Ultrasmall superparamagnetic iron oxide nanoparticles coated with ...

10 downloads 0 Views 6MB Size Report
Aim:We have designed ultrasmall superparamagnetic iron oxide (USPIO) nanoparticles associated with fucoidan (USPOI-FUCO), a natural sulfated ...
Research Article For reprint orders, please contact: [email protected]

Ultrasmall superparamagnetic iron oxide nanoparticles coated with fucoidan for molecular MRI of intraluminal thrombus

Aim: We have designed ultrasmall superparamagnetic iron oxide (USPIO) nanoparticles associated with fucoidan (USPOI-FUCO), a natural sulfated polysaccharide with high affinity for activated platelets, to visualize by MRI arterial thrombi. Materials & methods: USPIOs were prepared and sizes, zeta-potentials and relaxivities were measured. Elastase perfusion in the infrarenal aorta of Wistar rats induced intraluminal thrombus. They were scanned on 4.7 T MRI before and after injection of USPIO-FUCO or USPIO coated with anionic dextran. Results: Surface plasmon resonance evidenced that fucoidan and USPIO-FUCO bind in vitro to immobilized P-selectin. All intraluminal hyposignals detected by MRI after injection of USPIO-FUCO on animals (13 out of 13) were correlated by histology with thrombi, whereas none could be identified with control USPIOs (0 out of 7). No signal was seen in absence of thrombus. Thrombi by MRI were correlated with P-selectin immunostaining and USPIO detection by electron microscopy. Conclusion: In vivo thrombi can thus be evidenced by MRI with USPIO-FUCO. Original submitted 8 July 2013; Revised submitted 4 March 2014

Keywords:  aneurysm • iron oxide nanoparticle • MRI • P-selectin • thrombus

MRI is well suited for in vivo molecular imaging using specifically designed nanosystems [1] . Several nanoparticle agents have been reported for markers of cardiovascular disease, such as fibrin, angiogenesis, inflammation, fibrosis and apoptosis, among others. However, they are not yet applicable for clinical diagnosis [2,3] . In acute coronary syndrome and stroke, athero­sclerotic plaque disruption with superimposed thrombosis is the leading cause of morbidity and mortality worldwide. For example, the intraluminal thrombus (ILT), via its associated biological activities including ILT renewal by both platelet activation [4] , and neutrophil [5] and bacterial [6] trapping, is considered as a driving force in abdominal aortic aneurysm (AAA) evolution [7] . A similar role has been suggested for intraplaque hemorrhages in occlusive atherothrombotic disease [8] . Several authors have visualized

10.2217/NNM.14.51

ILT activities  [9–14] , using 99mTc annexin V in scintigraphy [15] or iron oxide particle trapping by phagocytes in MRI [16,17] , in animal and human AAA. Interestingly, in the model of elastase perfusion-­ induced experimental AAA in rats [18] , biological activities detrimental for the rat arterial wall predominate in the ILT, as observed in human aneurysms [19] . P-selectin is an adhesion molecule, expressed at the surface of endothelial cells and platelets upon activation, which mediates leukocyte rolling and leukocyte trapping within the thrombus [20] . P-selectin expression is involved in the pathophysiology of the renewal and growth of biologically active (‘at risk’) intraluminal thrombi [4] and is a molecular determinant of athero­thrombotic disease [21] . P-selectin has the unique ability to link innate immunity with coagulation in the cardiovascular system, representing

Nanomedicine (Lond.) (2014) 10(1), 73–87

Michimasa Suzuki1, Laure Bachelet-Violette1,2, François Rouzet1,3, Anne Beilvert1,2, Gwennhael Autret4, Murielle Maire1,2, Christine Menager4, Liliane Louedec1, Christine Choqueux1, Pierre Saboural1,2, Oualid Haddad1,2, Cédric Chauvierre1,2, Frédéric Chaubet1,2, JeanBaptiste Michel1, JeanMichel Serfaty‡,1,5 & Didier Letourneur*,‡,1,2 French Institute of Health & Medical Research (Inserm) U1148, Laboratory for Vascular Translational Science, CHU X Bichat, University Paris 7, 46 rue H Huchard, F-75877, France 2 Laboratory for Vascular Translational Science, Cardiovascular Bioengineering, University Paris 13, Villetaneuse, F-93430, France 3 Department of Nuclear Medicine, Assistance Publique Hôpitaux de Paris, CHU X Bichat, University Paris 7, F-75877, France 4 UMR 8234 French National Centre for Scientific Research (CNRS), Phenix, ESPCI, University Paris 6, F-75252, France 5 Department of Radiology, Assistance Publique Hôpitaux de Paris, CHU X Bichat, University Paris 7, F-75877, France *Author for correspondence: Tel.: +33 1 40 25 86 00 Fax: +33 1 40 25 86 02 didier.letourneur@ inserm.fr ‡ These authors equally participated in the supervision of this study. 1

part of

ISSN 1743-5889

73

Research Article  Suzuki, Bachelet-Violette, Rouzet et al. an important molecular target in acute and chronic cardio­vascular diseases [22] . Therefore, numerous efforts have been devoted to the development of contrast agents for imaging of P-selectin [14,23] . These agents were either P-selectin antibodies or synthetic mimics of sialyl Lewis X (SLe X), the natural ligand of P-selectin [24] . Although effective, the cost of synthesis of these agents, including raw materials, and the use of antibodies limit the potential translation to clinical use. Several sulfated polysaccharides have also been described to bind P-selectin. Among them, fucoidan, referring to a type of sulfated and fucosylated poly­ saccharide mainly derived from brown seaweed, is a naturally occurring mimic of SLe X [25,26] . Fucoidan exhibited in vitro a high affinity for immobilized P-selectin, two orders of magnitude greater than other polysaccharides, as well as the lowest nonspecific binding [27] . We recently demonstrated that 99mTc-radio­ labeled fucoidan enabled detection by scintigraphy of vascular thrombi and activated endothelium in rats [28] . In the present study, we propose a novel approach for molecular imaging of thrombus using fucoidan coated on ultrasmall superparamagnetic iron oxide nanoparticles (USPIOs). The ability of this new contrast agent to detect active intravascular thrombi was evaluated here in vivo in an experimental model in rats following elastase-induced vascular injury. Materials & methods Preparation of nanomaterials

A low-molecular-weight fucoidan (Algues et Mer, Ouessant, France) with a molecular weight of 7200 g/mol was used. USPIOs coated with carboxymethyldextran (CMD; USPIO-CMD) and fucoidan (USPIO-FUCO) were prepared as described below. The ferrofluid was synthesized by a classical method [29] and kept in nitric acid ([Fe] = 1.05 M). In a first step, a coating was performed with the acidic CMD (molecular weight: 15,000 g.mol-1, [COOH] = 1.3 mmol.g-1; SigmaAldrich, St Quentin Fallavier, France) as previously described [3] . USPIO-CMD (5 ml at [Fe] = 0.05 M) was treated for 15 min at room temperature with 20 mg of N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide and 3.0 mg of N-hydroxysuccinimide. Fucoidan at 15 mg/ml, previously aminated with diaminopropane at its reducing end [30] , was added to the mixture containing USPIO-CMD and maintained under agitation for 2 h. Purification was performed by dialyzing the suspension against NaCl 1 M (2×) and bidistilled water (5×) before ultrafiltration with a MicroSep® 100 kDa (Pall Life Sciences, VWR France, Fontenay-sousBois, France). Aliquots of 500 μl of USPIO-FUCO ([Fe] = 0.05 M) were prepared in 0.15 M NaCl and maintained at -80°C until use.

74

Nanomedicine (Lond.) (2014) 10(1)

Characterization of nanomaterials

Average diameters and zeta-potentials of USPIOCMD and USPIO-FUCO were determined with a Zetasizer Nano ZS (Malvern Instruments, Orsay, France). The mean diameter of USPIOs and surface compositions were also assessed by transmission electron microscopy (TEM) carried out with a Jeol 2011 microscope and energy filtered transmission electron microscopy (EFTEM; Jeol, Croissy-sur-Seine, France). Relaxation measurements of USPIO-CMD and USPIO-FUCO were performed with different concentrations of compounds in pure water at 37°C with relaxometers at 20 MHz (0.47 T; Minispec PC-120; Bruker, Karlsruhe, Germany) and at 60 MHz (1.42 T; Minispec mq-60; Bruker). Blood clearance of USPIO-FUCO was obtained after labeling [28] of fucoidan with 99mTc. Surface plasmon resonance

Binding interaction between P-selectin and fucoidan, USPIO-CMD or USPIO-FUCO was studied using surface plasmon resonance (SPR) with a Biacore™ X100 (GE Healthcare, Velizy-Villacoublay, France). The amine coupling kit, running buffer, sodium acetate buffer and CM4 sensor chip were supplied by GE Healthcare. The recombinant human P-selectin/ Fc chimera (146–160 kDa by SDS-PAGE) and antiFc antibodies were obtained from R&D Systems (Lille, France). P-selectin was immobilized through its Fc fragment on CM4 sensorchips. The surfaces of the flow cells were activated with a 1:1 mixture of 0.1 M N-hydroxysuccinimide and 0.4 M N-(3-dimethylaminopropyl)-N′ethylcarbodiimide for 420 s at a flow rate of 10 μl/min. Anti-Fc antibody dissolved in 10 mM sodium acetate (pH: 5.0) at a concentration of 25 μg/ml was injected on flow cell for 420 s at a flow rate of 5 μl/min. The surface of flow cells were blocked with a 420-s injection of 1 M ethanolamine (pH: 8.5) at 10 μl/min. P-selectin at a concentration of 25 μg/ml was then injected for 120 s at a flow rate of 5 μl/min. Kinetic analysis was performed using single-cycle kinetics, in which a concentration series of fucoidan (1.1–3.3–11–33–100 nM in the running buffer) is injected in a single cycle without regenerating the surface between injections. Fucoidan was injected at a flow rate of 30 μl/min and at 25°C. The association was monitored for 2 min and the final dissociation for 10 min. The flow cells were regenerated with 3 M MgCl 2 for 30 s. This procedure was repeated six times, and a buffer blank was flowed over the P-selectin and the reference surface (background sensorgram). The measurement sensorgam was subtracted from the reference to correct for nonspecific

future science group

Fucoidan-coated USPIOs for molecular MRI of intraluminal thrombus 

binding. The background signal was subtracted from the sensorgrams to correct for differences caused by the sample injection process. Biacore Evaluation software was used for kinetic evaluation, and sensograms were fitted using the 1:1 binding model (Langmuir binding model) to obtain an affinity constant. A similar protocol was used for USPIO-CMD and USPIO-FUCO (both at 0.1 and 1 μM concentrations) to evidence P-selectin binding. It is worth noting that no affinity constants could be measured for species of undefined molecular weights such as nanoparticles. Experimental model of AAA in rats

The procedure and the animal care complied with the ‘Principles of animal care’ formulated by the EU (Animal Facility Agreement 75-18-03; 2005), and animal experimentation was performed under the authorization 75-101 of the French Ministry of Agriculture after approval by the University ethical commitee. Under intraperitoneal pentobarbital anesthesia (4 mg/100 g bodyweight; Ceva Santé Animale, Paris, France), approximately 20 mm of the infrarenal aorta (beginning 2 mm below the left renal artery) was separated from the vena cava. Collateral arteries were dissected from surrounding connective tissue, ligated in two places and cut between them. The aorta was clamped and four units of porcine pancreatic elastase (E-1250; Sigma-Aldrich) in 550 μl NaCl 9% was perfused transmurally during 1 h, using an automatic pressure perfusion pump. The segment was then rinsed, the catheter removed, the entry hole closed by suture and flow was re-established [19] . A total of 26 male 7-week-old Wistar rats from CERJ (Le Genest, France) were allotted to the elastase-perfused group, and four were sham operated for biodistribution assay and two for sequential injection of USPIO-CMD followed with USPIO-FUCO. To localize the treated segment during the MRI session, the distances between the upper and lower points of the perfused segment and the origin of the left renal artery were measured with the microscope eyepiece, and surgical wounds were closed. This model is characterized by the presence of an ILT 1–2 weeks after aneurysm induction [19] . Four rats died before MRI sessions. Causes of death included rupture of abdominal aorta (n = 2), hindlimb paresis (n = 1) and unknown (n = 1). Blood clearance measurements

The biodistribution of USPIO-FUCO in vivo was assessed by radioactive counting of iterative blood samples. Fucoidan on USPIOs was radiolabeled according to a reduction reaction of pertechnetate [28] . Sodium

future science group

Research Article

pertechnetate (99mTcO4 -, approximately 740 MBq in