Very generally speaking, in a cell the protein pool detain the 90% of the executive power: proteins execute the algorithms encoded in the DNA genes (don't report this statement to a scientist or he will freak out in spontaneous combustion). The point is that protein functions rely on their ability to engage into specific protein-protein interactions. Several approaches (two-hybrid system, interaction trap) have therefore been developed by functional genomics to study these interactions (here a review in protein interactions). One of these approaches is Resonance Energy Transfer (RET), a proximity based assay that rely on the non-radiative transfer of energy between donor and acceptor molecules according to a so-called Forster mechanism. In my knowledge two RETs are actually developed: FRET (Fluorescence) and BRET (Bioluminescence), both with different advantages and disadvantages belonging to Fluorescence vs Bioluminescence (I'm writing a post about that).
Until now, BRET imaging development at the subcellular level was hampered by the low level of light intrinsic to the bioluminescent luciferase reaction compared to fluorescent ones, so Xiaodong Xu (Vanderbilt University) and Vincent Coulon (French INSERM), started to establish the appropriate experimental conditions to visualize and quantify protein-protein interactions with BRET. Recently they demonstrated respectively on the Proceeding of the National Academy of Science and on the Biophysical Journal, that BRET imaging offers enough resolution to detect signals that originate selectively from sub-cellular compartements. This is true also for plant and animal tissues, so now it is possible to track these interactions and have a knowledge if they occur in the nucleus or plasma membrane or endocytic vesicles directly by microscopy in alive cells, while in the past BRET was preferentially used in microplate readers of cell lysates.
Because BRET is made by simultaneously using of two reporter genes (Renilla luciferase and Yellow Fluorescent Protein) I can tell you once again that two is better than one.