Integral membrane protein complexes consisting of proteins and small molecules that act as cofactors have important functions in all organisms. To form functional complexes, cofactor biosynthesis must be coordinated with the production of corresponding apoproteins. To examine this coordination, we study bacteriorhodopsin (BR), a light-induced proton pump in the halophilic archaeon Halobacterium salinarum. This complex consists of a retinal cofactor and bacterioopsin (BO), the BR apoprotein. To examine possible novel regulatory mechanisms linking BO and retinal biosynthesis, we deleted bop, the gene that encodes BO. bop deletion resulted in a dramatic increase of bacterioruberins, carotenoid molecules that share biosynthetic precursors with retinal. Additional studies revealed that bacterioruberins accumulate in the absence of BO regardless of the presence of retinal or BR, suggesting that BO inhibits bacterioruberin biosynthesis to increase the availability of carotenoid precursors for retinal biosynthesis. To further examine this potential regulatory mechanism, we characterized an enzyme, encoded by the lye gene, that catalyzes bacterioruberin biosynthesis. BO-mediated inhibition of bacterioruberin synthesis appears to be specific to the H. salinarum lye-encoded enzyme, as expression of a lye homolog from Haloferax volcanii, a related archaeon that synthesizes bacterioruberins but lacks opsins, resulted in bacterioruberin synthesis that was not reduced in the presence of BO. Our results provide evidence for a novel regulatory mechanism in which biosynthesis of a cofactor is promoted by apoprotein-mediated inhibition of an alternate biochemical pathway. Specifically, BO accumulation promotes retinal production by inhibiting bacterioruberin biosynthesis.

The Saccharomyces cerevisiae Fkh1 protein has roles in cell-cycle regulated transcription as well as a transcription-independent role in recombination donor preference during mating-type switching. The conserved FHA domain of Fkh1 regulates donor preference by juxtaposing two distant regions on chromosome III to promote their recombination. A model posits that this Fkh1-mediated long-range chromosomal juxtaposition requires an interaction between the FHA domain and a partner protein(s), but to date no relevant partner has been described. In this study, we used structural modeling, 2-hybrid assays, and mutational analyses to show that the predicted phosphothreonine-binding FHA domain of Fkh1 interacted with multiple partner proteins. The Fkh1 FHA domain was important for its role in cell-cycle regulation, but no single interaction partner could account for this role. In contrast, Fkh1’s interaction with the Mph1 DNA repair helicase regulated donor preference during mating-type switching. Using 2-hybrid assays, co-immunoprecipitation, and fluorescence anisotropy, we mapped a discrete peptide within the regulatory Mph1 C-terminus required for this interaction and identified two threonines that were particularly important. In vitro binding experiments indicated that at least one of these threonines had to be phosphorylated for efficient Fkh1 binding. Substitution of these two threonines with alanines (mph1-2TA) specifically abolished the Fkh1-Mph1 interaction in vivo and altered donor preference during mating-type switching to the same degree as mph1Δ. Notably, the mph1-2TA allele maintained other functions of Mph1 in genome stability. Deletion of a second Fkh1-interacting protein encoded by YMR144W also resulted in a change in Fkh1-FHA-dependent donor preference. We have named this gene FDO1 for Forkhead one interacting protein involved in donor preference. We conclude that a phosphothreonine-mediated protein-protein interface between Fkh1-FHA and Mph1 contributes to a specific long-range chromosomal interaction required for mating-type switching, but that Fkh1-FHA must also interact with several other proteins to achieve full functionality in this process.