Isolation of secreted proteins from Drosophila ovaries and embryos through in vivo BirA-mediated biotinylation

Abstract

The extraordinarily strong non-covalent interaction between biotin and avidin (kD = 10-14-10-16) has permitted this interaction to be used in a wide variety of experimental contexts. The Biotin Acceptor Peptide (BAP), a 15 amino acid motif that can be biotinylated by the E. coli BirA protein, has been fused to proteins of interest, making them substrates for in vivo biotinylation. Here we report on the construction and characterization of a modified BirA bearing signals for secretion and endoplasmic reticulum (ER) retention, for use in experimental contexts requiring biotinylation of secreted proteins. When expressed in the Drosophila female germline or ovarian follicle cells under Gal4-mediated transcriptional control, the modified BirA protein could be detected and shown to be enzymatically active in ovaries and progeny embryos. Surprisingly, however, it was not efficiently retained in the ER, and instead appeared to be secreted. To determine whether this secreted protein, now designated secBirA, could biotinylate secreted proteins, we generated BAP-tagged versions of two secreted Drosophila proteins, Torsolike (Tsl) and Gastrulation Defective (GD), which are normally expressed maternally and participate in embryonic pattern formation. Both Tsl-BAP and GD-BAP were shown to exhibit normal patterning activity. Co-expression of Tsl-BAP together with secBirA in ovarian follicle cells resulted in its biotinylation, which permitted its isolation from both ovaries and progeny embryos using Avidin-coupled affinity matrix. In contrast, co-expression with secBirA in the female germline did not result in detectable biotinylation of GD-BAP, possibly because the C-terminal location of the BAP tag made it inaccessible to BirA in vivo. Our results indicate that secBirA directs biotinylation of proteins bound for secretion in vivo, providing access to powerful experimental approaches for secreted proteins of interest. However, efficient biotinylation of target proteins may vary depending upon the location of the BAP tag or other structural features of the protein.