This problem is relatively easy to resolve. You need antibodies labeled with a red (e.g. 635 nm emission) dye. In this wavelength range, proteins usually do not show any autofluorescence. However, you have to consider the available wavelengths (excitation / emission) of your fluorescence detector. If you know this, you can choose an appropriate dye or dye-antibody conjugate for your experiments. 

Thank you, this helps. I used so far a secondary antibody labelled with blue light, specifically alexa fluor 488. I am planning to use alexa fluor 555 as well, do you think may work?

re: you have to consider the available wavelengths (excitation / emission) of your fluorescence detector. 

I am not completely sure what you meant here, could you please tell me something more? is it related to the filter I use to capture the fluorescence?

Many thanks

In fluorescence, you have 2 relevant wavelengths, the excitation, and the emission. In some systems, you have a laser, a LED or a lamp for excitation. Depending on these light sources and their respective filters if present, you may be unable to use all dyes you want.