Background/Purpose: Current medical imaging technology detects structural rather than functional manifestations of disease.  Imaging agents designed to enhance signal based on molecular mechanisms might permit earlier diagnosis and personalized treatment. We examined whether optical contrast agents whose fluorescence properties may be switched from OFF to ON in response to specific biological stimuli can identify inflammation in vivo in the murine serum transfer-induced arthritis model.

Methods: Mice were injected with 150 ml of K/BxN sera on day 0. At day 5, mice were injected intravenously with an imaging agent and visualized by IVIS® in vivo imaging system. Clinical arthritis scores were assessed. The imaging agent consists of H₂O₂– and acid-responsive polymeric particles (size: 250 nm) packed with a high concentration of near infrared (NIR) dyes. The close proximity of the dye molecules quenches their fluorescence, thus creating an “off-state“. The signal of this agent is “turned on” through the cleavage of protecting moieties on the polymer backbones upon exposure to either H₂O₂ or acidic pH, which reveals hydroxyl groups and thus increases the material’s hydrophilicity. After the hydrophobicity switch, dye molecules diffuse out of the polymer matrix, relieving the particle from fluorescence quenching (“on-state”) (Figure 1).

Results: After injecting particles (dose: 100 mL of 0.7 mg/mL Hy-Dex-IR780 in DPBS) intravenously on day 5 after arthritis induction, mice were imaged by IVIS®. Regions of inflammation exhibited significantly higher fluorescence intensity measured as an average of radiant efficiency of paws between non-arthritic and arthritic joints (1.91E+07±1.5E+06 vs 4.44E+07±1.85E+07; p<0.05). This activation results from inflammation-triggered release of dye molecules, as fluorescence was insignificant in healthy animals, and a non-responsive control version poly(lactic-co-glycolic acid)-IR780 was not activated by inflammation. Radiant efficiency in each paw significantly correlated with their clinical score. For example, a paw with a clinical score of 4 had radiant efficiency of 6.95E+07 while a paw with clinical score of 1 had 2.58E+07.

Conclusion: These results suggest that a novel activatable bioimaging agent can detect inflammation in vivo for clinical diagnosis of inflammatory conditions or to assess inflammation in animal models.

Figure 1. Particles composed of H₂O₂– and acid-responsive polymers partially disassemble upon exposure to either stimulus, activating fluorescence by relieving the particle from self-quenching. Since the NIR dye is not fluorescent in aqueous environments, only the particles light up.