Opposites Attract: Some Phytochromes Do Not Form Homodimers

Phytochromes transduce red/far-red light signals into responses regulating such processes as germination, photomorphogenesis, flowering, and shade avoidance. Phytochrome holoproteins, which include covalently bound chromophores, are found as dimers, and it has been thought that they are usually present as homodimers (for a review, see Bae and Choi, 2008). Recently, Sharrock and Clack (2004) found that heterodimers occur in Arabidopsis thaliana, but their significance to the plant was not clear. Now, Clack et al. (pages 786–799) have followed up that work to show that some phytochromes appear to form only heterodimers.

They characterized interactions of the five phytochromes of Arabidopsis and could not find evidence of homodimerization of phyC or phyE. They tested this in planta using transgenic plants expressing epitope-tagged phyC or phyE. Independent of the light conditions, phyC coimmunoprecipitated phyB and some phyD, but not native phyC or any other phy. Similarly, phyE pulled down phyB and phyD, but no others. From this and their previous data, the authors conclude that phytochromes in Arabidopsis exist as an array of homodimers and heterodimers (see figure).

Figure 1.

The array of phytochrome dimers found in Arabidopsis includes homo- and heterodimers. A summary of the dimer forms based on data in Figure 4C of Clack et al. (2009) is shown. Phytochromes A to E are colored red, blue, yellow, green, and orange, respectively.

When they examined the effects of removing heterodimerization partners using mutant lines, they found that phyC and phyE respond differently. In the absence of phyB, phyC protein is present in much lower amounts, and phyC/phyD heterodimerization levels are somewhat increased. When phyD also is missing, the amount of phyC decreases even further. Even in the absence of both of its partners, there was no evidence of phyC homodimerization or of heterodimerization with other phys. PhyE also does not form homodimers or dimerize with phys other than phyB and phyD, but it is not destabilized by the absence of its binding partners. Instead, it is found as a monomer in phyBD mutants or when overexpressed.

If phyC and phyE are functional only as heterodimers, then mutants lacking them should not have different phenotypes than mutants lacking their partners. Clack et al. tested this and found that, on the contrary, the lack of phyE exacerbates the early flowering phenotype of the phyBD mutant. In addition, the phyABDE mutant, retaining only phyC, does not show a completely etiolated phenotype. Thus, both phyE and phyC must have some biological activity in the absence of their binding partners. Whether this activity is mediated by monomers or homodimers present at levels below the threshold for detection is not yet known.

Finally, the authors tested whether heterodimers could interact with PHYTOCHROME INTERACTING FACTOR3 (PIF3), one of the basic helix-loop-helix transcription factors known to be signaling partners of phytochromes. They found that epitope-tagged PIF3 could coimmunoprecipitate all Arabidopsis phys, with the possible exception of phyE. The fact that phyC was associated with PIF3 shows that phy heterodimers are capable of interacting with known phytochrome signaling pathways and supports the physiological relevance of these heterodimers.