Dual Nanoparticle Labeling Enables Real-Time Photoacoustic Imaging and Magnetic Retention of Mesenchymal Stromal Cells

Introduction: : Stem-cell therapy (SCT) offers a compelling approach to tissue repair and functional restoration, yet its efficacy is curtailed by inefficient delivery and rapid loss of transplanted cells. Across diverse injury models, acute ( < 24 h) retention rarely exceeds 10 %, whether cells are delivered intravenously, intra-arterially, or by direct injection. Magnetically assisted targeting where cells are loaded with iron-oxide nanoparticles (IONPs) and directed with an external field has been shown to enhance engraftment. However, this approach lacks dynamic feedback or real-time guidance. Evaluation of efficacy has primarily relied on end-point analysis methods, such as histology, which fail to reflect the real-time dynamics of the tissue region. Magnetic resonance imaging (MRI) has been commonly used for tracking transplanted cells, but it does not allow for image-guided delivery or real-time monitoring. Additionally, the strong magnetic fields can interfere with maintaining magnetically-targeted stem cells at the treatment site. Consequently, a non-invasive imaging modality that can monitor cell trafficking at depth without perturbing magnetic capture remains an unmet requirement.

Here we decouple targeting from visualization by combining magnetic retention with ultrasound/photoacoustic (US/PA) imaging. Mesenchymal stromal cells are dual-labelled with (i) PEGylated silica-coated IONPs, conferring robust magnetic responsiveness, and (ii) miniaturized, PEGylated silica-coated gold nanorods (AuNRs), providing US/PA image-guidance and monitoring (Figure 1A). US/PA imaging simultaneously provides sub-millimeter anatomical resolution and molecular imaging capabilities without ionizing radiation or interference from external magnetic fields. Utilizing our approach, we demonstrate image-guided magnetic retention and sustained localization of dual-labeled cells in vivo.

Materials and

Methods: : Iron-oxide nanoparticles (IONPs) were produced by a polyol reduction of FeCl3 in ethylene glycol, then sequentially coated with mesoporous silica and m-PEG-silane (MW = 2000). Miniature gold nanorods (AuNRs, λmax ~ 850 nm) were synthesized via seed-mediated CTAB reduction and similarly coated with mesoporous silica and PEG. UV-vis–NIR spectroscopy and transmission electron microscopy (TEM) confirmed morphology and optical properties. Mesenchymal stromal cells (MSCs) were dual-labelled with IONPs (2.5 OD) and AuNRs (1 OD), harvested, and assessed for viability (MTT assay), trilineage differentiation, and surface markers (CD14, CD31, CD44, CD90). Uptake and intracellular localization were visualized by bright-/dark-field microscopy, Prussian-blue/H&E staining, and cell TEM. Magnetic retention of single- and dual-labeled cells was tested in flow-tube constructs (250-10000 µL/min) and tissue-mimicking gelatin (2.5–10 % w/v) dome phantoms using a cube magnet (NdFeB, N52). Spectroscopic PA imaging (Vevo LAZR, LZ250; 680–970 nm, 2 nm steps) was performed using cell-laden gelatin‐dome phantoms to characterize the PA signal and limit-of-detection (500 cells/µL, 250 … 16, 8 cells/µL) of dual-labeled MSCs. Single-labelled (IONP-only or AuNR-only) MSCs, and unlabeled MSCs were used as control. For in vivo validation, female Nu/J mice (10–14 weeks) received intravenous dual-labeled MSCs with a hind-limb magnet for magnetic capture studies followed by 4 h magnet applications. 3D US/PA scans were acquired on days 0, 2, 4, 6 (LZ550). All data was exported, processed, and analyzed using VevoLAB 5.7.0 and ImageJ (1.53n).

Results, Conclusions, and Discussions:: AuNRs and IONPs were synthesized, silica-coated, PEGylated, and characterized. TEM verified monodisperse IONPs and AuNRs with silica-coating (Figure 1B). IONPs displayed negligible optical extinction in the NIR-I window, whereas AuNRs exhibited a sharp plasmon peak. NMR relaxometry confirmed strong magnetic responsiveness (r₂ = 138 mM-1s-1) for IONPs and minimal responsiveness for AuNRs. Both particles showed no change in PA signal after 6000 laser pulses (20 mJ/cm2), proving robust photostability.

Tissue-mimicking gelatin-dome phantoms (500 cells/µL) demonstrated that unlabeled MSCs and IONP-labeled MSCs (2.5 or 5.0 OD) generated negligible PA signals, whereas AuNR-labeled MSCs produced robust PA signals (Figure 1C). Spectroscopic PA imaging (680–970 nm) of the domes in 1C reproduced the expected spectra with no measurable contribution from IONPs (Figure 1D). Bright-field (Eosin and Prussian Blue staining) and dark-field microscopy corroborated dense intracellular uptake of both particles colocalized within the cytoplasm (Figure 1E). Dual-labeled cells exhibited comparable expression of mesenchymal surface markers (CD44⁺/CD90⁺/CD14⁻/CD31⁻) and trilineage differentiation indistinguishably with naïve stem cells. The dual-labeled cells exhibited cell density-dependent PA signals and were detected even at a cell density of 9 cells/µL (Figure 1F). Under controlled flow speeds that replicate venous flow rates (250-10,000 µL/min), dual-labeled cells displayed excellent magnetic capture efficiencies (~70% at 1,000 µL/min) (Figures 1G, 1H). AuNR-only and unlabeled controls showed no magnetic capture.

TEM of dual-labeled cells showed particle-rich endosomes evenly distributed in the cytoplasm; brief magnet exposure re-localized these vesicles toward the field source without nuclear penetration or membrane damage (Figure 1I). No significant differences in mitochondrial, nuclear, endosomal, or cell membrane morphology were observed across unlabeled and dual-labeled cells. Nu/J mice received dual-labeled MSCs intravenously (5x105 cells in PBS) while a magnet (NdFeB, N52) was fixed to the right hindlimb. The left hindlimb was used as contralateral control. Magnets were removed after four hours and 3D US/PA images acquired immediately post-magnet-removal showed significantly greater PA signal in the experimental right hindlimb indicating greater magnetic capture with localization confined to the magnet position (Figure 1J). Our approach promises to advance clinical stem cell therapy by enhancing cell retention and enabling precise, noninvasive tracking within the body.

Acknowledgements and/or References (Optional)::