Quantitative Tissue Spectroscopy of Near Infrared Fluorescent Nanosensor Implants

Implantable, near infrared (nIR) fluorescent nanosensors are advantageous for in vivo monitoring of biological analytes since they can be rendered selective for a particular target molecule while utilizing their unique optical properties and the nIR tissue transparency window for information transfer without an internal power source or telemetry. However, basic questions remain regarding the optimal encapsulation platform, geometrical properties, and concentration ranges required for high signal to noise ratio and effective detection through biological tissue. In this work, we systematically explore these variables quantitatively to optimize the performance of such optical nanosensors for biomedical applications. We investigate both alginate and polyethylene glycol (PEG) as model hydrogel systems, encapsulating d(GT)15 ssDNA-wrapped single-walled carbon nanotubes (SWNT) as model fluorescent nanoparticle sensors, responsive to riboflavin. Hydrogel sensors implanted 0.5 mm into thick tissue samples exhibit 50% reduction of initial fluorescence intensity, allowing an optical detection limit of 5.4 mm and 5.1 mm depth in tissue for alginate and PEG gels, respectively, at a SWNT concentration of 10 mg L(-1), and 785 nm laser excitation of 80 mW and 30 s exposure. These findings are supported with in vivo nIR fluorescent imaging of SWNT hydrogels implanted subcutaneously in mice. For the case of SWNT, we find that the alginate system is preferable in terms of emission intensity, sensor response, rheological properties, and shelf life.