Some biological samples exhibit naturally occurring fluorescence, even in the absence of applied labels. This type of fluorescence is known as autofluorescence, and arises from naturally-occuring fluorophores such as collagen, lipofuscin, and flavins (just to name a few).
When carrying out fluorescence microscopy, it is important to check whether fluorescence can be excited in the sample in the absence of applied labels. To highlight the importance of autofluorescence controls, consider the following scenario. You have identified an overlooked protein that you think is an important player in macrophage biology. You are keen to visualize the distribution of this protein within macrophages. Using immunofluorescence staining, you obtain preliminary results that reveal bright signal localized to large granules within each macrophage. Representative white-light and fluorescence images are shown in Figure 1.
Figure 1: Illustration of results obtained in first immunofluorescence experiment. Left panel, white-light imaging. Right panel, green fluorescence emission obtained with blue excitation. Created with BioRender.com
When sharing your results with a colleague, they recommend running autofluorescence controls. You take their advice and prepare samples with and without the fluorescent secondary antibody. You image the labeled samples first, and the signal is bright and granular, similar to the last experiment. Using the same acquisition settings, you then image the autofluorescent controls. When you compare both conditions, the staining looks identical. Your heart sinks a little, but you are relieved you noticed earlier rather than later.
This scenario highlights the importance of running autofluorescence controls. Fluorescence emission from naturally-occurring components within the sample can obscure the emission from the intended target. It can confuse you or worse, mislead you into thinking you have successful staining when you do not.
As you map out the next steps, you choose to give the primary antibody a few more tries, as it has been extensively used for immunofluorescence in other cell types. Upon further reading, you learn that macrophages sometimes contain high levels of lipofuscin that can be excited in the blue and emit in the green. Also, you note that in the first experiment you had used a secondary fluorescent antibody that has similar excitation and emission spectra to lipofuscin. Based on this realization, you decide to test whether replacing the secondary antibody with one that absorbs and emits in the far-red makes a difference.
In the next iteration, the fluorescent staining looks different. You image the experimental sample and then use the same settings to image the autofluorescence control (using two channels; the new far-red channel as well as the original green emission). The far-red channel of the experimental sample also reveals granular structures, but these are distinct and separate from the granules observed in the green emission channel (see Figure 2). Further, the far-red channel in the autofluorescent control has minimal intensity. You have set the stage for acquiring robust, rigorous, and exciting results.
Figure 2: Illustration of results obtained when the secondary antibody was switched for one that emits in the far red (magenta). The resulting staining is distinct from autofluorescence seen in the green channel. Created with BioRender.com
To summarize, when carrying out fluorescence imaging, it is important to check for endogenous fluorescence by imaging matched autofluorescence controls. It is important to use the same acquisition settings in the experimental and autofluorescence control samples (as well as in any other controls). If the signal levels are comparable in both conditions, then it is important to develop approaches to minimize the effects of the autofluorescence.
As highlighted above, the simplest approach is to move the imaging to longer (towards the red) wavelengths, as most biological samples fluoresce more strongly at shorter, high-energy wavelengths. Multiphoton microscopy can also be considered, as the two-photon excitation often leads to lower autofluorescence when compared to single-photon imaging (ie: confocal or wide-field microscopy). Other instrument-based approaches also include using spectral, lifetime, and time gating approaches.
If switching channels is not practical or feasible, such as when you need to image many fluorescence channels, you can consider chemical methods for reducing autofluorescence. This is known as “fluorescence quenching.” These chemical methods include treating the sample with reagents such as Sudan Black. Furthermore, many vendors now have commercial solutions available as well.
Thus far, we have treated autofluorescence as an unwelcome guest. While autofluorescence can create barriers, it also can provide exciting opportunities and reveal new insights about your biological questions. We will highlight how autofluorescence imaging can be a powerful and welcome addition to your experimental practice in the next post.