M 13-templated magnetic nanoparticles for targeted in vivo imaging of prostate cancer

D. Ghosh,Youjin V. Lee,Stéphanie Thomas,Aditya G. Kohli,Dong Soo Yun,A. Belcher,K. Kelly

Abstract:Molecular imaging allows clinicians to visualize the progression of tumours and obtain relevant information for patient diagnosis and treatment1. Owing to their intrinsic optical, electrical and magnetic properties, nanoparticles are promising contrast agents for imaging dynamic molecular and cellular processes such as protein-protein interactions, enzyme activity or gene expression2. Until now, nanoparticles have been engineered with targeting ligands such as antibodies and peptides to improve tumour specificity and uptake. However, excessive loading of ligands can reduce the targeting capabilities of the ligand3,4,5 and reduce the ability of the nanoparticle to bind to a finite number of receptors on cells6. Increasing the number of nanoparticles delivered to cells by each targeting molecule would lead to higher signal-to-noise ratios and improve image contrast. Here, we show that M13 filamentous bacteriophage can be used as a scaffold to display targeting ligands and multiple nanoparticles for magnetic resonance imaging of cancer cells and tumours in mice. Monodisperse iron oxide magnetic nanoparticles assemble along the M13 coat, and its distal end is engineered to display a peptide that targets SPARC glycoprotein, which is overexpressed in various cancers. Compared with nanoparticles that are directly functionalized with targeting peptides, our approach improves contrast because each SPARC-targeting molecule Authors for correspondence: Kimberly A. Kelly Department of Biomedical Engineering and the Robert M. Berne Cardiovascular Research Center University of Virginia PO Box 800759 Charlottesville VA, 22904 Kak3x@virginia.edu; Angela M. Belcher Departments of Materials Science and Engineering and Biological Engineering Koch Institute for Integrative Cancer Research Massachusetts Institute of Technology 77 Massachusetts Ave., 76-561 Cambridge, MA 02139 belcher@mit.edu. Author Contributions D.G., K.A.K, and A.M.B. conceived and designed the experiments. D.G., Y.L., and S.T. performed experiments. D.G., Y.L., S.T., and A.G.K. contributed new reagents. D.S.Y. conducted high-resolution transmission electron microscopy. D.G. and K.A.K. analysed data. D.G., K.A.K., and A.M.B. co-wrote the paper. Additional Information Supplementary information is available in the online version of the paper. Reprints and permission information is available online at http://www.nature.com/reprints. Competing Financial Interests The authors declare no competing financial interests. NIH Public Access Author Manuscript Nat Nanotechnol. Author manuscript; available in PMC 2014 June 16. Published in final edited form as: Nat Nanotechnol. 2012 October ; 7(10): 677–682. doi:10.1038/nnano.2012.146. N IH -P A A uhor M anscript N IH -P A A uhor M anscript N IH -P A A uhor M anscript delivers a large number of nanoparticles into the cells. Moreover, the targeting ligand and nanoparticles could be easily exchanged for others, making this platform attractive for in vivo high-throughput screening and molecular detection. M13 is a filamentous virus with five genetically modifiable coat proteins (p3, p6, p7, p8, and p9) for peptide or protein display. Previously, phage display has been exploited for epitope discovery7, gene delivery8, antibody delivery9, and in vitro and in vivo ligand discovery10,11. Although phage display allows rapid, amplifiable screening of lead molecules for targeting, the genetic and physical properties of M13 also make it an excellent biological building block for bottom-up fabrication and assembly of materials12. The highly ordered structure of the 2700 copies of the major coat protein p8 and the filamentous shape of M13 (~6.5 nm in diameter and ~880 nm in length) allows the phage to multivalently interact with nanomaterials. Our lab has engineered M13 for display of material-specific peptides on p3 and p8 as a template to grow and nucleate various inorganic materials for applications including high power Li ion batteries and photocatalytic water splitting13,14. Here, we extend this work and report the development of M13 bacteriophage as a vector incorporating a tumour-targeting peptide15 and a peptide motif for templated assembly of magnetic iron oxide nanoparticles (MNPs) towards nanoparticle-mediated, dark contrastbased magnetic resonance imaging for specific tumour detection (Fig. 1a). We investigate selective tumour targeting against prostate cancer xenograft models expressing different levels of secreted protein, acidic and rich in cysteine (SPARC), which is upregulated in various cancers (reviewed in 16) and correlated with poor prognosis16. The capsid organization of M13 spatially separates targeting and imaging moieties, circumventing any potential problems with ligand or probe functionality. Using p3 to display targeting ligands while assembling multiple MNPs along the p8 capsid (Fig.1a) achieves targeting and delivers a larger payload of MNPs per SPARC compared with directly functionalized nanoparticles. Previously, we engineered multiple glutamic acid residues on the N-terminus of the major coat capsid protein p8 on M13 for assembly of various metal oxides for anode battery and layer-by-layer polymer assembly applications17. At physiological pH, the triglutamate peptide-modified M13 has a net negative charge, allowing for electrostatic assembly with positively charged moieties, such as positively charged nanoparticles17. Further, to create a molecularly targeted M13, a SPARC-binding peptide (SPPTGIN, designated as SBP) identified using phage display15 was genetically engineered into the p3 minor coat protein of triglutamate modified M13 (Fig. 1a); in the presence of MNPs, the resulting phage is designated as M13-SBP-MNP. We then synthesized crystalline, monodisperse iron oxide nanoparticles to assemble along the modified M13 (Fig. 2 and Supplementary Note 1). We confirmed the monodispersity by transmission electron microscopy (TEM) and dynamic light scattering (DLS, Fig. 2). From high-resolution TEM, the particles demonstrated distinct lattice fringe patterns, suggesting a highly crystalline structure (Fig. 2, top row, left column, inset). X-ray diffraction confirmed the cubic spinel structure of the iron oxide nanoparticles, whereas a d-spacing comparison corroborated previous reports that iron oxide particles are either maghemite (γ-Fe2O3) or Ghosh et al. Page 2 Nat Nanotechnol. Author manuscript; available in PMC 2014 June 16. N IH -P A A uhor M anscript N IH -P A A uhor M anscript N IH -P A A uhor M anscript magnetite (Fe3O4) (Supplementary Fig. S1 and Table S1). These particles were watersolubilized with a PEG-lipid formulation (Supplementary Note 1) and were stable in water for at least two months (Supplementary Fig. S2). We characterized the assembly of the nanoparticles along the coat of M13. MNPs were mixed with M13-expressing triglutamate motif and allowed to assemble along the length of the virus. TEM confirmed multivalent MNP assembly along the virus (Fig. 2, top row, right column). Previous work has shown M13-templated materials with and without engineered p3 ligands have similar materials properties13. MNPs are aligned in a filamentous orientation along the p8 coat with a length of approximately ~ 800 nm, which is the approximate length of M13. The number of MNPs per M13 was quantified by iron content using inductive coupled plasma atomic emission spectroscopy (ICP-AES) (Supplementary Note 2). Using iron standards, we measured the emission intensities and made a calibration curve (Supplementary Fig. S3) to interpolate iron content of M13-templated MNPs from measured intensities and calculate the number of MNPs per virus. There are approximately 26 particles per M13, confirming multivalent assembly as observed by TEM. From dynamic light scattering, M13 and M13-assembled MNPs are monodisperse, with a radius of ~147.3 nm (Supplementary Fig. S4) and ~181.4 nm (Fig. 2, bottom row, right column), respectively. Importantly, M13-assembled MNPs have a single histogram population and lack multiple peaks, suggesting the nanoparticles stably complex with the virus and do not aggregate into multiple populations. The increase in size of M13-MNP compared with M13 can be attributed to multiple MNPs assembling on the 2700 p8 copies. M13-MNP is stable in other biological solutions including PBS (Supplementary Note 3 and Supplementary Fig. S5), culture media and serum for at least three weeks (Fig. 2) and is not cytotoxic by standard WST-1 cell proliferation assay (Supplementary Fig. S6). Virus-nanoparticle complexes are stable and non-toxic in physiological conditions, which makes them attractive for in vivo imaging. To investigate the magnetic resonance (MR) contrast potential of MNPs and M13-MNPs, we measured the transverse relaxation (T2). Proton relaxivity is a rate of proton relaxation in response to iron atom concentration. Paramagnetic iron oxide nanoparticles cause local inhomogenities in the magnetic field resulting in decoherence of the water proton precession of the spins. The resulting relaxation of the spins until the magnetization decays to zero causes a negative, or ‘dark’ contrast in the background of the light MR image. We measured the T2 of MNPs and M13-MNPs and determined relaxivity based on the iron concentration of the MNPs (Table 1 and Supplementary Fig. S7). MNPs and M13-MNP had comparable relaxivities of 65.1 mM−1 s−1 and 58.7 mM−1 s−1, respectively. Their relaxivities are an improvement over those of several superparamagnetic iron oxide particles including ferumoxtran and monocrystalline iron oxide nanoparticles (Table 1), which exhibit relaxivities of 35.3 and 34.8 mM−1 s−1, respectively19,20. These clinically used particles possess lower R2 relaxivities compared with our crystalline particles since they are synthesized in aqueous media and thus have poorer crystallinity21. Ferumoxides including Feridex (Table 1) demonstrate higher relaxivities (107 mM−1 s−1) than synthesized MNPs and M1