Quantifying Subcellular Protein Distribution

Prepared slides were mounted on the Leica TCS SP8 confocal microscope stage and viewed under 63× optical lens. Using lasers, samples were excited at 488, 552 and 638 nm to detect fluorescence from GFP, CrebA and TOTO3 DNA stain respectively. The whole cell area was chosen, and the fluorescence was recorded using ImageJ. The nuclear area from samples were chosen and fluorescence was recorded. Fluorescence from the nucleus was subtracted from fluorescence recorded from the whole area to get the fluorescence coming from the cytoplasm.
To calculate the distribution of JIG and CREB proteins between the nucleus and cytoplasm, we stained dissected salivary glands with DNA marker TOTO3 and using the appropriate antibodies. Confocal images of the whole organs were taken on Leica DMI8 confocal system and then analyzed using QuPath 0.4.0 software (54 (link)). We utilized the deep-learning neural network StarDist trained to detect fluorescently labeled nuclei (55 (link),56 (link)). Nuclei were detected in TOTO3 channel (633 laser), and fluorescence intensity was calculated for proteins (JIG and CREB) in 488 laser channel. We defined the cells as area 10mkm around each detected nucleus. The ratio in protein signal between the cytoplasm and nucleus was be calculated for the whole organ.