Two-hybrid system and false positives. Approaches to detection and elimination.

Abstract

The two-hybrid system has gained wide popularity in the last decade because of the relative ease and speed with which novel interacting proteins can be isolated. Because the two-hybrid approach is based on the interaction of hybrid proteins in a living yeast cell, it offers numerous advantages in comparison to traditional biochemical methods, including the avoidance of costly protein purifications or development of antibodies, and, consequently, great reduction in time required to identify novel protein partners. On the other hand, because two-hybrid screens rely on the indirect readout of transcriptional activation of reporter genes, they are intrinsically prone to distinct classes of artifacts related to the biasing of the transcriptional response. While two-hybrid assays have been quite successful at identifying real interactions (one estimate of success rate is ∼70% for usable baits; [1]), the vast majority of library screens also identify at least one protein that is presumed to be a false positive. The first part of this chapter discusses what classes of protein tend to isolate or be isolated as false positives, and the second part presents simple approaches to limit the occurrence of false positive.