Nanoparticle-based autoantigen delivery to Treg-inducing liver sinusoidal endothelial cells enables control of autoimmunity in mice.

BACKGROUND & AIMS: It is well-known that the liver can induce immune tolerance, yet this knowledge could, thus far, not be translated into effective treatments for autoimmune diseases. We have previously shown that liver sinusoidal endothelial cells (LSECs) could substantially contribute to hepatic tolerance through their ability to induce CD4+ Foxp3+ regulatory T cells (Tregs). Here, we explored whether the Treg-inducing potential of LSECs could be harnessed for the treatment of autoimmune disease. METHODS: We engineered a polymeric nanoparticle (NP) carrier for the selective delivery of autoantigen peptides to LSECs in vivo. In the well-characterized autoimmune disease model of experimental autoimmune encephalomyelitis (EAE), we investigated whether administration of LSEC-targeting autoantigen peptide-loaded NPs could protect mice from autoimmune disease. RESULTS: We demonstrate that NP-based autoantigen delivery to LSECs could completely and permanently prevent the onset of clinical EAE. More importantly, in a therapeutic approach, mice with already established EAE improved rapidly and substantially following administration of a single dose of autoantigen peptide-loaded NPs, whereas the control group deteriorated. Treatment efficacy seemed to depend on Tregs. The Treg frequencies in the spleens of mice treated with autoantigen peptide-loaded NPs were significantly higher than those in vehicle-treated mice. Moreover, NP-mediated disease control was abrogated after Treg depletion by repeated administration of Treg-depleting antibody. CONCLUSION: Our findings provide proof of principle that the selective delivery of autoantigen peptides to LSECs by NPs can induce antigen-specific Tregs and enable effective treatment of autoimmune disease. These findings highlight the importance of Treg induction by LSECs for immune tolerance.

A highly effective, nontoxic T1 MR contrast agent based on ultrasmall PEGylated iron oxide nanoparticles.

In this study we systematically developed a potential MR T(1) contrast agent based on very small PEGylated iron oxide nanoparticles. We adjusted the size of the crystalline core providing suitable relaxometric properties. In addition, a dense and optimized PEG coating provides high stability under physiological conditions together with low cytotoxicity and low nonspecific phagocytosis into macrophage cells as a part of the reticulo endothelial system at biologically relevant concentrations. The as developed contrast agent has the lowest r(2)/r(1) ratio (2.4) at 1.41 T reported so far for PEGylated iron oxide nanoparticles as well as a r(1) relaxivity (7.3 mM(-1) s(-1)) that is two times higher compared to that of Magnevist as a typical T(1) contrast agent based on gadolinium as a clinical standard.

The fate of a designed protein corona on nanoparticles in vitro and in vivo.

A variety of monodisperse superparamagnetic iron oxide particles (SPIOs) was designed in which the surface was modified by PEGylation with mono- or bifunctional poly(ethylene oxide)amines (PEG). Using (125)I-labeled test proteins (transferrin, albumin), the binding and exchange of corona proteins was studied first in vitro. Incubation with (125)I-transferrin showed that with increasing grade of PEGylation the binding was substantially diminished without a difference between simply adsorbed and covalently bound protein. However, after incubation with excess albumin and subsequently whole plasma, transferrin from the preformed transferrin corona was more and more lost from SPIOs in the case of adsorbed proteins. If non-labeled transferrin was used as preformed corona and excess (125)I-labeled albumin was added to the reaction mixtures with different SPIOs, a substantial amount of label was bound to the particles with initially adsorbed transferrin but little or even zero with covalently bound transferrin. These in vitro experiments show a clear difference in the stability of a preformed hard corona with adsorbed or covalently bound protein. This difference seems, however, to be of minor importance in vivo when polymer-coated (59)Fe-SPIOs with adsorbed or covalently bound (125)I-labeled mouse transferrin were injected intravenously in mice. With both protein coronae the (59)Fe/(125)I-labelled particles were cleared from the blood stream within 30 min and appeared in the liver and spleen to a large extent (>90%). In addition, after 2 h already half of the (125)I-labeled transferrin from both nanodevices was recycled back into the plasma and into tissue. This study confirms that adsorbed transferrin from a preformed protein corona is efficiently taken up by cells. It is also highlighted that a radiolabelling technique described in this study may be of value to investigate the role of protein corona formation in vivo for the respective nanoparticle uptake.

Intraperitoneal injection improves the uptake of nanoparticle-labeled high-density lipoprotein to atherosclerotic plaques compared with intravenous injection: a multimodal imaging study in ApoE knockout mice.

BACKGROUND: The aim of this study was to assess whether high-density lipoprotein (HDL) labeled with superparamagnetic iron oxide nanoparticles (SPIOs) and quantum dots was able to detect atherosclerotic lesions in mice after intravenous and intraperitoneal injection by multimodal imaging. METHODS AND RESULTS: Nanoparticle-labeled HDLs (NP-HDLs) were characterized in vitro by dynamic light scattering and size exclusion chromatography with subsequent cholesterol and fluorescence measurements. For biodistribution and blood clearance studies, NP-HDL(SPIOs) radiolabeled with (59)Fe (NP-HDL(59Fe-SPIOs)) were injected intravenously or intraperitoneally into ApoE knockout mice (n=6), and radioactivity was measured using a gamma counter. NP-HDL accumulation within atherosclerotic plaques in vivo and ex vivo was estimated by MRI at 7 Tesla, ex vivo confocal fluorescence microscopy, x-ray fluorescence microscopy, and histological analysis (n=3). Statistical analyses were performed using a 2-tailed Student t-test. In vitro characterization of NP-HDL confirmed properties similar to endogenous HDL. Blood concentration time curves showed a biexponential decrease for the intravenous injection, whereas a slow increase followed by a steady state was noted for intraperitoneal injection. Radioactivity measurements showed predominant accumulation in the liver and spleen after both application approaches. NP-HDL(59Fe-SPIOs) uptake into atherosclerotic plaques increased significantly after intraperitoneal compared with intravenous injection (P<0.01). In vivo MRI showed an increased uptake of NP-HDL into atherosclerotic lesions after intraperitoneal injection, which was confirmed by ex vivo MRI, x-ray fluorescence microscopy, confocal fluorescence microscopy, and histological analysis. CONCLUSIONS: In vivo MRI and ex vivo multimodal imaging of atherosclerotic plaque using NP-HDL is feasible, and intraperitoneal application improves the uptake within vessel wall lesions compared with intravenous injection.

A simple and widely applicable method to 59Fe-radiolabel monodisperse superparamagnetic iron oxide nanoparticles for in vivo quantification studies.

A simple, fast, efficient, and widely applicable method to radiolabel the cores of monodisperse superparamagnetic iron oxide nanoparticles (SPIOs) with (59)Fe was developed. These cores can be used as precursors for a variety of functionalized nanodevices. A quality control using filtration techniques, size-exclusion chromatography, chemical degradation methods, transmission electron microscopy, and magnetic resonance imaging showed that the nanoparticles were stably labeled with (59)Fe. Furthermore, the particle structure and the magnetic properties of the SPIOs were unchanged. In a second approach, monodisperse SPIOs stabilized with (14)C-oleic acid were synthesized, and the stability of this shell labeling was studied. In proof of principle experiments, the (59)Fe-SPIOs coated with different shells to make them water-soluble were used to evaluate and compare in vivo pharmacokinetic parameters such as blood half-life. It could also be shown that our radiolabeled SPIOs embedded in recombinant lipoproteins can be used to quantify physiological processes in closer detail than hitherto possible. In vitro and in vivo experiments showed that the (59)Fe label is stable enough to be applied in vivo, whereas the (14)C label is rapidly removed from the iron core and is not adequate for in vivo studies. To obtain meaningful results in in vivo experiments, only (59)Fe-labeled SPIOs should be used.