Collagen is weakly fluorescent at lower wavelengths and in the UV. If you have collagen matrices this fluorescence can be fairly significant, especially if you are using a widefield microscope. The easiest way to avoid this is to pick probes that fluoresce at longer wavelengths. Use something like TO-PRO-3 instead of DAPI, and for antibodies use fluorophores like Alexa 565. If you need 3 fluorophores and have a limited range of excitation wavelengths. Whatever channel you expect will be the brightest should be stained with a 488 fluorophore. If there is a large enough gap between background intensity and staining intensity you can subtract out the background without also removing the staining you are interested in.