2.5. Two-photon in vivo imaging

All animal procedures were performed in accordance with the University of Helsinki animal care regulations. Local authority (ELÄINKOELAUTAKUNTA-ELLA) approved the animal license (ESAVI/9071/04.10.07/2016) to conduct the procedures described in the study.

2 female C57BL/6JRccHsd WT mice were used for the imaging. 3–4 weeks before the start of imaging experiments, animals were anesthetized with ketamine/xylazine and operated for implantation of a cranial window. The cranial window was inserted over the somatosensory cortex at the following coordinates: AP −1.8, ML −2.0 from Bregma. Dental drill was used to remove a round shaped (d = 4 mm) piece of skull, and the hole in the bone was covered with a round cover glass (d = 5 mm). Mice were imaged with the FV1200MPE two-photon microscope (Olympus, Japan) with the 25X water immersion 1.05 NA objective specially designed for in vivo two-photon imaging. MaiTai Broad Band DeepSee laser tuned to 800 nm was used for excitation. In a preliminary set of in vitro tests 800 nm was found to be optimal for preferential excitation of FITC fluorophore. Unfortunately, isolated excitation of the two fluorophores was not possible due to the limitations of two-photon excitation and lasers. Emission light was collected using the band pass filters: 515–560 nm for FITC fluorescence and 590–650 for Dylight 549 fluorescence.

For in vivo imaging sessions, animals were anesthetized with ketamine/xylazine and head-fixed under two-photon microscope. Three-dimensional (3D) baseline autofluorescence image Z-stacks of the mouse somatosensory cortex was acquired. Stacks of images were collected with the vertical step of 3 μm with zoom factor one at 800 × 800 pixels aspect ratio. Line-scans were performed inside the lumen of superficial cortical vessels before and after the injection of biotin-labeled (or unlabeled) dextran and STV-labeled NPs (nSiO₂@mSiO₂@PEI@STV). 100 μL of 10 mg/mL solution of conjugated NPs was injected intravenously.

After collection of the data, images were analyzed using Fiji/ImageJ software. NPs were visually detected in the Dylight 549 channel and analyzed by ratiometric analysis. The advantage of using ratiometric analysis is that it is not sensitive to crosstalk between the dyes as crosstalk is constant between the experiments. Background fluorescence was subtracted from all the images. In total 19 NPs were analyzed. Graphs were plotted in Microcal Origin software.