Octreotide Functionalized Nano-Contrast Agent for ... - ACS Publications

Nov 16, 2016 - and Edward G. Robins. ‡,∥. †. Institute of Chemical and Engineering ... Island, Singapore, 627833. ‡. Singapore Bioimaging Cons...
0 downloads 0 Views 8MB Size
Download PDF
Article pubs.acs.org/Biomac

Octreotide Functionalized Nano-Contrast Agent for Targeted Magnetic Resonance Imaging Alexander W. Jackson,*,† Prashant Chandrasekharan,*,‡ Boominathan Ramasamy,‡ Julian Goggi,‡,§ Kai-Hsiang Chuang,‡,§,∥ Tao He,† and Edward G. Robins‡,∥ †

Institute of Chemical and Engineering Sciences, Agency for Science, Technology and Research (A* Star), 1 Pesek Road, Jurong Island, Singapore, 627833 ‡ Singapore Bioimaging Consortium, Agency for Science, Technology and Research (A* Star), 11 Biopolis Way, Helios, Singapore, 138667 § Department of Physiology, Yong Loo Lin School of Medicine, National University of Singapore, Singapore, 117456 ∥ Clinical Imaging Research Centre, Yong Loo Lin School of Medicine, National University of Singapore, Singapore, 117599 S Supporting Information *

ABSTRACT: Reversible addition−fragmentation chain transfer (RAFT) polymerization has been employed to synthesize branched block copolymer nanoparticles possessing 1,4,7,10tetraazacyclododecane-N,N,′N,″N,‴-tetraacetic acid (DO3A) macrocycles within their cores and octreotide (somatostatin mimic) cyclic peptides at their periphery. These polymeric nanoparticles have been chelated with Gd3+ and applied as magnetic resonance imaging (MRI) nanocontrast agents. This nanoparticle system has an r1 relaxivity of 8.3 mM−1 s−1, which is 3 times the r1 of commercial gadolinium-based contrast agents (GBCAs). The in vitro targeted binding efficiency of these nanoparticles shows 5 times greater affinity to somatostatin receptor type 2 (SSTR2) with Ki = 77 pM (compared to somatostatin with Ki = 0.385 nM). We have also evaluated the tumor targeting molecular imaging ability of these branched copolymer nanoparticle in vivo using nude/NCr mice bearing AR42J rat pancreatic tumor (SSTR2 positive) and A549 human lung carcinoma tumor (SSTR2 negative) xenografts.

■

INTRODUCTION Gadolinium-based contrast agent (GBCAs) are widely used in magnetic resonance imaging (MRI) for signal enhancement by shortening the T1 relaxivity (longitudinal or spin−lattice relaxation).1 Typically GBCAs include the FDA approved molecules of gadopentetic acid (Gd(DTPA)) and gadoteric acid (Gd(DOTA)).2−4 The r1 relaxivities of Gd(DTPA) and Gd(DOTA) are in the order of 3 mM−1 s−1 at higher clinical magnetic fields (3T) and are used at a clinically approved concentration of 0.1 mM/kg body weight.5 MRI imaging using T1 relaxing agent has the following relationship with respect to concentration:6

Nanotechnology has been widely investigated to develop MRI nanocontrast agents that are able to improve imaging sensitivity and biodistribution. Using various synthetic techniques, nanoparticle Gd3+-based contrast agents have been developed. Several architectures including dendrimers,8−12 liposomes,13−16 micelles,17−21 core-cross-linked star polymers7 and hyper branched polymers,7,22 have demonstrated improved pharmacokinetic profiles, increased resistance to enzymatic degradation and good chemical stability. The most important advantage of nanoparticles in the field of medicine lies in their ability to be functionalized with targeting moieties for site specific imaging or drug delivery,23,24 however, this potential benefit has yet to be fully exploited.25 While nanoparticle contrast agents can be very useful, their development is very challenging. The synthesis of dendrimers often suffers from challenging multistep synthesis and purification problems. Indirect assembly strategies utilizing amphiphilic block copolymers could lead to uncertainty during in vivo studies as the resulting micellar structures are not very

T1−1 = T1(0)−1 + r1·[CM]

where r1 is the relaxivity of the contrast agent, [CM] is the molar concentration of the contrast agent, T1(0) is the relaxivity of tissue before enhancement, and T1 is the relaxivity in the tissue post contrast agent administration. The required concentration of the contrast agent clearly depends on the relaxivity of the contrast agent. With this consideration, several contrast agents with improved T1 relaxivity have been designed. Vasovist is designed7 to bind to albumin in blood and has a typical r1 relaxivity of 10 mM−1 s−1 (at 3T), almost 3 times more sensitive than Gd(DTPA) or Gd(DOTA).6 © XXXX American Chemical Society

Received: August 22, 2016 Revised: November 14, 2016 Published: November 16, 2016 A

DOI: 10.1021/acs.biomac.6b01256 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules stable upon dilution. There is certainly a need for the development of more versatile and efficient synthetic routes to polymer-based nanoparticle contrast agents. Core crosslinked star and hyperbranched/nanogel polymer nanoparticles have very attractive synthetic procedures that can directly furnish covalently cross-linked architectures.7 More interestingly, it has been demonstrated that the precise molecular location of gadolinium atoms within nanoparticles has a significant influence on relaxivity and the efficacy of nanoparticle MRI contrast agents.26 Currently there is great interest in developing active targeting nanotechnologies for medical applications.27,28 The incorporation of a targeting moiety can decrease off target effects as the nanoparticle contrast agent or drug delivery vehicle accumulates predominately at a desired site. This concentration at the desired area of the body may also significantly increase the efficiency of the imaging probe or drug delivery system. It is even possible to adorn polymer-based nanoparticles with active cell-penetrating peptides to facilitate cellular uptake of the nanoparticle by receptor-mediated endocytosis, a process that requires a significantly lower concentration gradient across the plasma membrane than simple endocytosis. While these advantages are very appealing the synthetic challenge of preparing a highly functional material through a relatively simple process is a significant barrier to overcome. Non-site-specific T1 polymer nanocontrast agents can passively target tumor cells via the enhanced permeation and retention (EPR) effect. To improve the accumulation of contrast agents at precise sites, various targeting moieties such as peptides or antibodies have been conjugated onto polymer nanoparticle contrast agents to improve MRI efficiency through receptor-mediated active targeting.29−31 For example, Peptide T7 (HAIYPRH) has high affinity for Transferrin receptors which are highly expressed on liver cancer cells and brain glioma cells.32 GRGDS peptides have been conjugated onto polymer-based nanoparticle MRI contrast agents formed by self-assembly of linear comb-like polymers, and in vitro results clearly showed an increased uptake within U87MG glioblastoma cells.33 Folic acid has been conjugated onto the surface of hyper branched polymers and 19F MRI results indicate high uptake by B16 melanoma cells.34 Our previous research35−37 focused on developing a direct synthetic route to prepare covalently cross-linked branched copolymer nanoparticles without the need for self-assembly or subsequent cross-linking. We have illustrated a novel synthetic approach to amphiphilic branched copolymer nanoparticles, which comprise a hydrophilic corona and a hydrophobic covalently cross-linked core. Utilizing this method we have reported38 the direct synthesis of stable Gd3+ chelated branched copolymer nanoparticle contrast agents with a very high r1 relaxivity of 14 mM−1 s−1 (almost 4.5 times clinical GBCAs), via reversible addition−fragmentation chain transfer (RAFT) polymerization, for application in MRI. Our previous study reports that these nanoparticles are efficient MRI contrast agents. In vitro and in vivo experiments concluded that these nanoparticles are biocompatible, and have high relaxivity and long blood retention time. Xenograft experiments carried out at a dose of 0.045 mmol/kg further demonstrated that these nanoparticles are able to perfuse and passively accumulate in tumor cells, presumably by means of the EPR effect. Somatostatin receptors (SSTRs) are highly expressed on various tumor cells including breast cancer,39 small-cell lung cancer,40 gastric cancer, and hepatocellular carcinoma.41

Octreotide is a very efficient somatostatin analogue (SSTA),42 which displays high affinity for the SSTRs expressed on tumor cells, while being more synthetically amenable to conjugation. The current state of the art focuses on covalently linking octreotide to a macrocyclic chelator (DOTA or DTPA) to prepare small molecule targeting systems.43−45 These structures are commonly radiolabeled with 111In, 64Cu, and 68 Ga for application in SPECT/PET imaging. Although several small molecular SSTA PET radiotracers have been developed, there are very few reports of similar SSTA MRI studies reported in the literature. This is clearly an area that should be exploited, and we hypothesize that the advantages of modern polymer-based nanocontrast agents when combined with the interesting targeting capabilities of octreotide could lead to a useful system for investigating site-specific MRI. Herein, we report the direct synthesis of branched block copolymer nanoparticle MRI contrast agents. These polymerbased nanoparticles contain gadolinium chelated 1,4,7,10tetraazacyclododecane-N,N,′N,″N,‴-tetraacetic acid (Gd3+DO3A) macrocycles within their cores and octreotide peptides on their periphery. This is achieved through a RAFT polymerization mediated “arm-first” approach and we have evaluated their targeting potential both in vitro and in vivo. Overall these active-targeting nanoparticle contrast agents show promise as molecular imaging probes.

■

EXPERIMENTAL SECTION

Materials and Equipment. Lys(Boc)octreotide was prepared as previously described.46 Briefly, octreotide acetate was treated with 1 equiv of di-tert-butyl dicarbonate in dimethylformamide (DMF). Oligoethylene glycol methyl methacrylate monomethyl ether (Mw ≈ 300 g/mol) and n-butyl methacrylate were passed through a basic alumina column to remove the inhibitor prior to use. 1H and 13C NMR spectra were recorded on Bruker 400 Ultra Shield spectrometer. High-resolution mass spectrometry was performed on a Thermo Finnigan MAT 95XP HRSM spectrometer. Gel permeation chromatography (GPC) was conducted on a Waters 717 plus autosampler equipped with a Waters 515 pump and a Waters 2414 refractive index (RI) detector. Three columns; Styragel HR0.5 (0− 1000), Styragel HR3 (500−30 000) and a Styragel HR5E (2000− 4 000 000) were applied in sequence for separation. Dimethylformamide was used as the eluent at 0.8 mL/min. Dynamic light scattering analysis was carried out on a Malvern Zetasizer Nano-ZS machine. ICP-MS was calibrated for Gadolinium prepared from serial dilution of Gadolinium ICP standard solution (Sigma-Aldrich, Fluka 356220) in 1 M nitric acid. A known amount of polymer nanoparticles chelated with gadolinium was taken in a clean glass vial and digested using 70% concentrated nitric acid (200 μL). The vial was heated at 80 °C for 45 min; the sample was finally made up to 20 mL using DI water. The diluted sample was further filtered through a 0.45 μM disposable sterile polytetrafluoroethylene (PTFE) syringe filter unit. The sample was analyzed using an ICP-MS system (Agilent 7500). Gadolinium at molecular weight of 156 and 157 were chosen for detection. Thermo elemental analysis was performed on a Thermo Scientific Flash 2000 and analyzed using Eager 300 software. A combustion temperature of 950 °C, an oven temperature of 65 °C, and an analysis cycle of 720s were applied. Cryogenic transmission electron microscopy (cyroTEM) analysis was performed on Tecnai G2 Spirit-T12 with 120 kV acceleration, a Gatan UltraScan 4kx4k camera, and a Mark IV vitrobot. Succinimidyl Ester Poly(oligoethylene glycol methacrylate) (Poly(OEGMA)). 4-Cyano-4-(phenylcarbonothioylthio)pentanoic succinimide ester (1, 94.1 mg, 0.25 mmol, 1 equiv) and AIBN (4.1 mg, 25.0 μmol, 0.1 equiv) were added to a small Schlenk tube. Oligoethylene glycol methacrylate monomethyl ether (Mw ≈ 300 g/ mol, 4.50 g, 15.0 mmol, 60 equiv) was then added followed by DMF (15 mL). The reaction mixture was degassed via the freeze−pump− thaw technique and the vessel was backfilled with N2. The reaction B

DOI: 10.1021/acs.biomac.6b01256 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules mixture was then placed in an oil bath at 70 °C, and the polymerization was quenched after 20 h. The reaction mixture was dissolved in a minimal amount of THF and added dropwise to a large excess of ice cold hexane. The precipitation was repeated twice before the polymer (Poly(OEGMA)) was obtained as pink oil (3.90 g). 1H NMR (CDCl3): δ 4.06 (br, C(O)OCH2), 3.64 (br, CH2OCH2), 3.54 (br, CH2OCH3), 3.36 (br, CH2OCH3), 2.85 (br, C(O)CH2CH2(O)C), 1.76 (br, C(CH3)CH2), 0.86 (br, C(CH3)CH2). tert-Butyl DO3A Branched Copolymer Nanoparticle (N(tBuDO3A)). Poly(OEGMA) (3.60 g, 0.20 mmol, 1 equiv) and AIBN (3.3 mg, 20.0 μmol, 0.1 equiv) were added to a small Schlenk tube. Butyl methacrylate (0.85 g, 6.0 mmol, 30 equiv) and 1,3-ethyl methacrylate-5-methyDO3A tribenzamide (2, 0.96 g, 1.0 mmol, 5 equiv) were then added followed by DMF (15 mL). The reaction mixture was degassed via the freeze−pump−thaw technique and the vessel was backfilled with N2. The reaction mixture was then placed in an oil bath at 70 °C, and the polymerization was quenched after 20 h. The reaction mixture was dissolved in a minimal amount of THF and added dropwise to a large excess of ice cold hexane. The precipitation was repeated twice before the branched copolymer nanoparticle (N(tBuDO3A)) was obtained as light pink oil (4.80 g). 1H NMR (CDCl3): δ 4.06 (br, C(O)OCH2), 3.93 (br, CH2CH2CH2CH3), 3.63 (br, CH2OCH2), 3.55 (br, CH2OCH3), 3.36 (br, CH2OCH3), 2.85 (br, C(O)CH2CH2(O)C), 1.78 (br, C(CH3 )CH2), 1.59 (br, CH2CH2CH2CH3), 1.42 (br, C(CH3)3), 1.37 (br, CH2CH2CH2CH3), 0.93 (br, CH2CH2CH2CH3), 0.85 (br, C(CH3)CH2). DO3A Branched Copolymer Nanoparticle (N(DO3A)). N(tBuDO3A) (4.40 g) was dissolved in CH2Cl2 (40 mL), the solution was cooled to 0 °C in an ice bath. TFA (20 mL) was added and the reaction was stirred at 0 °C for 4 h. After this time the reaction was evaporated to dryness. The oil obtained was dissolved in THF (10 mL) and added dropwise to ice cold hexane. The precipitation was repeated twice before the branched copolymer nanoparticle (N(DO3A)) was obtained as light pink oil (4.10 g). 1H NMR (CDCl3): δ 4.05 (br, C(O)OCH2), 3.92 (br, CH2CH2CH2CH3), 3.66 (br, CH2OCH2), 3.55 (br, CH2OCH3), 3.36 (br, CH2OCH3), 2.85 (br, C(O)CH 2 CH 2 (O)C), 1.79 (br, C(CH 3 )CH 2 ), 1.59 (br, CH 2 CH 2 CH 2 CH 3 ), 1.38 (br, CH 2 CH 2 CH 2 CH 3 ), 0.91 (br, CH2CH2CH2CH3), 0.86 (br, C(CH3)CH2). Conjugation of Octreotide (N(DO3A)Oct). N(DO3A) (2.05 g) was dissolved in DMF (30 mL) followed by the addition of Lys(Boc)octreotide (0.43 g, 0.38 mmol, 5 eq per polymer chain/ succinimidyl functionality) and Et3N (77 mg, 0.76 mmol, 10 eq per polymer chain/succinimidyl functionality). The reaction was left to stir at room temperature for 24 h. After this time the solution obtained was purified extensively by dialysis (Mw cutoff 7 kDa) for 48 h. The obtained solution was evaporated to dryness and dissolved in THF (50 mL) and added dropwise to ice cold hexane and the Lys(Boc)octreotide functional branched copolymer nanoparticle isolated as a clear oil. This precursor was treated with TFA (10 mL) in CH2Cl2 (20 mL) at 0 °C for 4 h, after this the reaction was evaporated to dryness. The oil obtained was dissolved in THF (10 mL) and added dropwise to ice cold hexane. The precipitation was repeated twice before the branched copolymer nanoparticle (N(DO3A)Oct) was obtained as a clear oil (1.80 g). 1H NMR (CDCl3 + 1 drop of DMSO-d6): δ 7.8−6.8 (multiple octreotide peaks), 4.06 (br, C(O)OCH2), 3.94 (br, CH2CH2CH2CH3), 3.64 (br, CH2OCH2), 3.56 (br, CH2OCH3), 3.35 (br, CH 2 OCH 3 ), 1.80 (br, C(CH 3 )CH 2 ), 1.58 (br, CH 2 CH 2 CH 2 CH 3 ), 1.39 (br, CH 2 CH 2 CH 2 CH 3 ), 0.92 (br, CH2CH2CH2CH3), 0.85 (br, C(CH3)CH2). Chelation of Gadolinium within Nanoparticles (N(Gd3+DO3A) and (N(Gd3+DO3A)Oct). A typical chelation was conducted as follows: either N(DO3A) or N(DO3A)Oct (1.10 g, possessing approximately 80 mg and 0.20 mmol of DO3A units) was dissolved in isopropanol (15 mL). GdCl3.6H2O (1.86 g, 5.0 mmol, approximately 25 eq relative to DO3A units) was then added in 10 mM ammonium acetate (70 mL). The reaction was heated at 55 °C for 24 h. After this time, the solution obtained was purified extensively by dialysis (Mw cutoff 7 kDa) for 48 h. The obtained solution was evaporated to dryness and dissolved in THF (50 mL) and added

dropwise to ice cold hexane. The desired gadolinium chelated branched copolymer nanoparticles N(Gd 3+ DO3A) and N(Gd3+DO3A)Oct were obtained as clear oils (approximately 0.80 g each). 1H NMR (CDCl3): δ 4.06 (br, C(O)OCH2), 3.91 (br, CH2CH2CH2CH3), 3.68 (br, CH2OCH2), 3.54 (br, CH2OCH3), 3.37 (br, CH2OCH3), 1.80 (br, C(CH3)CH2), 1.57 (br, CH 2 CH 2 CH 2 CH 3 ), 1.39 (br, CH 2 CH 2 CH 2 CH 3 ), 0.92 (br, CH2CH2CH2CH3), 0.87 (br, C(CH3)CH2). Relaxivity Measurements of N(Gd 3+ DO3A) and N(Gd3+DO3A)Oct. Relaxivity was measured using a 300 MHz (7Tesla) Bruker ClinScan system with a Siemens interface. A mouse body coil was used for receive and transmit channel. Phantom of N(Gd3+DO3A) and N(Gd3+DO3A)Oct were prepared in different concentration (mM of gadolinium) in 1 mL syringe. r1 relaxivity was measured using a spin echo inversion recovery sequence TR/TE 8000/7 ms, and different inversion times TI = 31, 300, 700, 1400, 2500, 3500, 5500, and 7500 ms were used. T1 parametric maps were obtained using an in-house program (Matlab 7.7). Evaluation of Binding of N(Gd3+DO3A) and N(Gd3+DO3A)Oct to SSTR2 Receptor. Four nanoparticles were evaluated for their binding affinities to SSTR2 receptors. In brief, N(DO3A), N(Gd3+DO3A), N(DO3A)Oct, and N(Gd3+DO3A)Oct were prepared in 4 mL of sterile water. The octreotide concentration in solution was 0.5 mM. CHO-K1/SSTR2 cells (stable cell lines) were homogenized under frozen condition in cold buffer. The homogenate was centrifuged at 25 000g for 30 min. The pellets were resuspended in an assay buffer. The membrane protein was frozen at −80 °C. In a 96well plate, 120 μL of membrane (2 μg protein/well) were incubated with 15 μL of [125I]-somatostatin (500 pM final) and 15 uL of compound (prepared in various concentration) at 30 °C for 2 h. The binding reaction was then stopped by rapid filtration through Unifilter GF/B plates, the plates were then washed with ice cold water three times. The filtration plates were dried at 37 °C overnight and the radioactivity was determined by TopCount (PerkinElmer). The KD values were determined by a scatchard plot. Preparation of Xenograft Mouse Model. All animal work was approved by A*STAR ethics committee under the IACUC #140898. Xenografts were developed in nude/NCr mice female 5−6 weeks old (18−22g). In brief, A549 (negative for somatostatin receptors (SSTR2)) cells were mixed with Matrigel and injected at a cell number of 3 × 106 cells on the left flank and AR42J (positive for somatostatin receptors (SSTR2)) cells were injected at a cell number of 2.5 × 106 cells on the right flank. The tumors were allowed to grow to a volume of 120−150 mm3 before MRI scanning. Tracer Injection and DCE-MRI. Mice were scanned using a 7Tesla Bruker ClinScan, using a mouse body receive-transmit coil. Animals were kept anesthetized by supplying medical air and 1−1.5% isoflurane throughout the imaging experiment. Physiology of the animal was maintained at 36 °C. A tail vein catheter was introduced with heparinized needle. N(Gd3+DO3A) and N(Gd3+DO3A)Oct were prepared at a concentration of 10 mM (Gd) and 10 mM (Gd)/ 2 mM (octreotide), respectively. 120 μL of the N(Gd3+DO3A) or N(Gd3+DO3A)Oct was given to the mice as a bolus at a dose of 0.06 mM of Gd/kg during the dynamic contrast-enhanced MRI measurement (DCE-MRI) after 1 min baseline acquisition. Before and following bolus injection, T1 weighted anatomy images and a T1 map of the abdominal part of the mice including the xenografts were imaged at different times post contrast administration. DCE-MRI was carried our using a FLASH sequence with TR/TE = 10/1.2 ms; FA = 20°; with 8.9 s per acquisition, with a total of 75 acquisitions (total scan time of 11 min). T1 weighted images were acquired using a respiration triggered turbo-spin echo sequence with TR/TE = 913/8.9 ms; FA = 180°; FoV = 96.8 × 96.8 × 1000 um3 taken in the transverse plain. T1 parametric maps were acquired by using a respiration triggered spin−echo sequence with inversion−recovery using TR/TE = 5000/7.7 ms. FA = 180° and inversion times (TI) = 41, 200, 500, 1500, 2000, and 4000 ms. T1 parametric maps were obtained using inhouse program (Matlab 7.7). C

DOI: 10.1021/acs.biomac.6b01256 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

Scheme 1. Synthetic Route for the Preparation of Active Targeting Octreotide Functionalized and Gadolinium Chelated Branched Copolymer Nanoparticle N(Gd3+DO3A)Oct and Non-targeting Gadolinium Chelated Branched Copolymer Nanoparticle N(Gd3+DO3A)

■

pentanoic acid in the presence of dicyclohexylcarbodiimide.52 The dimethacrylate DO3A functional monomer 2 was synthesized as previously described.38 The overall synthesis of branched copolymer nanoparticles, via RAFT polymerization is illustrated in Scheme 1. It is well understood that nanoparticles are prone to sequestration by the mononuclear phagocyte system (MPS), which is triggered by the adsorption of plasma proteins onto nanoparticles.53 Therefore, we chose to employ poly(ethylene glycol) (PEG)-based monomers within the nanoparticle corona generating “stealth-like” nanoparticles to minimize protein adsorption and avoid MPS clearance. Initially, a linear succinimidyl ester end-functionalized oligoethylene glycol-based Poly(OEGMA) was synthesized via RAFT polymerization. Due to the living nature of the RAFT technique these linear copolymer building blocks retain their terminal dithioester functional groups allowing subsequent polymerization. Copolymerization in the presences of the DO3A containing dimethacrylate cross-linker (2) and a comonomer n-butyl methacrylate was then performed leading to highly branched but soluble macromolecules N(tBuDO3A). n-Butyl methacrylate was employed as a comonomer during the cross-linking polymerization due to its lower steric hindrance, we have also previously demonstrated38 that this core-forming

RESULTS AND DISCUSSION Initially, linear polymer chains are prepared, which possess succinimidyl ester-activated functionalities at their chain terminus. These linear polymer building blocks are then chain-extended and simultaneously cross-linked via the application of a dimethacrylate DO3A monomer (2), which incorporates the macrocycle chelator through two methacrylate functionalities increasing macrocycle stability. The resulting branched copolymer nanoparticles were conjugated with octreotide through amide bond formation at the previously mentioned succinimidyl ester-activated functional groups (Scheme 1). Gd3+ was then chelated, and the in vitro and in vivo characteristics of nanoparticles possessing octreotide were compared against nanoparticles without any targeting capabilities. Our RAFT-mediated approach to branched copolymer nanoparticles is synthetically similar to the previously described47−50 “arm-first” core cross-linked star polymers, an approach that has been recently reviewed51 by Davis, Boyer, and Qiao. Synthesis of Branched Copolymer Nanoparticles via RAFT Polymerization. The desired succinimidyl ester RAFT chain transfer agent 1 was synthesized by coupling Nhydroxysuccinimide with 4-cyano-4-(phenylcarbonothioylthio)D

DOI: 10.1021/acs.biomac.6b01256 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

Figure 1. (a) Gel permeation chromatography trace of Poly(OEGMA) and N(tBuDO3A) in DMF (0.8 mL/min). (b) Dynamic light scattering analysis in H2O of N(tBuDO3A), N(DO3A), and N(DO3A)Oct at 25 °C. (c) Cryo-TEM analysis of N(tBuDO3A) (left) and N(DO3A)Oct (right). (d) Relaxivity of N(Gd3+DO3A) and N(Gd3+DO3A)Oct nanoparticles of linear fitting of 1/T1 versus concentration.

Figure 2. (a) 1H NMR (CDCl3) of N(tBuDO3A). Inset displays the tBu deprotection, the black line shows N(tBuDO3A) before deprotection, and the red line shows N(DO3A) after treatment with TFA. (b) 1H NMR (CDCl3) of octreotide functional N(tBuDO3A)Oct, inset displays an expansion of the octreotide aromatic protons.

The succinimidyl ester RAFT chain transfer agent (1, 1eq) was used to mediate the polymerization of oligoethylene glycol methacrylate monomethyl ether (OEGMA, 60 equiv) at 70 °C in DMF to afford the linear succinimidyl ester end-functionalized polymer Poly(OEGMA) (Scheme 1), 1H NMR analysis (ESI, S1) displayed a peak at δ = 2.85 ppm confirming the presence of the terminal succinimidyl ester functionality. Gel permeation chromatography (GPC) analysis (Figure 1a) of Poly(OEGAM) displayed a single peak at 18.7 min which corresponds to Mn = 16 200 Da and Mw = 19 300 Da (PDI = 1.19). This result confirmed a relatively low polydispersity and suggested good control over the polymerization. The linear succinimidyl end-functionalized Poly(OEGMA) (1 equiv) was chain extended (Scheme 1) with n-butyl methacrylate (30 equiv) in the presence of the dimethacrylate

monomer can facilitate the solubilization of hydrophobic guest molecules which could permit hydrophobic drug loading for future work in theranostics. The dimethacrylate cross-linker (2) employed incorporates the DO3A macrocycles simultaneously during the branching process ensuring the DO3A macrocycles are predominantly linked to the branched copolymers through two methacrylate functionalities, increasing their stability upon undesirable ester hydrolysis. The tert-butyl protecting groups within the DO3A macrocycles were converted to the corresponding carboxylic acids furnishing N(DO3A). A portion of N(DO3A) was chelated with gadolinium to prepare the nontargeting control nanoparticle N(Gd3+DO3A). To synthesize the targeting nanoparticles N(DO3A) was conjugated with octreotide to yield N(DO3A)Oct which was then chelated with gadolinium to afford N(Gd3+DO3A)Oct. E

DOI: 10.1021/acs.biomac.6b01256 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

Figure 3. (a) In vitro competitive binding curve and (b) IC50 and Ki values of N(DO3A), N(Gd3+DO3A), N(DO3A)Oct, N(Gd3+DO3A)Oct and somatostatin control for SSTR2 receptor.

DO3A functional cross-linker 2 (5 equiv) at 70 °C in DMF to afford the branched copolymer nanoparticle N(tBuDO3A) composed of covalently cross-linked copolymer chains. The ratio of the branching dimethacrylate monomer 2 to the growing polymer chain is restricted to five monomer units per chain and the reaction carried out under relatively dilute conditions, thus the formation of a macromolecular gel is inhibited, and discrete soluble molecular species are formed. Dynamic light scattering analysis (Figure 1b) displayed a single peak with a hydrodynamic diameter of 43 nm (PDI = 0.207) and 1H NMR analysis (Figure 2a) displayed a peak at δ 2.89 ppm confirming the continued presence of the succinimidyl ester functionalities. Gel permeation chromatography (GPC) analysis (Figure 1a) of N(tBuDO3A) displayed a broad single peak at 16.1 min, which corresponds to Mn = 58 500 Da and Mw = 108 800 Da (PDI = 1.86). This data confirmed the successful chain extension of Poly(OEGMA) and the formation of a covalently cross-linked branched copolymer. Comparing cross-linked branched polymers to linear polymer standards does not generate accurate molecular weights, and it is likely that the molecular weight of N(tBuDO3A) is significantly higher than the values obtained. N(tBuDO3A) was further analyzed by cryo-TEM (Figure 1c) which displayed nanoparticles of sizes similar to those determined by dynamic light scattering. To permit chelation of gadolinium the tertbutyl groups present within the DO3A macrocycles had to be deprotected. This was achieved by treating the branched copolymer nanoparticle N(tBuDO3A) with TFA in CH2Cl2 (Scheme 1); the resulting branched copolymer nanoparticle N(DO3A) was characterized by 1H NMR spectroscopy, which confirmed, when overlapped with the 1H NMR spectra of N(tBuDO3A) (Figure 2a (inset)), that deprotection had successfully occurred. 1H NMR analysis also displayed a peak at δ = 2.89 ppm confirming that the conditions had not affected the terminal succinimidyl ester functionalities. Dynamic light scattering analysis in H2O of N(DO3A) (Figure 1b) displayed a slightly larger hydrodynamic diameter of 52 nm (PDI = 0.243) when compared to N(tBuDO3A) (43 nm), confirming that the deprotection process does not result in any undesirable nanoparticle decomposition. The increase in hydrodynamic diameter is likely due to the increased hydrophilicity of the nanoparticle core resulting in an influx of water and subsequent swelling. Conjugation of Octreotide to Branched Copolymer Nanoparticle. To facilitate active cell targeting the Somatostatin analog octreotide was conjugated to N(DO3A) via amide

bond formation at the succinimidyl ester functionalities present at the nanoparticle periphery. N(DO3A) was treated with an excess of Lys(Boc)octreotide in the presence of Et3N. The lysine residue plays a key role in bioactivity and the lysine amino group was therefore protected prior to conjugation to ensure amide bond formation occurred exclusively at the phenylalanine amino functionality. 1H NMR analysis (Supporting Information, Figure S2) of Lys(Boc)octreotide46 confirmed the successful protection of the lysine residue. After conjugation, the nanoparticles were purified from the excess Lys(Boc)octreotide by extensive dialysis (Mw cutoff 7 kDa) against water for 48 h and subsequently treated with trifluoroacetic acid in CH2Cl2 to facilitate the deprotection of the Lys(Boc) residue resulting in the desired octreotidedecorated branched copolymer nanoparticle N(DO3A)Oct. The presence of octreotide was confirmed by 1H NMR spectroscopy (Figure 2b) and by thermo elemental analysis which displayed an increase in sulfur content from 0.0% (N(DO3A)) to 0.57% (N(DO3A)Oct) and an increase in nitrogen content from 0.47% (N(DO3A)) to 1.74% (N(DO3A)Oct). The increased presence of sulfur and nitrogen confirms the successful conjugation of approximately 0.9 Octreotide peptides per polymer chain. Dynamic light scattering analysis in H2O of N(DO3A)Oct (Figure 1b) displayed a hydrodynamic diameter of 56 nm (PDI = 0.255), slightly larger than that of N(DO3A) (53 nm). N(DO3A)Oct was also analyzed by cryo-TEM (Figure 1c), which again displayed well dispersed nanoparticles of sizes similar to those determined by dynamic light scattering. Chelation and Relaxivity of Gadolinium within Branched Copolymer Nanoparticles. N(DO3A) and N(DO3A)Oct were treated with a large excess of GdCl3· 6H2O in a mixture of isopropanol: H2O (30:70) at 55 °C for 24 h to generate N(Gd3+DO3A) and N(Gd3+DO3A)Oct, respectively. The solutions obtained were extensively dialyzed (Mw cutoff 7 kDa) against water for 48 h to completely remove the excess Gd, the aqueous solutions obtained were evaporated to dryness, and the nanoparticles were further purified by precipitation. Known amounts of N(Gd3+DO3A) and N(Gd3+DO3A)Oct were analyzed by ICP, which confirmed chelation values of approximately 95%. The Gd(III) chelated DO3A macrocycle units within the nanoparticles N(Gd3+DO3A) and N(Gd3+DO3A)Oct improve the hydrophilicity of the nanoparticle core, thus ensuring the ability of surrounding water molecules to approach the core and directly exchange with the coordinated water in the Gd(DO3A) F

DOI: 10.1021/acs.biomac.6b01256 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

Figure 4. (a) Tracer blood kinetics from the vena cava after injection of N(Gd3+DO3A) or N(Gd3+DO3A)Oct, (b) perfusion pattern of N(Gd3+DO3A) into AR42J and A549 tumors, (c) perfusion pattern of N(Gd3+DO3A)Oct into AR42J and A549 tumors.

Figure 5. (a) r1 change within AR42J tumors (positive for SSTR2) and A549 (negative for SSTR2) after injection of N(Gd3+DO3A) and (b) r1 change within AR42J tumors (positive for SSTR2) and A549 (negative for SSTR2) after injection of N(Gd3+DO3A)Oct.

(Gd3+DO3A)Oct was injected as an I.V. bolus at a dose of 0.06 mM of Gd/kg. Following injection, T1 weighted anatomical images and T1 maps of the abdominal region of the mice including the xenografts were acquired. After injection the signal change from the vena cava (Figure 4a) for both N(Gd3+DO3A) and N(Gd3+DO3A)Oct indicates similar blood pharmacokinetics for both nanoparticles. For active targeting to occur the nanoparticles must passively perfuse into both the AR42J and A549 tumors. Both N(Gd3+DO3A) (Figure 4b) and N(Gd3+DO3A)Oct (Figure 4c) nanoparticles perfused into the AR42J tumors to a slightly greater extent than the A549 tumors. This finding may be due to small differences in vascularisation between the two tumor types. We also observed that N(Gd3+DO3A)Oct perfused into AR42J to a greater extent than N(Gd3+DO3A) (Figure 4b,c). After nanoparticle perfusion into the tumors it was hypothesized that the N(Gd3+DO3A) nanoparticles would be cleared from both tumors, whereas the Octreotide functional N(Gd3+DO3A)Oct nanoparticles should be specifically retained within AR42J tumors (positive for SSTR2). Change in r1 (Δr1) provides a measure of the accumulation of nanoparticle concentration in the tumor postadministration. The maximum relaxivity change (Δr1) induced by N(Gd3+DO3A) in both tumors (Figure 5a) was observed at 3 h postinjection, indicating the peak accumulation of the nanoparticle. Subsequently, complete clearance of the nanocontrast agent was observed by 24 h for both tumors. However, the octreotide functionalized nanocontrast agent, N(Gd3+DO3A)Oct, displayed peak accumulation in AR42J tumors at 6 h postinjection (Figure 5b). N(Gd3+DO3A)Oct was found to be retained within the AR42J tumor to a slightly greater extent than the A549 tumor, and also demonstrated marginally increased uptake in both tumors by comparison to N(Gd3+DO3A), At 24 h postinjection, N(Gd3+DO3A)Oct was observed to be retained in the AR42J tumor with a mean Δr1 = 0.075 s −1 , corresponding to a Gd 3+ concentration of approximately 0.57 mM Gd. At the same time point,

macrocycles. This was also understood from the relaxivity measurements in deionized water (Figure 1d) using a 7 T Bruker ClinScan determined the longitudinal relaxivity (r1) of N(Gd3+DO3A) and N(Gd3+DO3A)Oct to be 7.6 mM−1 s−1 and 8.3 mM−1 s−1, respectively. This confirms that the presence of octreotide on the nanoparticle periphery does not affect the relaxivity of gadolinium within the nanoparticle core, and also that the gadolinium chelate is not affected by the receptor induced magnetization enhancement. In vitro Affinity SSTR2 Binding Assay. Figure 3a shows the results of an SSTR2 receptor competitive binding assay. The octreotide-conjugated nanoparticles showed good affinity for SSTR2 receptors, with or without gadolinium chelated. The Ki values of N(Gd3+DO3A)Oct and N(DO3A)Oct were 0.077 nM and 0.255 nM, respectively, when compared to that of somatostatin, Ki = 0.385 nM. The Ki value of N(Gd3+DO3A)Oct was 5 times greater than that of somatostatin, while N(DO3A)Oct has a similar binding to that of somatostatin, indicating an improved targeting efficiency for N(Gd3+DO3A)Oct compared to somatostatin (Figure 3b). Nanoparticles without octreotide present at their periphery (N(DO3A) and N(Gd3+DO3A)) showed effectively no binding to the SSTR2 receptors. These preliminary in vitro results showed suitable affinity to proceed with in vivo analysis. In Vivo MRI Xenograft Analysis of Active Targeting Nanoparticles. To examine the in vivo targeting ability of octreotide functional branched copolymer nanoparticles a murine dual tumor model was generated. Mice were implanted with both A549 tumors (negative for SSTR2) on the left flank and AR42J tumors (positive for SSTR2) on the right flank. Each mouse was injected with either N(Gd3+DO3A) or N(Gd3+DO3A)Oct and nanoparticle retention within the two tumors was followed. Both N(Gd 3+ DO3A) and N(Gd3+DO3A)Oct were prepared with an equal Gd concentration of 10 mM, which corresponded to an octreotide concentration of 2 mM. After 1 min of DCE-MRI baseline acquisition, 120 μL of either N(Gd 3+ DO3A) or NG

DOI: 10.1021/acs.biomac.6b01256 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

Figure 6. (a) T1 map showing the xenograft tumors AR42J and A549 before and after injection of N(Gd3+DO3A) and (b) T1 map showing the xenograft tumors AR42J and A549 before and after injection of N(Gd3+DO3A)Oct.

Figure 7. (a) Change in Δr1 within the pancreas post administration of N(Gd3+DO3A) and N(Gd3+DO3A)Oct and (b) yellow arrow pointing to the T1 weighted image of the pancreas of mice injected with N(Gd3+DO3A) and N(Gd3+DO3A)Oct nanoparticles.

N(Gd3+DO3A)Oct was retained in the A549 tumor with Δr1 = 0.02 s−1, corresponding to an approximate Gd3+ concentration of 0.152 mM. Perfusion of both N(Gd3+DO3A) and N(Gd3+DO3A)Oct nanoparticles was observed for both AR42J and A549 tumors. N(Gd3+DO3A) nanoparticles were completely cleared from both tumors at 24 h postadministration. However, both tumor types displayed some degree of nanoparticle retention of N(Gd3+DO3A)Oct. The difference in Δr1 observed for N(Gd3+DO3A)Oct between the two tumor types may indicate some degree of specific binding. However, the differences observed are difficult to identify from the corresponding T1 image (Figure 6). Despite the high in vitro affinity of N(Gd3+DO3A)Oct for the SSTR2 receptor, the in vivo signal enhancement in MRI experiments targeting SSTR2 positive AR42J tumors was not as promising. During our investigation, differences between the two nanocontrast agents were observed in highly perfusive organs, in particular the pancreas, which is also known to express SSTR2 receptors in both β- and α-cells of the pancreatic islets.54,55 Postinjection, both nanoparticles demonstrated uptake within the first 10 min in the region of the pancreas consistent with the highly perfusive nature of the organ. Subsequently, N(Gd3+DO3A) was cleared within 6 h, whereas N(Gd3+DO3A)Oct demonstrated an initial high change in relaxivity at early time points (Δr1 = 0.275 s−1 at 10 min) that slowly reduced in magnitude (Δr1 = 0.100 s−1 at 3 h) before remaining constant up to the 6 h time point (Figure 7).

MRI nanocontrast agents both in vitro and in vivo. The octreotide functional nanoparticles (N(Gd3+DO3A)Oct) have been compared to analogous nanoparticles (N(Gd3+DO3A)) that do not possess any targeting functionalities. Our initial in vitro binding studies confirmed that octreotide functional nanoparticles bind to somatostatin receptor type 2 (SSTR2) to an equal or greater extent than the somatostatin control. In vivo, both N(Gd3+DO3A) and N(Gd3+DO3A)Oct nanoparticles are able to passively perfuse into AR42J (positive for SSTR2) and A549 (negative for SSTR2) xenograft tumors and highly perfusive organs, such as the pancreas. Small differences in tissue relaxivity (Δr1) were observed between the octreotide and non-octreotide functional nanoparticles in the AR42J tumors, necessitating further development of the N(Gd3+DO3A)Oct nanocontrast agents for SSTR2 positive tumor-specific imaging.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.biomac.6b01256.

■

■

Additional 1H NMR characterization of Poly(OEGMA) and Lys(Boc)Octreotide (PDF)

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected].

CONCLUSIONS Branched copolymer nanoparticles functionalized with the somatostatin analogue octreotide at their periphery and DO3A macrocycles within their core have been synthesized by a RAFT polymerization-mediated “arm-first” methodology. After chelation with Gd(III), these nanoparticles have been evaluated as

ORCID

Alexander W. Jackson: 0000-0001-6618-7654 Notes

The authors declare no competing financial interest. H

DOI: 10.1021/acs.biomac.6b01256 Biomacromolecules XXXX, XXX, XXX−XXX

Article

Biomacromolecules

■

(27) Davis, M. E.; Chen, Z.; Shin, D. M. Nat. Rev. Drug Discovery 2008, 7, 771−782. (28) Jokerst, J. V.; Gambhir, S. S. Acc. Chem. Res. 2011, 44, 1050− 1060. (29) Lee, S.; Xie, J.; Chen, X. Y. Chem. Rev. 2010, 110, 3087−3111. (30) Ye, Z.; Zhou, Z. X.; Ayat, N.; Wu, X. M.; Jin, E. L.; Shi, X. Y.; Lu, Z. R. Contrast Media Mol. Imaging 2016, 11, 32−40. (31) Villaraza, A. J. L.; Bumb, A.; Brechbiel, M. W. Chem. Rev. 2010, 110, 2921−2959. (32) Han, L. A.; Li, J. F.; Huang, S. X.; Huang, R. Q.; Liu, S. H.; Hu, X.; Yi, P. W.; Shan, D.; Wang, X. X.; Lei, H.; Jiang, C. Biomaterials 2011, 32, 2989−2998. (33) Shokeen, M.; Pressly, E. D.; Hagooly, A.; Zheleznyak, A.; Ramos, N.; Fiamengo, A. L.; Welch, M. J.; Hawker, C. J.; Anderson, C. J. ACS Nano 2011, 5, 738−747. (34) Rolfe, B. E.; Blakey, I.; Squires, O.; Peng, H.; Boase, N. R. B.; Alexander, C.; Parsons, P. G.; Boyle, G. M.; Whittaker, A. K.; Thurecht, K. J. J. Am. Chem. Soc. 2014, 136, 2413−2419. (35) He, T.; Adams, D. J.; Butler, M. F.; Yeoh, C. T.; Cooper, A. I.; Rannard, S. P. Angew. Chem., Int. Ed. 2007, 46, 9243−9247. (36) He, T.; Adams, D. J.; Butler, M. F.; Cooper, A. I.; Rannard, S. P. J. Am. Chem. Soc. 2009, 131, 1495−1501. (37) He, T.; Di Lena, F.; Neo, K. C.; Chai, C. L. L. Soft Matter 2011, 7, 3358−3365. (38) Jackson, A. W.; Chandrasekharan, P.; Shi, J.; Rannard, S. P.; Liu, Q.; Yang, C. T.; He, T. Int. J. Nanomed. 2015, 10, 5895−5907. (39) Sharma, K.; Srikant, C. B. Int. J. Cancer 1998, 76, 259−266. (40) Weiner, R. E.; Thakur, M. L. BioDrugs 2005, 19, 145−163. (41) Reynaert, H.; Rombouts, K.; Vandermonde, A.; Urbain, D.; Kumar, U.; Bioulac-Sage, P.; Pinzani, M.; Rosenbaum, J.; Geerts, A. Gut 2004, 53, 1180−1189. (42) De Jong, M.; Breeman, W. A. P.; Kwekkeboom, D. J.; Valkema, R.; Krenning, E. P. Acc. Chem. Res. 2009, 42, 873−880. (43) Anderson, C. J.; Dehdashti, F.; Cutler, P. D.; Schwarz, S. W.; Laforest, R.; Bass, L. A.; Lewis, J. S.; McCarthy, D. W. J. Nucl. Med. 2001, 42, 213−221. (44) Eiblmaier, M.; Andrews, R.; Laforest, R.; Rogers, B. E.; Anderson, C. J. J. Nucl. Med. 2007, 48, 1390−1396. (45) Lewis, J. S.; Lewis, M. R.; Srinivasan, A.; Schmidt, M. A.; Wang, J.; Anderson, C. J. J. Med. Chem. 1999, 42, 1341−1347. (46) Smith-Jones, P. M.; Stolz, B.; Bruns, C.; Albert, R.; Reist, H. W.; Fridrich, R.; Macke, H. R. J. Nucl. Med. 1994, 35, 317−325. (47) Syrett, J. A.; Haddleton, D. M.; Whittaker, M. R.; Davis, T. P.; Boyer, C. Chem. Commun. 2011, 47, 1449−1451. (48) Ferreira, J.; Syrett, J.; Whittaker, M.; Haddleton, D.; Davis, T. P.; Boyer, C. Polym. Chem. 2011, 2, 1671−1677. (49) Wei, X. H.; Moad, G.; Muir, B. W.; Rizzardo, E.; Rosselgong, J.; Yang, W. T.; Thang, S. H. Macromol. Rapid Commun. 2014, 35, 840− 845. (50) Li, Y.; Duong, H. T. T.; Laurent, S.; MacMillan, A.; Whan, R. M.; Elst, L. V.; Muller, R. N.; Hu, J. M.; Lowe, A.; Boyer, C.; Davis, T. P. Adv. Healthcare Mater. 2015, 4, 148−156. (51) Ren, J. M.; McKenzie, T. G.; Fu, Q.; Wong, E. H. H.; Xu, J. T.; An, Z. S.; Shanmugam, S.; Davis, T. P.; Boyer, C.; Qiao, G. G. Chem. Rev. 2016, 116, 6743−6836. (52) Zheng, Q.; Pan, C. Y. Macromolecules 2005, 38, 6841−6848. (53) Hume, D. A. Curr. Opin. Immunol. 2006, 18, 49−53. (54) Li, M.; Li, W.; Kim, H. J.; Yao, Q. Z.; Chen, C. Y.; Fisher, W. E. J. Surg. Res. 2004, 119, 130−137. (55) Kailey, B.; van de Bunt, M.; Cheley, S.; Johnson, P. R.; MacDonald, P. E.; Gloyn, A. L.; Rorsman, P.; Braun, M. Am. J. Physiol.Endocrinol. Metab. 2012, 303, E1107−E1116.

ACKNOWLEDGMENTS This research was supported by the JCO Grant (1131CFG002) of A* Star (Agency for Science Technology and Research).

■

REFERENCES

(1) Aime, S.; Caravan, P. J. Magn. Reson. Imaging 2009, 30, 1259− 1267. (2) Caravan, P.; Ellison, J. J.; McMurry, T. J.; Lauffer, R. B. Chem. Rev. 1999, 99, 2293−2352. (3) Aime, S.; Botta, M.; Terreno, E. Adv. Inorg. Chem. 2005, 57, 173− 237. (4) Caravan, P. Chem. Soc. Rev. 2006, 35, 512−523. (5) Murphy, K. J.; Brunberg, J. A.; Cohan, R. H. AJR, Am. J. Roentgenol. 1996, 167, 847−849. (6) Rohrer, M., MRI Contrast Media  Introduction and Basic Properties of the Blood Pool Agent Gadofosveset (Vasovist®). In Clinical Blood Pool MR Imaging, Leiner, T., Goyen, M., Rohrer, M., Schönberg, S., Eds. Springer: Berlin/Heidelberg, 2008; pp 3−15. (7) Li, Y.; Beija, M.; Laurent, S.; vander Elst, L.; Muller, R. N.; Duong, H. T. T.; Lowe, A. B.; Davis, T. P.; Boyer, C. Macromolecules 2012, 45, 4196−4204. (8) Wiener, E. C.; Brechbiel, M. W.; Brothers, H.; Magin, R. L.; Gansow, O. A.; Tomalia, D. A.; Lauterbur, P. C. Magn. Reson. Med. 1994, 31, 1−8. (9) Fu, Y. J.; Raatschen, H. J.; Nitecki, D. E.; Wendland, M. F.; Novikov, V.; Fournier, L. S.; Cyran, C.; Rogut, V.; Shames, D. M.; Brasch, R. C. Biomacromolecules 2007, 8, 1519−1529. (10) Langereis, S.; de Lussanet, Q. G.; van Genderen, M. H. P.; Meijer, E. W.; Beets-Tan, R. G. H.; Griffioen, A. W.; van Engelshoven, J. M. A.; Backes, W. H. NMR Biomed. 2006, 19, 133−141. (11) Jaszberenyi, Z.; Moriggi, L.; Schmidt, P.; Weidensteiner, C.; Kneuer, R.; Merbach, A. E.; Helm, L.; Toth, E. JBIC, J. Biol. Inorg. Chem. 2007, 12, 406−420. (12) Laus, S.; Sour, A.; Ruloff, R.; Toth, E.; Merbach, A. E. Chem. Eur. J. 2005, 11, 3064−3076. (13) Fossheim, S. L.; Fahlvik, A. K.; Klaveness, J.; Muller, R. N. Magn. Reson. Imaging 1999, 17, 83−89. (14) Lokling, K. E.; Fossheim, S. L.; Skurtveit, R.; Bjornerud, A.; Klaveness, J. Magn. Reson. Imaging 2001, 19, 731−738. (15) Wang, T. T.; Hossann, M.; Reinl, H. M.; Peller, M.; Eibl, H.; Reiser, M.; Issels, R. D.; Lindner, L. H. Contrast Media Mol. Imaging 2008, 3, 19−26. (16) Kabalka, G. W.; Davis, M. A.; Moss, T. H.; Buonocore, E.; Hubner, K.; Holmberg, E.; Maruyama, K.; Huang, L. Magn. Reson. Med. 1991, 19, 406−415. (17) Zhang, G. D.; Zhang, R.; Wen, X. X.; Li, L.; Li, C. Biomacromolecules 2008, 9, 36−42. (18) Reynolds, C. H.; Annan, N.; Beshah, K.; Huber, J. H.; Shaber, S. H.; Lenkinski, R. E.; Wortman, J. A. J. Am. Chem. Soc. 2000, 122, 8940−8945. (19) Turner, J. L.; Pan, D. P. J.; Plummer, R.; Chen, Z. Y.; Whittaker, A. K.; Wooley, K. L. Adv. Funct. Mater. 2005, 15, 1248−1254. (20) Gong, P.; Chen, Z. Y.; Chen, Y. Y.; Wang, W.; Wang, X. S.; Hu, A. G. Chem. Commun. 2011, 47, 4240−4242. (21) Kaida, S.; Cabral, H.; Kumagai, M.; Kishimura, A.; Terada, Y.; Sekino, M.; Aoki, I.; Nishiyama, N.; Tani, T.; Kataoka, K. Cancer Res. 2010, 70, 7031−7041. (22) Hu, X. L.; Liu, G. H.; Li, Y.; Wang, X. R.; Liu, S. Y. J. Am. Chem. Soc. 2015, 137, 362−368. (23) Kamaly, N.; Xiao, Z. Y.; Valencia, P. M.; Radovic-Moreno, A. F.; Farokhzad, O. C. Chem. Soc. Rev. 2012, 41, 2971−3010. (24) Nicolas, J.; Mura, S.; Brambilla, D.; Mackiewicz, N.; Couvreur, P. Chem. Soc. Rev. 2013, 42, 1147−1235. (25) Srikar, R.; Upendran, A.; Kannan, R. Wiley Interdiscip. Rev.Nanomed. Nanobiotechnol. 2014, 6, 245−267. (26) Li, Y.; Laurent, S.; Esser, L.; Elst, L. V.; Muller, R. N.; Lowe, A. B.; Boyer, C.; Davis, T. P. Polym. Chem. 2014, 5, 2592−2601. I

DOI: 10.1021/acs.biomac.6b01256 Biomacromolecules XXXX, XXX, XXX−XXX