"reverse translation" from protein to nucleotide does not yield a unique sequence. Your best bet is to use the BLAST feature that searches your protein sequence against the entire genbank DNA database translated into protein:
http://www.ncbi.nlm.nih.gov/BLAST/Blast.cgi?PROGRAM=tblastn&PAGE_TYPE=BlastSearch&LINK_LOC=blasthome This will find the natural genes (and the closest homologues).
