Counting cells in motion by quantitative real-time magnetic particle imaging

Abstract

Magnetic Particle Imaging (MPI) is an advanced and powerful imaging modality for visualization and quantitative real-time detection of magnetic nanoparticles (MNPs). This opens the possibility of tracking cells in vivo once they have been loaded by MNPs. Imaging modalities such as optical imaging, X-ray computed tomography (CT), positron emission tomography (PET), single photon emission computed tomography (SPECT), and magnetic resonance imaging (MRI) face limitations, from depth of penetration and radiation exposure to resolution and quantification accuracy. MPI addresses these challenges, enabling radiation-free tracking of MNP-loaded cells with precise quantification. However, the real-time tracking of MNP-loaded cells with MPI has not been demonstrated yet. This study establishes real-time quantitative tracking of MNP-loaded cells. Therefore, THP-1 monocytes were loaded with three different MNP systems, including the MPI gold standard Resovist and Synomag. The real-time MPI experiments reveal different MPI resolution behaviors of the three MNP systems after cellular uptake. Real-time quantitative imaging was achieved by time-resolved cell number determination and comparison with the number of inserted cells. About 95% of the inserted cells were successfully tracked in a controlled phantom environment. These results underline the potential of MPI for real-time investigation of cell migration and interaction with tissue in vivo.

Figure 1

Experimental setup for real-time cell tracking using a MPI scanner. MNP-loaded cells were located in a tubing and moved through the FoV (blue dashed line) using a high-pressure multichannel pump (Ismatec). The tube (inner diameter d_i = 1 mm, length l = 1 m.) was attached to an MPI sample holder and inserted into the mouse coil of the MPI scanner.

Figure 2

(a) Box plots of the internalized MNP amounts (iron mass/per cell) of SynP50, SynC30, and RES after an incubation time of 10 min. An iron mass of 28 µg was added to the cells (V = 1000 µL, c_Fe = 0.5 mmol/L). The uncertainty was calculated based on three repeated measurements and corresponds to the standard deviation (coverage factor k = 1). Note that for SynC30 and RES the individual three replicates are so close that they coincide within the symbol thickness. (b) Amplitude spectra A_k of SynC30, SynP50, and RES after cellular uptake (10 min incubation, exposed iron mass of 28 µg) and washing of the MNP-loaded cells. (c) Microscopy images of the cells (scale bar = 20 µm) incubated with SynP50, SynC30, and RES. The MNP and MNP aggregates were stained with Prussian blue, while the cell nucleus was stained with Nuclear Fast Red. Here, the cells were washed once. The first microscope image shows a control image of the cells before cellular uptake.

Figure 3

MPI spatial resolution determined by the SF-based two-voxel analysis using MNP-loaded cells. The separation quality Q_S as a function of distance (number of empty voxels) for SynP50 (yellow squares), SynC30 (blue spheres), and RES (magenta triangles) incubated with 5 $·$ 10⁶ cells is shown and fitted with a Heaviside function (black line) to determine the resolution r that is defined as the voxel distance x at which the separation quality Q is 50% (dashed line).

Figure 4

Images of the experimental setup showcasing the tube phantom filled with MNP-loaded cells positioned on the MPI sample holder (white box). Followed by MPI frames of the real-time flow measurements of cells loaded with (a) SynP50, (b) SynC30, and (c) RES at the time points t = 5 s, 20 s, and 30 s after injection. The dashed line represents the path of the tube.

Figure 5

Quantitative evaluation of real-time flow measurements of MNP-loaded cells with (a) SynP50 (yellow), (b) SynC30 (blue), and (c) RES (magenta). The gray area shows the signal maximum reached by the MNP-loaded cells.

Figure 6

Quantitative MPI image slices during the real-time tracking of (a) SynP50-, (b) SynC30- and (c) RES-loaded cells at the time of maximum cell number in FoV t_max = 24.3 s. The top row shows the x–y plan, where the tube was positioned. For SynP50, and RES the 5th layer was selected, whereas for SynC30 the 6th layer is shown, which is due to slight variations in the height of the tube phantom in the FoV during each experiment. The bottom row displays the z–y plane containing the cross section of the tube to determine the diameter of the MNP-loaded tube.

Figure 7

Signal maximum of the cell bolus for (a) SynP50, (b) SynC30 and (c) RES determined and tracked at each point in time.