Why is autofluorescence obscuring my Stellaris® RNA FISH signal?

Most cells and tissue types have natural autofluorescence that is more pronounced in the green region of the visible spectrum. It is easier to discern true signal from autofluorescence if longer wavelength fluorophores such as Quasar® 570/670 or CAL Fluor® Red 610 are used.  The signal from shorter wavelength dyes, such as fluorescein (emission near 520 nm), is more difficult to detect due to the autofluorescence.

Autofluorescence is generally more evident in tissue samples, is more pronounced in certain cultured cell types, and can have very broad emission that spans the visible spectrum. For example, fluorescence from lipofuscin bodies can be detected from 360 nm to 650 nm.  Fluorescence from the lipofuscin bodies is mostly seen as cytoplasmic perinuclear spots which can be distinguished from true single molecules of RNA by being slightly larger, brighter, and by having fluorescence over a broad range of the spectrum. When you’re searching for your RNA of interest in tissue, it is especially important to distinguish the Stellaris RNA FISH signals from autofluorescent features. One way to accomplish this is to collect the light in a secondary unused filter. This will let you confirm that your RNA of interest is only present when you specifically excite the fluorophore you labeled it with.  For example, if your probe is labeled with Quasar 670 that means you will excite this dye near 647 nm and it will emit fluorescence at 670 nm.  If you try to excite that same molecule at a shorter wavelength with an unused filter, say FITC near 470 nm, you should not see the fluorescence in the same location.  If you do see fluorescence using multiple filters or with filters which are incorrect for your dye, then you are likely seeing autofluorescence and not a true RNA molecule.  For more information about the prevalence and origin of autofluorescence, please visit the Nikon Microscopy U website.

In addition, the degree of autofluorescence will depend on the fixation method used. Formaldehyde tends to crosslink fluorescent enzyme co-factors, such as flavins, whereas fixation with methanol/acetic acid tends to release them and wash them away from the sample. Prolonged fixation at elevated temperatures tends to exacerbate the autofluorescence. To assess the overall background fluorescence and to determine the contribution of non-specific probe binding, we recommend imaging a second sample in parallel that has not undergone probe hybridization - a no probe control.