What I would first do is check for two things:

• will your combined protein contain the second one in the correct reading frame?
• does the sequence for you interstitial protein (the one between the two known ones) show any similarity to any other known protein (blast nr)

if the answer is no to both questions then I would doubt that your resulting protein is functional - although crazier things could happen in bacteria

Whenever something like that happens to me, I automatically think the assembly and/or annotation is to blame. However,... It is definitely possible that a fusion protein was made, and that could be really neat.
It is almost impossible to predict the effects of fusion proteins like that. It really depends on whether two domains were fused, what the functions of the individual domains are/were. E.g. imagine a DNA binding protein fusing to a transcription factor, this could have some funky consequences...

But as I mentioned, rule out errors/bugs in the annotation first!