DZF domain mediated homodimerization can be studied in the Bacterial one-hybrid

4.2.1 The Bacterial one-hybrid (B1H) system

In addition to the B2H system a B1H system has been described which can be used to study and identify dimeric proteins (Hu et al., 1990). This system is based on the observation that the N-terminal DNA binding domain (NTD) of the bacteriophage lambda repressor (λcI) requires a dimerization domain in order to bind efficiently to its specific DNA site, the λ operator. Thus, fusing a domain of interest (X) to the λcI NTD will permit the λcI-X fusion protein to bind to the λ operator as long as X can mediate dimerization. The dimerization can be detected by using an appropriate E. coli reporter strain. As shown in Figure 4.2, in this strain, the λ operator is placed in between the -10 and -35 region of a strong promoter and binding of the λcI-X fusion protein to its site will repress the promoter by blocking binding of the RNAP. By using lacZ as the reporter gene, repression of the promoter can be easily assessed performing quantitative β-galactosidase assays.

4.2.2 Validation of the B1H system for studying homodimeric DZF domain interactions

To initially test whether homodimeric interactions of the DZF domain can be detected in the B1H system, we generated plasmids encoding the DZF domains from Ikaros (work performed by R. Fang), Pegasus and Drosophila Hunchback fused to the C-terminus of the λcI DBD (consisting of the NTD and the linker that connects the NTD with the naturally C-terminal domain of λcI). Each of these plasmids (encoding the λcI-DZF fusion proteins) was transformed into a B1H reporter strain containing a λcI operator positioned between the -35 and -10 hexamers of the lac UV5 promoter. To test for homodimerization and DNA-binding mediated by these different DZF domains, β-galactosidase assays were performed. As expected, it was found that cells which expressed the λcI NTD+linker alone (lacking the C-terminal λ dimerization domain) showed high levels of transcription of the lacZ reporter gene. In contrast, transcription of lacZ was repressed in cells expressing the λcI-DZF fusion proteins suggesting that all three DZF domains mediate efficient homodimerization of the chimeric protein (Figure 4.3A).

λ operator

Figure 4.2 Schematic of the Bacterial one-hybrid (B1H) system. A dimeric protein X can mediate dimerization of the N-terminal DBD of λcI (λcI-NTD, blue circles) resulting in the binding of the λcI-X fusion protein to the λ operator site (lower panel, left side). In an appropriate B1H reporter strain the λ operator is placed between the -10 and -35 sequences. Thus, binding of the dimeric λcI-X fusion protein competes with binding of the RNAP which results in a repression of the reporter gene expression (lower panel, right side). In contrast, non-dimeric proteins fused to λcI-NTD will abolish the ability of λcI to bind to the λ operator site, thus resulting in an activation of reporter gene expression (upper panel). This figure was kindly provided by K. Joung.

To further test, if the repression of lacZ is dependent on the expression of the different λcI-DZF fusion proteins, an IPTG titration experiment was performed where the expression of the proteins was induced using various amounts of IPTG.

As shown in Figure 4.3B, repression of lacZ was directly related to the protein expression level as long as the IPTG concentration stayed below a certain threshold. Interestingly, the effect of increased IPTG concentration was different for the individual fusion proteins. The Pegasus DZF domain reached a maximum level of repression at an IPTG concentration of

fold-repression of β-galactosidase expression

IPTG concentration (μM)

fold-repression of β-galactosidase expression fold-repression of β-galactosidase expressionfold-repression of β-galactosidase expression

A

B C

D E

fold-repression of β-galactosidase expression

IPTG concentration (μM)

fold-repression of β-galactosidase expression fold-repression of β-galactosidase expressionfold-repression of β-galactosidase expression

A

B C

D E

Figure 4.3 Analysis of the Ikaros, Pegasus and Hunchback Drosophila DZF domains in the B1H system. (A) B1H reporter strains expressing the three DZF domains fused to λcI were assayed for β-galactosidase activity. A control expressing the λcI-NTD protein is also shown. In this assay an IPTG concentration of 50 μM was used for inducing the expression of the fusion proteins. (B, C, D and E) The activity of these wild-type constructs together with constructs harboring the D18Q mutation in the DZF domains were assessed using increasing concentrations of IPTG. The three type proteins (B), wild-type and D18Q Ikaros (C), wild-wild-type and D18Q Pegasus (D) as well as wild-wild-type and D18Q Hunchback (E) were group wise compared to each other.

100 μM, whereas for the Ikaros DZF domain the repression level increased linearly with the IPTG concentration even up to high levels of IPTG (Figure 4.3B and data not shown, R.

Fang). Overexpression of the Hunchback DZF domain was toxic to the bacterial strains and the fusion protein could not mediate repression when the IPTG concentration went above 25 μM. The toxicity may have forced the bacterial cells to select for a Hunchback mutant in order to survive and grow. This would explain why these cells did not repress lacZ when high levels of proteins were present. On the other hand, these hybrid proteins might have dimerized detached from the DNA and were therefore not able to mediate transcriptional repression at the promoter.

To further validate the B1H system as a genetic method to study DZF mediated homodimerization an additional series of constructs was designed where the previously described D18Q mutation (McCarty et al., 2003) was introduced into the DZF domains.

Plasmids encoding this mutation were transformed into the B1H reporter strain and β-galactosidase assays were performed using cultures grown in various concentrations of IPTG.

It was found that the three proteins showed a different response to the D18Q mutation at various concentrations of IPTG. For the Ikaros DZF domain, the ability of the protein to repress lacZ was completely abolished by the D18Q mutation even at very high levels of IPTG (Figure 4.3C, R. Fang). In contrast, overexpression of the mutated Hunchback DZF domain abolished repression only at IPTG concentrations below 10 μM, indicating that the effect of the mutation is very sensitive to the amount of expressed protein (Figure 4.3E). In the case of the Pegasus DZF domain, the D18Q mutation resulted in a small reduction of repression (Figure 4.3D) which is consistent with the data obtained in the B2H system (see section 3.4.1 and data not shown).

In summary, the results of these assays demonstrated that the repression observed in the B1H assay is due to expression of the different domains since the level of repression depends on the amount of protein expressed in the cell. Furthermore, the D18Q mutation affected dimerization of Ikaros and Hunchback but not of Pegasus which was also found when this mutation was tested in the B2H system (see section 3.4.1). These results confirm the validity of using the B1H system to analyze dimerization mediated by the DZF domain.