In vivo ultrasound and photoacoustic monitoring of mesenchymal stem cells labeled with gold nanotracers

Abstract

Longitudinal monitoring of cells is required in order to understand the role of delivered stem cells in therapeutic neovascularization. However, there is not an imaging technique that is capable of quantitative, longitudinal assessment of stem cell behaviors with high spatial resolution and sufficient penetration depth. In this study, in vivo and in vitro experiments were performed to demonstrate the efficacy of ultrasound-guided photoacoustic (US/PA) imaging to monitor mesenchymal stem cells (MSCs) labeled with gold nanotracers (Au NTs). The Au NT labeled MSCs, injected intramuscularly in the lower limb of the Lewis rat, were detected and spatially resolved. Furthermore, our quantitative in vitro cell studies indicate that US/PA imaging is capable of high detection sensitivity (1×10⁴ cells/mL) of the Au NT labeled MSCs. Finally, Au NT labeled MSCs captured in the PEGylated fibrin gel system were imaged in vivo, as well as in vitro, over a one week time period, suggesting that longitudinal cell tracking using US/PA imaging is possible. Overall, Au NT labeling of MSCs and US/PA imaging can be an alternative approach in stem cell imaging capable of noninvasive, sensitive, quantitative, longitudinal assessment of stem cell behaviors with high spatial and temporal resolutions at sufficient depths.

Figure 1. Diagram of the procedure for monitoring mesenchymal stem cells (MSCs) in vivo.

Once MSCs are loaded with nanotracers, the labeled MSCs are entrapped in the PEGylated fibrin gels and implanted at the ischemic region. The PEGylated fibrin gels promote MSC differentiation toward a vascular cell type, thus contributing to regeneration. Both MSC distribution and neovascularization can be monitored using the combined ultrasound and photoacoustic imaging of cells loaded with nanotracer such as gold plasmonic nanoparticles.

Figure 2. Gold nanotracer (Au NT) labeling of MSCs.

(A) The TEM image of 20 nm gold nanotracers and the dark field images (20× magnification) of MSCs without and with nanotracer loading (B and C, respectively). The normalized absorbance spectra of Au NTs, MSCs and the Au NT labeled MSCs (D,E, and F, respectively).

Figure 3. In vivo monitoring of Au NT labeled MSCs using combined ultrasound and photoacoustic (US/PA) imaging.

(A–D) In vivo ultrasound, photoacoustic, US/PA, and US/spectroscopic images of the LGAS in which PEGylated fibrin gel containing Au NT loaded MSCs (1×10⁵ cells/mL) was injected. PEGylated fibrin gel location is outlined with yellow dotted circle. Injection depth was about 5 mm under the skin. (E–H) Control at the region of the LGAS of the other hind limb without any injection. Photoacoustic images were acquired at the wavelength of 760 nm with a fluence of 11 mJ/cm². Spectral (650–920 nm) analysis of photoacoustic signal was able to differentiate skin (shown in yellow), oxygenated (red) and deoxygenated (blue) blood, and Au NT loaded MSCs (green). The images measure 23 mm laterally and 12.5 mm axially.

Figure 4. Quantification of Au NT labeled MSCs.

(A) Ultrasound (top), photoacoustic (middle), and US/PA images (bottom) of the gelatin phantom with inclusions containing different concentrations of MSCs. Photoacoustic images were obtained at a wavelength of 750 nm with a fluence of 5.1 mJ/cm². All images measure 98 mm laterally and 7.7 mm axially. (B) The quantitative analysis of photoacoustic signal. The mean and the standard deviation of the photoacoustic signal amplitude as a function of the nanoparticle concentration is shown in the outer graph in a semi-logarithmic scale. The inset graph presents the linear regression fit (with an R² value of 0.984) of the mean values of the photoacoustic signal amplitude as a function of the nanoparticle concentration in a linear scale.

Figure 5. Longitudinal in vitro photoacoustic imaging of MSCs labeled with Au NTs.

(A) Longitudinal photoacoustic images of MSCs loaded with Au NTs, or MSCs only, in a PEGylated fibrin gel at a wavelength of 750 nm with a fluence of 11 mJ/cm². The MSCs in the PEGylated fibrin gel were cultured in a 24 well plate over a one week time period. While the MSCs without Au NT loading did not produce any photoacoustic signal, strong photoacoustic signals were detected from the MSCs loaded with Au NTs over a one week time period. The images measure 14.1 mm laterally and 16.6 mm axially. (B) The quantitative analysis of photoacoustic signal. The mean and the standard deviation of the photoacoustic signal amplitude from the PEGylated fibrin gels containing the MSCs with and without Au NT loading were shown in green and yellow colors, respectively.

Figure 6. Longitudinal in vivo monitoring spatial distribution of MSCs labeled with Au NTs using US/PS imaging.

(A) 3-D combined ultrasound and spectroscopic images of the rat hind limb in which the PEGylated fibrin gel containing Au NT labeled MSCs were injected (day 3,7, and 10). The MSCs labeled with Au NTs were distinguished using spectral analysis and presented in green color. The bounding box for each image measures 23 mm laterally, 12.5 mm axially, and 25 mm elevationally. (B) The quantitative analysis of photoacoustic signal. The photoacoustic signals at three different time points obtained from the Au NT labeled MSCs and the background tissue were summed and displayed in green and yellow colors, respectively.