Effects of missense mutation on structure and function of photoreceptor

Abstract

Phytochromes (PHYs) are photoreceptors of the red (R ~660  nm) and far-red (FR ~730 nm) light, and they control a wide range of responses affecting crucial aspects of plant life. There are five genes PHYA-PHYE encoding for phytochromes of different but overlapping function. One of these, PHYA has the unique function controlling specific responses in high irradiance far-red, as well as in very weak light. Appropriate PHYA functioning requires not only the photoreversibility of molecule but also the proper nuclear localization and degradation of receptor. Recently, we identified and described a mutant PHYA allele (phyA-5) in Arabidopsis thaliana, which showed reduced binding affinity to FHY1/FHL, the proteins regulating its nuclear transport, resulting in impaired nuclear localization and altered signaling under certain conditions. We present here a hypothesis to explain how the identified amino acid substitution may lead to structural changes manifested as altered signaling and phenotype displayed by the phyA-5 mutant.

Keywords: :

• Arabidopsis
• photoreceptor
• photomorphogenesis
• phytochrome A
• nuclear import
• FHY1
• FHL
• 3D structure

This article refers to:

A functional PHY molecule is a dimer of two ~120 kDa monomers, and each of the monomers contains the N-terminal photosensory and the C-terminal regulatory domains (see Fig. 1A). The N-terminal part is necessary for the photosensory function, and it contains the NTE, PAS, GAF and PHY domains. The GAF domain harbors the linear tetrapyrrol chromophore attached to a conserved cysteine residue via thioether bond.^1,2 The C-terminal part, containing two PAS domains and a region similar to the histidine kinases (HKRD)¹ is essential for the nuclear translocation and dimerization of molecule.³

Figure 1. Location and nature of the phyA-5 mutation. (A) Schematic arrangements of the phytochrome domains (not to scale). The position of phyA-5 mutation is indicated. NTE, N-terminal extension; PAS, PER/ARNT/SIM domain; GAF, cGMP phosphodiesterase, adenylate cyclase, FhlA domain; PHY, phytochrome domain; HKRD, histidine kinase-related domain; black rectangle represents the chromophore. (B) Backbone conformation of the predicted structure, based on the sequence including the residues 20­50 for the PHYA protein. The helical segment is represented by yellow ribbon, while the other parts of molecule by blue ribbon. (C) Backbone conformation of the predicted structure, based on the sequence covering the residues 20­50 for the PHYA-5 protein. The helical segment is represented by yellow ribbon, while the other parts of molecule by blue ribbon. (D) Predicted probabilities regarding the α-helical structure for the amino acids, on the basis of sequence including the residues 1­68 for the PHYA (blue line) and PHYA-5 (red line) proteins.

PHYs are synthesized in Pr form which can be converted by R photon absorption to the Pfr form which can be transformed back to Pr by FR photon absorption. The biologically active phyA Pfr conformer binds directly to nuclear import shuttle proteins, such as FHY1 (for FAR-RED ELONGATED HYPOCOTYL1) and FHL (for FHY1-LIKE)^4,5 and it is transported to the nucleus. PHYA Pfr is ubiquitinated and undergoes rapid degradation.^6-8 Recent results, supported with mathematical modeling prove that not only the unperturbed Pr/Pfr photoconversion but also the nuclear import and degradation of photoreceptor are required for the normal PHYA functioning. These events are tightly linked to each other like interlacing “cycles”.⁹ Very recently, we identified and described a mutant (phyA-5) expressing PHYA with a missense mutation.¹⁰ This mutant line shows a fairly complex phyA-specific phenotype, including a reduced nuclear import at low intensities of FR indicating that the point mutation identified alters the FHY1/FHL-phyA Pfr association/dissociation “cycle”. The fact that PHYA-5 governs wild-type-like signaling when it is fused to a nuclear localization signal (NLS), strongly supported this conclusion.¹⁰

Our knowledge about the interaction of phyA with the FHY1 and FHL proteins is rather limited. Yeast two-hybrid assays showed that the Pfr form of the phyA406 (1–406 amino acids) specifically binds the FHY1/FHL transporter molecules.^5,11 However, the exact three-dimensional (3D) structure of plant PHYs in contrast to microbial PHYs is not yet resolved¹² and most of the published studies on microbial PHYs focused on the structural changes of Pr/Pfr conformation.^2,12-15 The phyA-5 mutation locates in the NTE domain, thus it is unlikely that it alters the structure of the chromophore bearing GAF domain (see Fig. 1A). Functional analyses of the NTE domain revealed that the serine residues located here are necessary for the PHY functionality.^16-22 These studies focused on the possible function of the phosphorylation of serine residues, examining mutant PHYs with serine-to-alanine substitutions or with deletion of certain residues. Interestingly, the alanine-to-valine replacement in the phyA-5 mutant receptor results in a complex PHYA-specific phenotype,¹⁰ despite the similar characteristics of these amino acids. To reveal the possible structural changes caused by this mutation, we performed protein structure prediction by the RaptorX program,²³ applying single- and multiple-template threading methods. For both PHYA and PHYA-5, we have chosen three sequences consisting of various numbers (from 30 to 140) of residues, in order to predict the 3D structural features of the N-terminal parts of proteins. Interestingly, the results indicate that the substitution of alanine by valine in position 30 results in shortening of the α-helical segments for all sequences, as it is demonstrated in the case of two representative structures in Figure 1B and 1C. Nevertheless, based on the predicted distributions of different secondary structures calculated for each residue, an additional observation can be made, concerning the effects of alanine-to-valine replacement. The predicted probabilities with regard to the α-helical structure for the amino acids are presented in the case of two representative samples in Figure 1D. Considering these data, it can be seen that for the PHYA-5, the amino acids located in the region covering the residues 28­38 can be characterized by a lower propensity to occupy the α-helical conformation, as compared with that found in the case of PHYA. These results suggest that the shorter α-helical segment, as well as the lower propensity to form this helical structure, in the NTE domain can disturb the binding of PHYA to FHY1/FHL. We note, that the data obtained by this modeling study make feasible to determine, by targeted and judicious substitution of amino acids located between 28­38, to what extent this short region of the phyA N-terminal domain affects protein-protein interactions crucial for proper phyA signaling.

Acknowledgments

Work in Szeged, Hungary was supported by the Hungarian Scientific Research Fund [OTKA-81399] to A.V., and by the János Bolyai Research Scholarship of the Hungarian Academy of Sciences to B.L. Work in Freiburg, Germany was supported by an Sonderforschungsbereich [SFB 592] grant to E.S.