Large magnetic nanoparticles or aggregates are advantageous in their magnetic resonance properties over ultrasmall superparamagnetic iron oxide (USPIO) nanoparticles (NPs), but the former are cleared faster from the blood pool. Therefore, the "smart" strategy of intracellular aggregation of USPIO NPs is required for enhanced T2-weighted MR imaging. Herein, employing an enzyme-instructed condensation reaction, we rationally designed a small molecule Ac-Asp-Glu-Val-Asp-Cys(StBu)-Lys-CBT (1) to covalently modify USPIO NPs to prepare monodispersive Fe3O4@1 NPs. In vitro results showed that Fe3O4@1 NPs could be subjected to caspase 3 (Casp3)-instructed aggregation. T2 phantom MR imaging showed that the transverse molar relaxivity (r2) of Fe3O4@1 NPs with Casp3 or apoptotic HepG2 cells was significantly larger than those of control groups. In vivo tumor MR imaging results indicated that Fe3O4@1 NPs could be specifically applied for enhanced T2 MR imaging of tumor apoptosis. We propose that the enzyme-instructed intracellular aggregation of Fe3O4 NPs could be a novel strategy for the design of "smart" probes for efficient T2 MR imaging of in vivo biomarkers.

Zebrafish central nervous system (CNS) possesses a strong neural regeneration ability to restore visual function completely after optic nerve injury (ONI). However, whether neurogenesis of retinal ganglion cell (RGC) contributes to functional recovery remains controversial. Our quantitative analysis of RGCs in different ONI models showed that almost all RGCs survived in optic nerve crush (ONC) model; while over 90% of RGCs survived in the first 2 weeks with 75% remaining after 7 weeks in optic nerve transection (ONT) model. Retrograde labeling from tectum revealed a surprising regeneration rate, with over 90% and over 50% of RGCs regrowing axons to tectum at the first week in ONC and ONT model respectively. In the latter one, the number of regenerative RGCs after 4 weeks had no significant difference from the control group. As for neurogenesis, newborn RGCs were rarely detected either by double retrograde labeling or BrdU marker. Since few RGCs died, microglia number showed a temporary increase at 3 days post injury (dpi) and a decrease at 14 dpi. Finally, myelin structure within retina kept integrity and optomotor response (OMR) test demonstrated visual functional restoration at 5 weeks post injury (wpi). In conclusion, our results have directly shown that RGC survival and axon regrowth are responsible for functional recovery after ONI in adult zebrafish.

Compared to (1)H MRI, (19)F MRI provides higher selectivity but lower sensitivity. Therefore, the need to inject high doses of the (19)F probe to improve its sensitivity for in vivo diagnosis remains a challenge. A "smart" strategy is needed that could locally concentrate a low-dose (19)F probe while avoiding the fast transverse relaxation of the probes. Locally self-assembling and disassembling (19)F nanoparticles may be an optimal measure to achieve this goal. Herein, we report a dual-functional probe 1 for glutathione (GSH)-controlled self-assembly and subsequent caspase 3/7 (Casp3/7)-controlled disassembly of formed nanoparticles (i.e., 1-NPs). Consecutive assembly and disassembly of 1-NPs translate to "off" and "on" (19)F magnetic resonance (MR) signal states, respectively. Employing this smart strategy, we successfully used 1 for the consecutive detection of GSH and Casp3/7 activity in vitro and in cells and imaging Casp3/7 activity in cells and in zebrafish at low doses with a 14.1 T magnetic field.

(19)F MRI has higher selectivity but lower sensitivity than (1)H MRI for in vivo diagnosis. Therefore, to avoid using a high injection dose of the (19)F probe while, in the meantime, maintaining the high sensitivity of (19)F MRI has remained challenging. Local self-assembly and disassembly of (19)F nanoparticles could be one of the "smart" strategies to achieve this goal. Herein, we report a dual-functional probe 1 for glutathione (GSH)-controlled self-assembly and subsequent legumain (Lgmn)-controlled disassembly of its nanoparticles (i.e., 1-NPs). Self-assembly and disassembly of 1-NPs confer (19)F magnetic resonance (MR) signals "off" and "on", respectively. Employing this strategy, we successfully applied 1 for consecutive detections of GSH and Lgmn in vitro and in cells, imaging Lgmn activity in HEK 293T tumors in zebrafish at a low dosage under 14.1 T.

For cancer diagnosis, 1H magnetic resonance imaging (MRI) is advantageous in sensitivity but lacks selectivity. Endogenous 19F MRI signal in humans is barely detectable and thus 19F MRI has very high selectivity. A combination of 1H and 19F MRI is ideal for precise tumor imaging but a protease‐controlled strategy of simultaneous T2 1H MRI enhancement and 19F MRI “Turn‐On” has not been reported. Here, used is a click condensation reaction to rationally project a dual‐functional fluorine probe 4‐(trifluoromethyl)benzoic acid (TFMB)‐Arg‐Val‐Arg‐Arg‐Cys(StBu)‐Lys‐CBT (1), which is further utilized to functionalize Fe3O4 nanoparticle (IONP) to achieve IONP@1. As such, the IONP aggregation can be activated by furin addition, thereby enhancing the T2 1H MRI signal and switching the 19F NMR/MRI signal “On”. Using this strategy, IONP@1 is successfully applied to detect the activity of the furin enzyme with “Turn‐On” 19F NMR/MRI and T2 1H MRI signals are enhanced. Moreover, IONP@1 is also applied for precise dual‐mode (1H and 19F) MR imaging of tumors in zebrafish under 14.1 T. The current approach, therefore, provides a feasible and robust means to reconcile the dilemma between selectivity and sensitivity of conventional MRI probes. More importantly, it is envisioned that, by substituting the TFMB moiety in 1 with a perfluorinated compound, this “smart” method could be of potential use for precise 1H MR and 19F MR imaging of tumor in mouse or in bigger rodents in near future.