Concentrate your secreted media via rotary evaporator, for better concentration, then load the sample in Tricine-PAGE with ladder for separation (if you think that the size of your desired protein is big then try 2D), cut every unknown size band, if needed all bands, and took them separate tubes, and purify them with protein purification kit, and then you can place them for MS analysis, otherwise you got haphazard data. Or you can try Agilent OFFGEL Fractionator for collecting every different fractions, and then try MS analysis. I try both methods, last one was good. Depend on your working flow.