Detection of Single Particles and Single Cells by MRI

E. M. Shapiro, S. Skrtic, J. M. Hill, C. E. Dunbar, A. P. Koretsky National Institutes of Health, Bethesda, MD, United States Synopsis Labeling cells with iron oxide is a useful tool for MRI based molecular and cellular imaging. Here we demonstrate, in vitro, that single, ~micron size superparamagnetic iron oxide particles can be robustly visualized with resolutions of between 50-200 microns. The enhancement effect of the particles is as high as several hundred microns and at high resolution the dipolar field of a single particle is clearly visible. Furthermore, we demonstrate in isolated cells, that individual cells with only single iron oxide particles can be readily detected. Due to this, these particles may be useful under conditions where labeling efficiencies are low. Introduction Current methods for molecular and cellular imaging involve introducing paramagnetic species into cells and tracking them using MRI. These include (but are not limited to) small inorganic chelates[1,2], nanometer size paramagnetic particles [3] or magnetodendrimers [4]. For cell labeling, much work has gone into developing strategies for increasing the efficiency of labeling with ultra-small paramagnetic iron oxide particles (USPIO) [5]. Some limitations of this method are the dilution of the signal when cells migrate and divide, the large concentration of contrast agent necessary to produce robust contrast, and efficient ways for intracellular translocation of the agent. Large, micron size, encapsulated, fluorescent, superparamagnetic iron oxide particles are efficiently endocytosed by many cell types [6]. The large size of these particles makes them effective T2 * agents and offers the unique ability to visualize single particles and single cells harboring just a single particle. Methods Six samples of different iron concentrations were made for four different size encapsulated superparamagnetic iron oxide particles. The concentrations were 1.0, 0.5., 0.1, 0.01, 0.001 and 0.0001 mM iron. The four samples were Feridex, an FDA approved iron oxide nanoparticle (iron core 5-10 nm), and three large iron oxide nanoparticles (0.7, 0.96 and 1.63 micron) ranging from 27-45% iron (Bangs Laboratories, IN). These ‘Bangs’ particles are magnetite cores encapsulated with styrene/divinyl benzene with dragon green fluorescent dye (480,520) soaked in. Samples were made as 4% agarose gels in microtubes. The microtubes were immersed in agarose as a six sample set of a single size particle. All MRI was performed on an 11.7 T horizontal bore Bruker Avance imaging system using a 35 mm birdcage coil. Fast gradient echo images were obatined on each set of six samples at 4 different resolutions (400, 200, 100 and 50 micron isotropic) and at 2 or 4 echo times (5, 10, 15 and 20 ms). Several cell types were harvested and/or cultured using standard methods. Cells were incubated with various amounts of the different size particles for various lengths of time. After rigorous washing, cells were plated on collagen coated cover slips at various densities. Imaging was then performed identical to the agarose samples. Confocal and light microscopy was used to detect the dragon green fluroecence from the particles in the cells to confirm the number of particles in the cells. Results Figure 1 shows the MRI obtained at 50 micron resolution with a TE of 10 msec on six concentrations of the 1.63 micron particles. As the number of particles decreases, the overall intensity increases, however small regions of low intensity can be detected which were attributed to single particles. The larger particles (1.63 microns) yielded higher susceptibility based contrast than the smaller particles (0.7 microns), with higher concentrations producing the greatest contrast within a set. The calculated number of particles per slice in the bottom three tubes is ~ 500, ~ 50 and ~ 5, from right to left. The number of dark spots in the images follow this general trend. With the largest particles, single particles could be detected in the more dilute samples with resolutions up to 200 microns at an echo time of ten ms. Single, 0.7 micron particles could be detected only at 100 micron resolution and lower. Figure 2a shows an MRI from very sparsely labeled hepatocytes. Dark regions in the MRI were attributed to cells labeled with the 0.96 micron particles. Figure 2b shows the white box in Figure 2a expanded. At 50 micron resolution, the dipolar field from each single particle can clearly be observed as a “barbell shaped” artifact. The total artifact volume of a single particle is several hundred microns. Confocal microscopy of the green fluorescence confirmed the presence of single particles within cells (Figure 2c). Discussion We have shown that single, micron size particles can be visualized using gradient echo based MRI. This allowed the detection of single cells harboring just a single particle. Advantages of using these particles are conditions where it may be difficult to get large labeling efficiency. Additionally, cell division can not dilute the contrast below a minimum observable threshold because only single particles are required to produce an effect on the order of a few hundred microns. Previously it has been demonstrated that single cells can be imaged when efficiently labeled with USPIO at 50 micron resolution [7]. Achieving 50 micron resolution in vivo is possible but requires long scan times. The large susceptibility artifact these micron sized particles produce allows single cell detection at up to 200 micron resolution. This should enhance the ability of MRI for stem cell tracking, cell fate monitoring and tumor and plaque labeling even on clinically relevant MRI scanners and clinical time scales. Strategies for in vivo targeting of these particles to specific cell types are under development. References [1] Jacobs, R.E. and Fraser, S.E., Science, 263:681-684 (1994). [4] Bulte, J.W.M., et al, Nature Biotech, 19:1141-1147 (2001). [2] Louie, A.Y., et al, Nature Biotech, 18:321-325 (2000). [5] Lewin, M., et al, Nature Biotech, 18:410-414 (2000). [3] Shen, T., et al, MRM, 29:599-604 (1993). [6] Hinds, A., et al, manuscript submitted (2002). [7] Dodd, S.J., Biophys J, 76:103-109 (1999).