Highly sensitive microplate β-galactosidase assay for yeast two-hybrid systems

The yeast two-hybrid system is a powerful tool to monitor protein-protein interactions. It relies on the expression of hybrid proteins in the yeast nucleus and the activation of reporter genes when a hybrid transcription factor (e.g., Gal4) is reconstituted via two interacting proteins (1). The method has been used at the genome level in organisms ranging from viruses to plants to assess protein-protein interaction networks using prototrophy markers as reporter genes (2–4). A second reporter gene, LacZ, offers the possibility of quantitative measurements of β-galactosidase activity by a colorimetric method (5). A β-galactosidase test in bacteria has been transposed to microplates (6). In the case of yeast, the conditions for cell lysis in microplates have been improved (7,8), but for truly high-throughput twohybrid tests (4,9,10), the β-galactosidase reporter activity is still assayed in test tubes. In this report, we have successfully adapted yeast culture, permeabilization, and β-galactosidase assay conditions to microtiter format, and the optimization of all these steps led us to increase the sensitivity of the enzyme assay by at least 50-fold compared to classical methods. Direct kinetic recording instead of end point measurements allows for the processing of highand low-activity samples in the same plate with good reproducibility. The colorimetric assay costs approximately $200/1000 plates, which is about 200-fold less expensive than a commercially available β-galactosidase assay kit. The method can be used for the various types of yeast two-hybrid systems currently in use. To set up the method, we used a twohybrid system in which the DNA-binding domain was the LexA protein from Escherichia coli that has been previously described (11) in L40 strain. We then extended the method to the other LexAbased system (12) in EGY48 strain with reporter plasmid pSH18:34 and to a system in which the DNA-binding domain was the DNA-binding domain of the yeast transcription factor Gal4 itself (BD Biosciences Clontech, Erembodegem, Belgium) in Y190 strain. The two-hybrid systems and yeast strains have been previously described (13). Yeast cells expressing interacting pairs were grown in glucose medium (galactose in the case of Reference 12) with appropriate selection. With the classical approach, to assess the reporter gene HIS3 (or LEU2), yeast patches were replicated on agar medium that lacked the appropriate amino acids, and growth was visualized after 48 h (Figure 1A). The reference method for the β-galactosidase assay was a filter test, using X-gal as a substrate (5), in which a blue coloration developed in 1–10 h (Figure 1B). To adapt the test to microplates, we modified several steps. Yeast culture was made in microplate format instead of culture tubes. Culture medium (0.5 mL), consisting of 2% glucose, 6.7 g/L yeast nitrogen base without amino acids, supplemented with an amino acid mixture lacking the appropriate amino acids for selection (Qbiogene S.A., Illkirch, France), was inoculated with a toothpick from a patch on agar plates, and yeast cells were cultured in deep-well, conical-bottom microplates (Eppendorf, Le Peck, France) in a thermoblock shaker with a microplate holder. We found the Thermomixer Comfort model (Eppendorf) provided the vigorous shaking (900 rpm) needed to prevent the yeast