Design, Synthesis and Evaluation of Dual-Modality Glyco-Nanoparticles for Tumor Imaging

d-Glucosamine (DG) was conjugated to a core-cross linked polymeric micelle (CCPM) system equipped with both a near-infrared fluorophore (NIRF) and a gamma emitter (111In). The resultant nano-scale tumor-targeting imaging tracer, 111In-DG-NIRF-CCPM, selectively accumulated in a human epithelial carcinoma A-431 xenograft model in mice. At 24 hrs post injection, the tumor uptake was 2.62 ± 0.80 % of the injected dose per gram of tissue (%ID/g). Tumors were clearly delineated in both single-photon emission computed tomography (SPECT) and optical imaging. The results suggest that the prepared imaging tracer is a promising agent for tumor diagnosis.


Introduction
Optical imaging is a relatively new molecular imaging modality that offers non-invasive, real-time and high-sensitivity imaging of fluorophores embedded in diseased tissues [1,2]. There has been an increased interest in the use of polymeric micelles to deliver tumor-imaging tracers, because polymeric OPEN ACCESS micelles with adequate size can accumulate within tumor tissue due to the enhanced permeation and retention (EPR) effect [3,4]. Specifically, polymeric micelles containing NIRF compounds have been extensively explored for non-invasive molecular imaging [5,6]. Previously, we reported the synthesis of core-cross linked polymeric micelles (CCPM) containing a Cy-7-like NIRF dye as well as a gammaemitter 111 Indium (T 1/2 = 2.83 d, 0.171 MeV (90%), 0.245 MeV (94%)). CCPM system exhibited prolonged blood half-life and enhanced uptake in tumor [7,8]. In this study, we investigated whether the surface of CCPM can be further modified with sugar molecules to improve its pharmacokinetics profile and minimize its non-tumor uptake [9].
Carbohydrate-mediated molecular recognitions have been applied to nano-medicine to improve drug delivery efficiency of nano-carriers [10,11]. Glyco-nanotechnology, as an innovative medicine, has been extensively used in basic and clinical research [12,13]. D-Glucosamine (DG) is an attractive scaffold as a glucosyl ligand. We recently reported the synthesis of 68 Ga-and 111 In-labeled DG for the positron emission tomography (PET) and SPECT imaging of tumors [14][15][16]. Those studies have showed that DG attached to a bulky moiety still preserved its biological functions. In this study, we conjugated DG to the CCPM surface via a succinate linker, and investigated its performances regarding the tumor targeting, in vitro and in vivo stability, biodistribution, and dual gamma and NIRF optical imaging. We found that DG-CCPM had excellent stability, could be taken up by tumor cells, and was visualizable by both gamma and optical imaging modalities. Further studies are needed to improve the drug targeting efficiency with dual-labeled CCPM for molecular imaging.

Synthesis and Characterization of CCPM Nanoparticles
DG-COOH was synthesized by the amidation of glucosamine to succinic anhydride. The identity of this product was confirmed by negative mode electrospray mass spectrometry (Agilent LC/ DTPA was conjugated to CCPM by simply mixing NCS-Bz-DTPA and CCPM under nitrogen overnight. Unreacted NCS-Bz-DTPA was removed by centrifugal filtration or by PD-10 column chromatography. About 70% of the NCS-Bz-DTPA was attached to the NIRF-CCPMs. Next, DG-COOH was coupled to CCPM using1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (EDC) as the coupling agent to form ester linkage. A UV-Vis spectrophotometer was used to measure the absorbance (optical density, OD) at 780 nm or 335 nm; which was converted into a concentration using the Beer-Lambert relation (path-length = 1mm). The standard curve of DTPA-NIRF-CCPM generated at 780 nm was used to determine the concentration of CCPM [ Figure 1(a)]. The standard curve generated at 335 nm after treatment with 2,4,6-trinitrobenzenessulfonate (TNBS) was used to quantify the amount of amines on the surface of CCPM [ Figure 1(b)] [17][18][19]. The DTPA-NIRF-CCPM-DG (DG-CCPM) nanoparticles were characterized with regard to composition, size, surface charge and optical properties. As shown in Table 1. DG-CCPM nanoparticles had approximately 19 DTPA and 15 DG molecules per nanoparticle. The volume average size of DTPA-NIRF-CCPM was 24 ± 8.9 nm, determined by dynamic light scattering. Prior to conjugation, the amine-abundant CCPM was weakly positively charged. After DTPA and DG conjugation, however, the surface potential changed to slightly negative, indicating the successful conjugation.

Radiolabeling of the Nanoparticles
Radiolabeling of the nanoparticles was accomplished through incubation of 111 InCl 3 with DTPA-NIRF-CCPM or DG-CCPM in 0.1 M NaOAc (pH 5.2) at room temperature, respectively. Under such conditions, a labeling efficiency of greater than 96% was achieved without further purification. As indicated by radio-TLC, most radioactivity was associated with 111 In-labeled DTPA-NIRF-CCPM or DG-CCPM nanoparticles (R f~ 0.0), with less than 5% detectable free 111 InCl 3 (R f ~ 0.9) ( Figure 2).

Microscopic Imaging of DTPA-NIRF-CCPM-DG
The binding of NIRF-CCPM and DG-CCPM to A431 cells was studied by fluorescence microscopy (Figure 3). At two hours after incubation, NIRF-CCPM did not uptake in these cells. However, at 4 hours after incubation, uptake of NIRF-CCPM in cells was clearly visualized. In comparison, uptake of DG-CCPM in A431 cells were observed at 2-hour time point, and the fluorescent signal intensity appeared to have increased over from 2 to 4 h ( Figure 3). The data indicated that both NIRF-CCPM and DG-CCPM could be internalized over time. Cell uptake of NIRF-CCPM could be attributed to nonspecific phagocytic activity of tumor cells. Many studies have demonstrated that 18 F-FDG and its fluorescent analogs are substrates of glucose transporters (GLUTs) [20]. Further studies are needed to establish whether uptakes of DG-CCPM in A431 cells are mediated through GLUTs.

Biodistribution in Tumor Bearing mice
Biodistribution studies of 111 In-DG-CCPM and 111 In-DTPA-NIRF-CCPM were carried out in mice bearing subcutaneous A431 tumor. As shown in Figure 4, the percentage of injected dose per gram of tissue (%ID/g) at 24 h after intravenous injection in the liver and the spleen were 4.27 ± 0.15% and 6.2 ± 0.14% for 111 In-DG-CCPM, and 10.79 ± 1.81% and 10.1 ± 0.74% for 111 In-DTPA-NIRF-CCPM, respectively. 111 In-DG-CCPM had significantly lower spleen uptake than that of the 111 In-DTPA-NIRF-CCPM (p = 0.0097). On the other hand, 111 In-DTPA-NIRF-CCPM had significantly higher activity in the blood (p = 0.0098) and the lung (p = 0.0067). 111 In-DTPA-NIRF-CCPM activity was also significantly higher in the tumor (p = 0.046). The higher activity observed with 111 In-DTPA-NIRF-CCPM as compared to 111 In-DG-CCPM in the liver, spleen, lung, and tumor may be attributed to higher blood pool activity of 111 In-DTPA-NIRF-CCPM.  However, the distribution patterns for each particle were similar, with the liver and the spleen being the major organs taking up the radio-labeled glyco-nanoparticles. Figure 5 compares the tumor-tonormal tissue (T/NT) uptake ratio of 111 In-DG-CCPM and 111 In-DTPA-NIRF-CCPM. 111 In-DTPA-NIRF-CCPM did not show significantly higher T/NT ratios compared with 111 In-DG-CCPM at 24 h post injection.  Figure 6 shows the representative gamma images obtained at various time points after injection of 111 In-DTPA-NIRF-CCPM (300µCi/2.6 × 10 14 particles/mouse) and 111 In-DG-CCPM (300µCi/4 × 10 13 particles/mouse). Both nanoparticles exhibited sustained blood pool activity, prominent and persistent accumulation in the liver and the spleen for up to 120 h, and gradual accumulation in the tumor. The cardiac area was clearly seen up to 48 h post injection, reflecting a long blood circulation time for CCPMs. Uptake in the tumor was clearly visualized by 120 h post injection with γ-scintigraphy. Tumors in the mice received intravenous injection of 111 In-DG-CCPM, but not in mice received 111 In-DTPA-NIRF-CCPM, were visualized at later time points (24-72 h post injection) ( Figure 6).

In vivo NIRF Imaging
The biodistribution of 111 In-DG-CCPM (4 × 10 13 particles) in A431-bearing mice was also determined by in and ex vivo analysis of the fluorescence signal intensities of excised tissues ( Figure 7A,B). Accumulation of 111 In-DG-CCPM nanoparticles in the tumor could be readily visualized with optical imaging ( Figure 7A). In consistent with gamma imaging finding, the highest activity was found in the tumor, the liver and the spleen ( Figure 7B). This indicates that glyco-CCPM nanoparticles displayed a strong and persistent fluorescence signal in vivo. Clearly, nuclear imaging and NIRF optical imaging techniques provided complementary information. Although our biodistribution data do not support active targeting of DG-CCPM to A431 tumors at 24 h after injection, the imaging data appears to show better tumor delineation with 111 In-DG-CCPM than with 111 In-DTPA-NIRF-CCPM at 24-72 h post injection. One possible explanation is the relatively small size of DG (MW = 278 Dalton), which could be easily masked by much larger PEG chains on nanoparticles' surface. Our finding in the current study differs from that of an earlier report which did validate active targeting when the same NIRF-CCPM was used but with a different targeting ligand. In that study, the targeting ligand is annexin V, which has a molecular weight of ~35,000 Dalton [21]. Larger ligand relative to CCPM may be less prone to the masking effect of PEG. Another possibility is heterogeneous intratumoral distribution of 111 In-DG-CCPM, which could lead to insignificant change in tumor uptake value because the radioactivity counting method averages out contribution from the entire tumor mass. Although our data do not conclusively support active targeting with 111 In-DG-CCPM, these data provide valuable insight that may help design better ligand-CCPM conjugate for targeted nanoparticle delivery.

Synthesis of D-Glucosamine-NH-SA
D-Glucosamine-NH-SA (DG-COOH) was synthesized by treating commercially available D-glucosamine hydrochloride with succinic anhydride using NMM as a base. Briefly, succinic anhydride (1.0 mmol) and NMM (0.2 mL) were added to a solution of D-glucosamine hydrochloride (1.0 mmol) in 10 mL of dimethylformamide (DMF)/H 2 O (9/1, v/v). The mixture was stirred at ambient temperature for 12 h. After evaporation of the solvents under vacuum, the residue was washed with ether. Residue was dissolved in water and purified by HPLC to yield a white powder.

Conjugation of DG-COOH to CCPM
DTPA-CCPM was synthesized according to previously reported procedures with some modification (Scheme 1) [7]. An aqueous solution of CCPM nanoparticles (2 mL, 2 mg dry weight/mL of water) was placed in an amber vial. The pH of the solution was adjusted to 7.5 using 2% sodium bicarbonate solution. NCS-Bz-DTPA (0.013 mg, 1.54 × 10 −3 M in PBS) was added to the vial. The reaction was stirred under nitrogen overnight. Unreacted NCS-Bz-DTPA was removed by centrifugal filtration using a membrane with a molecular weight cutoff of 30, 000. The unreacted NCS-Bz-DTPA was quantified by capillary electrophoresis. Approximately 70% of the DTPA used was attached to the particles.
To couple DG-COOH to the primary amines on the surface of CCPM, DG-COOH (17.0 mg in 2.0 mL 0.1 M PBS) was allowed to react with EDC (50 µL 2.232 mg/mL) at room temperature (RT) for 30 min. Then 0.5 mL CCPM (1.54 × 10 15 particles/mL) was added and the reaction mixture was stirred at RT overnight. The molar ration between amine groups in CCPM and DG-COOH was 1:3. The reaction mixture was purified with PD-10 column, which was eluted with 0.01 M PBS (pH7.4). The amount of amines before and after treatment with DG-COOH was were determined by 2,4,6-trinitrobenzenessulfonate (TNBS) assay.

Standard Curves of DTPA-NIRF-CCPM
The concentration of DTPA-NIRF-CCPM nanoparticles was quantified by comparing UV-Vis absorption at 780 nm with a standard curve. DTPA-NIRF-CCPM in 0.01 M pH 7.4 PBS was serially diluted with PBS to obtain concentrations ranging from 15.4 × 10 12 /mL to 2.4 × 10 11 /mL, and their UV/Vis spectra measured with a UV-Vis spectrometer.
2,4,6-Trinitrobenzenesulfonate (TNBS) was used to quantify the number of primary amines in CCPM. Briefly, 0.5 mL TNBS in 0.1 M pH 8.5 sodium bicarbonate (0.01%, w/v) was added to each 1 mL solution of phenylalanine standard in NaHCO 3 buffer ranging from 2 to 20 μg/mL. The mixtures were then incubate at 37 °C for 2 h, followed by addition of 0.5 mL of 10% sodium dodecyl sulfate (SDS) and 0.25 mL of 1 M HCl to each sample. The absorbance of the solutions was measured with UV/Vis spectrometer at 335 nm.

Radiolabeling
The radiolabeling of DG-CCPM was shown in Scheme 2. Into 0.2 mL aqueous solution of DG-CCPM containing 6.5 × 10 13 particles was added 10 µL 1M pH 5.2 NaOAc buffer, 40Μl 111 InCl 3 (520µCi). The mixture was allowed to stand at RT for 30 min. Radiolabeled nanoparticles were analyzed using a thin layer chromatography (radio-TLC) system. The Radio-TLC strips were developed with 0.01 M PBS (pH 7.4) containing 4 mM EDTA and quantified using a Bioscan IAR-2000 TLC Imaging Scanner (Washington, DC, USA). Free 111 In 3+ moved to the solvent front (Rf~0.9), and the nanoparticles remained at the original (Rf~0.0). The labeling efficiency for DG-CCPM was >96%. The radiochemical purity was > 96%. DTPA-NIRF-CCPM was labeled with 111 Inas a control using similar procedures.

Fluorescence Microscopy Study
Binding of NIRF-CCPM and DG-CCPM to tumor cells was studied using a human epithelial carcinoma A431 cell line. Cells were incubated with NIRF-CCPM (1.54 × 10 15 particles/mL) and DG-CCPM

Biodistribution Studies
All animal studies were approved by Institutional Animal Care and Use Committee (IACUC), and were handled and maintained according to IACUC guidelines. The mice were kept under specific pathogen free conditions. A-431 tumor cells were inoculated subcutaneously in the right thigh of nude mice (20-25 g; Harlan Sprague Dawley, Inc., Indianapolis, IN, USA) by injecting 1 × 10 6 viable tumor cells in suspension in PBS. When tumors had grown to 8-12 mm in average diameter, mice were randomly allocated into 3 groups, with each group consisting of three mice. Mice in group 1 and 2 were injected with 111 In-DG-CCPM and were euthanized at 4 h and 24 h after injection, respectively. Mice in group 3 were injected with 111 In-DTPA-NIRF-CCPMand were euthanized at 24 h after injection. Mice in each group were injected intravenously with 111 In-labeled DG-CCPM at a dose of 1.3 × 10 12 particles/mouse (10 μCi/mouse). Animals were euthanized by CO 2 exposure 48 h after injection. Blood, lung, muscle, spleen, kidney, liver, and tumor tissues were removed, weighed, and counted for radioactivity with a Packard Cobra gamma counter (Downers Grove, IL, USA). Uptakes of nanoparticles in various tissues were calculated as the percentage of the injected dose per gram of tissue (%ID/g).

In Vivo Gamma Imaging
Nude mice bearing subcutaneous A431 tumors on their upper arm, in two groups of three mice each, were injected intravenously with either 111 In-DG-CCPM or 111 In-DTPA-NIRF-CCPMat a dose of 300µCi/4 × 10 13 particles/mouse. For gamma imaging, mice were placed in the anterior position. Prior to imaging, mice were anesthetized with 2% isoflurane gas (Iso-Thesia, Rockville, NY, USA) in oxygen. During imaging, mice were maintained in an anesthetized state with 0.5%-1.5% isofluranein oxygen. gamma images were acquired using an M-Camgamma camera equipped with medium-energy collimatorand ICON software (Siemens, Hoffman Estates, IL, USA). Images were acquired with the following parameters: matrix: 512 pixels × 512 pixels; zoom: 3.20; energy peak: 247 keV (15%) and 172 keV (15%); total counts: 500 k.

In Vivo NIRF Imaging
Whole-body optical imaging was done at various time points up to 120 h after injection using Xenogen-IVIS ® Lumina II imaging system (Hopkinton, MA, USA).The field of view was 12.5 cm in diameter. For ex vivo imaging, each organ resected at the end of the imaging study (120 h after injection) were imaged with ICG filter sets (excitation, 745 nm; emission, 810-875 nm). The camera settings included CCD Size 1.3 × 1.3 cm, 1024pixel × 1024pixel resolution, 13 microns pixel size and an exposure time of 1.0 s.

Statistical Analysis
Biodistribution data were analyzed using two-tailed, unpaired Student's t tests, with p < 0.05 considered to be statistically significant. The in vitro percentage in vivo percentage of injected dose per gram (%ID/g) of wet tissue weight, and tumor-to-normal tissue ratios are presented as means ± standard errors of the means. All statistical computations were processed by using a computer software program (Excel; Microsoft, Redmond, WA, USA).

Conclusions
We successfully synthesized novel glyco-nanoparticles DG-CCPM.It was labeled with 111 In through radiometal chelator diethylentriaminepentaaceticacid (DTPA) with high radiolableing efficiency. There were no significant differences with respect to tumor-to-normal tissue uptake ratios between 111 In-DG-CCPM and its corresponding control 111 In-DTPA-NIRF-CCPM at 4 h and 24 h post injection. However, both nuclear imaging and optical imaging delineated tumor only in mice injected with 111 In-DG-CCPM at later time points (24-72 h post injection). Further studies to compare the pharmacokinetics and to validate specific tumor uptake of 111 In-DG-CCPM are needed in order to understand the behaviors of ligand-conjugated CCPM as a platform for targeted delivery of nanoparticles for dual nuclear/NIRF fluorescence imaging.