and resorption surfaces in bone [46]. In hydroxyapatite binding experiments with different crystallinity, P-D-Asp 8 -FITC showed preferential binding to hydroxyapatite of higher crystallinity compared with P-ALN-FITC. These observations indicated that bisphosphonate and aspartic acid peptides have different modes of hydroxyapatite binding. Accordingly, 68 Ga-DOTA-(Asp) 14 PET imaging may give different information than that obtained by 99m Tc-MDP bone scintigraphy methods. Since 99m Tc-MDP mainly accumulates osteoblastic lesions in bone, it has been known that sensitivity 99m Tc-MDP often shows false-negative in osteolytic bone metastases lesions, and consequently, its sensitivity is reduced [47]. On the contrary, 68 Ga-DOTA-(Asp) 14 PET imaging may have a potential to improve its sensitivity. Meanwhile, in a previous 99m Tc-MDP-bone scintigraphy study, treatments with bisphosphonate and alendronate may have caused false negative scintigraphy by producing competition between the drug and tracer, and blocking entrapment and accumulation of the tracer in bone [48]. In this study, pretreatment with alendronate inhibited bone accumulation of 67 Ga-DOTA-Bn-SCN-HBP more effectively than that of 67 Ga-DOTA-(Asp) 14 , suggesting that bisphosphonate-induced false negative scintigraphy is less likely to occur in 68 Ga-DOTA-(Asp) 14 PET. Radiogallium complexes of 1,4,7-triazacyclononane-triacetic acid (NOTA) or triazacyclononane-phosphinate (TRAP) may produce radiocomplexes with a higher specific activity than DOTA, allowing the use of much lower concentrations of precursor for labeling [49]. Despite this, DOTA was used in this study because unlike receptor imaging, much higher specific activity is not necessary for hydroxyapatite-targeted bone imaging. Moreover, DOTA is more versatile and could be developed for both imaging and therapeutics. Since the DOTA ligand forms a stable complex with not only gallium ( 67/68 Ga), but also lutetium ( 177 Lu) and yttrium ( 90 Y), which are beta particle emitters as radionuclides for therapy, furthermore, bismuth ( 213 Bi) as an alpha emitter could be applicable, its application to therapy from diagnosis could be made available. That is, radiometal complexes of DOTA-(Asp) n for radionuclide therapy could be useful as agents for the palliation of metastatic bone pain.     In conclusion, the 67 Ga-DOTA complex-conjugated aspartic acid peptides 67 Ga-DOTA-(Asp) n showed ideal biodistribution characteristics as bone scintigraphy agents. Therefore, these agents may facilitate the drug design of PET tracers with 68 Ga for the diagnosis of bone disorders, such as bone metastases. Further studies are required to determine whether 68 Ga-DOTA-(Asp) n can provide additional information to that of bone scintigraphy, and to develop these compounds for radionuclide therapy.