eTopic 12.1 Mapping Protein-Protein Interactions

Learning which proteins in a cell physically interact with each other is an important step toward understanding global aspects of cell metabolism and regulation. Two approaches were described in the text; the classic yeast two-hybrid system and a method in which affinity-tagged bait proteins and their attendant interacting partners are pulled from cytoplasmic extracts by affinity chromatography (see Section 12.3). An interesting yeast two-hybrid experiment used to identify yeast interacting proteins was carried out by Micheline Fromont-Racine and colleagues at the Pasteur Institute in Paris. They developed a high-throughput, genome-wide version of the yeast two-hybrid system to generate protein interaction maps for whole cells (Fig. 1). The automated version of the procedure is rapidly becoming the method of choice for mapping whole proteomes. With a yeast cell–mating procedure that increases screening efficiency, Fromont-Racine and colleagues used their complex yeast genome library of 5 ´ 106 clones to test 7 ´ 108 interactions against 15 proteins.

The method identified and classified 170 potential interactors, including approximately 70 proteins of previously unknown function. More than 25% of the interactors are probably biologically relevant, although one cannot be sure without additional in vivo testing. The other 75% may be spurious interactions due to the unnatural overexpression of the two proteins in the test cells. Overexpression of two proteins could magnify low-affinity interactions that would not take place if those proteins were present at lower, more physiological concentrations.

The achievements of this group opened the way to systematic analysis of the protein interaction networks of the 6,000 open reading frames of the yeast proteome. Another European team (Hybrigenics, in Paris) adapted the Fromont-Racine procedure to analyze a bacterial genome and linked half of the 1,600 proteins of the ulcer-provoking bacterium Helicobacter pylori into partial protein interaction maps.


Figure 1  Protein-protein interaction maps.  Representation of a protein interaction map for the most likely interactions in Haemophilus influenzae. The large black circle represents the genome. 0° corresponds to the first base pair and 360° to the last base pair of the genome. Predicted protein interactions are indicated by linking the circular map positions of the genes (each gene named by an arbitrary number) involved. In cases of neighboring genes (<5°), a small circle indicates the predicted interaction between two genes at that region; otherwise, an arc links the two genes in question.